journal for constitutional theory and philosophy of law
Ljubljana, 2023
                revija za ustavno teorijo in filozofijo prava
            časopis za ustavnu teoriju i filozofiju prava
           revija za ustavna teorija i filozofija na pravoto

3
journal for constitutional theory and philosophy of law
(2023) 51
  VSEBINA
V tokratni številki so objavljeni prispevki v angleščini in španščini. 
Za slovenske povzetke in ključne besede glej strani 169–177.
David Duarte
 7 Rights as formal combinations of normative 
variables
Miklós Könczöl
 73 Parental proxy voting and political representation
  Bartosz Brożek
 89 Ciencia cognitiva y naturaleza del derecho
Johan Hermansson
 107 Structuring concepts of legal personhood
  On legal personhood as a cluster property
  Ana Lúcia Costa
 131 Vulnerability, sexual violence and the law
  A conceptual analysis
  Kenneth Einar Himma
 151 In defence of the Coercion Thesis and modest 
conceptual methodology
  A reply to critics
 169 POVZETKI IN KLJUČNE BESEDE
revija za ustavn  teorijo in filozofijo prava 

5
journal for constitutional theory and philosophy of law
(2023) 51
  TABLE OF CONTENTS
This issue is composed of papers in English and Spanish. For synopses 
and keywords, see pages 169–177.
David Duarte
 7 Rights as formal combinations of normative 
variables
Miklós Könczöl
 73 Parental proxy voting and political representation
  Bartosz Brożek
 89 Cognitive science and the nature of law
  (in Spanish)
Johan Hermansson
 107 Structuring concepts of legal personhood
  On legal personhood as a cluster property
  Ana Lúcia Costa
 131 Vulnerability, sexual violence and the law
  A conceptual analysis
  Kenneth Einar Himma
 151 In defence of the Coercion Thesis and modest 
conceptual methodology
  A reply to critics
 169 ENGLISH SYNOPSES AND KEYWORDS

7
journal for constitutional theory and philosophy of law
(2023) 51
David Duarte*
Rights as formal combinations of normative 
variables
Hohfeld’s table of legal positions, though highly consistent under his own assumptions, 
seems to be vulnerable to the exact flaws assignable to those assumptions. Specifically, 
the assumptions that a norm is not a necessary condition of a legal position and that one 
single action in a correlativity line is sufficient to bring about the action’s result. With a 
simple proposal of norm individuation, this paper develops a totally norm-based table 
of legal positions in which co-action from correlated agents is also considered (without 
threatening atomicity). Since a legal position is just the result of a combination of norma-
tive variables, the present approach leads to a strictly formal composition of legal posi-
tions in a way that challenges the role played so far by the traditional theories of rights.
Keywords: Hohfeld (Wesley Newcomb), legal norms, legal positions, correlativity lines, 
correlated agents
1 THE LEGAL NORMATIVIST APPROACH: TWO BASIC 
PRESUPPOSITIONS
At the beginning of Normative Systems, which is probably the most sophisticat-
ed piece of legal science written in the 20th century, Alchourrón and Bulygin point 
out the hardware differences between moral and legal systems (at least when the 
contemporary legal systems, combining primary and secondary norms [Hartian 
style], are taken as their archetype).1 Among other reasons, this is because they 
stress the inexistence of a moral legislator, the inexistence of moral courts (which 
also means no official bodies for converting prima facie positions into definitive 
ones), and the significant fact that, as opposed to legal systems, it is particularly 
difficult to identify moral systems and, even more, the norms they entail.2
Accepting how incontestable these reasons seem to be, a relevant distance 
between moral and legal systems has to be recognized. And it follows from that 
distance that the transposition of normative propositions from one field to the 
* davidduarte@fd.ulisboa.pt | Lisbon Legal Theory. University of Lisbon (Portugal). I would 
like to thank my colleagues of the Lisbon Legal Theory, particularly Pedro Moniz Lopes and 
Jorge Silva Sampaio, for their helpful comments on a previous draft of this paper. A very 
special acknowledgment is due to Andrew Halpin, with whom I had wonderful conversations 
about many of the topics addressed in this paper. I would also like to thank two anonymous 
referees for their constructive remarks. Finally, I am also grateful to the Lisbon Public Law 
Research Centre and the Fundação para a Ciência e Tecnologia for their financial support.
1 Alchourrón & Bulygin 1971: 2. On the distinction, see Hart 1994: 79.
2 Alchourrón & Bulygin 1971: 3.
revus (2023) 51, 7–71
8
(2023) 51
journal for constitutional theory and philosophy of law
David Duarte
other is, for the intellectual honesty of the enterprise, particularly delicate: it is 
possible to have true propositions regarding moral norms that are false with 
legal norms (and vice-versa).3 Accordingly, and in order to keep a very precise 
object of inquiry, it is relevant to say that the next pages focus solely on legal 
norms. This means that the present approach to legal positions is strictly based 
on norms belonging to a legal system (no matter which).4
It is also relevant to say that a legal system is understood here as a set of 
norms or, equivalently, that such system is a set that has norms as its mem-
bers.5 And since membership to the set is defined by legality and deductibility, 
it follows that each legal system is a set containing only the norms produced 
through the exercise of competence and those that logically follow from them.6 
Furthermore, deductibility does not merely lead to the membership of other 
norms. Regardless of what composes a complete norm, from that criterion of 
membership equally follows that norms also include the deontic consequences 
somehow implied by a primitive one (to be addressed later on).7
A presupposition of a legal system as a set of norms entailing all its logical de-
ontic consequences is the basis upon which the analysis of legal positions should 
be carried out. It shows, specifically, that the positions “existent” in a legal system 
are limited to those conferred by its norms (and their deontic consequences).8 
Besides removing normative positions that are here irrelevant (such as the moral 
ones), this means that the recognition of a position in a legal system depends (as 
a necessary condition) on a specific deontic modalization of action.9 A table of 
legal positions must include, therefore, only and no other than the various (types 
of) positions that are given by the norms belonging to the set.
3 A good example (among many) is the statement (Marmor 1997: 16) by which one should go 
into the roots of a right in order to determine its limits. Rigorously, if this is acceptable in the 
moral realm, it seems unacceptable in the legal: here, limits of a right are given by the scope 
of the action deontically modalized (as linguistically enacted by the normative authority). The 
roots of a legal right are pre-normative data irrelevant for assessing what is inside or outside 
the scope of a linguistically enacted legal position.
4 Legal positions are a subject matter in which the contingency of legal systems plays almost 
no role (on this contingency, Bulygin 2010: 285, Mendonca 2000: 63). For instance, all the 
examples of norms given in this paper could belong to any legal system.
5 Bulygin & Mendonca 2005: 46, Kelsen 2005: 31.
6 Regarding these norms (criteria) of identification (recognition), Caracciolo 1986: 51, Bulygin 
& Mendonca 2005: 51. On the other hand, “exercise of a competence norm” here means any 
human-made production of norms, be it from the action of a normative authority or by the 
behaviour inherent to custom (and precedent, when it is the case).
7 Alchourrón & Bulygin 1971: 86, Moreso & Navarro 1998: 278.
8 Navarro & Rodríguez 2004: 117. The dependency that legal positions have from norms has 
strong methodological consequences: in some way, no accurate analysis of legal positions can 
be carried out without the frame given by the corresponding norms.
9 Also, strictly addressing the legal domain, Hohfeld 1919: 26, Kramer 1998: 8. On positions 
depending on deontic modalities and action, Sergot 2013: 363, Herrestad 1996: 56.
9
journal for constitutional theory and philosophy of law
(2023) 51
Rights as formal combinations of normative variables
2 WHY THE HOHFELDIAN STRATEGY DOES NOT 
WORK: TOO LITTLE AND TOO MUCH
Within the knowledge that the Hohfeldian analysis gave to legal science one 
specific point is a clear game changer: atomicity (understood as the irreducibility 
of a legal position). While searching for the disambiguation of the word “right”, 
Hohfeld arrived at a kind of “indivisible unit”, from which he was able to create 
a table of legal positions in which all can be predicated as atomic.10 Atomicity 
is, undeniably, an outstanding tool. It provides the means to understand com-
plex legal positions, to understand the components they are composed of and, 
fundamentally, to understand the connection between an irreducible normative 
position and the (single) deontic modality that stands as its ground.
Whatever the future of the Hohfeldian legacy, one thing seems certain: once 
atomicity is lost (or it is not taken into full account), no improvements what-
soever can be made. The reason is simple: losing atomicity means losing the 
unique basis capable of explaining the morphological diversity of a molecular 
legal position.11 Therefore, whenever no entailment is provided, assertions that 
claim-rights contain a claim to enforcement (for instance) are a way of playing 
a different game.12 So, when taken seriously, atomicity draws a line that cannot 
be crossed (at least if cognitive growth is to be considered by legal science as a 
relevant epistemic value within the present topic).
It is known that Hohfeld made some mistakes. That is the case with oppo-
sites instead of contradictories, when he says that a liberty is the opposite of a 
duty (as presented in his scheme of jural relations). Since, however, when one 
is true the other is false, a liberty is the contradictory of a duty (and not its 
opposite).13 That seems to be also the case when he says that a no-right is the 
correlative of a liberty (something said in the exact same sentence): bearing in 
mind the square of oppositions, the correlative of a liberty is a no-right not (a 
no-right is actually the correlative of a liberty not).14 As is also visible in other 
parts of his seminal work, Hohfeld uses unilateral permissions but not always 
with complete accuracy.
10 Atomicity, then, stands for irreducibility (or indivisibility), which is what Hohfeld actually 
meant with the predicate “fundamental”. Hohfeld 1919: 64, Halpin 1985: 436.
11 Since only atomicity explains the internal difference (through their atomic components) be-
tween distinct aggregated normative positions. Lindhal 2006: 331, Biasetti 2015: 954.
12 A problem transversally harming the whole will theory (Kramer 1998: 64). Examples of mo-
lecularity without considering atomicity appear, for instance, in Jaffey 2006: 140 or Steiner 
1994: 73. On the virtues of atomicity, Schlag 2015: 218.
13 Hohfeld 1919: 38. On this point, Kocourek 1921: 27, O’Reilly 1999: 275.
14 Hohfeld 1919: 38. Correcting the so-called Hohfeldian mistakes: Williams 1956: 1138; Robin-
son, Coval & Smith 1983: 277.
10
(2023) 51
journal for constitutional theory and philosophy of law
David Duarte
(i) the following squares show the Hohfeldian positions already adjusted; 
the squares show the known contradictories (ct) and correlatives (co) in 
Hohfeld’s table of first order legal positions:
These mistakes are not, however, particularly problematic. What is problem-
atic in the Hohfeldian understanding is his specific engineering of correlativity, 
designed with a strategy of negations and contradictions, leading to the insuf-
ficient outcome of deontic positions without any independent content: the case 
of claim-rights and no-rights (now only regarding first order legal positions). 
To put it differently, the point rests on a conception of correlativity that, as a 
material equivalence, says nothing about the content of those positions and the 
deontic status of their holders: claim-rights and no-rights have no content be-
sides being the correlatives of duties and liberties.15
(ii) a claim-right is “only” the correlative of a duty and has no deontic con-
tent: no one knows which deontic modality the holder is under; differently, 
the duty holder is in the position corresponding to being under an obliga-
tion; this might be expressed by the following scheme of correlativity:
claim-right to φ ↔ Oφ → duty to φ
(iii) a no-right is “only” the correlative of a liberty not and has no deontic 
content: no one knows which deontic modality the holder is under; differ-
ently, the liberty holder is in the position corresponding to being under a 
permission; this might be expressed by the following scheme of correlativity:
no right to φ ↔ Pφ → liberty to φ
15 Which is to say that duties and liberties have no content besides being the correlatives of 
claim-rights and no-rights (for instance, Kramer 1998: 13, Cruft 2004: 351). The main reason 
for addressing correlativity from the perspectives of claim-rights and no-rights rests on the 
(relevant) fact that, unlike duties (obligation) and liberties (permission), they do not stem 
from a deontic modality of their own. On the strict correlative content of the Hohfeldian legal 
positions, Eleftheriadis 2008: 115, Arriagada 2018: 14.
11
journal for constitutional theory and philosophy of law
(2023) 51
Rights as formal combinations of normative variables
Such emptiness in two of the four first order Hohfeldian legal positions has 
deeper causes, though. Besides a false symmetry in each line of correlativity, the 
outcome of legal positions without an independent content apparently comes 
from a simplistic and rather reductive conception of normative action (uncrit-
ically accepted).16 Actually, for Hohfeld, each line of correlativity ligates two 
holders sharing one single action that is assigned to only one of them. This is 
because Hohfeld thought that the exercise of a single action was sufficient to 
bring about its result, a premise that led him to disregard any other possible 
correlated action (by the holder of the correlated position).17
(iv) for Hohfeld, the duty ↔ claim-right line of correlativity is based on the 
single action φ (paying one-thousand dollars or staying off the land) assigned 
to the duty bearer; for Hohfeld, φ-ing is a sufficient condition to bring about 
the result (own the money or stay off the land).
(v) for Hohfeld, a liberty ↔ no-right line of correlativity is based on the sin-
gle action φ (eating a shrimp salad or occupying Whiteacre) assigned to the 
liberty holder; for Hohfeld, φ-ing is a sufficient condition to bring about the 
result (have eaten the salad or occupied Whiteacre).
Hohfeld’s understanding of action seems, however, to fall short. When con-
sidering that no result can be brought about without an interaction between 
both correlated holders, Hohfeldian correlativity is unable to accommodate the 
fact that some co-action from the no-right or the claim-right holders is (if pos-
sible) a necessary condition to achieve the result of the main action.18 And if 
we accept this, we probably have to admit that norms present correlativity as a 
structural scheme with space for co-action and, furthermore, with space for its 
specific deontic modalization.19 But if this is somehow correct, then Hohfeldian 
correlativity is clearly insufficient.
16 On that asymmetry, Halpin 2019a: 86, Lyons 1970: 45. This asymmetry has nothing to do with 
the one pointed out under non-atomic views of legal positions (Hart 1982: 184 or Simmonds 
1998: 221). On the other hand, it seems also out of target to criticize the Hohfeldian symmetry 
based on the alleged flaws of correlativity (as apparently Frydrych 2019: 6 does). One thing 
is to know if the Hohfeldian correlativity is symmetric; another to know if it is (universally) 
valid for all duties or all liberties.
17 It is important to note that by “result of action” it is meant (as in von Wright 1963: 39) the state 
of affairs produced by the exercise of the type of action at hand (end-state of change) and the 
state of affairs in which that performance consists of (change). Similarly, Hornsby 2004: 18.
18 In the Hohfeldian scheme, since action is only assigned to one agent (Finnis 2011: 378, 
D’Almeida 2016: 556), there is no such thing as action (or specifically co-action) carried out 
by the no-right or the claim-right holders (Kramer 1998: 14, Williams 1956: 1145).
19 A problem already posed and analyzed by Lindahl (1977: 126; 2001: 160) for claim-rights 
and widely discussed regarding interferences in liberties (for instance, Lyons 1970: 54, Curtis 
Nyquist 2002: 251).
12
(2023) 51
journal for constitutional theory and philosophy of law
David Duarte
(vi) the Hohfeldian duty ↔ claim-right line of correlativity does not accom-
modate the fact that φ-ing by the duty bearer might be insufficient to bring 
about the result; if co-action from the claim-right holder is also necessary 
(e.g., meeting to receive the money), then correlativity implies an independ-
ent space for co-action by the claim-right holder and a specific deontic mo-
dalization for such co-action.
(vii) the Hohfeldian liberty ↔ no-right line of correlativity does not accom-
modate the fact that φ-ing by the liberty holder might be insufficient to bring 
about the result; if co-action from the no-right holder is also necessary (e.g., 
not grabbing the salad eater), then correlativity implies an independent space 
for co-action by the no-right holder and a specific deontic modalization for 
such co-action.
Showing positions without content or disregarding co-action are not, 
though, the only problems in the Hohfeldian system of jural relations. Looking 
into the table of second order legal positions, some other problems become vis-
ible, namely, in his disability ↔ immunity line of correlativity (but not only). 
For Hohfeld, an immunity is the correlative of a disability, which is the absence 
of competence. As a kind of right, an immunity is, then, the position one has 
when facing someone without the power to change a legal position.20 It repre-
sents, consequently, the other Janus face of a not having power: it belongs to 
a “symmetry” between not being habilitated to produce deontic consequences 
and not being targeted by them.
Such a conception is, however, problematic. Independently of the concep-
tion of competence one may adopt, there is no understanding of such a notion 
that is able to justify that the contradictory of a position based on constitutivity 
(a disability) has to be seen as a legal position in itself. Accordingly, it is also dif-
ficult to understand how not having a competence can be the basis for a corre-
lated legal position (immunity) that, in the same way, has never been conferred 
by any norm of the set.21 Therefore, the strategy of negations and contradictions 
here seems to have taken Hohfeld too far. And one wonders whether he was not 
actually confusing norms with normative propositions.22
20 Hohfeld 1919: 60. On Hohfeld’s immunities, Edmundson 2004: 91, Corbin 1919: 170.
21 This point was already unveiled in Kocourek 1921: 34 and developed later on in Kocourek 
1923: 154. The question at hand is whether an immunity is not just a total absence (where 
“Alice does not have a pet” would be correct), differently from a position negatively described, 
the possible case of a no-right (where “Alice has a pet that is not a dog” would be correct). For 
this approach to Kocourek, Halpin 1985: 441.
22 On the difference between norms and normative propositions (the difference between pre-
scribing “X has a right to φ” and describing “there is a right to φ held by X”), Bulygin 2015a: 
196, Niiniluoto 1991: 368.
13
journal for constitutional theory and philosophy of law
(2023) 51
Rights as formal combinations of normative variables
3 NORM INDIVIDUATION: A BRIEF SCHEME
Assuming that a legal system is a set of norms implies that each norm is an 
element of the set and that “a norm” is the unit of the system. Considering that 
norms are enacted by normative authorities using a natural language and that 
no match exists between an utterance and a norm, the immediate problem of 
norm individuation is posed. In simple terms, the problem involves determin-
ing what a norm is or, more accurately, what a complete norm is. At stake is 
not only the issue of how to represent the legal system for some explanatory 
purposes, but to define under which criteria can one say that there is a norm. In 
other words, to list the necessary conditions to have a norm.23
The relevancy of norm individuation is unquestionable.24 If it is unavoid-
able for many tasks performed in legal science, namely, to classify norms or to 
understand how norms conflict with each other, it is in the field of legal posi-
tions that norm individuation becomes decisive: to know which legal positions 
are given by a norm depends on knowing what a (single) norm is. On the other 
hand, norm individuation is also a remarkable tool for eliminating usual mis-
understandings similar to those of breaching atomicity: withdrawing from one 
norm, without criteria, several disconnected positions.25 Individuation thus 
provides the key to knowing which positions (no less and no more) are con-
ferred by a (single) norm.
It is recognized that norm individuation faces the problem that law has no 
criteria for such demarcation: normative authorities enact norms without con-
sidering norms as units and the written materials they present are solely organ-
ized under some linguistic conventions.26 A single norm sentence matching one 
norm is, therefore, a matter of chance (under some criterion of individuation). 
Given that the law has no natural kind serving this specific purpose, it follows 
that norm individuation has to be constructed as a scientific convention.27 In 
other words, only an intersubjective agreement on what a complete norm is 
(within the scientific community) may allow us to mean the same thing when 
we speak about a norm.
23 Circumscribing norm individuation to those explanatory purposes, D’Almeida 2015: 159. On 
the broader scope of norm individuation, Bentham 1945: 247, Raz 1980: 74.
24 A relevancy inversely proportional to the attention the topic has deserved so far. On the rel-
evancy of norm individuation, Spaak 2003: 97, James 1973: 359.
25 A good example is the norm in the sentence “human life is inviolable” (included in some con-
stitutions): besides its reading as a prohibition to kill, it is sometimes seen as also including 
an imposition (directed to the legislator) to not revoke the criminalization of murder. How 
the very same norm can contain a prohibition regarding two entirely different actions (killing 
and revoking norms), which are moreover assigned to different addressees, remains, however, 
unexplainable.
26 On the frame of enacting norms, Frändberg 2018: 7. On its linguistic process, Duarte 2011: 111.
27 Raz 1980: 141. On norm individuation as a task of legal science, Spaak 2003: 98.
14
(2023) 51
journal for constitutional theory and philosophy of law
David Duarte
Among the criteria already presented to achieve such goal, it seems that the 
Razian strategy of looking for the minimal notion is by far the best.28 This is not 
only because it leads to an individuation closer to its usual understanding with-
in the legal practice, but, and primarily, because it implies all and no more than 
the necessary conditions for regulating behaviour: it is presented, consequently, 
as the atomistic kind or, in other words, as the “necessary morphology” when 
action has to be deontically modalized.29 Under this approach, a (complete) 
norm is (or it is suggested to be, for conventional purposes) the sum of the nec-
essary conditions for an action to be permitted, forbidden, or mandatory.
3.1 A norm: Material elements
With this frame, necessary conditions for regulating behaviour are: (i) an ac-
tion; (ii) a deontic modalization of that action; and (iii) the conditions under 
which the deontic modalization of the action depend. It follows that regulation 
of behaviour is impossible if one of these conditions is absent (some explanations 
for the third will be given later on). These three necessary conditions lead to the 
known elements of norm structure, which in their conditional order are: (i) an-
tecedent; (ii) deontic operator; and (iii) consequence.30 For present purposes, the 
complete analysis of each one is unnecessary. To understand how norms confer 
legal positions, only some features of those elements have to be considered.
(viii) the norm sentence (NS) “when one loses faith in humanity it is manda-
tory to listen to Mozart’s Magic Flute” has the norm (N) “when one loses faith 
[…]” (antecedent [lfh]), “it is mandatory” (deontic operator [O]) to “listen to 
Mozart’s Magic Flute” (consequence [lmf]): “lfh ⇒ O lmf ”.
(ix) any change in the content of each element creates a different norm (iden-
tity of norms based on the contingency of each element’s content); in the 
antecedent: “wr (when it rains) ⇒ O lmf ”; in the deontic operator: “lfh ⇒ F 
(forbidden) lmf ”; or in the consequence: “lfh ⇒ O s (sing)”.
Regarding the antecedent, the main point here is that, independently from 
other conditions expressed by the normative authority, a condition inherent 
to the possibility of exercising the action foreseen in the consequence has to 
be therein recognized.31 This means that, regardless of others (necessarily ex-
pressed), all norms contain the (usually unexpressed) condition of the opportu-
nity to perform the action deontically modalized.32 Thus, when this condition is 
28 Raz 1980: 72.
29 Something Raz points out in his relevant third guiding requirement (Raz 1980: 144).
30 Sartor 2009: 36, Nino 2003: 72.
31 von Wright 1963: 73.
32 As von Wright (1963: 73) puts it, it is the condition for performing the corresponding elemen-
tary acts (individual acts). On the conditions of the antecedent, Duarte 2012: 38.
15
journal for constitutional theory and philosophy of law
(2023) 51
Rights as formal combinations of normative variables
not fulfilled (the action is not possible), the norm at hand is not applied, despite 
being in force like all the others within the set; when the condition is fulfilled 
(the action is possible), it is a necessary (and possibly sufficient) condition for 
triggering the consequence.
(x) in “when one loses faith in humanity it is mandatory to listen to Mozart’s 
Magic Flute” (formally, lfh ⇒ O lmf), the antecedent has one expressed con-
dition: “when one loses faith […]”; yet, it also has an opportunity condition 
(o) inherent to the action “to listen”: it has to be read as “o ∧ lfh ⇒ O lmf ”.
(xi) in “o ∧ lfh ⇒ O lmf ”, the condition “o” is necessary but not sufficient to 
the obligation of listening to be applied (given “lfh”); however, in “everyone 
is allowed to sing” (“o ⇒ P s), the condition “o” is necessary and sufficient: 
whenever the action is possible, it is (effectively) permitted.
(xii) a N such as “o ⇒ P s” (everyone is allowed to sing), although in force at 
some point in space and time, it is only applied whenever the condition “o” is 
filled; there is no opportunity of singing when the addressee is a person in a 
coma: in such a situation, “o ⇒ P s” is not triggered.
Given its consequences on the analysis of legal positions, providing some 
precision to what is meant by the “opportunity” condition is of the utmost im-
portance. Rigorously, it is not a mere reference to an external state of affairs 
where the exercise of the action at hand becomes possible (which is already pre-
supposed). Differently, it is a condition entirely related to the mental and physi-
cal possibilities of acting, intrinsically linked to the “ought implies can” that un-
derlies law.33 Therefore, the deontic modalization of an action is only triggered 
whenever its performance is, in this sense, “internally possible”. Of course, it is 
not a matter of personal skills as well, but a strict matter of (human) ability.34
(xiii) the opportunity condition in “everyone is allowed to sing” (o ⇒ P s) is 
not a reference to an external context, such as a score to follow (if one wants 
to sing something not memorized); it is related to the internal possibility of 
acting (singing): it is, then, an internal possibility condition (ip).
(xiv) the internal possibility condition in “ip ⇒ P s” is not related with per-
sonal skills as well; if Josepha is not capable of singing “Der Hölle Rache” as 
“Queen of the Night”, this does not mean that the internal possibility condi-
tion is not met: she is nevertheless capable of singing.
Regarding the deontic operator, two points are relevant here. The first point 
concerns the fact that it is within this element that the deontic status of the ac-
33 On “ought implies can”, von Wright 1963: 108, Hintikka 1970: 84.
34 Which means that, from now on, “opportunity” only means “internal possibility”. On the 
topic, von Wright 1963: 50, Nordenfelt 1997: 153.
16
(2023) 51
journal for constitutional theory and philosophy of law
David Duarte
tion foreseen in the consequence is defined. So, and using the square of deontic 
modalities, it follows that action φ can be forbidden, mandatory, positively per-
mitted, or negatively permitted.35 Given the accepted meaning of these words, 
from a prohibition it follows that the agent has a duty to ~φ; from an obligation, 
a duty to φ; from a positive permission, a liberty to φ; and from a negative per-
mission, a liberty to ~φ.36 Of course, action is here always understood both as an 
action in the strict sense (e.g., singing) and as an omission (e. g., not singing).37
(xv) the deontic status of an action φ gives rise to specific legal positions; 
therefore, obligations to φ and to ~φ confer the duties to φ and to ~φ (con-
traries [c]) and permissions to φ and to ~φ confer the liberties to φ and to ~φ 
(implicated [i]).
(xvi) the following square shows the correspondence between the four deon-
tic modalities and the legal positions they confer with regard to some given 
action φ:
The second point concerns the deontic status of the “contradictory action” 
(the deontic status of ~φ when only φ is regulated).38 The point is that there are 
good arguments to sustain that, like prohibitions and obligations, permissions 
can also be understood as a complete deontic modality.39 That is, in the same 
35 Alchourrón & Bulygin 1971: 60, Frändberg 2018: 101.
36 O’Reilly 1995: 285, Hurd & Moore 2018: 309.
37 Hilpinen 1997: 85, Bach 2010: 50.
38 The predicate “contradictory” is used here because when φ is true ~φ is false. As it is known, 
the terminology is unstable (namely, “contrary action” or “complementary action”). Alchour-
rón & Bulygin 1971: 66, Ratti 2018: 87.
39 A complete deontic modality regarding action. It is, then, what can be designated as the “ac-
tion completeness” of permissions.
17
journal for constitutional theory and philosophy of law
(2023) 51
Rights as formal combinations of normative variables
way that the prohibition of φ entails ~φ to be mandatory, and the obligation of 
φ entails ~φ to be forbidden, the permission of φ can also be seen as ~φ to be 
permitted as well.40 An understanding of permissions such as this one means 
that being permitted to φ (or ~φ) entails both φ and ~φ and that the agent has 
the “privilege” of choice: the agent has to decide her own course of action.41
(xvii) a bilateral permission is a complete deontic modality since it aggre-
gates two unilateral permissions: with “everyone is allowed to sing” (ip ⇒ P 
s), both the act “singing” and the omission “not singing” are permitted; thus, 
the square of oppositions can be transformed into a triangle:
Two main reasons allow reading permissions as bilateral. The first reason 
regards the fact that, once the equivalences between O φ and F ~φ and between 
F φ and O ~φ are assumed, it turns out that permissions also ask for their own 
completeness. Consequently, one has to know if the permission of φ implies ~φ 
to be permitted as well or, differently, to be forbidden.42 However, since when 
φ is permitted we do not know if ~φ is permitted or forbidden, and it would be 
rather odd to assume that it is forbidden (because absence about ~φ naturally 
leads to a weak permission), at least one can say that if the normative authority 
40 Manero 2018: 44, Raz 1975: 161.
41 For the following triangle, Halpin 2003: 49. On the “facultative” operator (bilateral permis-
sion), Guibourg, Echave & Urquijo 2008: 135.
42 Action completeness is, naturally, relevant: given the unstable and also less informative nature 
of half permissions (Halpin 2014: 208), the agent has to know the deontic status of ~φ in order 
to manage her behaviour when under the permission of φ. D’Almeida (2016: 564) has recently 
criticized the Halpinian reduction (the triangle); as he says: since legal systems may fail to be 
consistent and that a conflict of legal duties may always arise, half liberties are not less stable 
or less complete than duties. It seems, though, that he misses the point. Stating the instability 
and the incompleteness of half permissions is a statement independent from conflicts of legal 
positions: differently from a duty, a half liberty is unstable and incomplete even if no conflicts 
arise. At the prima facie level, a prima facie duty is stable and complete and a prima facie half 
liberty is not.
18
(2023) 51
journal for constitutional theory and philosophy of law
David Duarte
does not explicitly state the prohibition, then ~φ has to be understood as per-
mitted.43
The second reason rests on the undisputed fact that the meaning of the 
word “permission” (or the meaning of the deontic verbs used to express it), 
when normative authorities enact norm sentences, is clearly bilateral. In oth-
er words, whenever such authority states action φ as permitted, the pragmatic 
understanding of the community of speakers is that ~φ is permitted as well. 
Accordingly, since norms are linguistically expressed, it would not make much 
sense to think that the shared meaning of “permission” could somehow be sub-
stituted by its logical significance.44 Accepting that permissive norm sentences 
confer bilateral permissions seems to be, then, pragmatically unavoidable.
(xviii) with “everyone is allowed to sing” (ip ⇒ P s) there is a permission to 
carry out the action “to sing”; the norm sentence pragmatically means “P s 
∧ P ~s”.
(xix) the triangle is now a line of contradiction; each imposition of φ or of ~φ 
contradicts the permission of φ and~φ; the permission of φ and ~φ contra-
dicts either with the imposition of φ or of ~φ:
O φ ⊻ φ ↔ P (φ ∧ ~φ)
As for the consequence, the initial observation is that it is necessarily related 
to action, the object of the deontic modalization at hand.45 Therefore, it makes 
no difference whether a norm authority simply describes a type of action (ought 
to do) or a state of affairs to be achieved (ought to be), as it is irrelevant whether 
it describes a specified action or a set of unspecified ones.46 The same is even 
valid for the more distant cases of a “tû-tû” type of consequence: one may find 
here not only the action of conferring (or protecting) a status, but also the ac-
tions foreseen in the (remitted) subset of norms the status contains.47 So, no 
matter the description, the consequence always regards action.
43 Ratti 2018: 92. Actually, for ~φ to be forbidden, φ would have to be mandatory, meaning that 
the normative authority was imposing something with a permissive deontic verb (which is, 
somehow, non-sensical). On weak and strong permissions, Hansson 2013: 201, Alchourrón 
& Bulygin 1991: 217.
44 Ratti 2018: 92, Guibourg, Echave & Urquijo 2008: 134.
45 Von Wright 1963: 71, Ross 1968: 107.
46 For the distinction between ought to do and ought to be, Castañeda 1972: 675. For the dis-
tinction between specified and unspecified actions, Lopes 2017: 476.
47 On “tû-tû”, Ross 1957: 816. Also, Brożek 2015: 16.
19
journal for constitutional theory and philosophy of law
(2023) 51
Rights as formal combinations of normative variables
(xx) an “ought to do” norm has a consequence about action: in “it is forbid-
den to lie (l)” (F l), the action is “lying”; an “ought to be” norm has a conse-
quence also about action(s): in “it is mandatory to be virtuous (v)” (O v), “v” 
amounts to “carry out any act suitable to reach the state of virtuousness”.
(xxi) a consequence with a specified action is about action: “everyone is al-
lowed to sing” (P s); but the same is valid with a consequence with unspeci-
fied actions: “everyone is allowed to make music (mm)” (P mm) refers to 
actions (to play the Glockenspiel ∨ to sing “Der Vogelfänger bin ich ja” ∨ …).
(xxii) a “tû-tû” consequence is also about action: “virtuous people have the 
right to belong to Sarastro’s Brotherhood” contains the action of conferring 
the “status (s)” (O s) and also remits to actions foreseen in the status subset of 
norms (O to tell the truth ∧ O to act with virtue ∧ …).
Actions foreseen in the consequence of a norm can be divided between de-
ontic and non-deontic. A deontic action is one that produces deontic conse-
quences; that is, the action that, operating at the legal level, changes the law or 
the existing legal positions (direct legal consequences).48 A non-deontic action 
does not produce those consequences: it only works at the level of the empirical 
world.49 When legal consequences follow from a non-deontic act, that is due to 
the fact that such act triggered the consequences foreseen in other norms (indi-
rect legal consequences). Accordingly, and only taking into account what they 
are, non-deontic acts never imply direct legal consequences.
(xxiii) when a normative authority states that “it is forbidden to sing inside 
Sarastro’s Temple”, the law has been changed: a deontic action (to forbid sing-
ing inside the Temple) has been carried out; it is an action that directly pro-
duces legal consequences (independently from triggering other norms).
(xxiv) when the Three Ladies offer Tamino the Magic Flute, all their legal 
positions have been changed: a deontic action (transfer of ownership) has 
been carried out; it is an action that directly produces legal consequences 
(independently from triggering other norms).
(xxv) when Tamino moves from Java to Egypt and becomes a resident there, 
he has just carried out a non-deontic act; moving did not produce any direct 
legal consequences: changes in his legal positions are due to the fact that he 
now fills the antecedents of several Egyptian norms.
48 A deontic action is, then, the exercise of a competence norm. Ross 1958: 216, Hage 2005: 222.
49 A non-deontic action is the action carried out under any other norm (besides competence 
norms). This distinction matches the one between “normative action” and “action” in von 
Wright 1963: 75.
20
(2023) 51
journal for constitutional theory and philosophy of law
David Duarte
A final observation regarding action is necessary. Regardless of being a gen-
eral or a particular norm, or containing a specified action or a set of unspecified 
ones, an action normatively foreseen is always a generic action: that is, it is a 
category (of action) that can be empirically instantiated (at some point in space 
and time).50 So, when a norm modalizes action φ as forbidden, it means that all 
individual acts of φ-ing are prohibited, no matter other contextual variables or 
properties. On the other hand, φ being mandatory and doing φ are facts from 
different worlds: it is only the individual act of φ (as a token) that will contribute 
to realizing the ideal state of affairs the obligation of φ is.
(xxvi) a NS such as “it is forbidden to lie (l)” expresses a N (ip ⇒ F l) foresee-
ing in its consequence the generic action of “lying”; whenever Papageno lies 
(in a space-time uniqueness [when he says that he was the one who killed the 
serpent]), he is actually instantiating (and violating) “ip ⇒ F l”.
(xxvii) “ip ⇒ F l” is an ideal state of affairs: whenever its conditions are 
met, some act ought to be omitted; realizing such ideal state of affairs de-
pends, however, on a match with the real state of affairs (for instance, when 
Papageno says that the serpent was effectively killed by the Three Ladies).
3.2 A norm: Subjective components
Any deontic content establishing that an action, under some conditions, is 
permitted, forbidden, or mandatory is incomplete if it does not demarcate a 
circle of addressees: someone has to be permitted, prohibited, or obliged to do 
something. Such triviality points, though, to the relevant fact that norms also 
have a subjective element or, in other words, that addressees are also a necessary 
condition for the regulation of behaviour.51 Independently of its specificities (for 
now), a set of people has to be identified at the connection between the deontic 
operator and the consequence: thus, “someone” has the duty to ~φ whenever φ 
is forbidden. This “someone” set may be designated as “primary addressees”.52
(xxviii) the NS “it is forbidden to steal the Magic Flute (smf)” contains a N 
such as “ip ⇒ F smf ”, a prohibition that seems to be addressed to everyone; 
however, and regardless of how to delimitate its addressees, this norm fore-
sees a set of primary addressees {PA}.
50 von Wright 1963: 36. Even action foreseen in an individual norm seems to be generic (dif-
ferently, von Wright 1963: 23). The fact that a “genus” has only one token does not prevent it 
from being a “genus” (Davidson 2001: 142).
51 Bulygin & Mendonca 2005: 16, Ross 1968: 107.
52 The subsequent notation of norms (and because of addressees) follows the suggestions made 
in Herrestad & Krogh 1995: 2011 namely to make correlativity more explicit.
21
journal for constitutional theory and philosophy of law
(2023) 51
Rights as formal combinations of normative variables
(xxix) “ip ⇒ F smf ” could be written as “∀x (xsmfip ⇒ F xsmf)” (for all x, 
if there is an internal possibility to smf to all x, it is forbidden to smf to all 
x); for the reasons already mentioned, another notation is used : “ip ⇒ F {PA} 
smf ”, where the brackets stand for the set of primary addressees.
Identifying only a set of addressees is still a reductive understanding of 
norms. Once correlativity is taken into account, it follows that norms also com-
prise a second set, specifically, the one containing those holding the correlative 
position. Recognizing a second set of addressees has various (and complex) con-
sequences on the analysis of norms, which must, however, be seen as an implica-
tion of accepting correlativity: if “someone” has the duty to φ (under the imposi-
tion of φ) and this means that “someone correlative” has the right to φ, then it 
follows that two positions are being given by the exact same norm.53 This “some-
one correlative” set may be designated as the set of “secondary addressees”.
(xxx) in “ip ⇒ F smf ” (it is forbidden to steal the Magic Flute), a set of pri-
mary addressees was already acknowledged: those that are forbidden from 
stealing the magical instrument the Three Ladies gave to Tamino; therefore, 
that N has to be read as “ip ⇒ F {PA} smf ”.
(xxxi) however, once there is a set of persons forbidden from stealing the 
magical instrument, there must be a second set of people correlatively con-
nected with the prohibition (the owners of the instrument): they are the sec-
ondary addressees {SA}; the norm can be formalized as: “ip ⇒ F {PA} smf {SA}”.
3.3 A complete norm: Adding up the elements
Adding the previous elements is enough to present what might be seen (for 
norm individuation purposes) as a complete norm. As just said, it comprises 
three material elements and two subjective ones, each one of them taken as a 
necessary condition for the regulation of behaviour. In the former, an action 
(with the mentioned variations), the conditions under which the deontic mo-
dalization depend (omnipresent due to the internal opportunity condition), 
and the deontic modalization itself. In the latter, the set of primary addressees 
and, given correlativity, the set of secondary addressees. Accordingly, regulation 
provided by a norm depends on the specific contents a normative authority as-
signs to each one of these elements.54
53 On addressees and correlativity, Frydrych 2019: 15, Lyons 1970: 54. Also, next point 9.
54 It should be noted that this norm morphology is a suggestion of norm individuation merely 
regarding legal positions. It is insufficient for broader purposes: known problems of norm 
individuation remain unconsidered and unexplained. For instance, what to do with excep-
tions (whether negative conditions are the antecedent of another norm or not), how to deal 
with disjunctive or conjunctive conditions in the antecedent (besides the internal possibility 
22
(2023) 51
journal for constitutional theory and philosophy of law
David Duarte
(xxxii) the N in “Sarastro’s Brotherhood has the duty to provide health care 
(hc) to citizens” has five elements: in “ip ⇒ O {Sarastro’s Brotherhood} hc {citizens}”, 
an antecedent, a deontic mode, a consequence, and two sets of addressees are 
recognizable (however, the antecedent is not expressed).
(xxxiii) the N in “everyone is allowed to sing” has five elements: in “ip ⇒ P 
{physical persons} s {SA}”, an antecedent, a deontic operator, a consequence, and 
two sets of addressees are recognizable (however, one set of addressees and 
the antecedent are not expressed).
4 FIRST ORDER POSITIONS: LIBERTY AND DUTY NOT 
TO DEFEAT THE INTERNAL POSSIBILITY
A norm with a permission to carry out an action confers the primary ad-
dressee the legal position commonly called a “liberty”.55 Therefore, and regard-
ing an action φ, this means that the addressee is allowed to do φ, which is also 
valid to the forbearance from acting if the permission is connected with the 
consequence ~φ. Whenever a permission is understood as bilateral (the com-
mon situation in contemporary legal systems, as mentioned above), a liberty 
comprises both φ or ~φ. In such a normative scenario, holding a permission 
involves the inherent choice of φ-ing or ~φ-ing. So, a liberty is a legal position 
of personal autonomy: it is the way to give the agent an option regarding her 
own course of action.
(xxxiv) in “everyone is allowed to sing” (ip ⇒ P {PA} s {SA}), a primary ad-
dressee, for instance Pamina, has the liberty to sing; since it contains a bilat-
eral permission, this norm allows Pamina to sing and not to sing (both are 
permitted by the same norm).
condition) or whether a consequence with alternative types of action signifies one or more 
norms (for these and other problems of individuation, D´Almeida 2009: 399, Duarte 2012: 
39). However, individuation options such as these are here less relevant: for legal positions, 
what has been said is enough. On the other hand, it should also be noted that a relevant 
epistemic value is obtained here by clarifying the methodological presupposition that norm 
individuation is: it offers the tools to distinguish what regards norms and what regards the 
positions given by norms. It becomes clear, for instance, that claim-rights including enforce-
ment is a problem of norm individuation and that liberties being the contradictory of duties 
is just a strict matter of legal positions.
55 For known reasons (Williams 1956: 1131, Wenar 2005: 226), “liberty” is effectively a better 
designation than the Hohfeldian “privilege” (Hohfeld 1919: 42).
23
journal for constitutional theory and philosophy of law
(2023) 51
Rights as formal combinations of normative variables
(xxxv) the bilateral permission in “ip ⇒ P {PA} s {SA}” gives each primary ad-
dressee a sphere of personal autonomy: thus, as one of them, Pamina may 
decide her course of action inside the scope of what is being permitted (a 
choice is given about singing).
The scope of a liberty is entirely demarcated by the generic action foreseen 
in the consequence. Accordingly, a specific liberty comprises the exercise of all 
possible variables of the action at hand: notwithstanding the difficulty in quali-
fying some individual acts as tokens of the generic action foreseen, this does 
not deny, despite such borderline cases, that to be permitted to do φ is to be 
permitted to do φ whatever φ-ing can be. No doubt, some variables of φ can 
intersect the antecedent of other norms and cause a conflict between norms; 
that is, however, a different problem.56 At the prima facie level of a liberty, what 
matters is that when φ is permitted, it is permitted regardless of “how” the ac-
tion is carried out.
(xxxvi) the scope of the liberty conferred by “ip ⇒ P {PA} s {SA}” is demarcated 
by all variables of the action “to sing”: this means that Pamina is allowed to 
“~s”, to “s1”, to “s2”, and so on, or to “s1a” or “s1b”, or “s2c” or “s2d”, and so forth; 
given the endless number of variables and their endless possible combina-
tions, variables of the action “to sing” are infinite.
(xxxvii) within that scope, some variables of action may intersect the anteced-
ent of other norms (with their own deontic modalization); thus, if Pamina 
sings something seriously attacking the honour of Monostatos, then she also 
triggers the N in “it is forbidden to harm personal reputation”; however, her 
singing is still a prima facie permitted act exclusively under “ip ⇒ P {PA} s {SA}”.
Accepting that the exercise of a liberty comprises all variables of the action at 
hand does not mean that the permissive norm gives a different legal position (a 
different liberty) for each one of the possible courses of action. Understanding 
the difference between the legal position itself and the variables of the action 
foreseen is crucial. While variables of action regard the various ways by which 
the agent can perform the action, the legal position conferred by the permissive 
norm is always unique: an agent is always exercising the exact same liberty to φ 
when she chooses to do φ in one way or another.57 It nearly goes without saying, 
then, that variables of action are not averse to the atomicity of a specific liberty.
56 On conflicts and the underlying distinction between prima facie and all norms considered 
rights, Borowski 2000: 43, Kumm 2007: 139.
57 Halpin 1997: 120, Ross 1958: 184. On the other hand, this is not to be confused with the 
Hohfeldian distinction between paucital and multital positions (Hohfeld 1919: 73). While the 
paucital position refers to the uniqueness of the same position due to a directed correlativ-
ity, the premise above refers to the uniqueness of the position despite all the possible tokens 
24
(2023) 51
journal for constitutional theory and philosophy of law
David Duarte
(xxxviii) with “ip ⇒ P {PA} s {SA}”, variables such as “~s”, “s1”, “s2”, “s1a” “s1b”, 
“s2c”, or “s2d”, for instance, are just different ways to exercise the liberty to “s”; 
no matter which is the combination of variables chosen by Pamina in the 
individual act she always exercises the same atomic position.
As follows from the internal possibility condition, the exercise of a liberty 
is dependent on such possibility: the holder of the liberty to φ can exercise the 
legal position if she is in a condition to do so. Such internal possibility is, then, 
a necessary condition for the exercise of a liberty. However, if an impossibility 
of exercising that position might arise from some limitations of the holder, it 
can also arise from the action of third parties whenever they defeat the internal 
possibility of the liberty holder to act.58 Accordingly, if in the former there is no 
connection between the holder and the secondary addressees, in the latter, it is 
a fact that impossibility is only due to how these behave regarding the holder’s 
act.
(xxxix) with “ip ⇒ P {PA} s {SA}”, Pamina has the liberty to sing; if Pamina is 
in a comatose state, she cannot sing, a scenario preventing the norm from 
being triggered: although in force, the permission is not applied at all; here, 
secondary addressees are, however, irrelevant for such effect.
(xl) however, Pamina can be prevented from singing because Monostatos has 
knocked her unconscious; such scenario prevents the norm from being trig-
gered as well: the permission is not applied to her; this impossibility, how-
ever, is totally due to the action of Monostatos.
An internal impossibility due to how secondary addressees act means that 
they control that condition of the antecedent: if they defeat possibility, there is 
an impossibility; if they do not defeat it, then the condition can be met and the 
holder of the liberty can choose whether to act or not. With this scheme, sec-
ondary addressees become decisive for the liberty in itself.59 And when a SA set 
agent defeats the internal possibility of a liberty holder, she is not just cancelling 
the exercise; she is, furthermore, denying the deontic fact that the liberty holder 
is permitted to act. Ultimately, such defeat turns out to be exactly the same as 
removing that specific liberty holder from the set of PA agents.
of the generic action foreseen. On paucital and multital rights in Hohfeld, Barker 2018: 592, 
Frydrych 2018: 330.
58 Von Wright 1963: 55. Also, Navarro 2019: 69, Spena 2012: 171.
59 It is important to note that defeating the internal possibility is not the result of a conflict be-
tween the permission and any other norm. Under the present norm individuation approach, 
only the permission at hand is being considered: therefore, it is the possibility to prima facie 
exercise the liberty that can be defeated.
25
journal for constitutional theory and philosophy of law
(2023) 51
Rights as formal combinations of normative variables
(xli) with “ip ⇒ P {PA} s {SA}”, Pamina has the liberty to sing; when Monostatos 
knocks her unconscious, preventing her from singing, he is preventing her 
from exercising the liberty at hand; however, Monostatos is also preventing 
“ip ⇒ P {PA} s {SA}” from applying on such occasion.
(xlii) when Monostatos defeats Pamina’s internal possibility to sing and, con-
sequently, prevents “ip ⇒ P {PA} s {SA}” to be applied, Monostatos affects much 
more than her singing; this becomes clear when seeing that his act amounts 
to changing the norm into “ip ⇒ P {PA unless Pamina} s {SA}”.
An example such as this, shows how the Hohfeldian liberty ↔ no-right line 
of correlativity is incomplete: φ being permitted is not a sufficient condition for 
the liberty to be applied (for a token of φ to be the case). And it is incomplete 
because another necessary condition is required: restricting the secondary ad-
dressees to not defeat the internal possibility to act.60 It is only with this im-
position that the normative authority is not enacting a “self-defeasible norm”: 
a norm that can be defeated by an agent belonging to the SA set (and without 
any conflict of norms). The “deontic completeness” of the permission depends, 
thenceforth, on taking in the duty not to defeat.61
(xliii) by defeating Pamina’s internal possibility to sing, Monostatos prevents 
“ip ⇒ P {PA} s {SA}” from being applied; however, by changing that norm into 
“ip ⇒ P {PA unless Pamina} s {SA}”, Monostatos shows the norm at hand as “self-
defeasible”: it can be defeated by an agent therein foreseen.
(xliv) thus, the completeness of “ip ⇒ P {PA} s {SA}” depends on a duty im-
posed on SA set agents to not defeat the internal possibility of PA set agents 
to act; this duty is, then, a necessary condition to such completeness: without 
it, applying a permission depends on the will of any SA set agent.
An example such as the one above also shows that the Hohfeldian liberty ↔ 
no-right line of correlativity amounts to a violation of Hume’s guillotine: law is 
effectively changeable (and without power) whenever any member of the SA set 
wants that to be the case.62 Any SA set agent is free (apparently under a weak 
permission) to remove any agent from the PA set, modifying the one originally 
60 Which is a way to not accept the strange consequence Sartor (2006:106) seems to accept: “the 
very fact that an action is permitted is often sufficient to ensure a possibility of performing 
that action”. In other words, it is a way to not accept that a permission sometimes confers a 
liberty and sometimes it does not.
61 From which it follows that the overall completeness of permissions depends on: (i) being 
bilateral, the previously mentioned “action completeness”, and (ii) being associated with a 
duty not to defeat, the now mentioned “deontic completeness”. Accordingly, the latter is a 
completeness regarding the sum of the necessary conditions for a permission to be deonti-
cally exercisable (and kept as enacted).
62 On Hume’s guillotine, Woleński 2008: 21, Black 1969:101.
26
(2023) 51
journal for constitutional theory and philosophy of law
David Duarte
formulated by the normative authority.63 The same cannot be said, however, 
when the liberty correlates with a duty not to defeat: with this correlation, law is 
not changed by a SA set agent; it is violated. So, when a permission is complete, 
a SA set agent preventing a PA set agent from exercising her liberty is simply 
breaching a duty.
(xlv) by changing the norm into “ip ⇒ P {PA unless Pamina} s {SA}” (or creating 
the same effect), Monostatos (an agent without power) also shows that the 
Hohfeldian “no-right” is incompatible with Hume’s guillotine: a mere fact 
(knocking Pamina unconscious) equals changing the law.
(xlvi) such assessment ceases to be valid if “O not to defeat” (addressed to 
SA set agents) is associated to “ip ⇒ P {PA} s {SA}”; with this duty, Monostatos 
does not change “ip ⇒ P {PA} s {SA}” when he knocks Pamina unconscious; he 
is simply violating the duty not to defeat.
All these considerations lead to what can be called “the permissions fork”. 
That is: (i) either it is assumed that normative authorities enact complete per-
missions (deontic completeness) and the liberty conferred to PA set agents cor-
relates with a duty not to defeat (held by SA set agents); or (ii), contrarywise, 
and for the reason of being almost deontically irrelevant, it is assumed that per-
missions are not “regulative norms” at all.64 This is so since: (i) any SA set agent 
is free to block or compel the exercise of a liberty (not breaching the law while 
doing so); and (ii) any SA set agent may effectively change any permission en-
acted by a normative authority (and without power to carry out such change).
(xlvii) thus, “ip ⇒ P {PA} s {SA}” is either a complete permission (and it is once 
“O not to defeat” is imposed upon SA set agents) or it is irrelevant regarding 
the regulation of behaviour: being within the system or not amounts to ex-
actly the same regarding the agents the norm itself foresees.
Therefore, if normative authorities enact complete permissions (actually, 
mere “regulative permissions”), then it follows that the liberty conferred to PA set 
agents necessarily correlates with a duty not to defeat the internal possibility (held 
by SA set agents).65 So, when the internal possibility condition and the counter-
63 On the no-right as a space of weak permission, Brown 2005: 344, Lindahl 2006: 335.
64 And “almost deontically irrelevant” since a positive permission will always contradict a prohi-
bition and a negative permission will always contradict an imposition (Alchourrón & Bulygin 
1991: 229). However, the occurrence of a contradiction is inconsequential for an effective 
regulation of action.
65 A proposition merely working at the conventional level of norm individuation and not pre-
supposing any logical implication between permissions and “prohibitions to defeat “. On the 
common non sequitur regarding the prohibition of interferences in liberties, Hohfeld 1919: 
43, Lyons 1970: 52.
27
journal for constitutional theory and philosophy of law
(2023) 51
Rights as formal combinations of normative variables
party’s co-action are taken into account, it becomes clear that, regarding other 
agents, tokens of φ are only exercisable if defeating is understood as forbidden.66 
Given the duality of necessity and sufficiency, and being such duty a necessary 
condition to the deontic completeness of permissions, it follows that a liberty to φ 
is a sufficient condition for the SA set agents to bear a duty not to defeat.67
(xlviii) if “O not to defeat” is a necessary condition to “ip ⇒ P {PA} s {SA}” to 
be applied, then it follows that “ip ⇒ P {PA} s {SA}” is a sufficient condition to 
“O not to defeat”; therefore, “ip ⇒ P {PA} s {SA}” implies that SA set agents are 
under a duty not to defeat the internal possibility of singing.
A scheme such as this has the consequence that a permissive norm, as the ba-
sis for a liberty, also has to take into account the deontic modalization governing 
the behaviour of SA set agents. Therefore, besides the original sequence of an an-
tecedent followed by a deontic operator and a consequence, a norm that confers 
a permission of φ also entails a sort of “second norm” expressing the (additional) 
necessary deontic condition for the completeness of the liberty at hand. It is im-
portant to note, however, that such complement depends exactly on the same 
condition: preventing the exercise of a liberty also depends on its own internal 
possibility and, for this reason, it becomes irrelevant whenever it is impossible.
(xlix) the N in “everyone is allowed to sing” can be formalized, associating 
the implied “O not to defeat”, as “(ip ⇒ P {PA} s) → (ip ⇒ O {SA} ~dip)”, where 
“dip” stands for “defeat the internal possibility” of exercising the main action 
(that is, regarding the action of PA set agents).
(l) as co-action, not to defeat also has its own “internal possibility”; when 
Monostatos is hidden inside the forest, he cannot knock Pamina unconscious 
(she is elsewhere); therefore, not to defeat is a necessary condition to bring 
about the result of action only whenever it is possible.
As follows from all these considerations, the position correlated with a liber-
ty is closely connected with how permissive norms are understood. And, once 
it is assumed that permissions also entail a correlativity line as any other norm, 
then the facts that the Hohfeldian “no-right” dissolves the difference between 
strong and weak permissions and presupposes pointless permissive norms must 
also be acknowledged. Hence, along with the solid reasons that lead to seeing 
permissions as “action complete”, there also seem to be compelling reasons to 
do the same regarding their “deontic completeness”: recognizing a duty not to 
66 By “co-action”, it is meant the interdependent interaction without which the result of the main 
action cannot be brought about (not joint action). Tuomela 1984: 270, Pörn 1977: 77.
67 On the necessity-sufficiency duality, Sanford 1989: 179, von Wright 1951: 66.
28
(2023) 51
journal for constitutional theory and philosophy of law
David Duarte
defeat is the only way to understand permissions as norms effectively regulating 
behaviour.68
(li) one can conceive a legal system where (time1) everything is weakly per-
mitted except singing (forbidden); however, when the normative authority 
enacts the N expressed by “everyone is allowed to sing” (time2), the forbid-
den action becomes strongly permitted.
(lii) since the Three Ladies do not want Papageno to sing (in order to punish 
him for his lies), they are free in time1 to put a padlock on his lips; the prob-
lem is that under the liberty ↔ no-right correlativity they are still free to do 
it in time2; so, the permission enacted is pointless.
(liii) and it is pointless because: (i) a possible revocation of the prohibition 
(creating a weak permission) would be exactly the same as enacting the 
strong permission; and (ii) the system shows itself pragmatically identical in 
time1 and in time2 (it is legitimate to use the padlock in both).
Therefore, a permissive norm creates a line of correlativity connecting a “lib-
erty” to a “duty not to defeat the internal possibility”. Within this correlativity, 
secondary addresses are under such duty, which means that their atomic posi-
tion also has: (i) a specific content, and (ii) a specific deontic modalization.69 
Regarding the latter, a permission connected to a set of agents puts the agents of 
the correlative set under the contradictory deontic modality: while members of 
the PA set are permitted to do φ (and ~φ), members of the SA set are under an 
obligation not to defeat the internal possibility of φ-ing (and ~φ-ing and even 
choosing between φ and ~φ).
(liv) the answer to the legal question, “how ought Monostatos behave regard-
ing Pamina’s act of singing” is given by the permission itself; from “(ip ⇒ P 
{PA} s) → (ip ⇒ O {SA} ~dip)” it follows that Monostatos is under an obligation 
not to defeat her internal possibility to do so; evidently, all SA set agents are 
under such obligation (the contradictory of the permission foreseen to PA 
set agents).
As regards the former (the content of the duty not to defeat), it is worth 
noting that it is inherently connected with the internal possibility condition 
68 The next example is a variation of the famous Guibourg’s “Toro Sentado” example (Guibourg, 
Echave & Urquijo 2008: 155). It is worth to note, though, that the original example seems to be 
partially misleading: the effect recognized to permissions when Toro Sentado appoints a subor-
dinated Minister (permissions prevent the enactment of contradictory norms by the Minister) 
actually does not come from the permission in itself. Differently, it comes from the superiority of 
the permission enacted by Toro Sentado given by the hierarchy between the two bodies. On the 
original example, Alchourrón & Bulygin 1991: 236, Atienza & Manero 1998: 91.
69 Unveiling the topic (but solving it outside atomicity), Lindahl 1977:126. Also, Sergot 2013: 400.
29
journal for constitutional theory and philosophy of law
(2023) 51
Rights as formal combinations of normative variables
(understood as above), from which it follows that it is strictly demarcated by 
the acts that somehow block or compel the liberty holder regarding the exer-
cise of her position (and just that).70 So, a duty not to defeat, being related to a 
main action φ only assigned to the first pole of the correlativity line, comprises 
a variable content (φ´) formed by whatever can cease the mental and physi-
cal possibilities of φ-ing and ~φ-ing (and choosing). Understandably, whatever 
might go beyond such scope has to be seen as already outside the content of this 
(adaptative) duty.
(lv) “defeating the internal possibility” is merely to defeat the internal pos-
sibility condition, which is the same to say that Pamina’s mental and physical 
possibilities to sing have to be kept intact; anything beyond this scope is not 
within the object of the duty.
(lvi) thus, if Pamina’s singing becomes unnoticed because Monostatos is si-
multaneously singing in a particularly brilliant way, he is not breaching his 
duty not to defeat her internal possibility; he is just exercising his own per-
mission to sing.
Accordingly, it is totally outside the content of the duty not to defeat the un-
derstanding by which this legal position involves the acts suitable to undermine 
the purposes intended by the liberty holder: conceivable acts that may prevent 
the success aimed by the holder with the exercise of her liberty have nothing to 
do with the internal possibility condition.71 A duty not to defeat is not about 
making the exercise of the liberty pointless, but making it impossible: it is only 
in this way that the duty bearer blocks the internal possibility. Therefore, the 
usual criticism based on examples that merely illustrate possible ways of mak-
ing the acts of the liberty holder unsuccessful seems totally off target.72
(lvii) when Papagena and Papageno both run to pick up the Glockenspiel 
that fell on the floor, Papagena is not defeating Papageno’s internal possibility 
to run if she runs faster and picks up the Glockenspiel first (under “freedom 
of movement”); she would be defeating his internal possibility if she grabbed 
him to a point that he could not effectively run.
70 Defeating the internal possibility comprises both actions and forbearances as it includes them 
regarding a positive or a negative exercise of the liberty (preventing φ or ~φ). It involves, 
accordingly, and in Makinson´s terminology, an agent indexed and non-modulated interfer-
ence (Makinson 1986: 417).
71 Spena 2012: 170 (although regarding the duty of non-interference).
72 Such as running faster (Simmonds 1998: 157) or speaking louder (Kramer 1998: 11). With 
similar examples, Williams 1956: 1144, Wellman 1997: 3.
30
(2023) 51
journal for constitutional theory and philosophy of law
David Duarte
4.1 Duty not to defeat and duty not to: Adaptative content
Bearing in mind the content of the duty not to defeat, as was roughly de-
scribed above, it is important to know whether there is any difference between 
such duty and the duty not to (correlated with a claim-right). An answer to this 
question must be clearly affirmative: any confusion between them seems mistak-
en and no overlapping can be recognized (except, of course, for the mere fact that 
they are both duties). The first reason is that, within the norm’s morphology (the 
position addressees have), a duty not to defeat is held by SA set agents while the 
holders of a “normal” duty not to are PA set agents (by definition). Rigorously, a 
duty not to defeat imposed on primary addressees just does not exist.
(lviii) “everyone is allowed to sing” foresees a duty not to defeat the inter-
nal possibility to SA set agents; “it is forbidden to lie” foresees a duty not 
to assigned to PA set agents; it is never the case that a duty not to defeat is 
imposed on PA set agents or that a duty not to is imposed on SA set agents.
It could be said that nothing prevents a normative authority from formulat-
ing a norm in a way that the duty not to defeat the internal possibility is assigned 
to PA set agents; for instance, “it is forbidden to interfere in the freedom to φ“, 
an utterance that, moreover, would include a consequence with unspecified ac-
tions: to not interfere.73 However, such an argument seems to confuse the sen-
tence (not written in the conditional order of a norm) and the norm itself: such 
wording does not prevent the facts that a liberty is being given (a liberty as any 
other) and that the duty is totally related to a token of φ.74 Irrespective of how 
it is written, the norm expressed imposes a duty not to defeat on SA set agents.
(lix) in “it is forbidden to interfere in the freedom to φ”, a duty not to de-
feat the internal possibility is apparently given to PA set agents; however, the 
wording of the NS does not disturb the conditionality of the N, which effec-
tively is “(ip ⇒ P {PA} φ) → (ip ⇒ O {SA} ~φ´)”; materially, to not interfere in φ 
is also a content-adaptative duty dependent on what is (or will be) the PA set 
agent token of φ.
As follows from the first reason, a second and decisive reason arises from the 
difference regarding the content of the duty. Accordingly, while a duty not to re-
gards an action somehow foreseen in a consequence, the content of a duty not to 
defeat the internal possibility is only apprehensible upon the act to be carried out 
by the liberty holder. It follows, then, that the duty not to defeat has an adaptative 
73 The example above is presented as a possible expression of the duty not to defeat imposed on 
PA set agents; and it is not a strict theoretical hypothesis: for instance, the norm foreseen in 
article 5/6 of the Brazilian constitution.
74 On mismatches between norm sentences and norms, Grabowski 2009: 130, Schauer 1991: 23.
31
journal for constitutional theory and philosophy of law
(2023) 51
Rights as formal combinations of normative variables
content, meaning that its scope is dependent on (and conditioned by) the pos-
sible individual acts to be taken by the PA set agent.75 In other words, the content 
of a duty not to defeat is “content-adaptative” and, for this reason, is shaped by 
the counterparty’s φ-ing or ~φ-ing (which is never the case with a duty not to).
(lx) in “it is forbidden to make music”, a duty not to is imposed on PA set 
agents (norm-shaped duty); its content is to forbear the exercise of the action 
foreseen (here, unspecified actions); it is not conditioned, though, by the pos-
sible course of action to be taken by SA set agents; thus, Papageno omitting to 
play the Glockenspiel or singing happens irrespective of how third parties act.
(lxi) in “it is allowed to make music”, a duty not to defeat is imposed on SA 
set agents (counterparties-shaped duty); its content regards the internal pos-
sibility condition, being adaptative to the PA set agent act; thus, the Three 
Ladies defeating Papageno’s internal possibility depends on his exercise of 
the liberty: if he sings, putting a padlock on his lips (not successful if he plays 
the Glockenspiel).
4.2 Liberties are protected: Weak and strong protections
Accepting that a liberty correlates with a duty not to defeat the internal pos-
sibility, merely on the basis of a single norm, means that there is no such thing 
as “naked” (or “unprotected”) liberties.76 As soon as a norm allows action φ, it 
is immediately the case that the liberty conferred is already protected by a duty 
not to defeat, preventing the exercise of the liberty to be legitimately disturbed 
in its internal possibility. It is surely a weak protection, since it merely covers a 
reduced (although decisive) spectrum of the set of possible accomplishments 
offered by the position. Nonetheless, beyond that scope, no protection is given 
by the duty not to defeat (solely from the permission at hand).
(lxii) “(ip ⇒ P {PA} s) → (ip ⇒ O {SA} ~dip)” stands for the N expressed in 
“everyone is allowed to sing”; as follows from its content, Papageno has a lib-
erty to sing and all the SA set agents have the duty not to defeat his internal 
possibility to do so (to not put a padlock on his lips, for instance); evidently, 
Papageno is not protected (only with that norm) against Papagena not listen-
ing to him or against Tamino singing so loud that Papageno actually fails to 
seduce Papagena with his marvellous voice.
Nothing prevents a normative authority from increasing the protection a 
liberty already grants by itself and from introducing into the set other norms 
forbidding specific (and non-adaptative) ways of interference: that is, enacting 
75 On the adaptative nature of some interaction, Tuomela 2007: 152. Also, Miller 2004: 5.
76 On naked (or unprotected) liberties, Bentham 1838: 218, Alexy 2002: 144.
32
(2023) 51
journal for constitutional theory and philosophy of law
David Duarte
prohibitions to act against any of the variables of the action permitted.77 With 
such third norms, the degree of protection given to a liberty increases, provid-
ing a higher level of protection to the liberty at hand. So, instead of distinguish-
ing between naked and vested liberties (or unprotected and protected), it seems 
that the proper distinction is between liberties with weak and strong protec-
tions: in the latter, other norms reinforce the protection already offered by the 
duty not to defeat.
(lxiii) by conferring a duty not to defeat, “(ip ⇒ P {PA} s) → (ip ⇒ O {SA} ~dip)” 
immediately gives some protection to the liberty: it is a weakly protected lib-
erty; when the normative authority enacts “absolute silence is mandatory 
when someone sings”, Papageno’s liberty to sing becomes strongly protected: 
besides the weak protection (as seen), he now also has the protection given 
by this latter norm (it becomes a strongly protected liberty [Tamino cannot 
sing when Papageno is singing]).
Since other legal positions are given by those norms reinforcing the protec-
tion the liberty already has, it follows that a strongly protected liberty is actually 
an aggregate of atomistic legal positions, specifically the liberty in itself and all 
the claim-rights conferred by each one of the prohibitions. If X has the liberty to 
φ and Y is forbidden to φ´ (an action against the exercise of that liberty), then 
the strongly protected liberty X has is the sum of her liberty with the claim-right 
towards Y not to do φ´. And this seems to be the proper scheme to explain Hart’s 
protective perimeter: its content is just the aggregate of claim-rights that legal 
systems confer when actions that might trouble liberties become forbidden.78
(lxiv) as seen, “everyone is allowed to sing” with “absolute silence is manda-
tory when someone sings” makes that liberty strongly protected; however, it 
is an aggregate of atomistic positions: the liberty to sing with the claim right 
to “absolute silence”.
(lxv) Papageno’s protective perimeter is formed, regarding his liberty to sing, 
by a compound of claim-rights coming from, and for instance, “absolute si-
lence is mandatory when someone sings”, “it is forbidden to cut one’s vocal 
cords”, or “it is forbidden to kill”, and so forth.
A unique exception to such a scheme is found in weak permissions: those that 
pragmatically cover an action not explicitly allowed (not regulated). Without a 
norm covering an action, it seems that such action may be carried out (oper-
ating the absence of law as a pragmatic permission) even though no intrinsic 
77 Manero 2018: 45, Ross 1958: 166.
78 Hart 1982: 171. On the other hand, this also shows why Hart’s protective perimeter is irrel-
evant within an atomic approach to legal positions.
33
journal for constitutional theory and philosophy of law
(2023) 51
Rights as formal combinations of normative variables
protection exists for such exercise.79 A main reason sustains this “exception” to 
the correlativity between liberties and duties not to defeat: it would be rather 
anomalous (and here rather incoherent) to accept that one may have a duty not 
stated by any norm of the set. Thus, if no duty can follow from the absence of a 
norm, then these “vacuum liberties” do not even have a weak protection.
(lxvi) within a legal system where no norm allows wearing bird-feather 
clothes (and no general permission of action exists, of course), Papagena’s 
“liberty” to dress as she does is a “vacuum liberty”: she may dress that way 
only because the absence of a norm stands for pragmatic permission.
(lxvii) Papagena’s “liberty” to wear bird-feather clothes, under the conditions 
now identified, does not correlate with a duty not to defeat: “vacuum liber-
ties” are “empty” in the sense that they are not based on norms that could 
support the ascription of duties to third parties.
5 FIRST ORDER POSITIONS: DUTY AND  
CLAIM-RIGHT
When a norm has an imposition or a prohibition, a duty is normatively con-
ferred to agents of the PA set; if it is an imposition, it is duty to φ (duty to), 
but if it is a prohibition, it is a duty to ~φ (duty not to).80 Independently of the 
generic action foreseen in the consequence, the duty is to be exercised through 
the individual act performed by an agent of the PA set: it is her conduct (an act 
in the strict sense or an omission) that leads to bringing about a certain result 
(somehow, the “service” to be provided to the holders of the correlated position 
[“claim-right” holders]).81 So, impositions and prohibitions give rise to the cor-
relativity line duty ↔ claim-right (the second one in first order positions).
(lxviii) “it is mandatory to pay 500 florins (£500) per month to personal 
employees” expresses a N imposing the duty of employers to pay £500 per 
month and the claim-right of employees to receive that amount of money; 
this N sustains a duty ↔ claim right correlativity line; so, as an agent of the 
PA set, Sarastro has the duty to pay £500 per month to Monostatos (his per-
sonal employee), who is, as a SA set agent, the (a) claim-right holder.
79 Besides the previous references, Bulygin 2010: 285, Mendonca 2007: 54. On the universal law 
of liberty, Bentham 1970: 120. On the other hand, it is exactly the presupposition that such 
universal law gives rise to legal positions that explains the completeness of the Hohfeldian 
normative system (Biasetti 2015: 955).
80 O’Reilly 1995: 285, Sreenivasan 2010: 466.
81 Borrowing the famous Benthamian expression (Bentham 1970: 57).
34
(2023) 51
journal for constitutional theory and philosophy of law
David Duarte
(lxix) “it is forbidden to enter into Sarastro’s Temple during a solemn assembly 
of the Brotherhood” expresses a N imposing on everyone the duty not to en-
ter into the Temple during those assemblies and conferring the Brotherhood 
the claim-right to others to stay out; this N sustains a duty ↔ claim right cor-
relativity line; so, as an agent of the PA set, Papagena has the duty to stay out 
of the Temple (in the condition mentioned), being the Brotherhood, as the 
SA set agent, the claim-right holder.
In the structure of imposition or prohibition norms, there is also an ante-
cedent; and, independently of other conditions possibly expressed, there is also 
an internal possibility condition. As happens with permissive norms, this is a 
necessary condition for the norm to be applied and, consequently, for the duty 
to be the case.82 In the same way as with permissions (as discussed above), the 
individual act that stands for the compliance of the duty cannot take place (no 
matter if an imposition or a prohibition) if the duty bearer faces an internal 
impossibility. It can be said, therefore, that the possibility of a token of the main 
action is also a necessary condition for the exercise of the duty.
(lxx) the N in “it is mandatory to pay 500 florins (£500) per month to person-
al employees” can be initially formalized as “ip ⇒ O {PA} £ 500/month {SA}”; 
naturally, if Sarastro has his hands tied when he is going to pay Monostatos, 
he cannot give the money to his personal employee.
(lxxi) the N in “it is forbidden to enter into Sarastro´s Temple (est) during a 
solemn assembly (sa) of the Brotherhood” can be initially formalized as “ip ∧ 
sa ⇒ F {PA} est {SA}”; naturally, if Papagena is knocked unconscious and taken 
into the Temple, she did not enter: her body has been put there.
There is, however, a significant difference in the role played by the internal 
possibility condition in permissions, on the one hand, and in impositions and 
prohibitions, on the other. The point is the following: while in permissions such 
possibility is a necessary condition to bring about the result of action, in impo-
sitions and prohibitions it is just a contributory condition; that is, the internal 
possibility is only a necessary condition to act, which is in itself another neces-
sary condition to bring about the result at hand.83 Putting differently: in norms 
that are impositions and prohibitions, the antecedent’s internal possibility con-
dition is merely the first step in a chain of necessary conditions.
82 Von Wright 1963: 74.
83 Contributory condition (in causation) as a condition within a set of necessary conditions that 
are jointly sufficient, von Wright 1951: 148, Broad 1944: 16.
35
journal for constitutional theory and philosophy of law
(2023) 51
Rights as formal combinations of normative variables
(lxxii) in “(ip ⇒ P {PA} s) → (ip ⇒ O {SA} ~dip)”, not to defeat the internal pos-
sibility (not putting a padlock on Papageno’s lips) is a necessary condition to 
bring about the result of action: if Papageno can open his lips, he can chose to 
sing and effectively sing (or even refraining from singing).
(lxxiii) in “ip ⇒ O {PA} £500/month {SA}”, the internal possibility (Sarastro not 
having his hands tied) is only, however, a necessary condition to another one 
(in order to bring about the result of action): besides his hands being free, 
Sarastro only pays Monostatos if he gives him the money.
(lxxiv) in “ip ∧ sa ⇒ F {PA} est {SA}”, the internal possibility (Papagena not be-
ing knocked unconscious) is only, however, a necessary condition to another 
one (in order to bring about the result of action): besides being conscious, 
Papagena only omits entrance if she does not enter.
As a mere contributory condition, the internal possibility loses its relevancy 
in impositions and prohibitions: given the direction of causation and its asym-
metry, the second necessary condition (to act or omit) succeeds the first one 
(internal possibility), which means that the first condition of the chain is con-
sumed by the exercise of the duty.84 As is clear in the following examples, any 
exercise of the duty imposed or prohibited necessarily presupposes the presence 
of the internal possibility condition. But such causal direction is not inconse-
quential: it implies to this correlativity line that SA set agents deal specifically 
with the most proximate condition (in this correlativity line, the individual act 
under the duty).
(lxxv) in “ip ⇒ O {PA} £500/month {SA}”, the result of action is “transferring 
the money to the SA set agent”; being free to pay is prior to “transferring the 
money to the SA set agent” (the latter presupposes the former, but not vice-
versa); thus, correlativity here connects SA set agents with the most proxi-
mate condition: their role in this correlativity line (co-action) directly faces 
the action foreseen.
(lxxvi) in “ip ∧ sa ⇒ F {PA} est {SA}”, the result of action is “staying off the prop-
erty of the SA set agent”; being conscious is prior to “staying off the property of 
the SA set agent” (the latter presupposes the former, but not vice-versa); thus, 
correlativity here connects SA set agents with the most proximate condition: 
their role in this correlativity line (co-action) directly faces the action foreseen.
84 On the direction of causation and its consequences, Sanford 1984: 57, Mackie 1965: 261.
36
(2023) 51
journal for constitutional theory and philosophy of law
David Duarte
(lxxvii) in “(ip ⇒ P {PA} s) → (ip ⇒ O {SA} ~dip)”, the result of action is to 
(choose to) sing; having a padlock on the lips was the obstacle to that result 
(no other condition is necessary); thus, correlativity here connects SA agents 
with that condition: the role played by SA set agents in this correlativity line 
(co-action) directly faces the internal possibility condition (not the main ac-
tion foreseen).
An analysis of action within the duty ↔ claim-right correlativity line shows 
that the result of action is also dependent on the specific co-action to be carried 
out by the holder of the claim-right: irrespective of the type of action foreseen, 
the result at hand cannot be brought about without the SA set agent interde-
pendent agency.85 An observation such as this points out that the SA set agent 
also plays a sort of “constitutive role” in this correlativity line (totally ignored in 
Hohfeld’s scheme): actually, it is a variable contribution to the result that varies 
from effective cooperation (in which the agent acts by way of response) to mere 
passive action (in which the agents just accepts some external course of action)86.
(lxxviii) under “ip ⇒ O {PA} £500/month {SA}”, Sarastro has to pay Monostatos 
£500 this month; for Monostatos to have the money, Monostatos probably 
has to meet his employer or, at least, if Sarastro goes to meet him, to put the 
money in his pocket (or under his mattress); if Sarastro pays by bank trans-
fer, Monostatos has the passive action of not cancelling his bank account: he 
agrees with a state of affairs (have a bank account) in order to have another 
one (own £500).
(lxxix) under “ip ∧ sa ⇒ F {PA} est {SA}”, Papagena has to stay out of Sarastro’s 
Temple; co-action is necessary to bring about the result; if a member of the 
Brotherhood compels Papagena to enter into the Temple, she will definitely 
enter (she probably does not if the member refrains from doing it); it is to 
be considered as well that any member of the Brotherhood might invite her 
to enter: in such case, and irrespective of anything else, a SA set agent is evi-
dently contributing to prevent the result of action.
As can been seen in these examples, SA set agents must co-act in order to 
bring about the result of the act carried out by the PA set agent. However, such 
co-action also depends on its own internal possibility: if the SA set agent can-
not co-act, it follows that her contribution to the result is absent.87 When this 
is the case, one of two consequences follows: either (i) the duty held by the PA 
85 On this interdependency (common effect requirement), Schmid 2009: 229, Pacherie 2011: 174.
86 On cooperative agency (i-intentions), Tuomela 2000: 4, Miller 2004: 5. On passive action, 
Mele 1997: 137, Zhu 2004: 299.
87 Impossibility to co-act might arise from the most different causes (e.g., the SA set agent is 
imprisoned or in a coma). However, it is always the case when the PA set agent does not act 
37
journal for constitutional theory and philosophy of law
(2023) 51
Rights as formal combinations of normative variables
set agent is not (fully) complied with, although as a result of the SA set agent’s 
impossibility; or (ii) the result is still achieved despite the impossibility (omis-
sion works). Be it as it may, whenever co-action is possible, the SA set agent 
contribution is decisive. So, and again, it can be said that co-action is necessary 
whenever it is possible.
(lxxx) under “ip ⇒ O {PA} £ 500/month {SA}”, Sarastro has to pay Monostatos 
£ 500 this month; if Sarastro does not meet Monostatos to pay him, it becomes 
irrelevant what Monostatos does (and he does not receive the money); but if 
Sarastro pays by bank transfer, impossibility to co-act does not block the result 
(yet, if there is possibility, Monostatos can always cancel his bank account).
(lxxxi) under “ip ∧ sa ⇒ F {PA} est {SA}”, Papagena has to stay off Sarastro’s 
Temple; if a member of the Brotherhood cannot compel Papagena to enter 
into the Temple, she either enters or not only depending on her judgment; 
however, if a member of the Brotherhood has the possibility to interfere in 
the compliance of the prohibition, then refraining from doing it is necessary 
to the result.
As just mentioned, whenever it is possible, SA set agents have to co-act to 
bring about the result of the (main) action assigned to PA set agents. And in a 
close parallel with the duty not to defeat (in permissions), the content of their 
position is also adaptative: it is not normatively foreseen, and it also depends on 
how the PA set agent exercises her duty.88 Accordingly, the position held by SA 
set agents in this correlativity line is only definable by the possible acts that co-
ordinate with the act of the duty bearer when she is complying with her duty. In 
one way or another, a SA set agent will act or omit as a way to “manage the result 
of action” that is to be brought about with (and not only by) the duty bearer.89
(lxxxii) “ip ⇒ O {PA} £500/month {SA}” and “ip ∧ sa ⇒ F {PA} est {SA}” both con-
fer a duty ↔ claim-right correlativity line; to bring about the result of action 
(to pay and to stay off), SA agents have to co-act in a variation coming from 
passive action to effective cooperation.
(violating the duty). Given the chain of necessary conditions, not complying with the duty 
amounts to surpass one (non-final) necessary condition.
88 Unpredictability about these co-acts decreases in proportion to what is normatively foreseen 
(the more the detail about the compliance of the duty, the less the unpredictability about the 
SA set agent co-action). Nonetheless, irrespective of the degree of detail, co-action is always 
adaptative.
89 By “managing the result of action” it is meant the adaptative co-acts carried out by the claim-
right holder towards the actual or future exercise of the main action by the duty bearer (re-
garding the result of such action).
38
(2023) 51
journal for constitutional theory and philosophy of law
David Duarte
(lxxxiii) with “ip ⇒ O {PA} £500/month {SA}”, the claim-right holder (Mono-
statos) “manages the result of action” by effectively receiving or not receiv-
ing the money; as seen, Monostatos can accept the money and put it in his 
pocket, refuse to accept it, cancel his bank account, and so on.
(lxxxiv) with “ip ∧ sa ⇒ F {PA} est {SA}”, the claim-right holder (the 
Brotherhood) “manages the result of action” by letting Papagena stay out; 
as seen, any member of the Brotherhood can compel Papagena to enter (or 
not), might invite her, or even agree with her entrance.
Given that the content of a claim-right is to “manage the result of action” (a 
structural feature of the whole correlativity line), it seems that the SA set agent 
is free regarding the possible co-acts that can consubstantiate her contribution 
to bring about the result of action (either acts or forbearances).90 And, rigor-
ously, since the SA set agent is under no duty whatsoever (neither enacted, nor 
implied), it follows that such agent may either co-act to bring about the result of 
action or may also refrain from doing it. Thus, a duty assigned to a PA set agent 
correlates with a permission held by the claim-right holder, showing that this 
correlativity line also presents two contradictory deontic modalizations.
(lxxxv) considering the object and the deontic modalization of the claim-
right position, the N expressed by “it is mandatory to pay 500 florins (£500) 
per month to personal employees” can be formalized as: “(ip ⇒ O {PA} £500/
month) → (P {SA} mra)”, where “mra” stands for “manage the result of action”; 
this means, consequently, that SA agents are prima facie allowed “to accept” 
and “not to accept” the result of the action to be carried out by PA agents.
(lxxxvi) considering the object and the deontic modalization of the claim-right 
position, the N expressed by “it is forbidden to enter the Sarastro’s Temple dur-
ing a solemn assembly of the Brotherhood” can be formalized as: “(ip ∧ sa ⇒ 
F {PA} est) → (P {SA} mra)”, where “mra” stands for to “manage the result of ac-
tion”; this means, consequently, that SA agents are prima facie allowed “to ac-
cept” and “not to accept” the result of the action to be carried out by PA agents.
5.1 Waiving: Weak and strong waiving and non-waivability
Within the duty ↔ claim-right correlativity line, the term “waiving” has 
been used with some ambiguity. It seems to be used as meaning either: (i) to 
authorize the duty bearer not to act (weak-waiving); or (ii) to extinguish the 
90 It is relevant to note that omissive co-action is not absence. Since there is always an inter-
nal opportunity condition (presupposing an external opportunity), omission has the “factual 
proximity” that differentiates one from the other (Bernstein 2016: 2577, Clarke 2012: 137). 
Accordingly, the Brotherhood is not omitting (in its co-action) when Papagena is at the forest 
with Papageno (the Brotherhood is just doing nothing simpliciter).
39
journal for constitutional theory and philosophy of law
(2023) 51
Rights as formal combinations of normative variables
legal position (strong-waiving).91 Such ambiguity, along with preventing accu-
racy and precision, also prevents the identification of a major difference: in fact, 
allowing not to act is something quite different from extinguishing the position. 
In the former, the claim-right holder agrees with the non-compliance of a token 
of the duty, but the position remains totally in force. In the latter, the legal posi-
tion itself is cancelled, implying a proper and effective modification of the law 
(the existent positions).
(lxxxvii) under “(ip ∧ sa ⇒ F {PA} est) → (P {SA} mra)”, Sarastro is just except-
ing her restriction from entering the Temple if he occasionally invites her to 
enter: such authorization does not mean, naturally, that “(ip ∧ sa ⇒ F {PA} est) 
→ (P {SA} mra)” is no longer in force and not applicable on other occasions.
(lxxxviii) under “(ip ∧ sa ⇒ F {PA} est) → (P {SA} mra)”, Sarastro is extinguish-
ing Papagena’s position (and the position of other SA set agents) if he revokes 
“(ip ∧ sa ⇒ F {PA} est) → (P {SA} mra)”; in this case, Papagena (and the other 
SA set agents) are no longer under any duty to stay out of the Temple.
Accordingly, the tools required to realise each one of these “re-actions” towards 
the duty are different: while in weak-waiving the claim-right holder is just exercis-
ing her correlative permissive position (the deontic modalization governing the 
claim-right), in strong-waiving the claim right holder is extinguishing the legal 
position (and effectively changing the law), something she can only accomplish 
if she has the power to do so.92 Hence, the duty cannot be extinguished without 
such power (the law cannot be modified). However, even without it, the claim-
right holder may still always weakly waive the PA set agent from her individual 
act of compliance (as seen, she is permitted to “manage the result of action”).
(lxxxix) under “(ip ∧ sa ⇒ F {PA} est) → (P {SA} mra)”, Sarastro is just weak-
waiving Papagena from the duty to stay out of the Temple if he occasion-
ally invites her to enter: when Sarastro waives Papagena from her duty, he is 
merely exercising his permission to manage the result of action.
(xc) under “(ip ∧ sa ⇒ F {PA} est) → (P {SA} mra)”, Sarastro is strong waiving 
Papagena’s position (and the position of other SA set agents) if he revokes 
“(ip ∧ sa ⇒ F {PA} est) → (P {SA} mra)”; however, revoking the N at hand pre-
supposes that Sarastro has the power to do so.
This difference is crucial for understanding the waiving of constitutional 
claim-rights (or even statutory or administrative rights whenever conferred by 
general norms). Understandably, a SA set agent of a duty given by a constitution-
91 On the difference between waiving and nearby concepts, Wellman 1997: 45. Strictly connect-
ing waiving with power, Kramer 2001: 60, Preda 2015: 410. Differently, Martin 2003: 42.
92 On removing norms from the set (and power), Mendonca 2000: 146, Ferrer 2000: 144.
40
(2023) 51
journal for constitutional theory and philosophy of law
David Duarte
al norm does not have the power to change the constitution: this competence is 
usually assigned to parliaments and under very specific and formal procedures.93 
Yet, when the SA set agent waives a fundamental claim-right, she is not extin-
guishing the legal position in itself; she is merely exercising the permission she 
holds as a claim-right holder. So, it almost goes without saying that the duty is 
not extinguished and that it remains exigible in other space and time conditions.
(xci) “access (a) to third parties correspondence (c) is forbidden” is a NS 
in the constitution expressing the N “(ip ⇒ F {PA} ac) → (P {SA} mra)”; when 
Monostatos and the Queen of the Night allow the Three Ladies to read the let-
ter he wrote to her explaining why he wants now to be at her service, they are 
not extinguishing the constitutional claim-right to secrecy of correspondence.
(xcii) if “(ip ⇒ F {PA} ac) → (P {SA} mra)” is a constitutional norm, it is evident 
that Monostatos and the Queen of the Night do not have the power to eliminate 
such a norm from the legal system (and to extinguish this constitutional claim-
right); however, if this does not prevent them from weak-waiving, it also does 
not mean that everyone ceases to be under the duty constitutionally foreseen.
A distinction between weak and strong waiving becomes less clear when the 
duty is only exercisable by one individual act: that is, when allowing the duty 
bearer to not exercise a token of the duty simultaneously means that the duty 
ceases to exist. However, such overlapping between weak and strong waiving 
does not deny that power is indispensable to changing the law: even if the claim-
right holder is free to weak-waive, she cannot extinguish the duty without having 
the proper power. But this is not particularly complex: an overlap between weak 
and strong waiving necessarily means the uniqueness of the legal position at 
hand and this only happens in private law (where power is usually aggregated).94
(xciii) when the Three Ladies agree with Papageno that they will give him the 
Glockenspiel, he is subsequently permitted to not accept the instrument; by 
doing this, Papageno is weak-waiving the duty the Three Ladies have under 
such agreement; however, while waiving, he is also extinguishing the Three 
Ladies’ duty as well; anyway, Papageno may waive the duty (and actually ex-
tinguish it) only because he has the power to enter into contracts and to cease 
them (as usually conferred in private law).
93 For instance, Mendonca 2018: 51, Guastini 1993: 72.
94 Since legal systems usually have a general competence norm to “private agents” (Lindahl 
2017: 161, Ross 1968: 203).
41
journal for constitutional theory and philosophy of law
(2023) 51
Rights as formal combinations of normative variables
(xciv) if there was a N imposing that a Glockenspiel be given to people wear-
ing bird-feather clothes, a possible weak-waive would not extinguish the le-
gal position: the claim-right to Glockenspiels would remain in the system; if 
a public body decided to give Papageno a Glockenspiel, his possible weak-
waive would not extinguish the legal position: such donation presupposes a 
previous (general) norm foreseeing such position (rule of law) and he would 
be just waiving a token of that duty.
A duty with only one token without a norm conferring power is not the sole 
normative situation where the claim-right holder is limited regarding her per-
mission to weak-waive: two more normative situations must be taken into ac-
count.95 The second one comes from the normative scenario where the result of 
the main action is not divisible by each SA set agent. When the result of the action 
has this conjunctive effect in SA set agents, given that it cannot be individually 
experienced by them, it follows that only a collective weak-waiving (by the whole 
set) would be admissible as an exercise of the claim-right holder’s position. For 
this reason, individual weak-waiving by each claim-right holder is inadmissible.
(xcv) with “Sarastro has the duty to promote a healthy environment (he)”, the 
legal system has a N such as “(ip ⇒ O {Sarastro} he) → (P {SA} mra)”; this N im-
poses on Sarastro the adoption of various acts suitable to achieve that result; 
however, none is individually experienced by each agent of the SA set.
(xcvi) given that the unspecified actions foreseen in “(ip ⇒ O {Sarastro} he) → 
(P {SA} mra)” have this conjunctive effect, only a collective waiving from all 
the SA set agents would be admissible; it follows, then, that each of the SA set 
agents may not individually weak-waive Sarastro’s duty.
A third normative situation where the claim-right holder is limited to weak-
waiving is the one arising from the normative scenario where her permission is 
in conflict with a prohibition to waive the exact same duty: as a matter of fact, 
nothing prevents a normative authority from enacting a norm by which SA set 
agents are forbidden to waive.96 If a norm such as this is enacted, it follows that 
a conflict arises between the permission (to weak-waive) and the prohibition (to 
weak-waive): they have exactly the same object. Accordingly, a balancing has to 
be carried out and the all norms considered position of SA agents depends on 
the prevailing norm: if the prohibition prevails, waiving is inadmissible.97
95 It seems that theories of rights do not have a systematic enumeration of the conditions by which 
a right is or becomes unwaivable. As far as it is known, the topic has been treated solely on the 
basis of specific rights and without any analytical justification. For instance, Cruft 2013: 199.
96 Which is not so unusual (e.g., the minimum wage case). Wenar 2013: 218, Steiner 2013: 241.
97 It is difficult to understand how a prohibition of waiving could lead to a definitive unwaiv-
ability: a normative conflict is unavoidable, and the issue is only to know whether it is solved 
42
(2023) 51
journal for constitutional theory and philosophy of law
David Duarte
(xcvii) under “(ip ∧ sa ⇒ F {PA} est) → (P {Sarastro} mra)”, Papagena is forbid-
den to enter into Sarastro’s Temple whenever a solemn assembly is going 
on; however, Sarastro may weak-waive her from her duty; when the norma-
tive authority enacts “Sarastro is forbidden to waive (w) entrance into the 
Temple”, a N such as “(ip ⇒ F {Sarastro} w) → (P {SA} mra)” enters into force, 
conflicting with the first one: Sarastro is simultaneously under a permission 
to weak-waive and a duty not to weak-waive.
(xcviii) since these norms are deontically incompatible and no norm of con-
flicts is applicable, only a balancing can provide a solution: if the prohibi-
tion prevails, the duty to not enter is unwaivable; one could argue, though, 
that the second norms is useless: if a SA set agent may waive the prohibition 
to waive, then it would follow that Sarastro may waive again; however, that 
seems to be wrong: the second duty is either only collectively waivable or 
waivable by the State.
6 SECOND ORDER POSITIONS: POWER AND DUTY 
NOT TO DEFEAT
A competence norm is a norm that foresees in its consequence the (generic) 
deontic act of “producing deontic consequences”. It is an action as any other, 
as explained above, only with the specificity of being an action that directly 
modifies the law by introducing or removing norms from the set (thus chang-
ing previous legal positions).98 A competence norm is, therefore, a very specific 
norm of conduct; besides its regulative component (on the deontic action), it 
comprises a constitutive one as well: precisely, the bestowing of the possibility 
to produce deontic consequences (that is, conferring a power to a PA set agent). 
Without this constitutivity, no agent can act at the ought to be level.99
As a consequence of its constitutivity, a norm conferring a power creates an 
action that did not exist before its enactment: to produce deontic consequences 
is a generic action (allocated to a person or a body, public or private) beyond 
human biological capacity.100 Even if one can express a command, it does not 
work as such if that utterance is not based on competence (particularly evident 
when we are dealing with legal systems). A competence norm, accordingly, and 
besides everything else, also performs the task of conferring possibility to an ac-
by a norm of conflicts or by balancing. Regarding the conditions for these conflicts, Zorrilla 
2007: 87, Sardo 2018: 57.
98 Guastini 2016: 108, Frändberg 2018: 44.
99 Ruiter 1998: 474. For competence norms as norms of conduct, despite posterior changes, 
Alchourrón & Bulygin 1971: 151, Ross 1958: 50.
100 On constitutivity, Searle 1969: 34. In the legal field, Peczenik 2008: 226, Ross 1968: 130.
43
journal for constitutional theory and philosophy of law
(2023) 51
Rights as formal combinations of normative variables
tion that would otherwise be impossible: with such a norm, it becomes possible 
for someone, and concerning some topic, to act inside the law.101
(xcix) also dependent on an internal possibility condition, a norm of com-
petence confers possibility to the generic action of “producing deontic con-
sequences” (dc); more specifically, such norm does so with regard to some 
subject matter, be it “everything” or just “padlocks”.
(c) the NS “Sarastro is competent with padlocks” expresses the norm of 
competence (NC) “ip ∧ p ⇒ ◊ {Sarastro} dc {SA}”; this NC makes it possible for 
Sarastro to exercise the generic deontic action of “producing deontic conse-
quences” on the topic “padlocks”.
It seems to make no sense, however, to think that a norm can create a new type 
of action without allowing such action to be carried out (prima facie): although 
possibility and permission have nothing to do with each other (worlds apart), it 
would be absurd to not recognize that at least some tokens of such action were 
effectively permitted.102 Some justification can sustain this merely intuitive ap-
proach. When considering that nothing supports the assumption that producing 
deontic consequences is a forbidden or a mandatory generic action (given the 
usual utterances of normative authorities), it follows that, under such absence, 
the action at hand must be seen, ultimately, as permitted in the weak sense.
(ci) the usual linguistic expression of a competence norm is something as 
“Sarastro is competent with padlocks” or “the power to regulate padlocks is 
assigned to Sarastro”; no reasons whatsoever sustain that the generic action 
is here being forbidden or imposed.
(cii) however, if such action is not forbidden or imposed (and that such sen-
tences refers to an action), it follows that producing deontic consequences 
(i.e., regulating behaviour) is a generic action that is at least weakly permitted.
Understanding competence norms as creating the weakly permitted generic 
action of producing deontic consequences does not, however, match with the 
fact that those norms are effectively enacted, becoming members of the set. For 
this reason, such understanding implies either that a competence norm is mute 
regarding its deontic modalization (a peculiar format for a weak permission) or 
101 This “possibility” has nothing to do with the internal possibility condition, which is also pres-
ent in competence norms: while the latter is related to having the mental and physical pos-
sibilities to produce individual acts with deontic consequences, the former regards the overall 
ability to perform a generic action otherwise impossible. On the topic, but on different terms, 
Lindahl 2017: 165, Spaak 2003: 91.
102 von Wright 1963: 189. On those worlds apart, Johns 2014: 371, von Wright 19511: 3.
44
(2023) 51
journal for constitutional theory and philosophy of law
David Duarte
that we have to reconsider what we mean by strong permissions.103 However, if 
the second solution is pointless, the first is no better: if a normative authority 
enacts a norm with an action that it is weakly permitted, what it is likely really 
happening is that such authority is just saying that such action is (after all) per-
mitted.104
(ciii) when the normative authority enacts the NS “Sarastro is competent 
with padlocks”, a power on a certain topic is assigned to Sarastro; and, if that 
authority has enacted a norm that turns out to be permissive, that norm is 
actually a strong permission.
(civ) a competence norm, therefore, is a dual composition of the possibility 
to act and its own permission; following the same notation used here to for-
malize norms, Sarastro’s power conferring norm can be initially presented as 
“ip ∧ p ⇒ ◊ ∧ P {Sarastro} dc {SA}”.
The distinction between the generic action and the individual acts by which 
the law is effectually changed is, however, decisive to understand a competence 
norm as a constitutive permission. And this distinction is decisive because be-
ing (generically) permitted to produce deontic consequences does not neces-
sarily mean that all the tokens of such action are permitted as well: nothing 
inhibits the system from having other norms forbidding some exercises of the 
power at hand. Therefore, if a power conferring norm (generically) allows the 
production of deontic consequences, each of these consequences can be limit-
ed by the eventual existence (within the set) of other norms prohibiting tokens 
of that action.105
103 It goes without saying that conceiving norms of competence (as conceived by Bulygin 1991: 
497 and Ross 1968: 130) as norms without any deontic modality (strong or weak) is seen 
here as leading to an inescapable cul-de-sac. It implies discharging the production of deontic 
consequences as an action, something hardly understandable (when X says to Y “close the 
window!” she is doing something (Austin 1955: 6)).
104 It follows that a power conferring norm is also a strong permission regarding the (consti-
tuted) generic action of producing deontic consequences. Accordingly, written formulas such 
as “the body X is competent on Y” are linguist expressions to say “the body X on the topic Y 
can and may produce deontic consequences”. Although the conception of competence norms 
presented here is quite proximate from the one defended by Ferrer (2000: 165), it must be said 
that his claim that a sentence of competence expresses two norms cannot be accepted. On the 
one hand, this is because it has no “linguistic resemblance” whatsoever with such sentences, 
presupposing an uncanny criterion of norm individuation (mainly regarding the constitu-
tive norm); on the other hand, because nothing prevents the combination of analytical and 
synthetic propositions within a norm structure; that is what happens with a simple normative 
utterance such as “it is mandatory that A ≡ A”.
105 Which seems to be the explanation for the “thief example” (Lindahl 2001: 160, Spaak 2009: 
74): the following example (cvii) tries to demonstrate why that example is misleading.
45
journal for constitutional theory and philosophy of law
(2023) 51
Rights as formal combinations of normative variables
(cv) a norm such as “ip ∧ p ⇒ ◊ ∧ P {Sarastro} dc {SA}” gives Sarastro the possi-
bility and the permission to exercise the generic action of producing deontic 
consequences (here, on padlocks); therefore, he is allowed to “deontically do” 
whatever he wants on the topic.
(cvi) however, if the system has a norm such as the one in “it is forbidden to 
put padlocks on people’s lips”, then Sarastro may not enact a norm allowing 
padlocks on people’s lips: he is forbidden from producing this specific deon-
tic consequence (limiting superior norm).
(cvii) if Sarastro has enacted “everyone can sell padlocks”, then the Three 
Ladies can sell padlocks (they have the power to do so); but they cannot sell 
the padlock owned by the Queen of the Night if Sarastro also enacted “it is 
forbidden to sell another’s padlocks” (limiting superior norm).
A competence norm gives the PA set agent a position called power, usu-
ally inserted, under the Hohfeldian tradition, in a correlativity line with a li-
ability. Correspondingly, a liability is seen as the position of being unshielded 
from a change in previous legal positions: the liability holder, thus, is subject 
to the changes realized by the power holder.106 However, this seems empty. 
Being exposed to a normative change has no content of its own. It either just 
describes that a possibility of deontic change exists, or it results from a failure 
to distinguish competence from what is done with it. In the former, there is 
no independent content; in the latter, it seems evident that “subjection” comes 
from the norm enacted.107
It is clear that a norm of competence introduced into the system gives its 
holder a power to produce deontic consequences. For SA set agents, however, 
it seems that nothing happens besides the mere possibility of a change in their 
legal positions. So, if a liability stands for this mere possibility, it becomes dif-
ficult to unveil how it can be seen as a deontic position in itself: if power is not 
exercised (and while it is not), SA set agents remain deontically static.108 With a 
content such as the mere possibility of change, a liability only expresses what SA 
set agents already have just for being within the correlativity line: that they cor-
relate with agents that have power over them (there is no independent content).
106 Hohfeld 1919: 50, Kramer 1998: 20. It is worth noting that the notion of liability is strongly 
dependent on the concept of power adopted. An example is, for instance, the case of X being 
liable because Y is going to sue her (Kramer 2013: 251). It is quite difficult to understand, 
though, how suing a counterparty is the exercise of a power since no change is operated until 
the court decides (and the court can discharge the plaintiff).
107 Using the word “subjection”, Celano 2019: 34, Schlag 2015: 202.
108 So, when they were allowed to φ and forbidden to φ ´, they keep on being so. For the Hohfeld-
ian liability, see (Hohfeld 1919: 96). See also Kramer 1998: 20, Markovich 2019: 147.
46
(2023) 51
journal for constitutional theory and philosophy of law
David Duarte
(cviii) a norm such as “ip ∧ p ⇒ ◊ ∧ P {Sarastro} dc {SA}” gives Sarastro the pos-
sibility and the permission to produce deontic consequences regarding pad-
locks; this means that he can and may change legal positions of others (the 
Three Ladies’ and everybody else’s).
(cix) under such a norm, Sarastro can enact “it is forbidden to put padlocks 
on people’s lips”; however, until the utterance of this norm, nothing happens: 
the Three Ladies (as all the others) remain exactly with the same deontic set 
of positions as they had before.
(cx) with “ip ∧ p ⇒ ◊ ∧ P {Sarastro} dc {SA}”, the Three Ladies (and everybody 
else) are SA set agents of such norm: enacting that norm gave them a new 
position; however, describing this new position as “Sarastro has power over 
them” is merely to describe Sarastro’s position.
It is also clear, on the other hand, that a possibility of change signifies that 
an effective change occurs when power is exercised by its holder, which means 
that SA set agents are subject to whichever deontic consequences may be pro-
duced. However, if being subject is different from being under the mentioned 
possibility of change, then there is no room for any content other than be-
ing under the deontic consequences effectively produced: from this particular 
angle, subjection is the situation one is in when an action becomes imposed, 
forbidden, or permitted (and precisely for that reason).109 From this perspec-
tive, a liability would be simply an inaccurate way to describe the deontic con-
sequences produced.
(cxi) under “ip ∧ p ⇒ ◊ ∧ P {Sarastro} dc {SA}”, Sarastro enacted “it is forbidden 
to put padlocks on people’s lips”; with this norm, the legal position on the 
matter held by the Three Ladies ceased to be a liberty and it is now a duty not 
to: that is, they are now under a prohibition (new duty).
(cxii) being under a prohibition is the result of an exercise of that compe-
tence norm; thus, if subjection means the new duty, then a liability would 
be a confusion between the result of the action foreseen in the competence 
norm and the position given to SA set agents with the exercise of power.
What has been said is totally valid whenever power is exercised through 
contracts, dominant in the private sphere, where agents, for the reason of pri-
vate autonomy, only create deontic consequences through agreements and very 
109 Regarding SA set agents of the competence norm, that can also be PA or SA sets agents of the 
norms produced under the power. The very fact that Hohfeld spoke about favorable liabilities 
(Hohfeld 1919: 60) is a good piece of evidence that something is wrong here. If a power is 
correlated with a liability in the way he says, then: (i) either a liability would be the mentioned 
“standby” position and could not be favorable or unfavorable; (ii) or it would be the result of 
the exercise of power and would overlap with the (new) positions created with it.
47
journal for constitutional theory and philosophy of law
(2023) 51
Rights as formal combinations of normative variables
specific unilateral acts.110 And it is valid because change only occurs with the 
exercise of power, meaning that nothing happens before and that deontic con-
sequences only come from that exercise. This is immediately visible with a pro-
posal: when an agent makes an offer, she is exercising her power to do so; but 
she is only creating the self-binding duty to a posterior joint exercise of power 
on the terms she proposed. Before the proposal there was nothing; afterwards, 
there are new positions.111
(cxiii) under “ip ⇒ ◊ ∧ P {PA} dc {SA}” (a private autonomy general compe-
tence norm), Papageno can and may sell his Glockenspiel; when he proposes 
to sell it to the Three Boys for £50, he is exercising that power and creating a 
self-duty to jointly exercise the power on the terms of his proposal; natural-
ly, the Three Boys hold the correlated claim-right; the proposal just created 
these two positions.
An agreement following the acceptance of the proposal is, then, another 
exercise of power; specifically, a joint exercise of the competence norm that 
will create different positions from those already given by the offer (and ex-
tinguishing these).112 By concluding the agreement, the parties confer a set of 
legal positions to each other that did not exist before. And the main point is 
that there is no room for any sort of liability: the change manifests itself with 
new positions (duties, liberties, and so forth; or aggregates of them). So, again, 
no change takes place before the exercise of a private autonomy general com-
petence norm, whether until the enactment of a proposal or, afterwards, until 
the agreement is concluded.113
110 On the private exercise of power, Hage 2018: 98, MacCormick 2005: 250.
111 None of them being a liability. Actually, Hohfeld construction is perplexing here (Hohfeld 
1919: 55). In his specific view, X (the offeror) becomes liable with the offer and creates a power 
to Y (the offeree). However, in order to make the offer, X had to be exercising a power (oth-
erwise X could not change Y’s position). But if that is the case, then Y already held a liability 
correlated with that “original” power. So, and besides presupposing that an agent can confer 
power to another (which seems to be, in this example, another mistake (Hohfeld 1919: 50)), 
this leads to the awkward normative scenario where the exercise of power, by definition, puts 
PA set agents in the position of SA set agents and vice-versa. On the other hand, the text above 
shows that the usually called “potestative right” is actually a claim-right to a joint exercise of 
power. For the opposite view, Sartor 2006: 124.
112 On contracting as a case of joint action, Tuomela 2010: 85, Miller 2004: 54.
113 The following example (cxvi) is a remake of the “watch situation” that Hohfeld presented to 
show another aspect of his conception of liability (Hohfeld 1919: 60). The remake tries to 
show that a liability is also a wrong shot in such situation.
48
(2023) 51
journal for constitutional theory and philosophy of law
David Duarte
(cxiv) based on Papageno’s proposal (creating a duty and a claim-right to 
jointly exercise power under the terms of his offer), the subsequent conclu-
sion of the agreement is another (different) exercise of power, leading to new 
positions: now, Papageno owns £50 (the aggregated positions typical to own-
ership) and the Three Boys own the Glockenspiel (idem).
(cxv) accordingly, be it the self-duty and its correlated claim-right, resultant 
from the offer (unilateral exercise of power) or the ownership of £50 and 
a Glockenspiel, resultant from the agreement (bilateral exercise of power), 
neither Papageno nor the Three Boys had had any position other than those 
now described: there is no space for any liability.
(cxvi) if Papageno decides to abandon his Glockenspiel, he exercises his pow-
er to extinguish his ownership; but such exercise does not create any liability; 
first, a res nullius Glockenspiel is a mere factual condition to a future owner-
ship; second, if Papageno had the power to abandon his Glockenspiel, all the 
others would already have a liability before Papageno’s exercise of power.
All considered, it seems that a liability cannot be understood as meaning nei-
ther the deontic consequences produced under the competence norm nor an al-
ternative way to describe the position held by SA set agents while facing a possible 
legal change. As a matter of fact, being a power holder is exactly having power 
over a set of agents, which is precisely what is being designated with a Hohfeldian 
liability. Accordingly, the correlativity between a power and a liability only works 
with a conception of correlativity as a material equivalence, something that, as 
seen before, does not accurately describe the law. As with any norm of the set, the 
position given to SA set agents in a competence norm also involves some co-ac-
tion (an independent content), having a specific deontic modalization of its own.
An attentive observation of competence norms shows that they are permis-
sions as any others (under the present approach), except for the fact that those 
norms additionally confer the possibility of producing deontic consequences. 
This feature is, however, normatively secondary, since it is not related to the 
regulation of behaviour: it is strictly linked to the ability to act. It seems, then, 
that there are no reasons whatsoever to correlate power with a different position 
than the duty to not defeat the internal possibility (as follows from any permis-
sive norm). Although regarding tokens of a deontic action, SA set agents are 
here in the exact same position as any SA set agent is within a permission.
(cxvii) under “ip ∧ p ⇒ ◊ ∧ P {Sarastro} dc {SA}”, Sarastro has the power (possi-
bility and permission) to produce deontic consequences; thus, SA set agents 
are under a duty not to defeat his internal possibility to produce them: e.g., 
they may not grab or immobilize him when he is enacting a norm.
49
journal for constitutional theory and philosophy of law
(2023) 51
Rights as formal combinations of normative variables
(cxviii) the same is valid with the Sarastro’s Brotherhood or even Papageno 
(a body and a private individual) as PA set agents: SA set agents may not pre-
vent the Brotherhood assembly from deliberating just as much as they may 
not physically block Papageno from abandoning his Glockenspiel.
(cxix) conceiving competence norms as constitutive permissions as any oth-
ers leads to seeing them as follows: “(ip ∧ e ⇒ ◊ ∧ P {PA} dc) → (ip ⇒ O {SA} 
~dip)”, with “e” for “everything”, P regulating PA set agents’ deontic action 
and O regulating the SA set agents’ correlated (adaptative) co-action.
6.1 Without power: Disability and immunity are not legal 
positions
Accepting that power is the position connected with the type of action of 
producing deontic consequences and that this action is constituted by a com-
petence norm necessarily signifies that no power exists without the latter: such 
a norm is a necessary and sufficient condition of power.114 It follows, then, that 
it is only with a norm of this sort that power is assigned to a PA set agent, giv-
ing her a (new) legal position she would not hold without it. From the opposite 
perspective, it is also clear that a power with a specific content can be legally 
modified: when the normative authority revokes a power conferring norm or 
partially derogates its scope, some change occurs and power either disappears 
or has a new content.
(cxx) when the normative authority enacted “(ip ∧ p ⇒ ◊ ∧ P {PA} dc) → (ip 
⇒ O {SA} ~dip)”, Sarastro became the holder of a power regarding padlocks; 
now, he can and may produce deontic consequences on the topic, something 
“inexistent” to him (as the holder of power) before such norm.
(cxxi) if the normative authority revokes that norm, Sarastro returns to the 
original standpoint: he has no power; if that authority states that Sarastro is 
only competent with “wood padlocks”, his power has been changed; yet, in 
both these scenarios deontic consequences have been produced.
A totally different scenario occurs when a normative authority enacts what 
can be wrongly called a “norm of incompetence”: when a normative authority 
114 Which is, then, a necessary condition for (directly) producing deontic consequences. Actu-
ally, deontic consequences produced without a competence norm come from (indirectly) the 
mere triggering of already existing norms of the set (or from the impossibility of triggering 
them): that is, those consequences were already within the set. It is for this reason that the 
expression “quasi-powers” (Kurki 2017: 44) sounds quite unhappy. A flood extinguishing a 
property right signifies that it is definitely impossible to trigger property norms (or that a 
norm conferring that effect to natural disasters exists). The same applies to the parallel ex-
amples presented by Kramer 2001: 59.
50
(2023) 51
journal for constitutional theory and philosophy of law
David Duarte
states that a body or a person does not have power (and never did).115 A signifi-
cant aspect justifies the said difference: such “norm of incompetence” is not a 
norm at all. Since no deontic consequences are being produced (and no previ-
ous positions are being changed), an enactment of this sort is merely a descrip-
tive sentence about the law: it is saying that the body or the person at hand lacks 
power.116 Thus, officially stating “incompetence” is just a descriptive utterance 
about the absence of a competence norm (signalling “no power”).
(cxxii) if the power conferring norm “(ip ∧ p ⇒ ◊ ∧ P {PA} dc) → (ip ⇒ O {SA} 
~dip)” had never been enacted, Sarastro never had the power on the topic 
at hand; if, in this scenario, the normative authority enacts “Sarastro has no 
power regarding padlocks”, no norm is being produced.
(cxxiii) therefore, “Sarastro has no power regarding padlocks” is no other 
than a description of a state of affairs: the normative authority did not change 
the law, no deontic consequences were produced, and no ideal state of affairs 
has been conceived (by such authority).
It is precisely for the reason that the absence of power has no normative 
basis (and cannot be assigned to any legal norm whatsoever) that it is claimed 
here that Hohfeldian disabilities and immunities are no other than an expres-
sion of a confusion between norms and normative propositions: when Hohfeld 
makes reference to a disability he is not pointing out a legal position, but merely 
describing its absence (in this case, power).117 And this seems to be quite clear. 
If one accepts that there are no positions without norms and that no norm can 
confer a position opposed to one dependent on constitutivity, one must also 
accept that the word “disability” just plays the role of a normative proposition.
(cxxiv) if without “(ip ∧ p ⇒ ◊ ∧ P {PA} dc) → (ip ⇒ O {SA} ~dip)” Sarastro 
never had power regarding padlocks, then he (allegedly) has a Hohfeldian 
disability: he lacks power; lacking power, though, is a legal nothing, from 
which it follows that the word “disability” is merely a description of such 
emptiness; and this is so since power is constituted (and if not constituted it 
just does not exist).
Given Hohfeld’s conception of correlativity as a material equivalence, the 
“disappearance” of disabilities as legal positions has to imply the same conse-
quence to immunities. And such consequence seems to be right. As a matter 
of fact, it is so deontically empty to not have power as it is to be in the correla-
tive position: there is no possible normative basis regarding not producing and 
115 On these “norms”, Ferrer 2000: 162. Also, Ruiter 1993: 157.
116 On descriptive enactments, Amselek 1988: 195, Ross 1968: 71.
117 Hohfeld 1919: 60. On Hohfeld’s disabilities, Lindahl 1977: 25, Cook 1919: 727.
51
journal for constitutional theory and philosophy of law
(2023) 51
Rights as formal combinations of normative variables
not being targeted by deontic consequences.118 It is exactly for this reason that, 
when some agent claims an inexistent power and allegedly produces deontic 
consequences, it is common to have a third norm foreseeing a specific sanction 
for such action: often, it means the immediate removal of those “consequences” 
from the system.119
(cxxv) if Sarastro has no power regarding padlocks, he cannot and may not 
produce deontic consequences on the matter; consequently, no one can be af-
fected by something he cannot do; regarding power, Sarastro is as relevant to 
the Three Ladies (and everybody else) as the King of Portugal is.
(cxxvi) if even without power Sarastro enacts “it is forbidden to put padlocks 
on people’s lips”, the most probable outcome (contingently depending on the 
system) is that a norm such as “norms enacted by incompetent bodies or 
persons are null and void” will be applied.
Accepting that a legal position such as an immunity simply does not exist 
(because it cannot, for the given reasons) seems to confront its wide usage in the 
legal field: it is rather recurrent to affirm that an agent has an immunity when she 
is somehow safeguarded against another agent (a powerless one).120 This seems 
to be, however, a confusion either with an effective claim-right or with noth-
ing. Actually, the large majority of the usually called “immunities” are no other 
than claim-rights coming from prohibitions of exercising competence towards a 
specific set of persons. And all the remaining ones, on the other hand, are mere 
external descriptions of someone being outside the scope of a certain power.121
(cxxvii) under “(ip ∧ p ⇒ ◊ ∧ P {PA} dc) → (ip ⇒ O {SA} ~dip)” Sarastro has 
enacted “it is forbidden to put padlocks on people’s lips”; but it turns out that 
Sarastro also has the power (given by another competence norm) to apply a 
fine of £300 whenever the prohibition he enacted is breached.
(cxxviii) if the normative authority forbids Sarastro from applying that fine 
to members of the Queen of the Night’s court (forbidding a specific exercise 
of power), then each member of the court does not have an immunity: each 
one has a claim-right not to be fined (correlated to the duty; not the power).
118 Kocourek 1921: 34. And this might be, speculatively, the reason why Bentham disregarded 
these positions (Bentham 1970: 251). On the topic, Hart 1982: 164, Arriagada 2018: 23.
119 On norms with sanctions, Hage 2018: 207, Calsamiglia 1994: 763.
120 Examples of the common usage of immunity in Wellman 2016: 123, Edmundson 2004: 82. 
With an extremely broad conception of immunities, Zucca 2007: 41.
121 Distinguishing precisely between these two cases (prohibition of exercising competence and 
not having competence), Mendonca 2000: 137.
52
(2023) 51
journal for constitutional theory and philosophy of law
David Duarte
(cxxix) if the normative authority enacts “Sarastro’s power to apply the £300 
fine is partially derogated regarding peasants”, then Sarastro’s competence 
has been narrowed; however, peasants do not have immunities: Sarastro is as 
legally irrelevant for them as the King of Portugal is.
(cxxx) it can be difficult to grasp whether a norm sentence contains a prohi-
bition of exercising competence or a partial derogation of power (a hard case 
example could be “peasants are exempt from the £300 fine”); however, it is 
either one or the other and the corresponding consequence follows.
7 A TABLE OF LEGAL POSITIONS: FIVE ATOMIC 
POSITIONS
A table of atomic legal positions can now be presented. A table with positions 
that are, moreover, comprehensive, sufficient, and irreducible. Comprehensive 
since there are no other positions besides these (without aggregation). Sufficient 
in the sense that there are no positions (with aggregation) that cannot be as-
signed to a combination of them. And irreducible because there are no posi-
tions in which they can be broken into.122 On the other hand, the present ta-
ble of atomic legal positions is entirely norm-based: each atomic type follows 
from a specific deontic modalization of action (or constituted action, in power). 
Thus, each one has a specific content (independent) and a specific deontic sta-
tus (autonomous).
As seen, and at the first order level, a permissive norm creates two legal po-
sitions: (i) a liberty; and (ii) a duty not to defeat the internal possibility. Since 
strong permissions have to be understood as bilateral, the former is character-
ized both by the faculty of exercising or not the main action foreseen and by 
having the corresponding token of such action as content. On the other side of 
the correlativity line, the duty not to defeat the internal possibility regards the 
variable co-acts that can make the exercise of the liberty internally impossible 
(an adaptative content under an obligation not to). Clearly, these legal positions 
always follow whenever a permission is enacted by a normative authority.
(cxxxi) a permissive norm confers to PA set agents a liberty to φ and ~φ and 
to SA set agents (with an obligation) a duty not to defeat the internal possibil-
ity of φ and ~φ (co-act: φ´); it is what follows from a norm such as “(ip ⇒ P 
{PA} φ) → (ip ⇒ O {SA} ~φ´)”; the first correlativity line is:
Pφ → liberty to φ ∧ ~φ ↔ O~φ’ → duty not to dip φ ∧ ~ φ
122 Halpin 1985: 436, Hage 2018: 231.
53
journal for constitutional theory and philosophy of law
(2023) 51
Rights as formal combinations of normative variables
As also seen, and still at the first order level, prohibitions and impositions 
create two legal positions: (i) a duty; and (ii) a claim-right (negative or positive 
depending on being given by a prohibitive or a mandatory norm). A duty is the 
legal position regarding the main action foreseen in the consequence of those 
norms, having such action as its own content (regardless of being an action in 
the strict sense or an omission). On the other side, a claim-right is the position 
formed by the adaptative co-acts (including passive action) that connect the 
holder to the bringing about of the main action’s result. Such co-acts are de-
ontically permitted (since they are in contradiction with the status of the duty 
bearer).
(cxxxii) a mandatory norm confers to PA set agents a duty to φ (or ~φ) and 
to SA set agents a claim-right to φ (or to ~φ) under a permission to manage 
the result of action (co-act: φ´); it is what follows from a norm such as “(ip ⇒ 
O {PA} φ) → (ip ⇒ P {SA} φ´)”; the second correlativity line is:
Oφ (or ~ φ) → duty to φ (or ~ φ) ↔ P~φ‘ → claim-right to φ (or ~ φ)
(cxxxiii) a table of first order legal positions entails two correlatives (co): (i) 
duty with claim-right; and (ii) liberty with duty not to defeat; it also entails 
two contradictories (ct): (i) duty with liberty; and (ii) claim right with duty 
not to defeat; it is as follows:
At the second order level (where the dynamic of the system comes from), 
there are only power conferring norms. Since these norms are permissions as any 
others, yet with the specificity of constituting a type of action impossible without 
them, their distinction from first order permissions is only felt in the first posi-
tion of the correlativity line: precisely the point where the mentioned constitutiv-
ity makes some difference. Accordingly, a competence norm creates: (i) a power; 
and (ii) a duty not to defeat the internal possibility. If the latter has the same 
adaptative content as in any permission (being under an obligation), the former 
is the already known (permitted) possibility to produce deontic consequences.
54
(2023) 51
journal for constitutional theory and philosophy of law
David Duarte
(cxxxiv) a competence norm confers to PA set agents a power to φ and ~φ 
and to SA set agents (with an obligation) a duty not defeat the internal pos-
sibility of φ and ~φ (co-act: φ´); it is what follows from a norm such as “(ip ∧ 
e ⇒ ◊ ∧ P {PA} dc) → (ip ⇒ O {SA} φ´)”; the second order correlativity line is:
◊ ∧ P φ → power to φ (or ~ φ) ↔ O~φ‘→ duty not to dip φ ∧ ~ φ
(cxxxv) a complete table of legal positions shows the previous four with the 
single differentiated position given by competence norms: power; since these 
norms are permissions as any others, it is only on the PA set side that they 
show that difference; it is as follows:
A table of atomic legal positions entirely norm-based connects such posi-
tions with the two sets of addressees included in any norm. A taxonomy of such 
positions based on that criterion can follow. Therefore, one can distinguish be-
tween “primary legal positions”, those held by agents of a PA set, and “secondary 
legal positions”, those held by agents of a SA set. In the former we have: (i) liber-
ties; (ii) duties; and (iii) powers. In the latter we have: (i) duties not to defeat the 
internal possibility; and (ii) claim-rights. As was discussed above, while primary 
legal positions always regard the main action normatively foreseen, the second-
ary ones regard the adaptative co-action inherent to the correlativity line.
(cxxxvi) “primary legal positions” are positions held by PA set agents {PA}: 
liberties, as follows from “(ip ⇒ P {PA} φ) → (ip ⇒ O {SA} ~φ´)”; duties, as fol-
lows from (ip ⇒ O {PA} φ) → (ip ⇒ P {SA} φ´); and powers, as follows from “(ip 
∧ e ⇒ ◊ ∧ P {PA} dc) → (ip ⇒ O {SA} ~φ´)”.
(cxxxvii) “secondary legal positions” are positions held by SA set agents {SA}: 
duties not to defeat the internal possibility, as follows from “(ip ⇒ P {PA} φ) → 
(ip ⇒ O {SA} ~φ´)” and from “(ip ∧ e ⇒ ◊ ∧ P {PA} dc) → (ip ⇒ O {SA} ~φ´)”; 
and claim rights, as follows from (ip ⇒ O {PA} φ) → (ip ⇒ P {SA} φ´).
55
journal for constitutional theory and philosophy of law
(2023) 51
Rights as formal combinations of normative variables
Since norms have the dispositional property of being defeasible (at least first 
order norms), the positions they give, either to agents of the PA set or to agents 
of the SA set, are only prima facie positions: that is, their definitiveness regard-
ing some case depends on an all norms considered assessment recognizing the 
inexistence of a conflict (between two or more norms) or, if such conflict takes 
place, the prevalence of the norm at hand.123 And this is valid for both positions 
of a norm: irrespective of undercutting or rebutting defeasibility, the prevalence 
of a norm over another (imposed by a norm of conflicts or by balancing) signi-
fies the twofold prevalence of its correlated positions.124
(cxxxviii) Sarastro enacted “it is forbidden to put padlocks on people’s lips” 
with {all} in the PA set and “it is allowed to put padlocks on people’s lips” 
with {the Three Ladies} in the PA set; these norms are in conflict: the Three 
Ladies are forbidden to act by the first norm, but permitted by the second; 
since the system has lex specialis, their prima facie liberty prevails over their 
prima facie duty (undercutting defeasibility); but this also means that the 
correlated duty not to defeat prevails as well.
(cxxxix) Sarastro enacted “everyone is allowed to sing” and “it is forbidden 
to harm personal reputation”; when Papagena sings something attacking 
Monostatos’ honour, she is prima facie permitted by the first norm, but prima 
facie forbidden by the second; since the system has no norm of conflicts ap-
plicable (rebutting defeasibility), only a balancing solves the conflict; yet, the 
prevalence of the liberty also signifies that the correlated duty not to defeat 
prevails as well.
8 SETS OF ADDRESSEES: WHO, HOW MANY AND 
HOW
An analysis of norm sentences usually enacted by normative authorities 
shows that an explicit enunciation of both sets of addressees is uncommon. As 
a matter of fact, the large majority of norm sentences mention only one set: 
whether the PA or SA set (but sometimes neither).125 Although it is of no sur-
prise that normative authorities do not consider correlativity as a criterion for 
norm sentences drafting (as they could, among others), the fact is that when-
ever only one set is presented (almost always), a complex problem arises. Since 
correlativity is not text-dependent, the interpreter has to realize who the agents 
belonging to the set the normative authority did not make explicit are.
123 On defeasibility as a dispositional property of all norms, Ratti 2013: 122, Brożek 2004: 169.
124 On undercutting and rebutting defeasibility (and the way systems solve the conflicts that give 
rise to them), Prakken & Sartor 2004: 121, Duarte 2009: 172.
125 On expressing addressees, Frändberg 2018: 86, Pino 2016: 46.
56
(2023) 51
journal for constitutional theory and philosophy of law
David Duarte
(cxl) the NS “Sarastro’s Brotherhood has the duty to provide health care 
to citizens” is a less usual case of explicit reference to both sets: the PA set 
{Sarastro’s Brotherhood} and the SA set {citizens}; thus: “(ip ⇒ O {Sarastro’s 
Brotherhood} φ) → (ip ⇒ P {citizens} φ´).
(cxli) the NS “everyone is allowed to sing” is an usual case of a NS that only 
mentions the PA set {everyone}: nothing is written about who belongs to the 
SA set; on the basis of what is explicit, that NS expresses a N such as “(ip ⇒ P 
{everyone} φ) → (ip ⇒ O {SA} ~φ´)”; (or {everyone} = {physical persons}).
(cxlii) the NS “it is forbidden to harm personal reputation” is an usual case 
of a NS that only mentions the SA set {people}: nothing is written about who 
belongs to the PA set; on the basis of what is explicit, the N expressed is: “(ip 
⇒ F {PA} φ) → (ip ⇒ P {people} φ´)”; (or {people} = {physical persons}).
(cxliii) the NS “it is forbidden to use £1 coins” is a less usual case of a NS that 
does not mention the PA set nor the SA set: nothing is written about who 
belongs to each of the sets; on the basis of what is explicit, the N expressed is: 
“(ip ⇒ F {PA} φ) → (ip ⇒ P {SA} φ´)”.
As in many areas of legal knowledge, the task of realizing the members of an 
unexpressed set can also be an easy or a hard case. When there are some duties 
imposed on a set of addressees, identifying who the members of the omitted set 
are might turn out to be a simple task: a norm conferring a claim-right to “em-
ployees” (explicitly mentioned) most probably is imposing a duty on the agents 
belonging to an unwritten set of “employers”. It can be said that in such cases 
both sets are easily conceivable: one because it is explicitly mentioned, the other 
because it is connected with the first by a clear case of empirical correlativity.126 
Nevertheless, and as usual, things might turn out to be not that easy.
(cxliv) the NS “it is mandatory to pay 500 florins (£500) per month to per-
sonal employees” only mentions the SA set {employees}; however, it seems 
clear that the PA set is {employers}; thus, despite the insufficiency, one may 
reach “(ip ⇒ O {employers} £500/month) → (ip ⇒ P {employees} φ´)”.
(cxlv) the NS “freedom of the press is recognized” may be qualified as a hard 
case; on the one hand, because neither the PA set nor the SA set are explicitly 
mentioned; on the other hand, because defining the liberty holders depends 
on the (complex) open texture of the word “press”.
It must be stressed at this point that this is actually an interpretative problem 
related to the norm sentence (although with some specificities): it can be con-
nected to some semantic indeterminacy of the text (“press” in the last example), 
but it is mainly an interpretative problem arising from the insufficient informa-
126 On empirical correlativity, Halpin 2019b: 86, Radin 1929: 902.
57
journal for constitutional theory and philosophy of law
(2023) 51
Rights as formal combinations of normative variables
tion given by the sentence effectively enacted by the normative authority (sets 
of addressees that are not even mentioned).127 Therefore, the interpreter must 
somehow fill those gaps, using the contextual linguistic elements available, in-
ferences to the best explanation about who acts and who faces action, or even 
consider some pragmatic methodological guidelines.128
(cxlvi) “it is forbidden to use £1 coins” mentions neither the PA nor SA set; 
using the interpretative criterion coming from equality (generality of general 
norms), it seems that each set should be conceived as extended as it is inter-
pretatively appropriate; since both physical and legal persons can use coins, 
it seems that the N is: “(ip ⇒ F {all} φ) → (ip ⇒ P {all} φ´)”; ({all = physical and legal 
persons}).
Independent of such epistemic insufficiencies, the fact is that a set of address-
ees can contain a variable number of elements. On this basis and considering the 
universe of possible agents within the territorial scope of the norm at hand, they 
could be measured with the quantifier expressions {all}, {some}, and {one}.129 
Therefore, inside that domain, the set with {all} is the universal one and the sets 
with {some} and {one} are just proper subsets. In the former, membership to the 
set is assigned to a category of agents, while in the latter the set is a singleton. So, 
whichever the way a normative authority designs addressees, they will necessar-
ily amount to one of these three quantifiers: {all}, {some} and {one}.
(cxlvii) with a N such as “(ip ⇒ O {all} φ) → (ip ⇒ P {all} φ´)” both PA and 
SA sets contain the whole universe of agents under the present domain; this 
means that both the PA and SA sets entail all physical and all legal persons 
covered by the territorial scope of that norm.
(cxlviii) with a N such as “(ip ⇒ O {some} φ) → (ip ⇒ P {some} φ´)” both PA and 
SA sets contain a subset category of agents under the present domain; this 
would be the case of a PA set with {employers} or {legal persons} and a SA set 
with {employees} or {physical persons}.
(cxlix) with a N such as “(ip ⇒ O {one} φ) → (ip ⇒ P {one} φ´)” both PA and SA 
sets contain just one element of the universal set under the present domain; 
this would be the case of a PA set with {the State} or {Sarastro} and a SA set 
with {Papagena} or {the Queen of the Night}.
127 On addressees and interpretation, Florczak-Wator 2015: 24, Herrestad 1996: 20.
128 Two methodological guidelines can be proposed. The first regards legal and physical persons: 
whenever the performance of a generic action is incompatible with the legal personhood, 
then the members of the unstated set cannot be legal persons (e.g., to marry). The second is of 
a more general character: since general norms express a formal equality between individuals, 
a set should be extended as much as is interpretatively appropriate.
129 On these expressions, Iacona 2015: 129, Peters & Westerståhl 2006: 10.
58
(2023) 51
journal for constitutional theory and philosophy of law
David Duarte
When the PA and the SA sets contain {all} or {some}, each member of each 
one of those sets is a singular holder of the legal position the set stands for. 
Consequently, regarding the prohibition of φ, for instance, each member of the 
PA set is the holder of a duty to ~φ and each member of the SA set is the holder 
of a claim right to ~φ. Naturally, this also applies to permissions: each agent of the 
PA set holds a liberty to φ and ~φ and each agent of SA set holds a duty not to 
defeat the internal possibility of φ and ~φ.130 When the correlativity line has no 
other empirical or legal limits, each member of one set holds the position towards 
each of the agents that belong to the correlated set (under the norm at hand).
(cl) in “(ip ⇒ P {all} φ) → (ip ⇒ O {all} ~φ´)”, PA = {u, v, w, x, …} and SA = {u, 
v, w, x, …}; here, u holds a liberty towards v, and w, and x, and so forth; in the 
same way, x holds a duty not to defeat towards u, and v, and w, and so forth; 
in this case, there are as many correlativity lines (tokens) as the possible in-
stantiations of N; the same holds for a norm with {some} instead of {all}.
A setting such as this provides a scheme that must be taken as a mere default, 
namely since it presupposes that the exercise of the action φ by PA set agents is 
disjunctive: that is, each member of the PA set, taking into account the type of 
action foreseen, can exercise it individually. However, it does not have to be so. 
A norm can foresee a type of action that has to be exercised collectively (joint 
action): when this is the case, and irrespective of the specific way in which SA set 
agents perform their position, the exercise of the main action is jointly carried 
out by all (or a significative number of) the members of the set. As is known, this 
is one of the many senses in which one speaks about collective rights.131
(cli) in “(ip ⇒ P {physical persons} s) → (ip ⇒ O {all} ~φ´)”, each physical person is 
allowed to sing (s); since it is an action that can be exercised individually, the 
PA set is disjunctive: {∨ physical persons}.
(clii) in “(ip ⇒ P {physical persons} a) → (ip ⇒ O {all} ~φ´)” each physical person is 
allowed to assemble (a); since it has to be exercised collectively, the PA set is 
conjunctive: {∧ physical persons}.
(cliii) in “(ip ⇒ F {all} k) → (ip ⇒ P {physical persons} φ´)”, everyone is forbidden 
to kill (k); since it is an action (omission) that can be exercised individually, 
the PA set is disjunctive: {∨ all}.
130 Which is the (general) normative way to express multitalcity (Hohfeld 1919: 72, Barker 2018: 
592). Somehow mixing multitalcity with molecularity, Wellman 1997: 67.
131 When it is a right (in the following examples, with mandatory norms, it is a collective duty). 
On this kind of collective rights (collectively exercised [joint action]), Jovanović 2012: 115, 
Miller 2004: 216.
59
journal for constitutional theory and philosophy of law
(2023) 51
Rights as formal combinations of normative variables
(cliv) in “(ip ⇒ O {physical persons} snatime1) → (ip ⇒ P {all} φ´)”, all members of 
the set are obliged to sing the national anthem at time1 (snatime1); the PA set 
here is also conjunctive: {∧ physical persons}.
Regarding how the result of action impacts SA set agents’ co-action (not the 
individual or collective way in which it is exercised by PA set agents, as above), 
an equivalent distinction can be designed within the set of secondary address-
ees. Hence, it is possible to distinguish between types of (main) actions that im-
ply a disjunctive or a conjunctive co-action from SA set agents. In the latter, it 
is collective: since the main action is experienced by all once it is experienced 
by one, SA set agents’ co-action has to be common.132 In the former, the main 
action is only individually experienced by the members of the SA set: it follows 
that co-action from SA set agents is an individual task of each member of the set.
(clv) in “(ip ⇒ P {physical persons} s) → (ip ⇒ O {all} ~φ´)”, the main action (sing) 
is experienced by each one within the SA set; SA set agents’ co-action is dis-
junctive: {∨ all}.
(clvi) in “(ip ⇒ F {all} k) → (ip ⇒ P {physical persons} φ´)”, the main action (kill-
ing) is experienced by each one within the SA set; SA set agents’ co-action is 
disjunctive: {∨ physical persons}.
(clvii) in “(ip ⇒ O {State} he) → (ip ⇒ P {all} φ´)”, the main action (a healthy 
environment) is experienced by the whole SA set; SA set agents’ co-action is 
conjunctive: {∧ all}.
Since the content of any normative element is entirely dependent on the 
will of the normative authority, the quantifier {some} reflects the most differen-
tiated kinds of categories. A normative authority can confer liberties or claim-
rights to categories as different as “padlock manufacturers”, the “Sarastro’s 
Brotherhood”, or the “Sumerian speaking minority”.133 And, irrespective of the 
ties between their members (and their shared beliefs), each of these categories 
is solely a different configuration of {some}. So, with any of them we can have 
legal positions that are disjunctive or conjunctive exactly in the same terms as 
above. Analytically, there seems to be no difference between different possibili-
ties of {some}.134
132 Which is another one of the many senses in which we speak about collective rights (related to 
common goods, as it is usually mentioned (Miller 2004: 217, Newman 2011: 102)).
133 On groups (or collectives) and agency, Tuomela 2010: 13, Schmid 2009: 13.
134 Campbell 1998: 110. Differently, Jovanović 2012: 125, Wall 2007: 235.
60
(2023) 51
journal for constitutional theory and philosophy of law
David Duarte
(clviii) in “(ip ⇒ P {Sumerian speaking minority} sol) → (ip ⇒ O {all} ~φ´)”, each 
physical person member of the PA set is allowed to speak their own language 
(sol); since it is an action that can be exercised individually, the PA set is dis-
junctive: {∨ Sumerian speaking minority}.
(clix) in “(ip ⇒ O {State} rds) → (ip ⇒ P {Sumerian speaking minority} φ´)”, the 
main action (recognizing the decision to secede) is experienced by the whole 
SA set; accordingly, SA set agents’ co-action is conjunctive: {∧ Sumerian 
speaking minority}.
On the other hand, the PA and SA sets occupy different locations in the 
structure of norms, which means that they stand for different legal positions. 
Taking into account that each set results from a contingent decision taken by 
the normative authority, it is a matter of mere logic that the PA and SA set can 
possibly connect as follows: (i) PA and SA have the same agents; (ii) PA and 
SA do not share agents; (iii) PA is a subset of SA or vice-versa; and (iv) PA 
intersects with SA.135 For present purposes, these relations among sets are not 
relevant: they are purely informative. What is relevant is that, with equality, 
subsection, or intersection, an agent can be simultaneously a member of the PA 
and the SA sets.
(clx) if PA = {u, v, w, x} and SA = {u, v, w, x}, then PA = SA; it follows: (u, v, 
w, x) ∈ (PA ∧ SA).
(clxi) if PA = {v, w, x} and SA = {u, v, w}, then PA ⋂ SA; it follows:(v, w) ∈ 
(PA ∧ SA).
(clxii) if PA = {u, v} and SA = {u, v, w, x}, then PA ⊂ SA; it follows: (u, v) ∈ 
(PA ∧ SA).
(clxiii) if PA = {u, v} and SA = {w, x}, then PA ⋂ SA = Ø; no overlap exists.
In normative situations such as these (when the PA and SA set share agents), 
it is important to note that the agent does not correlate with herself. Such prem-
ise follows from the assumption that a line of correlativity between the same 
agent leads to a deflation of both legal positions. In a liberty ↔ duty not to 
defeat the internal possibility, this is because breaching the duty amounts ex-
actly to not exercising the liberty.136 In a duty ↔ claim right line, it is because 
135 When one set is or both sets are {one}, the third and the fourth relations are not possible: 
subsection presupposes at least a set with two members and intersection presupposes that 
both sets have at least three members. Nevertheless, for the present purposes, it is assumed 
that only {all} and {some} are being considered.
136 Which is also valid for a competence norm. However, this is not to be confused with the dif-
ferent case of a power holder inserting herself into the PA or SA sets of the norm she enacts. 
Actually, this is the very core of the idea of the rule of law (to submit power holders to their 
own law (MacCormick 2005: 25, Kelsen 2005: 292)).
61
journal for constitutional theory and philosophy of law
(2023) 51
Rights as formal combinations of normative variables
the (single) holder may weak-waive the duty, and a duty from which one may 
release cannot be qualified as duty at all.137 An agent member of both sets, thus, 
only correlates with other agents of the correlated set (PA to SA and vice-versa).
(clxiv) if PA = {u, v, w, x} and SA = {u, v, w, x}, when u holds a PA position she 
only correlates with v, w, and x; naturally, when u holds a SA position, u only 
correlates with v, w, and x.
(clxv) if u ∈ PA ∧ u ∈ SA, u cannot correlate with herself; if she holds a duty, 
she may weak waive it; if she holds a liberty, defeating herself is the same as 
not exercising the liberty.
9 CORRELATIVITY: ASSUMING ASYMMETRY AND 
DIFFERENTIATING DIRECTEDNESS
As mentioned, Hohfeld conceived correlativity as a necessary reflexive 
connection between two legal positions (material equivalence): since each of 
them is just a different perspective of the same ligation, a duty implies a claim-
right in the exact same way that the latter implies the former (effective in all 
correlativity lines).138 Such conception of correlativity was already criticized 
here: it is sustained on a reductive understanding of action, from which it fol-
lows that, although almost untouchable in its own scheme, it does not describe 
law correctly: since a duty often requires co-action from the correlated agent 
(e.g.), positions have different contents. And this is clearly incompatible with 
Hohfeld’s correlativity.
This criticism of Hohfeldian correlativity does not lead, however, to discharg-
ing correlativity itself. Quite on the contrary: correlativity is the exact expression 
for the relation between two poles that can be found in each and every norm of 
a legal system. Irrespective of which action is regulated, any norm of the system 
connects a PA set to a SA set. Given that law regulates the behaviour of more 
than one agent, the deontic modalization of an action gives rise to a position that 
is necessarily “relative”: any action permitted, imposed, or prohibited, if not cor-
related with {some} or {one}, ultimately expresses the connection an agent has 
(or a set of agents have) with all the others or the community as a whole {all}.139
137 Singer 1959: 203. With a different account, Hills 2003:135.
138 Hohfeld 1919: 36. Also, Kramer 1998: 26. It should be said, nonetheless, that Kramer´s slope 
metaphor (as an illustration to apparently point out a symmetric correlativity) is quite disput-
able: since only one of the correlated positions has the burden of action, it seems that Hohfel-
dian correlativity shows itself as asymmetric (Halpin 2019a: 230).
139 Even though this {all} can also be a very specific case of {one} (the State {State}) or an extend-
ed version of {some}, such as {physical persons}. In the final instance, and regarding norms 
62
(2023) 51
journal for constitutional theory and philosophy of law
David Duarte
Correlativity has been challenged, particularly in the duty ↔ claim-right 
line. However, and as Kramer has exemplarily shown, there is no single case 
that cannot be explained under a correlativity scheme.140 In fact, the main prob-
lem related to such challenges is that most of the examples presented are cases 
where quantifiers of the SA set are not taken into account. Norms of criminal 
or administrative law confer duties that, if not correlated with {one} or {some}, 
necessarily correlate with {all}.141 Regardless of how difficult it can be to grasp 
who SA set agents are, the fact is that to impose a duty on an agent would be 
pointless if even the community were not to experience the result of action.142
(clxvi) Sarastro enacted “(ip ⇒ F {PA} pl) → (ip ⇒ P {SA} φ´)”, a prohibition to 
put padlocks on people’s lips; the duty everyone has to refrain from doing it 
correlates with a claim-right every physical person has (disjunctively): there-
fore, Sarastro enacted “(ip ⇒ F {∨ all} pl) → (ip ⇒ P {∨ physical persons} φ´)”.
(clxvii) Sarastro also enacted “(ip ∧ bpp ⇒ O {breachers} pay £300) → (ip ⇒ P 
{SA} φ´)”, a £300 fine on whoever breaches the prohibition (“bpp” as breach-
ing the padlock prohibition); the duty correlates with a claim-right held by 
all (conjunctively): “(ip ∧ bpp ⇒ O {∨ breachers} pay £300) → (ip ⇒ P {∧ all} φ´)”.
(clxviii) both in the first N “(ip ⇒ F {∨ all} pl) → (ip ⇒ P {∨ physical persons} φ´)” 
and in the second one “(ip ∧ bpp ⇒ O {∨ breachers} pay £ 300) → (ip ⇒ P {∧ all} 
φ´)” there is an asymmetric correlativity: the mere adaptative content of the 
positions given to SA set agents is enough to verify it.
An asymmetric correlativity is recognized here, then, as inherent to any 
norm of the system, which is enough to claim it as a universal proposition. 
However, and given the way in which it is conceived, such asymmetric correla-
tivity also provides an answer to a very inconsequential topic: to know if duties 
are prior to rights or the other way around.143 Actually, and coherently with the 
normative structure that sustains each type of right, as seen before, it depends 
exactly on the right at hand: rights do not have or lack “priority” over duties just 
for being rights. So, while it seems clear that liberties and powers are prior to 
duties, claim-rights are duty-based and thus only justified on their basis.
enacted by normative authorities, there is always a relation between the State (as norm-giver) 
and the position wanted for an agent.
140 Kramer 1998: 58.
141 On those norms as denying correlativity, Hart 1982: 185, D’Almeida 2016: 560.
142 So, any norm correlates a primary position (main action) with a secondary one (co-action), 
a proposition about the law considered here as universal (valid to all norms of a system) and 
meaning that: (i) correlativity is recognizable in the four categories of norms that exhaust the 
deontic realm (permissions, obligations, prohibitions, and competence norms); and that (ii) 
correlativity is asymmetric (since both poles do not mirror each other).
143 Discussing the topic, MacCormick 1977: 200, Raz 1984: 211. Generally, Penner 1997: 304.
63
journal for constitutional theory and philosophy of law
(2023) 51
Rights as formal combinations of normative variables
(clxix) in “(ip ⇒ P {PA} φ) → (ip ⇒ O {SA} ~φ´)”, the right (liberty) is prior to 
the duty: the main action is assigned to the liberty holder (PA set).
(clxx) in “(ip ⇒ O {PA} φ) → (ip ⇒ P {SA} φ´)”, the right (claim-right) is not 
prior to the duty: the main action is assigned to the duty bearer (SA set).
(clxxi) in “(ip ∧ e ⇒ ◊ ∧ P {PA} dc) → (ip ⇒ O {SA} ~φ´)”, the right (power) is 
prior to the duty: the main action is assigned to the power holder (PA set).
Still within the present context, mixing correlativity with directedness seems 
to be heuristically pointless, at least once correlativity is accepted as univer-
sal.144 This is to say that there is no other conceptual space for “directedness” 
than to stick to the determinability of the correlated agent: using the duty ↔ 
claim-right correlativity line, a duty is directed when who the SA set agent is, is 
known (or, more precisely, determinable).145 So, if with the quantifier {one} the 
addressee is most probably determinable, with the quantifiers {some} and {all} 
determinability of the addressees is only the case if, for some reason, the norm 
at hand restrains its applicability to a specific occasion.146
(clxxii) in “(ip ⇒ O {Sarastro’s Brotherhood} hc) → (ip ⇒ P {citizens} φ´), where “hc” 
stands for health care, the duty is not directed: the claim-right holder is in-
determinable.
(clxxiii) in “(ip ⇒ O {Sarastro’s Brotherhood} hc) → (ip ⇒ P {Pamina} φ´), the duty is 
directed to Pamina: the claim-right holder is determinable (and it would also 
be with {the Queen of the Night’s daughter}).
10 THE FORMALITY OF RIGHTS: COMBINING 
NORMATIVE VARIABLES
As Hohfeld’s table of atomic legal positions already did, the table presented 
here also shows that the word “right” is used with different meanings; specifi-
cally, one of the following four: (i) a liberty; a (ii) a claim-right; (iii) a power; and 
(iv) an aggregate of atomic positions including at least one of the previous three. 
As is well known, Hohfeld’s initial purpose was, precisely, to “grasp the irreduc-
144 For instance, Sreenivasan 2010: 467. For various conceptions of directedness, Hedahl 2013: 
24, Steiner 2013: 232, Múrias 2018: 26.
145 Thus, “determinable” means here what Bentham meant with “assignable” (Bentham 1970: 
313). Also, Hart 1982: 178. On the determinability of addressees, Nino 2003: 77.
146 With the quantifier {one} indeterminability is only the case if just one unknown person can 
match a specific condition (e.g., the winner of a competition before its end); on the other 
hand, with the quantifiers {some} and {all}, determinability depends on a given space-time 
uniqueness (naturally, determinability with {all} is particularly rare).
64
(2023) 51
journal for constitutional theory and philosophy of law
David Duarte
ible” by drawing the differences between those rights.147 His table provided the 
means to understand, for instance, that holding a liberty or a power is to hold en-
tirely different legal positions (even though both are usually called rights). And 
clarifying those differences is, undeniably, a purpose that Hohfeld fully achieved.
However, a norm-based approach to atomic legal positions seems to be able 
to unveil a little bit more; specifically, that to hold a right (no matter if a liberty 
or any other) is solely to be in a particular position given by a specific combina-
tion of normative variables. Since each one of the three atomic kinds of rights 
has a specific design in the structure of norms, precisely the one provided by the 
connection between the deontic modality and the membership to the PA or SA 
sets, it turns out that a right is no other than a formal “standpoint” in a norm: ir-
respective of what is the substantive content given to such norm, a legal right is a 
position resultant from a combination of variables within the norm’s structure.148
Thus, to hold a liberty is just to be a member of the PA set in a norm that de-
ontically modalizes the main action as permitted: in a permissive norm, anyone 
belonging to that set holds the right we call a “liberty”. In the same way, to be the 
claim-right holder immediately follows from being a member of the SA set in a 
norm that deontically modalizes the main action as mandatory (or forbidden): 
here, anyone belonging to such set holds the right we call a “claim-right”. This is 
also the case for legal power: anyone (a body or a person) belonging to the PA 
set in a constitutive permission (a permission regarding the action of producing 
deontic consequences) is the holder of a “power”.
(clxxiv) a PA set agent of a permissive norm is the holder of a liberty: irre-
spective of its content, in “(ip ⇒ P {PA} φ) → (ip ⇒ O {SA} ~φ´)” such PA set 
agent always holds a liberty.
(clxxv) a SA set agent of a mandatory norm (or a prohibition) is the holder 
of a claim-right: irrespective of its content, in “(ip ⇒ O {PA} φ) → (ip ⇒ P {SA} 
φ´)” such SA agent always holds a claim-right.
(clxxvi) a PA set agent of a competence norm is the holder of a power: irre-
spective of its content, in “(ip ⇒ ◊ ∧ P {PA} dc) → (ip ⇒ O {SA} ~φ´)” such PA 
agent always holds a power.
A norm-based approach to atomic legal positions leads, consequently, to a 
completely formal conception of rights, which is even more evident when con-
sidering that the content of norms is strictly contingent: the way the different 
elements of a norm are filled with some content is totally dependent on the will 
147 Hohfeld 1919: 28.
148 Specifically, the variables related to: (i) the sets of addressees (which are the PA and the SA sets 
members); (ii) the deontic modality (if it is a permission, an imposition or a prohibition); and 
(iii) the type of action foreseen.
65
journal for constitutional theory and philosophy of law
(2023) 51
Rights as formal combinations of normative variables
of a normative authority.149 Therefore, and irrespective of other norms that may 
constrain the exercise of power, it is totally up to that authority to establish not 
only the main action and how it is deontically modalized, but also who belongs 
and does not belong to the PA and SA sets. By doing so, that authority is confer-
ring rights regardless of the content inserted in the norm’s elements.
(clxxvii) the norm structure has to be filled by the normative authority when 
enacting any norm: “(ip ⇒ DM {PA} φ) → (ip ⇒ DM {SA} φ´); accordingly, 
such authority has: (i) to select a type of action (if it is to “produce deontic 
consequences”, it also gives the possibility to act); (ii) to choose how such ac-
tion is deontically modalized (DM), from which follows the deontic modali-
zation for the correlated co-action (φ´); (iii) to choose the PA set members; 
and (iv) to choose the SA set members.
The understanding of rights as a mere formal combination of normative 
variables, sustained under the present norm-based approach to atomic legal 
positions, also shows that any right (no matter if a liberty or any other) is com-
pletely value-free: since it is a strictly contingent matter, a right in a legal sys-
tem is not necessarily an expression of a moral quality, or an interest of the 
holder, or whatever might provide it a metaphysical ground.150 So, in the same 
way that a normative authority can confer rights “favourable” (to the holder), 
nothing prevents that authority (besides other norms limiting its power) from 
conceiving rights that are “unfavourable”. In sum, legal rights (per se) do not 
have any axiological rationale.
(clxxviii) when the normative authority enacts “(ip ⇒ O {Monostatos} t) → (ip 
⇒ P {Pamina} φ´)”, expressed by “Pamina has the right to be tortured (t) by 
Monostatos”, it follows that Pamina has a very clear claim-right to be tor-
tured by the Moor; being tortured by Monostatos is obviously “unfavorable” 
to Pamina, it is repugnant to the community and it does not protect any as-
pect of Pamina’s situation (or of any typical bearer); however, her claim-right 
is a claim-right as any other.
149 Qualifying a theory of rights as formal just for their formalization, Herrestad 1996: 6.
150 Which has been the consolidated tradition in legal science. For instance, to Bentham, the 
benefit conferred (Bentham 1970: 56); to Austin, an advantage given (Austin 1875: 193); to 
Salmond, an interest recognized and protected (Salmond 1913: 181); to Hart, an exclusive 
control over another, the small-scale sovereignty (Hart 1982: 183); to Raz, the individual well-
being (Raz 1984: 195); to Sumner, the protection of individual autonomy and welfare (Sum-
ner 1987: 98); to Wellman, a domain of dominion (Wellman 1997: 213); to Kramer, to protect 
some aspects (typically beneficial) of the holder´s situation (Kramer 2013: 246).
66
(2023) 51
journal for constitutional theory and philosophy of law
David Duarte
(clxxix) an example such as this cannot be confused, however, with the dis-
tinct case of a norm sentence like “Monostatos is under the duty to torture 
Pamina (tP)”, expressing a norm such as “(ip ⇒ O {Monostatos} tP) → (ip ⇒ P 
{State} φ´)” in which the claim-right holder is the State and Pamina is only the 
object of the main action; this second example shows, though, that the pre-
vious one refers to a norm that effectively confers an “unfavourable” claim-
right (contrary to any interest whatsoever).
10.1 Why theories of rights are almost useless: Facing 
epistemic insufficiency
In an imaginary legal system where the normative authority would enact 
norms with all the information needed to fully understand the rights given by its 
norms, theories of rights would be somehow useless. Such uselessness would be 
the result of norm sentences explicitly showing the conditions, the members of 
PA and SA sets of addressees, or the conjunctive or disjunctive exercise of a right. 
With this data, there would be no doubts about who faces the liberty holder or 
who holds a claim-right when only the duty bearer is known (doubts that are, 
actually, la raison d´être of such theories).151 In other words, there would be no 
room for theories of rights because no epistemic insufficiency existed as well.
This is so because the main purpose of those theories, while attempting to 
provide an explanation of rights, is to offer the data that was omitted by the nor-
mative authority in the norm sentence: theories of rights live from epistemic 
insufficiency and such epistemic insufficiency is what makes them continually 
alive. Excluding some minor details, decades of confrontation between interest 
and will theories have been all about their degree of explanatory power, measured 
by how much they answer such questions.152 Yet, the assessment of epistemic in-
sufficiency may have some value for both theories of rights: it provides a frame 
for what to expect from those theories and shows how viable each one can be.
It seems clear that the will theory faces serious problems. The main reason 
being that its explanatory power is low (or even null). Assuming that a theory 
is an answer to a “why-question” formulated regarding some reliable data, the 
problems undermining the will theory reveal themselves instantly by offering 
answers that contradict what we already know. For instance, that a right be-
ing enforceable depends on a norm conferring enforceability, that there are 
unwaivable rights and, mainly, that a right is not necessarily an aggregate of 
151 Kramer has explicitly exposed this when he stated that “accounting for the directionality of 
legal duties is precisely what the Interest Theory and Will Theory seek to do” (Kramer 2013: 
246), an answer to Sreenivasan’s statement by which “almost no one goes on to raise the ques-
tion of what accounts for the direction of these duties” (Sreenivasan 2010: 467).
152 On the (endless) debate, Weissinger 2019: 200, Cruft 2019: 170.
67
journal for constitutional theory and philosophy of law
(2023) 51
Rights as formal combinations of normative variables
atomic positions.153 And these contradictions with how the law is, as becomes 
visible with a norm-based approach, seem to be fatal to the will theory.
On the other hand, the interest theory, although much more equipped to 
face the mentioned scenario of epistemic insufficiency, has to deal with the 
fact that such a scenario is, all things considered, an interpretative problem: 
epistemic insufficiency is given by the incompleteness of norm sentences, be-
ing strictly connected, therefore, with how normative authorities (or any agent 
exercising power) use language. A possible move to be taken by proponents of 
the interest theory is, correspondingly, to rethink the theory within the broader 
frame of interpretation and, in such a way, to conceive its explanatory purpos-
es within the processes any agent carries out in order to withdraw meaningful 
norms from norm sentences.
It must be said, though, that the interest theory faces an additional (and 
somehow unexpected) problem; specifically, its inconsistency with positivism 
(the separation thesis and the inherent assumption that, irrespective of content, 
norms belong to the set since they are produced under the system’s criteria of 
identification).154 As the previous (clxxviii) example shows, the interest theory 
fails to explain norms that confer claim-rights (or liberties) without serving the 
interest of the holder (or of any typical bearer). So, while not recognizing such 
rights as “rights”, the interest theory is either (i) distorting those criteria, or (ii) 
imposing the (moral) interest of the (typical) bearer as a criterion of validity.
153 On the other hand, if the will theory is just a terminological claim, then it is pointless to dis-
cuss it. On the criticism regarding the will theory, Kramer 2013: 262, MacCormick 1982: 157. 
On theories and “why-questions”, Hempel 1965: 335.
154 Pino 2014: 200, Guastini 1994: 224.
Bibliography
Alchourrón, C., & Bulygin, E. (1971). Normative Sys-
tems. Springer Verlag.
Alchourrón, C., & Bulygin, E. (1991). Permisos 
y Normas Permisivas. In C. Alchourrón & E. 
Bulygin, Análisis Lógico y Derecho (pp. 215-238). 
Centro de Estudios Constitucionales.
Alexy, R. (2002). A Theory of Constitutional Rights. 
Oxford University. Press.
Amselek, P. (1988). Philosophy of Law and the 
Theory of Speech Acts. Ratio Juris, 1 (3), 187-223.
Atienza, M., & Manero, J. (1998). A Theory of Legal 
Sentences. Dordrecht: Springer.
Arriagada, M. B. (2018). Fundamentality, Inter-
definability, and Circularity Three Ideas on Hoh-
feld Examined. Revus. 35. 1-37.
Austin, J. (1875). Lectures on Jurisprudence. The Phi-
losophy of Positive Law. Henry Holt And Company.
Austin, J. L. (1955). How To Do Things with Words. 
Clarendon Press.
Bach, K. (2010). Refraining, Omitting, and Negative 
Acts. In T. O’Connor & C. Sandis (eds.), A 
Companion to the Philosophy of Action (pp. 50-
57). Wiley-Blackwell.
Barker, K. (2018). Private Law, Analytical Philosophy 
and The Modern Value of Wesley Newcomb 
Hohfeld: A Centennial Appraisal. Oxford Journal 
of Legal Studies, 38 (3), 585-612.
Bentham, J. (1970). Of Laws in General. The Athlone 
Press.
Bentham, J. (1838). The Works Of Jeremy Bentham. 
Volume III. William Tait.
Bentham, J. (1945). The Limits of Jurisprudence De-
fined. Columbia University Press.
Bernstein, S. (2016). Omission. Impossible. Philo-
sophical Studies, 173, 2575-2589.
Biasetti, P. (2015). Hohfeldian Normative Systems. 
Philosophia, 43, 951-959.
Black, M. (1969). The Gap Between “Is” and “Should”. 
In W. Hudson, The Is/Ought Question (pp. 99-
112). MacMillan.
68
(2023) 51
journal for constitutional theory and philosophy of law
David Duarte
Borowski, M. (2000). La Restricción de los Derechos 
Fundamentales. Revista Española De Derecho 
Constitucional, 20, 29-56.
Broad, C. (1944). Von Wright on the Logic of In-
duction. Mind, 53, 1-24.
Brown, V. (2005). Rights, Liberties and Duties: Re-
formulating Hohfeld’s Scheme of Legal Relations? 
Current Legal Problems, 58, 343-367.
Brożek, B. (2015). On Tû–Tû. Revus, 27, 15-23.
Brożek, B. (2004). Defeasibility of Legal Reasoning. 
Zakamycze.
Bulygin, E. (2015a). Norms, Normative Propositions 
and Legal Statements. In C. Bernal, C. Ochoa & 
others, Essays in Legal Philosophy (pp. 188-206). 
Oxford University Press.
Bulygin, E. (2015b). Kelsen on the Completeness and 
Consistency Of Law. In C. Bernal, C. Ochoa & 
others, Essays in Legal Philosophy (pp. 337-353). 
Oxford University Press.
Bulygin, E. (2010). Sobre la Equivalencia Pragmática 
Entre Permiso y No Prohibición. Doxa, 33, 283-
296.
Bulygin, E. (1991). Sobre las Normas de Compe-
tencia. In C. Alchourrón & E. Bulygin, Análisis 
Lógico y Derecho (pp. 485-498). Centro de Estú-
dios Constitucionales.
Bulygin, E. & Mendonca, D. (2005). Normas y 
Sistemas Normativos. Marcial Pons.
Calsamiglia, A. (1994). Geografia de las Normas de 
Competencia. Doxa, 15/16, 747-767.
Campbell, C. (1996). The Myth of Social Action. 
Cambridge University Press.
Caracciolo, R. (1986). La Noción de Sistema en la 
Teoría del Derecho. Fontanamara.
Castañeda, H-N. (1972). On the Semantics of the 
Ought-To-Do. In D. Davidson & G. Harman, 
Semantics of Natural Languages (pp. 675-694). 
Reidel Publishing.
Celano, B. (2019). Los Derechos en el Estado 
Constitucional. Palestra Editores.
Clarke, R. (2012). What Is an Omission? Philo-
sophical Issues, 22, 127-143.
Cook, W. (1919). Hohfeld’s Contribution to the 
Science of Law. The Yale Law Journal, 28, 721-
738.
Corbin, A (1919). Legal Analysis and Terminology. 
The Yale Law Journal, 29, 163-173.
Cruft, R. (2019). The Circularity of the Interest and 
the Will Theories of Rights. In M. McBride, New 
Essays in the Nature of Rights (pp. 169-186). Hart 
Publishing.
Cruft, R. (2013). Rights and the Direction of Duties. 
Ethics, 123, 195-201.
Cruft, R. (2004). Beyond Interest Theory and Will 
Theory ? Law and Philosophy, 23 (4), 347-397.
D’Almeida, L. D. (2016). Fundamental Legal Con-
cepts: The Hohfeldian Framework. Philosophy 
Compass, 11, 554-569.
D’Almeida, L. D. (2015). Allowing For Exceptions. 
Oxford University Press.
D’Almeida, L. D. (2009). Normas Jurídicas Com-
pletas. In Liber Amicorum Sousa Brito (pp. 395-
410). Almedina.
Davidson, D. (2001). Individuation of Events. In 
Essays on Actions and Events (pp. 137-151). 
Oxford University Press.
Duarte, D (2012). An Experimental Essay on the 
Antecedent and Its Formulation. I-Lex, Rivista 
Di Scienze Giuridiche, Scienze Cognitive Ed 
Intelligenza Artificiale, 16, 37-60.
Duarte, D. (2011). Linguistic Objectivity in Norm 
Sentences: Alternatives in Literal Meaning. Ratio 
Juris, 24 (2), 111-138.
Duarte, D. (2009). Rebutting Defeasibility as 
Operative Normative Defeasibility. In Liber Ami-
corum Sousa Brito (pp. 161-174). Almedina.
Edmundson, W. (2004). An Introduction to Rights. 
Cambridge University Press.
Eleftheriadis, P. (2008). Legal Rights. Oxford Uni-
versity Press.
Ferrer, J. (2000). Las Normas de Competencia. Un 
Aspecto de la Dinámica Jurídica. Marcial Pons.
Finnis, J. (2011). Rights: their logic restated. Phi-
losophy of Law. Collected Essays, Volume 4 (375-
388).
Florczak-Wator, M. (2015). Applying the Constitution 
of the Republic of Poland in Horizontal Relations. 
Jagellonian University Press.
Frändberg, Å. (2018). The Legal Order. Springer.
Frydrych, D. (2019). Rights Correlativity. In S. 
Balganesh, T. Sichelman & H. Smith, The Legacy 
of Wesley Hohfeld: Edited Major Works, Selected 
Personal Papers, and Original Commentaries (pp. 
1-35). Cambridge University Press.
Frydrych, D. (2018). Hohfeld vs. The Legal Realists. 
Legal Theory, 24, 291-344.
Grabowski, P. (2009). Enactment, Provision, 
Norm: Reflections on the Normativeness of 
Provisions Regulating the Process of Legislation. 
Investigationes Linguisticae, XVII, 129-140.
Guastini, R. (2016). La Sintaxis del Derecho. Marcial 
Pons.
Guastini, R. (1999). Distinguiendo. Barcelona: Gedisa.
Guastini, R. (1994). Invalidity. Ratio Juris, 7 (2), 212-
226.
Guastini, R. (1993). Le Fonti Del Diritto e 
L’Interpretazione. Giuffrè.
Guibourg, R., Echave, D. & Urquijo, M. (2008). 
Lógica, Proposición y Norma. Astrea.
Hage, J. (2018). Foundations and Building Blocks of 
Law. Eleven International Publishing
Hage, J. (2005). Studies on Legal Logic. Springer
Halpin, A. (2019a). Choosing Axioms of Correlati-
vity. The American Journal of Jurisprudence, 64 
(2), 225–258.
Halpin, A. (2019b). Correlativity and Its Logic: 
69
journal for constitutional theory and philosophy of law
(2023) 51
Rights as formal combinations of normative variables
Asymmetry Not Equality in the Law. Canadian 
Journal of Law and Jurisprudence, 32 (1), 83-108.
Halpin, A. (2014). Bentham’s Limits and Hohfeld. 
In Guillaume Tusseau, The Legal Philosophy 
And Influence Of Jeremy Bentham (pp. 196-223). 
Routledge.
Halpin, A. (2007). Rights, Duties, Liabilities, and 
Hohfeld. Legal Theory, 13, 23-29.
Halpin, A. (2003). Fundamental Legal Concepts 
Reconsidered. Canadian Journal of Law and 
Jurisprudence, 16 (1), 41-54.
Halpin, A. (1997). Rights and Law Analysis and 
Theory. Hart Publishing.
Halpin, A. (1985). Hohfeld’s Conceptions: From 
Eight to Two. Cambridge Law Journal, 44 (3), 
435-457.
Hansson, S. (2013). The Varieties of Permission. 
In D. Gabbay, J. Horty & others, Handbook of 
Deontic Logic and Normative Systems (pp. 195-
240). College Publications.
Hart, H. (1994). The Concept of Law (2nd edition). 
Oxford University Press.
Hart, H. (1982). Legal Rights. In Essays on Bentham 
(pp. 162-193). Oxford University Press.
Hedahl, M. (2013) The Significance of a Duty’s 
Direction: Claiming Priority Rather Than Pri-
oritizing Claims. Journal of Ethics & Social Phi-
losophy, 7 (3), 1-29.
Hempel, C. (1965). Aspects of Scientific Explanation. 
The Free Press.
Herrestad, H. (1996). Formal Theories of Rights. 
Juristforbundets Forlag.
Herrestad, H., & Krogh, C. (1995). Obligations 
Directed from Bearers to Counterparties. In 
Proceedings From Icail’95 (pp. 210-218). ACM 
Press.
Hilpinen, R. (1997). On States, Actions, Omissions 
and Norms. In G. Holmström-Hintikka & R. 
Tuomela, Contemporary Action Theory, I (pp. 83-
107). Springer.
Hintikka, J. (1970). Some Main Problems of Deontic 
Logic. In R. Hilpinen, Deontic Logic: Introductory 
and Systematic Readings (pp. 59-104). Reidel 
Publishing.
Hills, A. (1003). Duties and Duties to the Self. Ame-
rican Philosophical Quarterly, 40 (2), 131-142.
Hohfeld, W. (1919). Fundamental Legal Concepts. 
Yale University Press.
Hornsby, J. (2004). Agency and Actions. In J. Hyman 
& H. Steward, Agency and Action (pp. 1-24). 
Cambridge University Press.
Heidi Hurd, H., & Moore, M. (2018). The Hohfeldian 
Analysis of Rights. The American Journal of 
Jurisprudence, 63 (2), 295-354.
Iacona, A. (2015). Quantification and Logical Form. 
In A. Torza, Quantifiers, Quantifiers, Quantifiers: 
Themes in Logics, Metaphysics, and Language (pp. 
125-141). Springer.
Jaffey, P. (2006). Duties and Liabilities in Private Law. 
Legal Theory, 12, 137-156.
James, M. H. (1973). Bentham on the Individuation 
of Laws. Northern Ireland Legal Quarterly, 24, 
357-382.
Johns, C. (2014). Leibniz and the Square: A Deontic 
Logic for the Vir Bonus. History And Philosophy 
of Logic, 35 (4), 369-376.
Jovanović, M. (2012). Collective Rights. A Legal 
Theory. Cambridge University Press.
Kelsen, H. (2005). Pure Theory of Law (2nd edition). 
The Lawbook Exchange.
Kocourek, A. (1923). Non-Legal Content Relations 
Recombated. Illinois Law Quarterly, 5 (3), 150-
159.
Kocourek, A (1921). The Hohfeld System of Funda-
mental Legal Concepts. Illinois Law Review, 15, 
24-39.
Kramer, M. (2019). On No-Rights and No Rights. 
The American Journal of Jurisprudence, 64 (2), 
213-223.
Kramer, M. (2013). Some Doubts About Alternatives 
to the Interest Theory of Rights. Ethics, 123 (2), 
245-263.
Kramer, M. (2001). Getting Rights Right. In M. 
Kramer, Rights, Wrongs And Responsibilities (pp. 
28-95). Palgrave.
Kramer, M. (1998). Rights Without Trimmings. In 
A Debate Over Rights (pp. 7-111). Oxford Uni-
versity Press.
Kumm, M. (2007). Political Liberalism and the 
Structure of Rights: On the Place and Limits of 
the Proportionality Requirement. In George 
Pavlakos, Law, Rights And Discourse (pp. 109-
129). Hart Publishing.
Kurki, V. (2017). Legal Competence and Legal Power. 
In Mark McBride, New Essays In The Nature Of 
Rights (pp. 31-48). Hart Publishing.
Lindhal, L. (2017). Legal Power: The Basic Defi-
nition. Ratio Juris, 30 (2), 158-185.
Lindhal, L. (2006). Hohfeld Relations and Spielraum 
for Action. Análisis Filosófico, 26 (2), 325-355.
Lindhal, L. (1977). Position and Change. A Study in 
Law and Logic. Reidel Publishing Company.
Lopes, P. M. (2017). The Syntax of Principles: Ge-
nericity as a Logical Distinction Between Rules 
and Principles. Ratio Juris, 30 (4), 471-490.
Lyons, D. (1970). Correlativity of Rights and Duties. 
Noûs, 4 (1), 45-55.
MacCormick, N. (2005). Rhetoric And the Rule of 
Law. Oxford University Press.
MacCormick, N. (1982). Children’s Rights: A Test-
Case for Theories of Rights. In Legal Right and 
Social Democracy: Essays in Legal and Political 
Philosophy (pp. 154-166). Oxford University 
Press.
MacCormick, N. (1977). Rights in Legislation. In 
P. Hacker & J. Raz, Law, Morality, and Society. 
70
(2023) 51
journal for constitutional theory and philosophy of law
David Duarte
Essays in Honour of Herbert Hart (pp. 189-209). 
Clarendon Press.
Mackie, J. (1965). Causes and Conditions. American 
Philosophical Quarterly, 2 (4), 245-264.
Makinson, D. (1986). On the Formal Representation 
of Rights Relation: Remarks on the Work of Stig 
Kanger and Lars Lindahl. Journal of Philosophical 
Logic, 15 (4), 403-425.
Manero, J. (2018). Diez Observaciones y un Quadro 
Final Sobre Permisos y Normas Permisivas. Re-
vus, 36, 43-54.
Markovich, R. (2019). Understanding Hohfeld 
and Formalizing Legal Rights: The Hohfeldian 
Conceptions and Their Conditional Conse-
quences. Studia Logica, 108, 129-158.
Marmor, A. (1997). On the Limits of Rights. Law and 
Philosophy, 16 (1), 1-18.
Martin, R. (1993). A System of Rights. Clarendon 
Press.
Mele, A. (1997). Passive Action. In G. Holmström-
Hintikka & R. Tuomela, Contemporary Action 
Theory, I (pp. 135-144). Springer.
Mendonca, D. (2018). Análisis Constitucional. 
Una Introducción. Centro de Estudios Consti-
tucionales.
Mendonca, D. (2007). Sobre el Princípio de Per-
misión. In J. J. Moreso & C. Redondo, Un Diálogo 
Con la Teoría del Derecho de Eugenio Bulygin (pp. 
57-60). Marcial Pons.
Mendonca, D. (2000). Las Claves del Derecho. Gedisa.
Miller, S. (2004). Social Action. A Teleological 
Account. Cambridge University Press.
Moreso, J. J., & Navarro, P. (1998). The Reception of 
Norms and Open Legal Systems. In S. Paulson 
& B. Paulson, Normativity and Norms (pp. 273-
291). Oxford University Press.
Múrias, P. (2018). Personal Value in the Analysis of 
Rights. Católica Law Review, 2 (2), 25-40.
Navarro, P. (2019). Tipos de Permisos y Discreción 
Judicial. In E. Bulygin, El Libro De Los Permisos 
(pp. 61-85). Marcial Pons.
Navarro, P., & Rodríguez, R. (2004). Deontic Logic 
and Legal Systems. Cambridge University Press.
Newman, D. (2011). Community and Collective 
Rights. Hart Publishing.
Niiniluoto, I. (1991). Norm Propositions Defended. 
Ratio Juris, 4, 367-373.
Nino, C. S. (2003). Introducción Al Análisis Del 
Derecho (13ª edición). Ariel.
Nordenfelt, L. (1997). On Ability, Opportunity and 
Competence: An Inquiry Into People’s Possibility 
for Action. In G. Holmström-Hintikka & R. 
Tuomela, Contemporary Action Theory, I (pp. 
145-158). Springer.
Nyquist, C. (2002). Teaching Wesley Hohfeld’s 
Theory of Legal Relations. Journal of Legal 
Education, 52, 238-257.
O’Reilly, D. (1999). Are There Any Fundamental 
Legal Conceptions? University of Toronto Law 
Journal, 49 (2), 271-279.
O’Reilly, D. (1995). Using The Square of Opposition 
to Illustrate the Deontic and Alethic Relations 
Constituting Rights. University of Toronto Law 
Journal, 45 (3), 279-310.
Pacherie, E. (2011). Framing Joint Action. Review of 
Philosophy and Psychology, 2, 173-192.
Peczenik, A. (2008). On Law and Reason. Springer.
Penner, P. (1997). The Analysis of Rights. Ratio Juris, 
10 (3), 300-315
Peters, S., & Westerståhl, D. (2006). Quantifiers in 
Language and Logic. Clarendon Press.
Pino, G. (2016). Teoria Analítica del Diritto I. La 
Norma Jurídica. Edizione ETS.
Pino, G. (2014). Positivism, Legal Validity, and the 
Separation of Law and Morals. Ratio Juris, 27 (2), 
190-217.
Pörn, I. (1977). Action Theory and Social Science. 
Reidel Publishing Company.
Prakken, H., & Sartor, G. (2004). The Three Faces of 
Defeasibility in the Law. Ratio Juris, 17 (1), 118-
139.
Preda, A. (2015). Rights: Concept and Justification. 
Ratio Juris, 28 (3), 408-415.
Radin, M. (1929). Correlation. Columbia Law Re-
view, 29 (7), 901-905.
Ratti, G. (2018). Tres Addenda a Sistemas Norma-
tivos. Doxa, 41, 85-99.
Ratti, G. (2013). El Gobierno de las Normas. Marcial 
Pons.
Raz, J. (1984). On the Nature of Rights. Mind, 93, 
194-214.
Raz, J. (1980). The Concept of a Legal System (2ª edi-
tion). Clarendon Press.
Raz, J. (1975). Permissions and Supererogation. 
American Philosophical Quarterly, 12 (2), 161-168.
Robinson, R., Coval, S., & Smith, J. (1983). The Logic 
of Rights. The University of Toronto Law Journal, 
33 (3), 267-278.
Ross, A. (1968). Directives and Norms. Routledge & 
Kegan Paul.
Ross, A. (1958). On Law and Justice. Stevens and 
Sons.
Ross, A. (1957). Tû-Tû. Harvard Law Review, 70, 
812-825.
Ruiter, D. (1998). Legal Powers. In S. Paulson & B. 
Paulson, Normativity and Norms (pp. 471-491). 
Clarendon Press.
Ruiter, D. (1993). Institutional Legal Facts. Legal 
Powers and their Effects. Springer.
Sanford, D. (1989). If P, then Q: Conditionals and the 
Foundations of Reasoning. Routledge.
Sanford, D. (1984). The Direction of Causation 
and the Direction of Time. Midwest Studies In 
Philosophy, 40, 53-75.
Salmond, J. (1913). Jurisprudence (4th edition). 
Stevens and Haynes.
71
journal for constitutional theory and philosophy of law
(2023) 51
Rights as formal combinations of normative variables
Sardo, A. (2018). Let’s Talk About Antinomies. Nor-
mative Systems Reloaded. Revus, 36, 53-79.
Sartor, G. (2009). Understanding and Applying Legal 
Concepts: An Inquiry on Inferential Meaning. In 
J. Hage & D. Pforden, Concepts in Law (pp. 35-
54). Springer.
Sartor, G. (2006). Fundamental Legal Concepts. 
Artificial Intelligence and Law, 14, 101-142.
Searle, J. (1969). Speech Acts. An Essay in the Phi-
losophy of Language. Cambridge University Press.
Sergot, M. (2013). Normative Positions. In D. 
Gabbay, J. Horty & others, Handbook Of Deontic 
Logic And Normative Systems (pp. 353-405). 
College Publications.
Schauer, F. (1991). Playing by the Rules. Clarendon 
Press.
Schlag, P. (2015). How To Do Things with Hohfeld. 
Law And Contemporary Problems, 78, 185-233.
Schmid, H. (2009). Plural Action. Springer.
Simmonds, N. (1998). Rights at the Cutting Edge. 
In A Debate Over Rights (pp. 113-231). Oxford 
University Press.
Singer, M. (1959). On Duties to Oneself. Ethics, 
69(3), 202-205.
Spaak, T. (2009). Explicating the Concept of Com-
petence. In J. Hage & D. Pforden, Concepts in 
Law (pp. 67-80). Springer.
Spaak, T. (2003). Norms That Confer Competence. 
Ratio Juris, 16 (1), 89-104.
Spena, A. (2012). The Strange Case of the Protective 
Perimeter: Liberties and Claims to Non-Inter-
ference. Law and Philosophy, 31, 161-184.
Sreenivasan, G. (2010). Duties and Their Direction. 
Ethics, 120 (3), 465-494.
Sreenivasan, G. (2005). A Hybrid Theory of Claim-
Rights. Oxford Journal of Legal Studies, 25 (2), 
257-274.
Steiner, H. (2013). Directed Duties and Inalienable 
Rights. Ethics, 123 (2), 230-244.
Steiner, H. (1994). An Essay on Rights. Blackwell.
Sumner, L. (1987). The Moral Foundations of Rights. 
Clarendon Press.
Tuomela, R. (2010). The Philosophy of Sociality. 
Oxford University Press.
Tuomela, R. (2000). Cooperation. A Philosophical 
Study. Springer.
Tuomela, R. (1984). A Theory of Social Action. Reidel 
Publishing Company.
Wall, S. (2007). Collective Rights and Individual 
Autonomy. Ethics, 117 (2), 234-264
Weissinger, J. (2019). Content and Competence. A 
Descriptive Approach to the Concept of Rights. 
Mohr Siebeck.
Wellman, C. (2016). Constitutional Rights. What 
They Are and What They Ought to Be. Springer.
Wellman, C. (1997). An Approach to Rights. Springer.
Wenar, L. (2013). The Nature of Claim-Rights. Ethics, 
123 (2), 202-229.
Wenar, L. (2005). The Nature of Rights. Philosophy & 
Public Affairs, 33 (3), 223-252.
Williams, G. (1956). The Concept of Legal Liberty. 
Columbia Law Review, 56 (8), 1129-1150.
Woleński, J. (2008). Applications of Squares of 
Oppositions and their Generalizations in Phi-
losophical Analysis. Logica Universalis, 2, 13-29.
Wright, G. von (1963). Norm and Action. Routledge 
& Kegan Paul.
Wright, G. von (19511). An Essay in Modal Logic. 
North-Holland Publishing Company.
Wright, G. von (19512). A Treatise on Induction and 
Probability. Harcourt, Brace & Co.
Zhu, J (2004). Passive Action and Causalism. Philo-
sophical Studies, 119, 295-314.
Zorrilla, D. (2007). Conflictos Constitucionales, Pon-
deración y Indeterminación Normativa. Madrid: 
Marcial Pons.
Zucca, L. (2007). Constitutional Dilemmas. Conflicts 
of Fundamental Rights in Europe and the USA. 
Oxford University Press.

73
journal for constitutional theory and philosophy of law
(2023) 51
revus (2023) 51, 73–88
Miklós Könczöl*
Parental proxy voting and political 
representation
This paper reviews the proposal to give parents extra votes that they can cast as proxies 
on behalf of their children. Justifications of parental proxy voting (PPV) are examined 
with a focus on various interpretations of the concept of ‘proxy’. The first part of the 
paper assesses the notion that PPV does not violate the principles of equal and direct suf-
frage. Contrary to proponents of PPV, I argue that parents voting on behalf of their chil-
dren cannot be considered as merely expressing children’s political preferences, and that 
persons who are taken to be unable to make a decision themselves cannot be represented 
in this way. Thus, PPV actually allocates extra voting rights to parents, giving additional 
weight to their preferences in decision-making. The second part turns to parents as pos-
sible proxies for children’s interests, with their extra votes being meant to overweigh 
those of the elderly or of non-parents. PPV thus understood could be supported by the 
claim that parents are better situated to represent their children’s interests than the aver-
age voter. Proposals of PPV usually refer to parents’ better access to information, their 
shared interests with their children, and/or their selflessness. These arguments are, how-
ever, either irrelevant or questionable, and do not therefore actually speak in favour of 
the introduction of PPV. In conclusion, while PPV is usually depicted as making politi-
cal decisions simultaneously more democratic and more prudent, it does neither. Since 
these aims cannot be achieved through a single institution, different methods to achieve 
each aim need to be explored.
Keywords: political representation, voting, suffrage, age discrimination, parental proxy 
voting
1 INTRODUCTION
This paper deals with a proposal, formulated repeatedly in several countries 
(but most prominently in Germany) since the 1970s,1 to have parents vote as 
* konczol.miklos@jak.ppke.hu | Research fellow at the Centre for Social Sciences, Institute for 
Legal Studies, and associate professor at Pázmány Péter Catholic University, Faculty of Law, 
Budapest (Hungary). This paper summarises research undertaken within two projects, ‘Tak-
ing Age Discrimination Seriously’ (grant ID: 17–26629S) at the Institute of State and Law of 
the Academy of Sciences of the Czech Republic, Centre for Law and Public Affairs (CeLAPA), 
and ‘Institutional Reforms in Ageing Societies: Legal and Political Perspectives’ (grant ID: 
OTKA K 112900) at the Centre for Social Sciences, Budapest.
1 First by Löw 1974. On the subsequent debate, see the overview by Westle 2006. While the 
idea has found support across German parliamentary factions (cf. Jesse 2003: 8–9), such a re-
form has never been adopted. The discussion continues, however (see, more recently, Adrian 
2016). In Hungary, the proposal has been formulated several times since the democratic tran-
sition (for a brief overview, see Reimer & Schanda 2016: 70). In 2011, shortly before enacting 
the new constitution, the government sent questionnaires to the electorate, which included 
a question about parental proxy votes. The majority of the respondents rejected the idea. For 
74
(2023) 51
journal for constitutional theory and philosophy of law
Miklós Könczöl
proxies on behalf of their children until they become of legal voting age. While 
the proxy proposal is part of efforts to find institutional solutions for problems 
related to sustainability,2 the arguments formulated in the discourse go beyond 
the question of efficiency, i.e., whether the proposed reform could contribute 
to more sustainable policies, and raise fundamental problems of constitutional 
law, as well as legal and political theory.
While its supporters tend to regard it as a kind of panacea for several issues 
concerning sustainability and democracy, the proposal has provoked quite some 
controversy. It has been challenged from at least three perspectives: first, as vio-
lating the democratic (and constitutional) principles of equal and direct (person-
al) suffrage;3 second, as unnecessary, since the lack of suffrage does not mean a 
lack of representation for children;4 and third, as unhelpful, either because extra 
parental votes may not, or at least not only, be used to promote sustainability or 
children’s interests,5 or because these votes may not suffice to make a difference 
in terms of voting outcomes.6 Related to this latter point, a ‘reform paradox’ may 
arise in the sense that wherever such an institution would receive the necessary 
support within the electorate to be enacted, it would not be necessary.7
The arguments anticipating or in response to the above objections can be clas-
sified as either ‘deontological’ or ‘consequentialist’.8 While a clear distinction be-
the discussion in constitutional scholarship since 2011, see Fröhlich 2011 and Schanda 2012 
who are in favour of the proposal, Jakab (2017, 2020) who maintains some practical reserva-
tions, and Bodnár 2014, M. Balázs (2017, 2018) and Kurunczi 2018 who are against it. For 
an overview of similar proposals in other countries, see van Parijs 1998: 309–310 and Reimer 
& Schanda 2016: 63–64, as well as de Briey, Héraut & Ottaviani 2009 on the Belgian plural 
voting system at the turn of the 19th and 20th centuries, and Bertaux 2011: 126–127 on the 
proposal for a family vote in France in the 1910s and 1920s.
2 Cf. van Parijs (1998), Jakab (2019, 2020). Sustainability is used rather broadly in the related 
literature, including environmental, financial, and demographic aspects. A natalist perspec-
tive seems to play an important role in many proposals (cf. van Parijs 1998: 314, and, for a 
more recent example, Löw 2005), with a clear focus on fertility rates in the works of Paul 
Demeny (see e.g., Demeny 1986, 2012), after whom the term ‘Demeny voting’ is used for PPV 
(cf. Sanderson & Scherbov 2007: 548).
3 See, within the context of German constitutional law, Müller-Franken 2013, with further ref-
erences.
4 See, e.g., Müller-Franken 2013: 59, with reference to Pechstein 1996: 84–85.
5 See, e.g., de Briey 2007.
6 On the lack of convincing empirical arguments, see Müller-Franken 2013: 104, quoting Wes-
tle 2006, and Kahl 2009. See also Goerres & Tiemann 2009, based on data from German gen-
eral elections, arguing that while age may be relevant for voting behaviour, the introduction of 
PPV would not have made any considerable difference for the election results between 1994 
and 2005. The possibility of conflicts of interests even among the group of parents is empha-
sised by Wernsmann 2005: 66.
7 Cf. Offe 1994, quoted by Goerres & Tiemann 2009: 58.
8 For that usage, see, e.g., Reimer 2004, de Briey 2007, Goerres & Tiemann 2009, and Reimer 
& Schanda 2016. In addition to these two groups, Westle 2006 distinguishes constitutional 
75
journal for constitutional theory and philosophy of law
(2023) 51
Parental proxy voting and political representation
tween the two may be difficult to make in some cases, they do clearly differ in their 
scope in that the former are meant to explain why children should have suffrage,9 
whereas the latter are aimed at justifying parental voting on behalf of children.
In what follows, I focus on the concept of, and justifications offered for, pa-
rental proxy voting (PPV). In doing so, I shall distinguish between what seem 
to me two different sets of arguments and examine each in turn. The first, the 
‘Equal Suffrage Account’, sticks to the notion that PPV actually means suffrage 
for children, yet with their right to vote being exerted by their parents on their 
behalf. The second, the ‘Unequal Suffrage Account’, is meant to provide justi-
fication for the extra votes that are to be cast by parents. While most propo-
nents of PPV combine these in some form, the two sets seem to be relatively 
independent, at least in the sense that arguments used to support the Unequal 
Suffrage Account may work even if those formulated in favour of the Equal 
Suffrage Account do not hold.10
I am going to tackle each of the accounts by looking at the interpretations of 
‘proxy voting’ they are (or could be) based on. In the case of the Equal Suffrage 
Account, neither of the possible interpretations seems to be adequate for defend-
ing PPV, for conceptual reasons. While there is a different interpretation un-
derlying the Unequal Suffrage Account that may provide strong arguments for 
some kind of a proxy, it does not work for parental proxies specifically. What fol-
lows for children’s representation in political decision-making is that they either 
ought to be able to participate directly, or their interests should be represented 
by a more reliable proxy. These two possibilities do not exclude one another, but 
their functions cannot be merged in one institution, and certainly not in PPV.
2 THE EQUAL SUFFRAGE ACCOUNT
Given that children under a certain age do not currently have suffrage, an 
argument for PPV requires at least the following steps:
(1) Children should be enfranchised.
(2) Children cannot cast votes.
(3) Parents should vote on behalf of their children.
arguments.
9 There may be at least one exception to that, in case one regards suffrage as a good (see Her-
mann 2011: 44–45) and argues that parents deserve the extra votes (as they produce public 
goods from private goods). A similar argument (among others) is put forth by Vanhuysse 
2013. I am not going to address that kind of claim directly here, but it seems to be countered 
by the argument that on the same account, parents may not be the only candidates for allocat-
ing extra votes (see Westle 2006).
10 As a parallel, see the well-known proposal to abandon equality for proportionality in demo-
cratic decision-making by Brighouse & Fleurbaey 2010.
76
(2023) 51
journal for constitutional theory and philosophy of law
Miklós Könczöl
The proposal, in its above form, presupposes that the vote belongs to the 
child. Children, qua citizens, are regarded as having the right to vote (1),11 but 
also as needing a proxy to cast the vote for them (2). This solution is often com-
mended as being in line with the constitutional principle of ‘one person, one 
vote’.12 Proponents of PPV stress that even though parents would be able to cast 
more than one vote this way, the additional votes would not be their own.13 
This is important, since the most obvious objection to PPV, already mentioned 
above, is that it actually amounts to plural vote.14
As mentioned above, my focus is on the justification of (3), i.e., the question 
of whether it is the parents, rather than either average voters or experts, who 
should act as proxies for their children in a political decision-making process. 
Thus, I am not dealing here with the validity of either (1) or (2), but accept them 
for the sake of argument, and confine myself to examining what interpretations 
of ‘proxy’ can be used to support arguments in favour of (3). It seems that there 
are two such conceptions, which I discuss in turn below. The first regards the 
proxy as someone who merely expresses the child voter’s decision (1.1), where-
as in the second, the proxy makes the decision on behalf of the child (1.2).
2.1 Casting the vote
To counter the objection that PPV would violate the principle of direct 
suffrage,15 the institution of voting by intermediary is often mentioned, e.g., the 
British and French solutions of ‘voting by proxy’ or ‘vote par procuration’, which 
are offered as an alternative to voting by mail for people who are unable to at-
tend voting in person.16 One has to apply for proxy voting, and entrust some-
one who has the right to vote to cast one’s vote.17 In that usage, then, a ‘proxy’ is 
an intermediary for performing the physical actions related to voting. It is im-
11 This claim is most often supported by referencing the concept of popular sovereignty: since 
children are part of the people, so the argument goes, they should not be excluded from 
the exertion of sovereignty through suffrage (see e.g., Peschel-Gutzeit 1999: 560, Merk 2009: 
531–534, Schanda 2012: 81, Reimer & Schanda 2016: 73). A similar argument links suffrage 
to the recognition of human dignity (see e.g., Knödler 1996: 559–561, Schreiber 2004: 1344, 
Löw 2005: 34–35, Merk 2009: 534).
12 See e.g., Reimer & Schanda 2016: 64.
13 That is why the institution is sometimes referred to as ‘children’s vote vicariously exercised’ 
(abbreviated as ChiVi/KiVi, see e.g., de Briey 2007), or as ‘Stellvertreterwahlrecht’ (see e.g., 
Müller-Franken 2013: 6–7). I am going to use the term ‘parental proxy voting’ to emphasise 
the role of parents, which is the focus of the following arguments.
14 See e.g., de Briey, Héraut, & Ottaviani 2009. For the same reason, opponents of PPV argue that 
there is no practical difference between parents casting their own extra votes or those of their 
children, see e.g. Wernsmann 2005: 54–55.
15 Cf. Art. 38(1) of the Grundgesetz.
16 See e.g., Löw 2002, Fröhlich 2011.
17 See Representation of People Act 2000, section 12(1), and Code électoral, article 147bis, re-
spectively.
77
journal for constitutional theory and philosophy of law
(2023) 51
Parental proxy voting and political representation
portant to see here, however, that the parallel with PPV is not a very close one. 
Voting by proxy is reserved for persons who would be able to cast a vote at the 
time were it not for their specific conditions,18 while proponents of PPV assume 
that children qua children cannot vote themselves, other conditions aside (that 
is, children who would be able to attend voting are equally excluded).
In the above case, unlike the person entitled to the vote, a reliable proxy 
would be both physically and legally able to cast the vote and be expected to cast 
the actual vote of the incapable voter, i.e., the vote expressing the decision of the 
latter, similar to an agent acting on behalf of a principal. The ‘principal voter’, to 
be sure, can instruct the proxy to cast the vote according to her own best judge-
ment.19 The institution nevertheless requires that the former can at least make a 
decision to be expressed through the vote cast by the latter. Thus, in the case of 
PPV, acting as a proxy seems to refer not only to the actual casting of the vote 
(in the sense of placing the ballot into the box or performing a similar action), 
but to include the moment of decision-making too.20 Apparently, even propo-
nents of PPV focus on that latter interpretation, using the parallel of ‘voting by 
proxy’ only to argue that direct suffrage does not require that the vote be cast 
personally.21 Yet, in the case of children, it is not their presence that is lacking, 
but, arguably, their capability to make a decision.
2.2 Making the decision
The idea of PPV implies, by definition, that children (or at least some chil-
dren) cannot make the decision themselves, but also that someone else can do 
so on their behalf. The question is, again, whether that departs from the princi-
ple of direct suffrage. A parallel with marriage is sometimes drawn here. Those 
opposing PPV argue that there are certain highly personal decisions that cannot 
be made by any other than those concerned. In modern times, so the argument 
goes, both marriage and electoral votes belong to this category.22 In response, 
proponents of PPV claim that the nature of voting differs from marriage, since 
the latter is a personal decision with consequences that affect one’s personal life 
to a definitive extent, while the consequences of one’s vote appear in a very dif-
18 Cf. Müller-Franken 2013: 78, also pointing out that ‘proxy voting’ in the above sense depends 
on the explicit will of the voter, with the law only providing for the possibility to have a proxy 
cast the vote on one’s behalf.
19 See Löw 2005: 38–39, quoting the example of Lionel Jospin, who, having failed to receive 
enough votes to enter the second round of the 2002 presidential elections, was reported to 
say he would have one of his friends vote for him, as he could not decide which one of his 
competitors to support.
20 Cf. Gaa 1997: 345.
21 E.g., Reimer & Schanda 2016: 182.
22 See, e.g., Gaa 1997: 345, Hinrichs 2002: 52.
78
(2023) 51
journal for constitutional theory and philosophy of law
Miklós Könczöl
ferent way.23 For proponents of PPV, what needs to be expressed is one’s inter-
ests. Even if one does not believe, unlike people across many historical periods 
and many cultures, that one’s spouse could adequately be chosen by someone 
else, we have good reason to think that one’s interests related to political deci-
sion-making can be fully taken into account by another person.
On this account, what makes parents not only suitable proxies but the best 
possible candidates to act as proxies, is their closeness to the person they are 
representing. They are close to their children both in terms of information and 
motivation. They may be best situated to know their children’s personal inter-
ests not only due to the fact that they spend a considerable amount of time with 
them, but also because they are emotionally motivated to take interest in con-
sidering the possible effects that public decisions would have on their children. 
Thus, in case their children do actually have some kind of electoral preference, 
they may be both able and willing to discuss the matter with them. A possible 
parallel here may be decisions concerning children’s education. Although par-
ents choose their children’s schools, they are meant to, and possibly do, take 
their children’s opinion into account.24 That motivation can also extend to the 
act of voting: even though there may be conflicts of interests between parents 
and children, parents are the most likely to act selflessly, or at least to pay due 
respect to their children’s interest, even by splitting the votes they cast.25
A similar argument highlights that the special link between parents and 
children is appreciated by most modern legal systems, which make parents the 
guardians of their children’s financial and other rights. Children actually do 
have rights they cannot exert themselves, as well as duties they cannot fulfil by 
their own actions.26 They may have the legal capacity to acquire property, e.g., 
by inheritance, but the related decisions are beyond that capacity. It is parents 
who have the legal power, and duty, to decide and act in the best interest of 
their children.27 Thus, the argument is not just one of consistency (voting rights 
should be regulated in the same way as property rights, or educational rights, 
etc.), but also one of analogy: legislators, on this account, express their trust in 
parents as proxies for their children’s interests and preferences in a number of 
contexts, which should be followed for voting as well.28
23 See Hattenhauer 1996: 16, Reimer 2004, also Schanda 2012: 83, pointing out that the personal 
character of suffrage is not among the constitutional requirements.
24 See, in general, UN Convention on the Rights of the Child, Art. 12(1) on the right of the child 
to ‘express [his or her own] views freely in all matters affecting the child, the views of the child 
being given due weight in accordance with the age and maturity of the child’.
25 Cf. Schanda 2012: 83, Reimer & Schanda 2016: 68.
26 Cf. Steffani 1999: 792.
27 Cf. Merk 2009: 535.
28 See e.g., Nopper 1999: 151.
79
journal for constitutional theory and philosophy of law
(2023) 51
Parental proxy voting and political representation
Leaving a more detailed discussion of the above considerations concerning 
the suitability of parents as proxy voters to the next section, it is worth men-
tioning that the idea of PPV entails regarding parents as proxies in terms of 
making, rather than just expressing, voting decisions on behalf of children. The 
latter interpretation of ‘proxy’ is only mentioned to show already existing and 
constitutionally admissible parallels for the institution proposed. Yet, the no-
tion that parents make decisions on behalf of their children raises the question 
of whether (2) and (3) are compatible, i.e., whether it is at all possible to make 
decisions on behalf of persons who are assumed not to be capable of making 
those decisions themselves.
PPV is usually argued for by saying that it improves the political representa-
tion of children. Yet, it seems there are at least two different interpretations of 
representation at play here. If one urges that it is the suffrage of the child that is 
exerted by the parent, then the parent must do more than just take into account 
the interests or preferences of the child. Within the limits of the Equal Suffrage 
Account, which is meant to comply with the principles of equal and direct suf-
frage, the proxy voter would be expected to represent the child in a substantive 
way, such as by making the decision the child would make, reflecting their exact 
preferences.29 Here, the problem may be that children may not be interested 
in making political decisions at all. Even though they may have preferences in 
terms of, e.g., schools or religion at a relatively early age,30 that may not be the 
case with politics. Yet, if one assumes that children in general start having po-
litical preferences at an earlier stage than acknowledged by current regulation, 
for instance at 14 or 16 rather than 18 years, then, in the absence of compelling 
arguments against extending suffrage, a lowering of the voting age limit would 
seem adequate.31 If, however, there are children who can be safely assumed on 
the basis of their age not to have political preferences,32 that leaves no place for 
proxy voting in the former sense either.
Most proponents of PPV, however, do not seem to conceive of voting as the 
expression of individual decisions, but concentrate on the effects of the pro-
posed changes on the composition of the electorate, and consequently on voting 
outcomes. For that reason, their actual focus is not on the equality of suffrage 
but the inequality of voting power. Such an inequality is necessary, they argue, 
29 For the distinction between formal and substantive representation, see Böckenförde 1982: 
318–322.
30 In terms of religion, the German Act on the Religious Education of Children, Art. 5 provides 
that children above 12 can decide on denominational changes in their education, and above 
14 they are completely free to choose their religion.
31 On the proposal to lower the age limit for suffrage at the German federal elections, see e.g., 
Knödler 1996, and Hoffmann-Lange & De Rijke 1996. See, more recently, Lőrincz 2018 on the 
relationship between age and rational deliberation.
32 See e.g., Rosenberg 2016 on the justification of exclusion due to inability.
80
(2023) 51
journal for constitutional theory and philosophy of law
Miklós Könczöl
for the sake of an adequate representation of children’s interests (i.e., making 
decisions favourable for them and their future perspectives), and for public de-
cisions supporting more sustainable policies. We shall now turn to this claim.
3 THE UNEQUAL SUFFRAGE ACCOUNT
An interpretation of ‘parental proxies’ as proxies of interests thus requires a 
different argument, replacing (1) and (3) as follows:
(1’) The interests of children should be taken into account with a greater 
weight than is currently given at elections/referenda.
(3’) Parents should have additional votes to be able to represent these inter-
ests more efficiently.
(1’) is actually the assumption underlying most proposals for PPV, and does 
not in itself contradict the Equal Suffrage Account either. (3’), in turn, is a more 
explicit (and therefore less frequent) formulation of the claim regarding the de-
sirable outcome. What is more important for us here, however, is that the argu-
ments supporting parental proxies (3), mentioned briefly in section 1.2 above, 
can be used to support that reformulated proposal (3’) as well. I shall now ex-
amine these in turn: first the ‘access to information’ argument, then the one of 
‘altruism’. I then turn to a third possible argument, according to which even 
without being particularly well-informed or selfless, parents may represent chil-
dren’s interests better than other candidates since they share much of these.
3.1 Information
As for the first argument, we need to ask what makes parents better in-
formed in terms of their children’s interests than other possible candidates 
(such as teachers, NGOs, or experts in various fields). Starting from their close-
ness, the obvious answer seems to be that they have direct access to a wide ar-
ray of information concerning the personal needs of their children. Even if it is 
not parents who spend the greatest average amount of time with their children, 
they spend time together in a range of situations. Moreover, their attention may 
be less divided than that of, e.g., teachers, who usually need to pay attention to 
several children at the same time. Similarly, it is parents who may best know 
their children’s preferences, especially political preferences, as these are most 
often shaped by parental influence rather than by peer groups (who could not, 
according to the proposal, act as proxies anyway).
Here we may have another look at the possible parallels for parental de-
cision-making mentioned above. Schooling choices, for one, can be made by 
parents because they are best situated to take the child’s individual needs and 
81
journal for constitutional theory and philosophy of law
(2023) 51
Parental proxy voting and political representation
preferences as well as the available options into account. The same may be said, 
albeit less cogently, about financial decisions. A common feature of these paral-
lels is that the decisions are made in situations that are intimately related to the 
personal future of individual children: their physical and intellectual develop-
ment, health, social network, financial situation, etc. With respect to the alleged 
parallel with marriage, however, the objection is made that it differs from voting, 
since the impact of voting is not linked to the personality of the decision-maker.
Once we regard parents as proxies for children’s interests, i.e., as decision-
makers rather than decision-expressers, the represented preferences and inter-
ests lose their strictly personal character. Moreover, political decision-making, 
as conceptualised by proponents of PPV, is impersonal in the sense that even 
though voting is motivated by personal interests, what the votes express is the 
voters’ preferences informed by the perceived interests of specific groups to 
which the represented individuals belong. Although such a claim is controver-
sial in itself, more important for us is that while parents as proxies may have 
privileged access to information regarding the personal interests of their chil-
dren, they cannot be expected to perform above the average when it comes to 
representing the best interests of children qua children (as opposed to the el-
derly). They may, of course, be competent voters, but that is not due to their 
personal links to their own children. In other words, if one claims that PPV 
does not violate the principle of personal suffrage and wishes to promote chil-
dren’s group interests by way of giving extra votes to parents, then the kind of 
information necessary to become a competent proxy voter is not what parents 
qua parents have access to. Indeed, other possible proxies with access to expert 
information may seem to be stronger candidates.
3.2 Altruism
The second argument is based on the notion that parents would make better 
proxies than other people by virtue of being more altruistic towards their chil-
dren.33 In general, the problem of generational egoism seems to be at the very 
source of the proposal. Gerontocracy has no inherent negative connotations 
and has even been praised by several authors addressing the history of political 
ideas.34 It becomes problematic only if one looks at the relation between genera-
tions as a conflictual one. In this section, I first consider whether the conflictual 
approach is adequate, before raising the question of whether PPV can actually 
build on parental selflessness.
Speaking of an opposition of generations may seem justified based on 
changes in age-specific policies on the one hand, and political attitudes on the 
other. As for age-specific policies, the diverging trends in child poverty and the 
33 Cf. Reimer & Schanda 2016: 182.
34 Cf. Cupti 1998, Palmore 1999: 45.
82
(2023) 51
journal for constitutional theory and philosophy of law
Miklós Könczöl
poverty of the elderly are often highlighted,35 together with the impacts of aus-
terity measures on pensions as opposed to education. Attitude surveys further 
show a difference in support for pro-elderly and pro-youth policies among dif-
ferent age groups.36
Yet, such data are often far from unequivocal, leaving them open to vari-
ous interpretations. Moreover, arguments based on them are criticised because 
they fail to take private transfers into account. Grandparents, for instance, may 
support their children and grandchildren financially, thus spending part of the 
pro-elderly expenditure in a pro-youth way, which, however, remain invisible if 
one only focuses on state policies. A more careful survey of transfers would sug-
gest a more nuanced picture of inter-generational relations. Further, in terms of 
attitudes, it is less than evident that elderly people would lack intimate relation-
ships with younger age groups, through their children and grandchildren for 
instance. This makes such a clear-cut opposition, as suggested by the frequent 
use of the concept of ‘gerontocracy’, difficult to support.
Now, even if one accepts that there is a conflict of interests between different 
age groups, there are problems that result from the above considerations. PPV 
gives extra votes to parents, and the most common justifications for that refer 
to their broader time horizon, due to the links they have to their children. That 
constructs a twofold opposition: (1) parents versus the elderly, and (2) parents 
versus non-parents. The problem here is that if the elderly are regarded as hav-
ing a less future-oriented perspective, and if PPV is meant to remedy gerontoc-
racy, then it is hardly more than just a second-best option to disenfranchising 
the elderly. In the sense of (1), the extra votes go to parents not qua parents but 
as presumably younger people. In that case, however, inequality is created, in 
the sense of (2), among people belonging to the same age group: those who have 
children can cast more votes than those who do not.
As for the selflessness of parents as proxies of their children, the first thing 
to note is the difference between altruism towards one’s children and altruism 
towards younger age groups in general. Parental partiality towards their chil-
dren is indeed legitimate,37 but if amplified through extra votes, it may bring 
about inequality among children, as the burdens of political decisions may not 
be evenly distributed within the same cohort. The same applies to differences 
among families: parents with many children would have more votes than those 
with say one or two, in which case their extra votes may lead to family poli-
cies favouring larger families to a disproportionate degree. While the chances of 
such outcomes depend on the actual demographic situation of the given society, 
35 See, e.g., Peterson 1992, Hinrichs 2002.
36 Cf. Gründinger 2014, quoted by Schickhardt 2015: 192.
37 Cf. Brighouse & Swift 2009.
83
journal for constitutional theory and philosophy of law
(2023) 51
Parental proxy voting and political representation
they cannot be neglected in assessing the power of arguments yielded by the 
Unequal Suffrage Account.
Second, due to the fact that they have to care for their children and not just 
themselves, average per-capita family income may be considerably lower for 
parents than non-parents in the same age group.38 That may shorten their time 
horizon, if not in terms of attitudes, then in practice, as they may not be able to 
afford being selfless39 in terms of social sustainability. Moreover, in comparison 
to elderly people, young parents may have more to expect from generous pro-
elderly policies in the present as well as the future. On the one hand, public 
support for older age groups alleviates the individual burden of looking after 
one’s own parents. On the other hand, younger people may have more years left 
to spend being supported themselves, especially given the increasing life expec-
tancy in Western societies.40
The question of egoism and altruism can also be raised from a different per-
spective. It has been argued that the current system of voting, which assumes 
that each voter takes the interests of non-voters into account, raises too high 
of a moral standard for those not having emotional ties to non-voters as in the 
case of non-parents. PPV would then absolve voters, at least in terms of chil-
dren’s interests, from the moral burden attached to voting in an egoistic way.41 
From the above considerations it seems clear that parents may have such a bur-
den themselves, as they may have competing interests42 or emotional ties that 
point in different directions when it comes to political decision-making.
Finally, there is a more practical, and perhaps weightier, caveat against la-
belling groups of voters as egoistic or generationally short-sighted. On the one 
hand, such a distinction weakens the sense of moral equality among the mem-
bers of the political community. On the other hand, and perhaps even more 
importantly, such a perceived discrimination (be it age-based or status-based) 
may provoke a bloc voting behaviour in groups that are now diverse regarding 
their political preferences.43
The ‘selfless proxy’ argument, as we have seen, presupposes a clear distinc-
tion and a conflict between the interests (and corresponding attitudes) of differ-
ent (age) groups. While the grounds and practical consequences of that concep-
tion may be open to doubt themselves, we have also seen that parents’ interper-
sonal altruism may not easily translate into an intergenerational one. In the last 
38 Schreiber 2004: 1344 and Vanhuysse 2013 regard that as something that should be compen-
sated for by means of giving them extra votes.
39 See Weber 1958, quoted by de Briey 2007.
40 See Van Parijs 1999, p. 323, quoting Andrew Williams ibid., n. 72.
41 Adrian 2016.
42 As taken into account by dual-interest approaches to parental rights (see e.g. Brighouse & 
Swift 2006, Shields 2019).
43 De Briey 2007.
84
(2023) 51
journal for constitutional theory and philosophy of law
Miklós Könczöl
part of this section, I look at a possible counterargument, according to which 
parents may not need to have access to information concerning children’s inter-
ests or to be particularly altruistic.
3.3 Shared interests
The above considerations notwithstanding, one may still argue that parents 
make the best proxies available, not because of their ability to know their chil-
dren better or because of their altruistic motivations, but because of their situ-
ational proximity to their children’s interests. Their interests may be the clos-
est to their children’s from some relevant perspective, either because they share 
the same interests or because their interests are somehow related to (or derived 
from) those of their children.44 An example for the former may be family al-
lowances: both parents and children benefit from these, hence their interests 
coincide, and, one may add, that is what separates them from other groups not 
receiving such benefits. Thus, parents need neither specific information nor a 
selfless attitude when deciding about a policy affecting family benefits, since by 
following their own interests they also advocate those of their children. What 
needs to be made sure is only that they use those resources properly. The lat-
ter kind of case is where there is no coincidence but rather some close relation 
between parents’ and children’s interests. Educational spending may be an ex-
ample, as it later results in higher payments for pensioners.
While such an argument may succeed in raising doubt about whether par-
ents need to know children’s specific interests and to have an altruistic attitude, 
it seems that some kind of general information may nevertheless be necessary. 
For such an ‘invisible hand’ to work, parents need knowledge not (only) of chil-
dren’s interests but their own (as well). Part of today’s sustainability issues result 
not from generational egoism, but from shortsightedness regarding one’s own 
interests: one may think of ill-advised and yet popular policies, the sustainability 
of which does not go beyond electoral cycles. Moreover, shared interests are not 
limited to parents: any person belonging to an older generation may profit from 
higher pensions and, consequently, may regard the public financing of education 
as an investment. The same applies to emotional ties, which may similarly exist 
in relationships beyond those between parents and children. Thus, while parents 
do not seem to be in a privileged position in terms of motivation, one cannot 
argue, either, that shared interests can compensate for the lack of competence.
44 De Briey 2007 rightly opposes this argument to the ‘selfless proxy’ one, arguing that the for-
mer is preferable to the latter.
85
journal for constitutional theory and philosophy of law
(2023) 51
Parental proxy voting and political representation
4 CONCLUSION
In this paper I discussed arguments for and against parental proxy voting 
(PPV), trying to shed light on the logical relations between these, while critically 
examining them. In doing so, I did not follow the usual division of deontological 
and consequentialist arguments, but instead distinguished between arguments 
remaining within the framework of equal suffrage, and those not grounded in it. 
Equal-suffrage arguments emphasise the constitutional construction underly-
ing PPV, according to which parents, when acting as proxy voters, cast the votes 
of their children, meaning they do not have a plural vote. That interpretation, 
however, is difficult to maintain, and proponents of PPV often bring in unequal-
suffrage arguments as well. Within the latter group, I distinguished three types 
of arguments: those related to the notion that parents would make informed 
proxies, those based on the assumption that they would use their extra votes in 
a selfless manner, and, third, that shared interests make parents capable of repre-
senting children’s interests, even if they do not know it. I have argued that each 
of these arguments is open to doubt and some important objections.
Without repeating what has been said on the preceding pages, it seems im-
portant to note that my observations regarding the two lines of argument have 
different consequences. The conceptual problems resulting from the two inter-
pretations of proxy voting within the Equal Suffrage Account suggest that there 
is no way to defend PPV against the charge of violating the principle of equal 
suffrage. Not all proponents of PPV are, however, equally sensitive to that kind 
of objection, and some may contend that the adequate representation of chil-
dren’s interests should prevail over currently held constitutional principles. In 
the second part of the paper, an examination of arguments in favour of parents’ 
(as proxies of interests) having extra votes has shown that these, too, are open 
to serious objections. Thus, even if one would be willing to favour some kind 
of plural vote, parents may not be the strongest candidates for allocating the 
additional votes. Yet, perhaps in trying to address the perceived inadequacy of 
political decision-making, one should focus less on the issue of voting power, 
together with the underlying decisionist image of the political process, and in-
stead focus more on how to make political deliberation more robust, with long-
term perspectives receiving due consideration.
From the perspective of children’s political representation, proposals of PPV 
indirectly highlight a dilemma. The Equal Suffrage Account is partly based on 
arguments in favour of extending suffrage to children, which is then bracketed 
by the claim that children cannot be expected to contribute to political deci-
sion-making. Taking the conceptual problems discussed above seriously, a more 
convincing version of that approach would examine the possibility of actually 
having children vote. Even though an age limit may be in order, a small exten-
sion of genuine suffrage seems a more important step here than giving adults 
86
(2023) 51
journal for constitutional theory and philosophy of law
Miklós Könczöl
plural votes. As for the Unequal Suffrage Account, it is based, partly again, on 
a negative assessment of the quality of decisions. Here, my argument points 
towards different ways of introducing checks into the procedure of decision-
making. The dilemma is due to the fact that the first way, extending suffrage, 
would allow for a more democratic model, while the latter, building in checks, 
would make it more aristocratic. That said, the dilemma only emerges if one 
promises, as most proponents of PPV, to make political decision-making both 
better informed and more democratic at the same time. Apart from such pre-
tences, the two ways are not exclusive. Indeed, the one may compensate for the 
other, both in terms of quality and democratic legitimacy.
Bibliography
Adrian, A. (2016). Grundsatzfragen zu Staat und 
Gesellschaft am Beispiel des Kinder-/Stell-
vertreterwahlrechts: Eine rechtliche Untersuchung 
mit Bezügen zu Demographie, Demoskopie, Psy-
chologie und Philosophie. Duncker & Humblot.
Bertaux, S. (2011). Reproduce or Perish? The Artefact 
of the Fertility Concept and the French School of 
Demography. Historical Social Research, 36(2), 
120–139. https://doi.org/10.12759/hsr.36.2011.2 
.120-139
Bodnár, E. (2014). A választójog alapjogi tartalma és 
korlátai [The Content and Limits of Suffrage as a 
Fundamental Right]. HVG-Orac.
Böckenförde, E-W. (1982). Mittelbare/repräsen-
tative Demokratie als eigentliche Form der 
Demokratie: Bemerkungen zu Begriff und 
Verwirklichungsproblemen der Demokratie 
als Staats- und Regierungsform. In G. Müller 
(Ed.), Staatsorganisation und Staatsfunktionen 
im Wandel: Festschrift für Kurt Eichenberger 
zum 60. Geburtstag (pp. 301–328). Helbing & 
Lichtenhahn.
Brighouse, H., & Swift, A. (2006). Parents’ Rights 
and the Value of the Family. Ethics, 117(1), 80–
108. https://doi.org/10.1086/508034
Brighouse, H., & Swift, A. (2009). Legitimate Parental 
Partiality. Philosophy and Public Affairs, 37(1), 
43–80. https://doi.org/10.1111/j.1088-4963.2008 
.01145.x
Brighouse, H., & Fleurbaey, M. (2010). Democracy 
and Proportionality. Journal of Political Philo-
sophy, 18(2), 137–155. https://doi.org/10.1111/
j.1467-9760.2008.00316.x
Cupit, G. (1998). Justice, Age, and Veneration. Ethics, 
108(4), 702–718. https://doi.org/10.1086/233848
De Briey, L. (2007). Droit de vote des enfants par 
procuration. In DicoPo: Dictionnaire de théorie 
politique. https://www.dicopo.org/spip_article88.
html
De Briey, L., Héraut, A., & Ottaviani, E. (2009). On 
Behalf of Children? The Plural Voting System in 
Belgium – from 1893 to 1919. Intergenerational 
Justice Review, 9(4), 144–145. https://doi.org/ 
10.24357/igjr.4.4.512
Demeny, P. (1986). Pronatalist Policies in Low-
Fertility Countries: Patterns, Performance, and 
Prospects. Population and Development Review, 
12(Suppl.), 335–358. https://doi.org/10.2307/280 
7916
Demeny, P. (2012). Geopolitical Aspects of Po-
pulation in the Twenty-First Century. Po pulation 
and Development Review, 38(4), 685–705. https:// 
doi.org/10.1111/j.1728-4457.2012.00532.x
Fröhlich, J. (2011). Alapkérdések a családi választó-
jog vitájában [Some Fundamental Questions in 
the Debate on Family Suffrage]. Pázmány Law 
Working Papers, 2011/20. https://plwp.eu/docs/
wp/2012/2011-20.pdf
Gaa, M. (1997). Familienwahlrecht bei den nächsten 
Bundestagswahlen? Zeitschrift für Rechtspolitik, 
30(8), 345–346.
Goerres, A., & Tiemann, G. (2009). Kinder an 
die Macht? Die politische Konsequenzen des 
Stellvertretenden Elternwahlrechts. Politische 
Vierteljahresschrift, 50(1), 50–74. https://doi.org/ 
10.1007/s11615-009-0125-6
Gründinger, W. (2014). Scheinargumente gegen das 
Kinderwahlrecht. In K. Hurrelmann & T. Schulz 
(Eds.), Wahlrecht für Kinder? Politische Bildung 
und die Mobilisierung der Jugend (pp. 24–36). 
Beltz Juventa.
Hattenhauer, H. (1996). Über das Minderjährigen-
wahlrecht. JuristenZeitung, 51(1), 9–16. [Repr. in 
Chr. Palentien & K. Hurrellmann (Eds.), Jugend 
und Politik: Ein Handbuch für Forschung, Lehre 
und Praxis (pp. 238–259). Luchterhand.]
Hermann, U. (2011). Ökonomische Analyse eines 
Kinderwahlrechts. Doctoral dissertation, Freie 
Universität Berlin. http://www.diss.fu-berlin.de/
diss/servlets/MCRFileNodeServlet/FUDISS_der 
ivate_000000015340/veroeffentlichung_herma 
nn_online.pdf
87
journal for constitutional theory and philosophy of law
(2023) 51
Parental proxy voting and political representation
Hinrichs, K. (2002). Do the Old Exploit the Young? Is 
Enfranchising Children a Good Idea? European 
Journal of Sociology, 43(1), 35–58. https://doi.
org/10.1017/S0003975602001017
Hoffmann-Lange, U., & de Rijke, J. (1996). 16jährige 
Wähler – erwachsen genug? Die empirischen 
Befunde. Zeitschrift für Parlamentsfragen, 27(4), 
572–585.
Hurrellmann, K. (1997). Für eine Herabsetzung des 
Wahlalters. In Chr. Palentien & K. Hurrellmann 
(Eds.), Jugend und Politik: Ein Handbuch für 
Forschung, Lehre und Praxis (pp. 280–289). 
Luchterhand.
Jakab, A. (2020). Kinderwahlrecht für Nachhaltig-
keit? Rechtsdogmatische Einordnung und recht-
spolitische Effektivität. Journal für Rechtspolitik, 
28(1), 27–37. https://doi.org/10.33196/jrp202001 
002701
Jakab, A. (2019). An Emerging Key Concept in 
European Constitutional Law: Sustainability. 
Hungarian Journal of Legal Studies, 60(4), 332–
365. https://doi.org/10.1556/2052.2019.00020
Jakab, A. (2017). Is Suffrage for Children the Consti-
tutional Solution for the Sustainability Challenge 
Faced by Democracies? In I. Motoc, P. De Pinto 
de Albuquerque, & K. Wojtyczek (Eds.), New 
Developments in Constitutional Law: Essays in 
Honour of András Sajó (pp. 213–224). Eleven.
Jesse, E. (2003). Reformvorschläge zur Änderung des 
Wahlrechts. Aus Politik- und Zeitgeschichte, 52(22 
December), 3–11.
Kahl, W. (2009). Staatsziel Nachhaltigkeit und 
Generationengerechtigkeit. Die Öffentliche Ver-
waltung, 62 (1), 2–13.
Kiesewetter, B. (2009). Dürfen wir Kindern das 
Wahlrecht vorenthalten? Archiv für Rechts- und 
Sozialphilosophie, 95(2), 252–273. https://doi.org/ 
10.25162/arsp-2009-0018
Knödler, Chr. (1996). Wahlrecht für Minderjährige – 
eine gute Wahl? Zeitschrift für Parlamentsfragen, 
27(4), 553–571.
Kurunczi, G. (2018). Az általános választójog elvé-
nek aktuális kihívásai [Current Challenges to the 
Principle of Universal Suffrage]. Államtudományi 
Műhelytanulmányok, 2018/3. http://hdl.handle.
net/20.500.12944/7296
Kymlicka, W., & Donaldson, S. (2017). Inclusive 
Citizenship Beyond the Capacity Contract. In A. 
Schachar, R. Bauböck, I. Bloemraad, & M. Vink 
(Eds.), The Oxford Handbook of Citizenship (pp. 
838–859). Oxford University Press.
Lőrincz, V. (2018). Fenntarthatóság, életkor és 
cselekvőképesség – néhány döntéselméleti és 
pszichológiai szempont [Sustainability, Age 
and Capacity – Some Viewpoints of Decision 
Theory and Psychology]. Magyar Tudomány, 
179(3), 383–393. https://doi.org/10.1556/2065.1 
79.2018.3.8
Löw, K. (1974). Das Selbstverständnis des Grund-
gesetzes und wirklich allgemeine Wahlen. Poli-
tische Studien, 25(213), 19–29.
Löw, K. (2002a). Kinder und Wahlrecht. Zeitschrift 
für Rechtspolitik, 35(10), 448–450.
Löw, K. (2002b). Ist das Wahlrecht übertragbar? 
Beispiele europäischer Nachbarstaaten. Das Par-
liament, 37–38(17–23 September).
Löw, K. (2005). Kann ein allgemeines Wahlrecht 
die demographische Katastrophe abbremsen? 
Medizin und Ideologie, 27(4), 34–41.
M. Balázs, Á. (2017). Családi választójog és 
fenntarthatóság [Family Suffrage and Sustain-
ability], Parlamenti Szemle, 2(1), 47–65.
M. Balázs, Á. (2018). A plurális választójog fogalma 
és megengedhetősége napjainkban [The Concept 
and Permissibility of Plural Vote Today]. Állam-
tudományi Műhelytanulmányok, 2018/6. http://
hdl.handle.net/20.500.12944/7299
Marschall, St. (2000). Wer vertritt wen? Volks-
entscheide und die Funktionslogik parlamenta-
rischer Repräsentation. Kritische Anmerkungen 
zu einem Beitrag von Winfried Steffani in Heft 
3/99 der ZParl. Zeitschrift für Parlamentsfragen, 
31(1), 182–187.
Meixner, G. (2013). Plädoyer für ein „höchstpersön-
liches Elternwahlrecht zugunsten der Kinder“. 
Zeitschrift für Parlamentsfragen, 44(2), 419–426.
Merk, P. (1997). Wahlrecht ohne Altersgrenze? In 
Chr. Palentien & K. Hurrellmann (Eds.), Jugend 
und Politik: Ein Handbuch für Forschung, Lehre 
und Praxis (pp. 260–279). Luchterhand.
Merk, K.-P. (2009). Das Wahlrecht von Geburt an 
und seine politische Bedeutung. Diskurs Kind-
heits- und Jugendforschung, 4(4), 525–538.
Müller-Franken, S. (2013). Familienwahlrecht und 
Verfassung: Veränderungen des Wahlrechts zu-
gunsten von Familien als Reaktion auf den de-
mographischen Wandel auf dem Prüfstand des 
Verfassungsrechts. Mohr Siebeck.
Niebel, D., & Zypries, B. (2008). Wahlrecht für 
Kinder? Zeitschrift für Rechtspolitik, 41(8), 271.
Nopper, K. M. A. (1999). Minderjährigenwahlrecht: 
Hirngespinst oder verfassungsrechtliches Gebot in 
einer grundlegend gewandelten Gesellschaft? Köh-
ler.
Offe, C. (1994), Zusatzstimmen für Eltern – ein 
Beitrag zur wünschenswerten Reform von De-
mokratie und Wahlrecht? In G. Grözinger & H. 
Geiger (Eds.), Zukunft wählen: Zusatzstimmen 
für Eltern? (pp. 1–21). Evangelische Akademie 
Bad Boll.
Palentien, Chr., & Hurrellmann, K. (Eds.) (1997). 
Jugend und Politik: Ein Handbuch für Forschung, 
Lehre und Praxis. Luchterhand.
Palentien, Chr. (1997). Pro- und Contra-Diskussion 
zu einer Veränderung des Wahlrechts. In Chr. 
Palentien & K. Hurrellmann (Eds.), Jugend und 
88
(2023) 51
journal for constitutional theory and philosophy of law
Miklós Könczöl
Politik: Ein Handbuch für Forschung, Lehre und 
Praxis (pp. 290–299). Luchterhand.
Palmore, E. B. (1999). Ageism: Negative and Positive 
(2nd ed.). Springer.
Pechstein, M. (1996). Diskussionsbeitrag. In K. H. 
Fell & B. Jans (Eds.), Familienwahrecht – pro und 
contra (p. 84). Vektor.
Peschel-Gutzeit, L. M. (1999). Das Wahlrecht von 
Geburt an: Ein Plädoyer für den Erhalt unserer 
Demokratie. Zeitschrift für Parlamentsfragen, 
30(2), 556–563.
Peterson, P. E. (1992). An Immodest Proposal. Dae-
dalus, 121(4), 151–174.
Rehfeld, A. (2011). The Child as Democratic Citizen. 
Annals of the American Academy of Political 
and Social Science, 633, 141–166. https://doi.
org/10.1177/0002716210383656
Reimer, F. (2004). Nachhaltigkeit durch Wahlrecht? 
Verfassungsrechtliche Möglichkeiten und 
Grenzen eines „Wahlrechts von Geburt an“. 
Zeitschrift für Parlamentsfragen, 35(2), 322–339.
Reimer, F., & Schanda, B. (2016). Improving 
Representation by Parental Proxy Voting? In 
L. Komáromi & Z. T. Pállinger (Eds.), Good 
Governance – Enhancing Representation (pp. 63–
74). Pázmány Press.
Rosenberg, J. H. (2016). The All-Affected Principle 
and its Critics: A Study on Democratic Inclusion. 
Acta Universitatis Upsaliensis.
Sanderson, W. C., & Scherbov, S. (2007). A Near 
Electoral Majority of Pensioners: Prospects and 
Policies. Population and Development Review, 
33(3), 543–554. https://doi.org/10.1111/j.1728-
4457.2007.00184.x
Schanda, B. (2012). A jog lehetőségei a család 
védelmére [The Means of Law for Protecting 
Families]. Iustum Aequum Salutare, 8(2), 77–88.
Schickhardt, C. (2015). Kinder im Wahlrecht 
und in Demokratien. Für eine elterliche 
Stellvertreterwahlpflicht. Zeitschrift für Prak tische 
Philosophie, 2(1), 191–248. https://doi.org/10.226 
13/zfpp/2.1.7
Schreiber, W. (2004). Wahlrecht von Geburt an – Ende 
der Diskussion? Deutsches Verwaltungsblatt, 
119(21), 1341–1348.
Schroeder, W. (2003). Familienwahlrecht und 
Grundgesetz. JuristenZeitung, 58(19), 917–922.
Shields, L. (2019). Parental rights and the importance 
of being parents. Critical Review of International 
Social and Political Philosophy, 22(2), 119–133. 
https://doi.org/10.1080/13698230.2016.1262316
Steffani, W. (1999). Das magische Dreieck demok-
ratischer Repräsentation: Volk, Wähler und 
Abgeordnete. Zeitschrift für Parlamentsfragen, 
30(3), 772–793.
Toulemon, A. (1948). Influence du vieillissement de 
la population sur la composition du corps elec-
toral. In Actes des journées pour l’étude scienti-
fique du vieillisement da la population (pp. 107–
115). Alliance nationale contre la dépopulation.
Vanhuysse, P. (2013). Measuring Intergenerational 
Justice – Toward a Synthetic Index for OECD 
Countries. In D. Schraad-Tischler & N. Azahaf 
(Eds.), Intergenerational Justice in Aging Societies: 
A Cross-national Comparison of 29 OECD Coun-
tries (pp. 10–61). Bertelsmann Stiftung.
Van Parijs, Ph. (1998). The Disfranchisement of 
the Elderly, and Other Attempts to Secure 
Intergenerational Justice. Philosophy and Public 
Affairs, 27(4), 292–333. https://doi.org/10.1111/ 
j.1088-4963.1998.tb00072.x
Weber, M. (1958). Wahlrecht und Demokratie in 
Deutschland. In Gesammelte politische Schriften 
(pp. 233–279). Winckelmann.
Wernsmann, R. (2005). Das demokratische Prinzip 
und der demographische Wandel: Brauchen wir 
ein Familienwahlrecht? Der Staat, 44(1), 43–66.
Westle, B. (2006). „Wahlrecht von Geburt an“ – 
Rettung der Demokratie oder Irrweg? Zeitschrift 
für Parlamentsfragen, 37(1), 96–114.
89
journal for constitutional theory and philosophy of law
(2023) 51
Bartosz Brożek*
Ciencia cognitiva y naturaleza del derecho
El objetivo del trabajo es considerar el posible impacto en la ontología jurídica de los 
descubrimientos en la ciencia cognitiva. Empiezo refutando el argumento de que nues-
tro esquema conceptual —y por tanto nuestra ontología básica— es a priori en relación 
con cualquier teoría científica. Luego esbozo una imagen del surgimiento de la cultura, 
tal como se encuentra en los escenarios evolutivos recientes y las teorías neurocientíficas. 
En este contexto, sostengo que no hay, y no puede haber, una comprensión correcta de 
lo que es el derecho, lo que explica por qué es posible desarrollar distintas ontologías 
jurídicas que sean igualmente aceptables.
Parablas claves: derecho, ciencia cognitiva, ontología del derecho, imitación, naturaleza 
del derecho, actuar según las reglas
1 INTRODUCCIÓN
Los recientes avances en la ciencia cognitiva han reformado en gran medi-
da nuestra comprensión del comportamiento humano.1 Por lo tanto, no es de 
sorprenderse que los abogados hayan estado estudiando los hallazgos de la neu-
rociencia2 y las disciplinas relacionadas, para determinar si la nueva ciencia de 
la mente3 puede contribuir al funcionamiento del derecho.4 Este proceso culmi-
nó en el establecimiento de una nueva disciplina jurídica, usualmente referida 
como ‘neuroderecho’ o ‘derecho y neurociencia’.5 Hasta ahora, los intereses de 
los ‘neuroabogados’ han sido más prácticos que teóricos o filosóficos.6 Según 
un artículo basado en un reciente sondeo, las principales áreas del ‘derecho y 
* bbroze@yahoo.com | Professor of Law and Philosophy, Jagiellonian University, Krakow (Po-
land). Translated by Francisco J. Campos Zamora and Alejandro Guevara Arroyo.
1 Este artículo es una traducción de un original inglés, escrito y publicado hace varios años. 
Como no he alterado considerablemente mis puntos de vista, solo he introducido pequeños 
cambios para que el argumento sea más fácil de seguir. Por ello, no he actualizado (con una 
excepción) la bibliografía, que ha crecido en los últimos años. Sin embargo, la investigación 
más reciente no influye en el eje central del argumento que aquí ofrezco.
2 El término ‘neurociencia’ se entiende aquí de manera muy amplia y se refiere a todo tipo de 
estudio interdisciplinario del sistema nervioso, abarcando varios aspectos del mismo, desde el 
molecular al conductual, e incluyendo, entre otros, neurociencia del comportamiento, neuro-
ciencia celular, neurociencia clínica, neurociencia cognitiva, neurociencia computacional, neu-
rociencia cultural, neurociencia del desarrollo, neurociencia molecular, neuroimagen, neuroin-
geniería, neuroinformática, neurolingüística, neurociencia social y neurociencia de sistemas.
3 Tomo prestada esta frase de Rowlands 2010.
4 Cf. Garland 2004. Para una descripción crítica véase Pardo & Patterson 2013.
5 Cf. Spranger 2012.
6 Algunas excepciones pueden encontrarse en Brożek, Hage & Vincent 2021.
revus (2023) 51, 89–106
90
(2023) 51
journal for constitutional theory and philosophy of law
Bartosz Brożek
neurociencia’ incluyen cuestiones jurídicamente relevantes como la muerte ce-
rebral, las lesiones cerebrales, el dolor y la angustia, la memoria, las emociones, 
la detección de mentiras, el juzgar, el desarrollo del cerebro y las adicciones.7
Estos problemas no son meramente interesantes desde el punto de vista ju-
rídico: cualquier enfoque razonable del derecho requiere tener una compren-
sión actualizada y científicamente sólida de los temas antes mencionados. Sin 
embargo, también hay estudios que sugieren que la neurociencia (y la ciencia 
cognitiva en general) puede contribuir al desarrollo de teorías doctrinales per-
tenecientes a la responsabilidad penal o al derecho contractual.8 Sin duda, tal 
enfoque es mucho más complejo y está conectado a numerosas trampas me-
todológicas. El derecho opera dentro de un cierto esquema conceptual que es 
bastante diferente del aparato conceptual de la ciencia cognitiva.9
La ontología jurídica se encuentra aún más arriba en la escala de la abs-
tracción—entendida aquí como la cuestión de la naturaleza del derecho y de la 
existencia de diversos fenómenos relacionados con el derecho— y de ahí surge 
la pregunta de si los descubrimientos de la ciencia cognitiva pueden de alguna 
manera concebible informar la investigación ontológica. Mi objetivo en este tra-
bajo es considerar este problema. Empiezo refutando un argumento de Maxwell 
Bennett y Peter Hacker en el sentido de que nuestro esquema conceptual —y 
por tanto nuestra ontología básica— es a priori en relación con cualquier teoría 
científica. Luego esbozo una imagen del surgimiento de la cultura, tal como 
se encuentra en los escenarios evolutivos recientes y las teorías neurocientífi-
cas. En este contexto, sostengo que no hay, y no puede haber, una comprensión 
correcta de lo que es el derecho, lo que explica por qué es posible desarrollar 
distintas ontologías jurídicas que sean igualmente aceptables.
2 LA ONTOLOGÍA JURÍDICA SE ENCUENTRA CON LA 
CIENCIA COGNITIVA
Cuando se considera la relación entre ciencia cognitiva y ontología, el pri-
mer problema a abordar es si los hallazgos de la ciencia cognitiva u otras dis-
ciplinas relacionadas tienen alguna relevancia para la investigación ontológica. 
Quizás la respuesta más famosa a esta pregunta la dieron Maxwell Bennett y 
Peter Hacker en su célebre libro Philosophical Foundation of Neuroscience.10 
Bennett y Hacker insisten, en primer lugar, en que uno debe distinguir clara-
mente entre dos tipos de preguntas, conceptual y empírica:
7 Cf. Jones, Schall & Shen 2014: 12-14.
8 Cf. Pardo & Patterson 2013.
9 Cf. Pardo & Patterson 2013.
10 Bennett & Hacker 2003. Esta sección se encuentra basada parcialmente en Brożek 2013a.
91
journal for constitutional theory and philosophy of law
(2023) 51
Ciencia cognitiva y naturaleza del derecho
El distinguir las cuestiones conceptuales de las empíricas es de suma importancia (...) 
Las cuestiones conceptuales anteceden a las cuestiones de verdad y falsedad. Son cues-
tiones relativas a nuestras formas de representación, no cuestiones relativas a la ver-
dad o falsedad de los enunciados empíricos. Estas formas se encuentran presupuestas 
por proposiciones científicas verdaderas (y falsas) y por teorías científicas correctas (e 
incorrectas). No determinan qué es empíricamente verdadero o falso, sino qué tiene 
y qué no tiene sentido. Por tanto, las cuestiones conceptuales no son susceptibles de 
investigación y experimentación científica ni de teorizaciones científicas. Porque los 
conceptos y las relaciones conceptuales en cuestión son presupuestos por tales investi-
gaciones y teorizaciones (Bennett, Dennett, Hacker & Searle 2007: 4).
La frase final del pasaje citado es de especial interés. Bennett y Hacker afir-
man que los conceptos son a priori a cualquier investigación científica. Creen, 
además, que el hecho de no darse cuenta de este hecho conduce a menudo a 
graves errores, y en particular a la llamada falacia mereológica, común —como 
subrayan— en la ciencia cognitiva contemporánea. Esta consiste en referirse al 
cerebro o sus partes mediante conceptos que son correctamente aplicables solo 
a una persona en su conjunto. Observan:
[hablando] de lo que el cerebro percibe, piensa, adivina o cree, o de que un hemisferio 
del cerebro sabe cosas que el otro hemisferio ignora, está muy extendido entre los 
neurocientíficos contemporáneos. Esto a veces se defiende como si no fuera más que 
una trivial façon de parler, pero es un grave error. Pues la forma característica de ex-
plicación en la neurociencia cognitiva contemporánea consiste en adscribir atributos 
psicológicos al cerebro y sus partes para explicar la posesión de atributos psicológicos 
y el ejercicio (y deficiencias en el ejercicio) de los poderes cognitivos por parte de los 
seres humanos (Bennett, Dennett, Hacker & Searle 2007: 7).
Cabe preguntarse si el problema que identifican Bennett y Hacker es real. Se 
puede argumentar, por ejemplo, que afirmaciones como ‘el cerebro piensa’ o ‘el 
hemisferio derecho es responsable de la toma de decisiones’ no deben tomarse 
literalmente. Algunas intuiciones lingüísticas fundamentales y el conocimiento 
básico del lenguaje son suficientes para darse cuenta de que tal utilización de 
las palabras ‘pensar’ o ‘decidir’ es metafórica o analógica. Bennett y Hacker son 
plenamente conscientes de esta estrategia para defender el lenguaje científico 
existente y aclaran que la evidencia de que los científicos cognitivos cometen 
la falacia mereológica no reside en el hecho de que en ocasiones utilizan tér-
minos psicológicos ‘inadecuados’ para describir el funcionamiento del cerebro, 
lo que fácilmente puede entenderse como aprovechar la analogía, la metáfora, 
el homónimo o el uso de un concepto con un significado derivado. La falacia 
mereológica se produce cuando los científicos cognitivos transfieren complejos 
enteros de conceptos del ‘discurso psicológico’ al ‘neurocientífico’ y, sobre la 
base de atribuciones tan inadecuadas, extraen conclusiones.
¿Es sostenible el argumento de Bennett y Hacker? Yo creo que no, y la razón 
es su visión fundacional del conocimiento. Hay dos interpretaciones del funda-
cionalismo de Bennett y Hacker. La interpretación más fuerte, que John Searle 
92
(2023) 51
journal for constitutional theory and philosophy of law
Bartosz Brożek
les atribuye, es que creen que el lenguaje natural determina la única ontología 
aceptable. Searle dice que cometen una falacia al confundir las reglas para usar 
las palabras con la ontología. Así como el antiguo conductismo confundía la 
evidencia de los estados mentales con la ontología de los estados mentales, este 
conductismo criterial wittgensteiniano construye los fundamentos para hacer 
la atribución con el hecho que se atribuye. Es una falacia decir que las condicio-
nes para el exitoso funcionamiento del juego del lenguaje son las condiciones 
para la existencia del fenómeno en cuestión.11
Esta lectura encuentra alguna evidencia textual. Curiosamente, al elaborar 
la doctrina de la falacia mereológica, Bennett y Hacker citan a Aristóteles como 
uno de los primeros en condenar este modo de pensar erróneo. Aristóteles ob-
servó que ‘señalar que el alma está enojada es como si se dijera que el alma teje 
o construye, porque seguramente es mejor no decir que el alma se compade-
ce, aprende o piensa, sino que un hombre hace esto con su alma’. Sin embar-
go, es necesario recordar que hay una cierta visión metafísica detrás de esta 
afirmación. La metafísica de Aristóteles es esencialista: cree que toda entidad 
pertenece a alguna categoría natural, determinada por la esencia (forma) de la 
entidad; además, cree que las esencias pueden ser captadas por las denomina-
das definiciones esenciales.12 Por lo tanto, el uso incorrecto o metafórico de las 
palabras no es un mero error, sino uno que puede arruinar nuestros intentos 
de construir los fundamentos del conocimiento, capturado por las definiciones 
esenciales. Esta doctrina está, por supuesto, lejos de la práctica científica actual. 
La historia de la ciencia muestra claramente que no deben asumirse tales fun-
damentos, ya que es muy probable que prohíban el progreso científico. Pero si 
es así, lo mismo vale para el punto de vista de Bennett y Hacker: si realmente 
creen que el esquema conceptual del lenguaje ordinario determina ‘la única’ 
ontología, su concepción es irremediablemente defectuosa.
También es posible leer a Bennett y Hacker de una manera más moderada; 
esta interpretación más débil consiste en solo subrayar que el esquema concep-
tual que constituye el marco del lenguaje ordinario no determina ninguna onto-
logía única, pero que, sin embargo, es independiente de cualquier práctica cien-
tífica, en el sentido de que para comunicar cualquier descubrimiento científico 
se necesita emplear conceptos de acuerdo con algunos criterios preexistentes. 
Si no lo hace, se corre el riesgo de seguir caminos equivocados y de pronunciar 
declaraciones sin sentido: el uso incorrecto del lenguaje puede llevarnos por 
mal camino. El esquema conceptual del lenguaje ordinario constituye, como 
mínimo, la base para comunicar las teorías científicas.
Esta opinión, incluso en la lectura moderada, es problemática. En primer 
lugar, y lo que es menos importante, Bennett y Hacker se equivocan cuando 
11 Bennett, Dennett, Hacker & Searle 2007: 105.
12 Véase Popper 1966.
93
journal for constitutional theory and philosophy of law
(2023) 51
Ciencia cognitiva y naturaleza del derecho
afirman que un uso excesivo de metáforas, y en particular grupos de metáfo-
ras, es destructivo para cualquier esfuerzo neurocientífico. Ciertamente, puede 
conducir a callejones sin salida, pero hay poco peligro de que las consecuen-
cias de tal forma de expresión sean perjudiciales. La razón es que la ciencia 
cognitiva, como cualquier otra ciencia, tiene algunos mecanismos correctivos 
incorporados que en última instancia nos ayudan a distinguir el progreso y las 
hipótesis fructíferas de los meros errores y conjeturas inútiles. La presencia de 
este mecanismo es evidente una vez que se consideran los recientes éxitos de 
la ciencia cognitiva, y de la neurociencia en particular. Una ciencia que hace 
un uso excesivo de las metáforas y no conduce a predicciones o explicaciones 
serias es simplemente una mala ciencia; el mero hecho de cometer u omitir la 
falacia mereológica no posee aquí ninguna importancia.
En segundo lugar, debemos considerar el panorama más amplio, que se re-
sume en la afirmación de Bennett y Hacker de que el esquema conceptual del 
lenguaje ordinario es a priori en relación con la práctica científica. Esta es parti-
cularmente problemática con respecto a la neurociencia. Debe tenerse en cuen-
ta que el idioma psicológico, característico del lenguaje ordinario, no solo está 
conformado por nuestra experiencia interior, sino también por las teorías desa-
rrolladas a lo largo de la historia que pretendían capturar conceptualmente los 
fenómenos mentales. El problema es que el esquema conceptual del lenguaje 
ordinario se caracteriza por cierta inercia: las concepciones científicas actuales 
necesitan mucho tiempo para ‘infiltrarse’ en nuestro esquema conceptual ordi-
nario. Por lo tanto, es seguro asumir que el lenguaje ordinario de hoy ‘abraza’ al-
gunas teorías psicológicas de ayer, o mejor aún: una mezcla de esas teorías e ideas 
de sentido común. Ahora bien, decir que los conceptos ordinarios son a priori 
relativos a la neurociencia equivale a decir que la psicología popular es a priori a 
las teorías neurocientíficas, lo cual es un sinsentido total: es uno de los principa-
les objetivos de la neurociencia contemporánea, uno que cumple con vigor y con 
mucho éxito, el revisar nuestras viejas nociones psicológicas de sentido común.
Esto muestra claramente que el fundacionalismo conceptual de Bennett y 
Hacker no es fiel ni a los mecanismos de la práctica científica ni a la forma 
en que evolucionan nuestros esquemas conceptuales: estos nunca son defini-
tivos ni independientes de las teorías que desarrollamos. Este punto es bastan-
te general y pertenece a cualquier proyecto filosófico fundacional: las fuentes 
de la reflexión filosófica se basan siempre, al menos parcialmente, en algunas 
concepciones científicas, aunque a menudo obsoletas. En el caso del tomismo, 
la visión aristotélica del mundo —o la ciencia aristotélica— constituye el fun-
damento del esquema conceptual. De manera similar, en el caso de aquellas 
filosofías que encuentran confirmaciones o refutaciones en el funcionamiento 
del lenguaje ordinario, es el conocimiento encapsulado allí (por ejemplo, una 
especie de psicología popular que mezcla las observaciones del sentido común 
94
(2023) 51
journal for constitutional theory and philosophy of law
Bartosz Brożek
y algunas teorías psicológicas antiguas) lo que determina en última instancia las 
doctrinas filosóficas de los seguidores de Austin y Strawson. En otros proyectos 
fundacionales, como la fenomenología, el conocimiento científico internaliza-
do por una persona determina de manera crucial sus experiencias y, por tanto, 
sus puntos de vista filosóficos. Con todo, no existe una fuente de conocimiento 
filosófico que sea independiente de algún tipo de ciencia, y el punto clave es que 
esta ‘ciencia oculta’ puede estar en desacuerdo con lo que la ciencia contem-
poránea tiene que decir. El espléndido aislamiento de los filósofos del lenguaje 
común es una ilusión: no hay escapatoria al enfrentamiento con los bárbaros 
del otro lado del Canal.
Por supuesto, esto no significa que las interacciones mutuas entre la ciencia 
cognitiva y la ontología sean sencillas y se encuentren libres de problemas. Es 
difícil imaginar que algún descubrimiento o teoría científica pueda tener una 
relación directa con las deliberaciones ontológicas. En este sentido, Hacker y 
Bennett plantean un punto válido: la filosofía y la ciencia cognitiva utilizan di-
ferentes métodos, diferentes esquemas conceptuales, además de abordar dife-
rentes tipos de problemas. Sin embargo, como intenté enfatizar, esos diferentes 
discursos no están aislados unos de otros; más bien, interactúan de muchas for-
mas. Es difícil mapear esas interacciones con precisión y aún más difícil des-
cubrir su estructura. Tienen lugar en varios niveles: conceptual (los conceptos 
migran de la filosofía a la ciencia y viceversa, a menudo alterando su significa-
do), presuposicional (las teorías y métodos científicos a menudo presuponen 
tesis estrictamente filosóficas), del problema (la filosofía y la ciencia a menudo 
abordan problemas similares) y funcionales (por ejemplo, las teorías científicas 
pueden en algunos contextos reemplazar las teorías filosóficas, mientras que las 
doctrinas filosóficas pueden desempeñar un papel heurístico en la ciencia).13 
El alcance del presente artículo es demasiado limitado para proporcionar una 
descripción más completa de estas interacciones.14 La moraleja es, sin embargo, 
que dada la existencia de tales interacciones, es posible relacionar ontologías 
con teorías desarrolladas en las ciencias cognitivas.
Las ontologías que se desarrollan pueden ser más o menos coherentes con 
determinadas concepciones de la ciencia cognitiva. Por ejemplo: si una teoría 
científica presupone alguna forma de reduccionismo fuerte (es decir, la tesis 
de que todos los fenómenos mentales pueden explicarse completamente por la 
actividad cerebral), entonces dicha teoría es más coherente con alguna versión 
del materialismo que con una ontología dualista, la que subraya la estricta se-
paración entre mente y cuerpo. La concepción aristotélica de la mente es más 
coherente con el paradigma de la mente encarnada que con la opinión de que la 
mente es una especie de máquina de Turing. Estos atrevidos ejemplos muestran 
13 Analizo con más detalle los cuatro niveles en Brożek 2011.
14 Para más detalles Brożek en manuscrito.
95
journal for constitutional theory and philosophy of law
(2023) 51
Ciencia cognitiva y naturaleza del derecho
que la ontología y la ciencia cognitiva no son completamente ajenas entre sí. Si 
uno adopta el estricto dualismo mente-cuerpo en su concepción ontológica, 
mientras cree en el reduccionismo estricto subyacente a sus consideraciones 
científicas, simplemente sería incoherente en sus creencias. Nuestro conoci-
miento no consiste en ‘islas’ aisladas: la de la ontología y la de la ciencia cogni-
tiva; más bien se parece a la red de creencias de Quine, donde las concepciones 
ontológicas y científicas están interconectadas de diversas formas.
Al investigar la relación entre la ciencia cognitiva y la ontología jurídica, uno 
tiene al menos dos estrategias a mano. La primera puede denominarse como 
análisis de presuposiciones. Consiste en descubrir las presuposiciones relevantes 
de una ontología dada y comprobar si son coherentes con las teorías de la cien-
cia cognitiva. Por ejemplo, la visión del derecho propuesta por Thomas Hobbes 
asume una visión peculiar de la naturaleza humana, según la cual los humanos 
actúan de una manera puramente egoísta. Esto es incompatible con los hallaz-
gos de la teoría evolutiva y la neurociencia contemporáneas. Al mismo tiempo, 
la visión de John Locke de la naturaleza humana, que sirve de base a su concep-
ción del derecho, es mucho más coherente con la imagen de los mecanismos 
motivacionales humanos que se encuentran en la literatura científica. Por tanto, 
se puede concluir que la teoría del derecho de Locke es más coherente con los 
puntos de vista adoptados por las ciencias cognitivas contemporáneas que la 
concepción de Hobbes.15
La segunda estrategia es genealógica. Aquí, uno no comienza con una onto-
logía, sino reconstruyendo la historia evolutiva del desarrollo de las habilida-
des relevantes de la especie humana. Consideremos el derecho o cualquier otra 
institución social. Una narrativa genealógica no conduce a una ontología del 
derecho en particular, sino que descubre los mecanismos que hicieron posible 
el surgimiento del derecho en primer lugar. Pero esta es una idea importante, 
una que excluye ciertas ontologías sociales (como incoherentes con el escenario 
evolutivo reconstruido), mientras que permite una serie de otras posturas onto-
lógicas (como coherentes con el escenario en mayor o menor grado).
En lo que sigue, adopto la segunda estrategia.
3 EL SURGIMIENTO DE LA CULTURA
Con miras a investigar los orígenes del derecho —o, para ponerlo en otros 
términos, para responder la pregunta ‘¿qué hace posible al derecho?’— uno debe 
tener en cuenta dos preceptos metodológicos. Primero, no tiene sentido limitar 
el análisis sólo a los descubrimientos de las ciencias cognitivas. Por un lado, la 
ciencia cognitiva es difícil de definir en forma tal que pueda ser distinguida níti-
15 Cf. Załuski 2009.
96
(2023) 51
journal for constitutional theory and philosophy of law
Bartosz Brożek
damente de ciencias emparentadas (v.g. teoría evolucionista, primatología, psi-
cología del desarrollo, etc.). Por otro lado, tal limitación haría mucho más débil 
nuestra búsqueda de una teoría de los orígenes del derecho. Entre más argumen-
tos desde distintas disciplinas puedan formularse en favor de un cierto escenario 
evolutivo, estará mejor justificado.16 El segundo precepto es el siguiente: no tiene 
sentido investigar per se en los orígenes del derecho, como si estuvieran com-
pletamente separados de otras formas de comportamiento cultural. No parece 
posible que el derecho evolucionara separadamente de la moralidad y otras ins-
tituciones sociales. La pregunta ‘¿cómo es posible el derecho?’ es sólo un aspecto 
pequeño de una pregunta más general: ‘¿cómo es posible la cultura?’
El surgimiento de la cultura es un enigma por una simple razón. Michael 
Tomasello señala:
Los 6 millones de años que separan al ser humano de los grandes simios es, evolutiva-
mente, un periodo muy corto de tiempo, con los humanos modernos y los chimpan-
cés compartiendo más o menos un 99 por ciento de su material genético – es el mismo 
grado de parentesco que otros géneros hermanos como leones y tigres, caballos y ce-
bras y ratas y ratones. Nuestro problema es, por ende, uno de tiempo. El hecho es que 
simplemente no ha habido suficiente tiempo para un proceso normal de evolución 
biológica que involucre variación genética y selección natural para crear, una por una, 
cada una de las habilidades cognitivas necesarias para que los humanos modernos 
inventen y mantenga complejas industrias y tecnologías que utilizan herramientas, 
complejas formas de comunicación y representación simbólica y complejas organiza-
ciones e instituciones sociales (Tomasello 2019: 2).
Y agrega:
Sólo hay una solución posible a este enigma. Esto es, hay únicamente un mecanismo 
biológico conocido que podría llevar a esta clase de cambios en el comportamiento y 
en la cognición en un tiempo tan corto (…). Este mecanismo biológico es la transmi-
sión social o cultural, que trabaja en escala de tiempo más rápidamente en órdenes de 
magnitud que aquellos de la evolución biológica (Tomasello 2019: 4).
Así, Tomasello afirma que la riqueza y complejidad de la cultura humana 
no puede ser explicada sólo por los mecanismos de la evolución biológica; en 
cambio, uno necesita postular que en nuestro pasado filogenético algunas adap-
taciones biológicas relativamente menores pavimentaron la vía para el surgi-
miento del mecanismo de evolución cultural. El aspecto crucial de este meca-
nismo es lo que Tomasello llama el efecto trinquete cultural [cultural ratchet]: la 
transmisión cultural es acumulativa, i.e. los patrones comportamentales descu-
biertos por una generación son transmitidos a las subsiguientes generaciones. 
Por esto, no hay necesidad de ‘reinventar la rueda’– nacemos en una sociedad 
que ya posee un arsenal sustancial de formas de conceptualizar la experiencia, 
de patrones de comportamiento y de herramientas.
16 Brożek 2015a.
97
journal for constitutional theory and philosophy of law
(2023) 51
Ciencia cognitiva y naturaleza del derecho
La pregunta esencial es: ¿cuáles son las habilidades condicionadas biológi-
camente de la especie humana que permiten la transmisión cultural? Una res-
puesta bien fundamentada apunta hacia el mecanismo de aprendizaje social de 
la imitación.
Robin Dunbar nota:
Los bebes humanos son máquinas imitativas que parece absorber cualquier cosa y 
todo lo que se encuentren que involucre el imitar el comportamiento de otro indi-
viduo. La enseñanza ayuda a guiar esta absorción, pero sin la aparentemente infini-
ta capacidad de imitar de los niños humanos, es dudoso que cualquier cantidad de 
enseñanza de parte del progenitor ayudaría en la absorción de tanto comportamiento 
en un periodo tan corto de tiempo. En contraste, jóvenes chimpancés parecen más 
proactivos y avispados encontrando cosas por sí mismos (Dunbar 2004: 159).
Es común el diferenciar diversos mecanismos de aprendizaje social. Dado el 
objetivo de este artículo, es importante distinguir entre imitación y emulación. 
La imitación consiste en copiar tanto el objetivo (el cambio en el ambiente) del 
comportamiento de alguien, así como la forma de actuar, mientras que la emu-
lación se circunscribe a intentar alcanzar el mismo objetivo. Los primatólogos 
están de acuerdo en que los primates no humanos emulan más que imitan. Por 
ejemplo, en un experimento conducido por Tomasello, Nagel y Olguin, a dos 
grupos de chimpancés y dos grupos de infantes de dos años, se les mostró dos 
formas de utilizar una herramienta semejante a un rastrillo, para alcanzar comi-
da, de las cuales una era más eficiente que la otra. Cada grupo vio sólo uno de 
los métodos. Resulta que los chimpancés de ambos grupos no replicaron exacta-
mente el comportamiento de los instructores, sino que utilizaron el rastrillo de 
diversas formas. Según Tomasello, esto puede interpretarse como una muestra 
de que los chimpancés aprenden por emulación: la forma de utilizar la herra-
mienta no es importante, mientras que sí lo es el cambio en el ambiente —al-
canzar la comida—. Mientras tanto, los infantes humanos que participaron en el 
experimento, repitieron las acciones de la persona instructora, incluso si seguían 
el método menos eficiente de utilizar el rastrillo. Tomasello cree que esto es evi-
dencia de que los humanos –en contraste con lo grandes simios- no aprenden 
por medio de la emulación, sino a través de la imitación. Agrega, que el apren-
dizaje por medio de la emulación puede ser en algunas circunstancias ser más 
eficiente que la imitación; no obstante, esta última tiene una importancia social 
potencial, pues la imitación requiere prestar atención no sólo a los cambios en el 
ambiente exterior, sino también en el comportamiento de los otros.17
Las ventajas evolutivas de la imitación sobre la emulación y otras estrategias 
de aprendizaje social se vuelven claras si se consideran los siguientes hechos. 
Primero, la imitación es un mecanismo que lleva a la acumulación de conoci-
miento sobre patrones de comportamiento. La emulación o el aprendizaje in-
17 Cf. Tomasello 1999.
98
(2023) 51
journal for constitutional theory and philosophy of law
Bartosz Brożek
dividual por prueba y error no garantizan la transmisión intergeneracional de 
las formas de actuar compartidas comunitariamente. Segundo, la imitación es 
menos costosa comparada con el aprendizaje individual. Tercero, es el único 
mecanismo de aprendizaje social que admite la recombinación, i.e. utilizar los 
mismos medios para alcanzar objetivos distintos o lograr el mismo fin con me-
dios distintos. En otras palabras, la imitación lleva a una especie de ‘explosión 
combinatoria’, aumentando considerablemente el conjunto de las herramien-
tas comportamentales que un individuo posee. Cuarto, la imitación admite 
modificaciones sencillas: debido al hecho de que requiere que aprendamos el 
modo de actuar, abre el camino para introducir modificaciones en los patrones 
de comportamiento existentes. Quinto, la imitación es finamente detallada [fi-
ne-grained], i.e. es el único mecanismo que habilita la existencia de patrones de 
conducta muy similares, aunque distintos. Consideremos el lenguaje: con fre-
cuencia utilizamos expresiones muy similares, aunque para alcanzar objetivos 
completamente distintos. Si la emulación constituyera la estrategia de aprendi-
zaje responsable por la comunicación, el lenguaje sería imposible.18
¿Cuáles son las adaptaciones biológicas que permiten la imitación? Es ha-
bitual el hablar sobre la habilidad humana para imitar y su tendencia a imitar: 
aquella es cognitiva y esta un mecanismo motivacional. El aspecto cognitivo de 
la imitación está fundamentalmente conectado con la habilidad de mindreadin-
g,19 i.e. adscribir creencias e intenciones a otros (este mecanismo es frecuente-
mente llamado ‘teoría de la mente’). Para distinguir entre la forma del acto y 
el objetivo de la acción, uno necesita entender lo que el otro individuo estaba 
tratado de hacer. Hay dos abordajes principales sobre cómo trabaja el mindrea-
ding: la teoría de la teoría y la teoría de simulación. De acuerdo con la primera, 
la adscripción de estados mentales o intenciones a otros procede por medio de 
una especie de razonamiento, desde relaciones en primera persona a relacio-
nes en tercera persona está basada en el esquema: ‘Me comporto de la forma 
x si siento dolor; por ende, si otro se comporta de la forma x, quiere decir que 
siente dolor’.20 La teoría de la simulación, por otro lado, sugiere el funciona-
miento de un mecanismo distinto. Por ejemplo, Alvin Goldman afirma que la 
adscripción de estados mentales a otros procede de la siguiente forma. En la 
primera etapa, a través de ‘mindreading’ sobre el estado mental de otra persona, 
el cerebro genera estados similares al estado de la mente de la otra persona. La 
segunda etapa consiste en procesar los datos obtenidos, con el uso del sistema 
neuronal propio, pero trabajando ‘fuera de línea’ [off-line]. Goldman cree que 
18 Brożek 2015b.
19 Nota de los traductores: “Mindreading” es un concepto complejo que implica la capacidad 
de atribuir estados mentales, predecir y explicar el comportamiento y pensamiento propio y 
ajeno a partir de tales atribuciones. Es por ello que, según los especialistas, en textos en espa-
ñol se utiliza en inglés por lo difícil de su traducción.
20 Cf. Meltzof 2005: 55-77.
99
journal for constitutional theory and philosophy of law
(2023) 51
Ciencia cognitiva y naturaleza del derecho
el cerebro utiliza los mismos circuitos cuando uno experimenta algo y cuando 
uno lee algo similar en las experiencias de otros. El resultado de la simulación 
fuera de línea es adscrito, en la última etapa, a la persona observada. Por ende, 
de acuerdo con la teoría de la simulación, el mindreading no requiere la reali-
zación de razonamiento alguno.21 Curiosamente, en la primera etapa del mo-
delo de Goldman, ocurre alguna forma muy básica de imitación (generación 
de un estado mental similar al estado mental de la otra persona). Esto lleva a la 
pregunta, qué es filogenética y ontogenéticamente anterior –mindreading o la 
imitación. Susan Gurley y Nick Chater sugieren que puede que la cuestión esté 
mal planteada: mindreading y la imitación son dos habilidades que se desarro-
llan simultáneamente– la imitación requiere capacidades plenamente desarro-
lladas de mindreading, y viceversa. Pero ambas capacidades están construidas 
sobre algunas habilidades rudimentarias: la muy temprana imitación puede que 
muestre una semejanza fundamental: yo-otro, mientras que la capacidad distin-
tivamente humana de aprendizaje mediante interpretación, con su estructura 
flexible: medio-fines, a su vez contribuye al desarrollo de la distinción yo-otro y 
a capacidades más avanzadas de mindreading. 22
El descubrimiento más importante que nos acerca a entender cómo la imita-
ción y el mindreading son posibles es posiblemente el hallazgo de las neuronas 
espejo. por parte de neurocientíficos de Parma a principios de la década de los 
noventa.23 Las neuronas espejo se activan tanto cuando una acción es ejecutada 
como cuando es observada.24 Desde entonces, han provisto algún apoyo a la 
teoría de la simulación: hay mecanismos neuronales que involucran los mismos 
circuitos cerebrales cuando una acción es realizada y cuando es sólo observada. 
De esta forma, no es necesario ningún razonamiento de primera persona a ter-
cera persona para comprender la acción de alguien más, dado que su acción es 
automáticamente simulada por el cerebro del observador. De forma similar, la 
existencia del sistema espejo podría contribuir a la explicación del mecanismo 
de la imitación. Por ejemplo, Giacomo Rizolatti afirma que hay dos tipos de 
resonancia de las neuronas espejo y, como resultado, dos clases de imitación. El 
nivel alto de resonancia es utilizado por reflejar el objetivo de una acción, mien-
tras que el nivel bajo de resonancia copia la forma de actuar. Según Rizolatti, 
únicamente el cerebro humano saca provecho de ambos mecanismos, lo que 
habilita la imitación. En simios y (posiblemente) otros animales sólo la reso-
nancia de nivel alto es utilizada y esto explica la habilidad de los simios para 
aprender sólo mediante emulación.25
21 Cf. Meltzof 2005: 55-77.
22 Hurley & Chater 2005: 33.
23 Cf. Di Pellegrino, Fadiga, Fogassi, Gallese & Rizolatti 1991: 176–180.
24 Hay, sin embargo, algunas reservas sobre el significado real del sistema espejo véase Hickok 
2014.
25 Cf. Rizolatti 2005: 55-76.
100
(2023) 51
journal for constitutional theory and philosophy of law
Bartosz Brożek
Prestemos ahora atención a la tendencia humana a imitar. Se insiste frecuen-
temente que mientras los primates no-humanos muestran alguna capacidad 
para la imitación, muy raramente imitan debido a la ausencia de mecanismos 
emocionales y motivacionales relevantes. Michael Tomasello nota:
Hay algún paso inicial en la evolución humana separándose de los grandes simios, 
que involucra el lado emocional y motivacional de la experiencia, que propulsó a los 
humanos hacia un nuevo espacio adaptativo en el cual pudieron ser seleccionadas 
complejas capacidades y motivaciones para la actividad colaborativa y la intencionali-
dad compartida (Tomasello 2009: 85).
Tomasello, incluso, afirma que la disposición humana para cooperar e imitar 
a otros es solo una manifestación de una adaptación biológica más básica: el 
mutualismo.26 Él observa que la cooperación entre simios está usualmente ba-
sada en parentesco y reciprocidad. El pasaje evolutivo desde estas dos formas de 
comportamiento a las capacidades cooperativas humanas fue condicionado por 
tres procesos: el desarrollo de la cognición social y los mecanismos motivacio-
nales, habilitando la coordinación de actividades cooperativas y comunicación 
compleja; el incremento de la tolerancia y la confianza en relación con otros, 
primordialmente en el contexto de obtención de alimento; y el desarrollo de un 
grupo de prácticas institucionales basadas en normas sociales.27
En el final neurocientífico de la historia, se especula que los mecanismos 
motivacionales y emocionales requeridos para generar comportamiento huma-
no colaborativo son también condicionados por la existencia del sistema espe-
jo. Por ejemplo, Marco Iacoboni observa —basándose en hechos anatómicos y 
datos obtenidos mediante neuroimagen— que una parte de la ínsula, el cam-
po disgranural [dysgranural field], está conectado tanto con el sistema límbico, 
como con el parietal posterior, inferior y frontal y con el cortex superior tem-
poral. Esto lleva a la hipótesis que es la ínsula la que provee a las áreas límbicas, 
responsables de la respuesta emocional, con la información sobre la acción ob-
servada y ejecutada. La hipótesis recibe más apoyo por descubrimientos expe-
rimentales que indican que el proceso de imitar una expresión facial emocional 
aumenta la actividad de la amígdala, i.e. la parte del sistema límbico que juega 
un rol crucial en los procesos emocionales.28
Estas y otras observaciones llevaban a Iacoboni a la conclusión de que el 
sistema neuronal espejo —y la imitación— son necesarias para experimentar 
empatía:
Entendemos los sentimientos de otros por medio de un mecanismo de acción-repre-
sentación que da forma al contenido emocional, de tal manera que nuestra resonancia 
26 Tomasello 2009: 85.
27 Silk afirma que el altruismo es anterior al mutualismo; véanse sus comentarios en Tomasello 
2009: 111–124.
28 Cf. Iacoboni 2005: 95.
101
journal for constitutional theory and philosophy of law
(2023) 51
Ciencia cognitiva y naturaleza del derecho
empática se basa en la experiencia de nuestro cuerpo en acción y en las emociones 
asociadas a movimientos específicos. (…) Para empatizar, dependemos de la medi-
ación de la representación de las acciones asociadas con las emociones que estamos 
presenciando y en la red cerebral que incluye estructuras que apoyan la comunicación 
entre los circuitos de acción representación y los circuitos dedicados al procesamiento 
emocional. (Iaboconi 2005: 98).
Una posición semejante adopta Vittorio Gallese, quien remarca:
El descubrimiento de neuronas espejo en los individuos adultos muestra que el mi-
smo sustrato neuronal se activa cuando algunos de estos actos expresivos son tanto 
ejecutados como percibidos. Así, tenemos un espacio común instanciado subperso-
nalmente. Depende de los circuitos neuronales involucrados en el control de acciones. 
La hipótesis que presento aquí es que un mecanismo similar podría apuntalar nuestra 
capacidad de compartir sentimientos y emociones con otros. Mi propuesta es que las 
sensaciones y emociones exhibidas por otros pueden también ser empatizadas con, y 
por ende implícitamente entendidas, un mecanismo espejo de coincidencia (Gallese 
2005: 133).
La habilidad humana de imitar, en conjunto con nuestra tendencia de ser 
como los otros e imitarlos, condicionan el surgimiento de patrones comporta-
mentales compartidos dentro de las comunidades humanas y transmitidos de 
generación en generación. Ya en esta etapa de la filogénesis humana, uno puede 
hablar del desarrollo de alguna forma rudimentaria de práctica de seguimiento 
de normas.29 Los patrones de comportamiento que emergen espontáneamen-
te, propagados a través de los mecanismos de imitación y compartidos por la 
comunidad, tienen cierto nivel de objetividad. Es más, fueron aplicados colec-
tivamente: los trasgresores fueron castigados o corregidos por otros miembros 
de la comunidad. Notablemente, este marco cognitivo y motivacional también 
constituye una plataforma para el surgimiento del lenguaje. Tomasello apunta:
La comunicación cooperativa humana (…) evolucionó primero porque estas activida-
des proveen de la base necesaria para establecer temas conjuntos, y porque generaron 
los motivos cooperativos.30
Hay un número de escenarios evolutivos que subrayan que las fuentes de 
las habilidades lingüísticas yacen en la tendencia humana a cooperar, y están 
basadas en último término en la estructura neuronal provista por las neuronas 
espejo. Por ejemplo, Merlin Donald identifica cuatro tipos principales de repre-
sentación mimética, que son la clave para la transmisión y propagación de la 
cultura: (1) mímica reenactiva [reenactive mime], característica del juego de ro-
les; (2) la imitación precisa medios-fines (como en el aprendizaje de cómo freír 
un huevo); (3) la práctica y refinamiento sistemático de una habilidad (como al 
aprender cómo manejar un auto); y (4) gestos no-lingüíticos (como al aprender 
29 Brożek 2013b.
30 Tomasello 2009: 73.
102
(2023) 51
journal for constitutional theory and philosophy of law
Bartosz Brożek
cómo bailar).31 Además, afirma que estas habilidades miméticas fueron el fun-
damento para el surgimiento del lenguaje y todas las otras formas de cultura. 
Enfatiza que su propuesta difiere de los escenarios tradicionales que condicio-
nan el surgimiento de la cultura en el surgimiento previo del lenguaje (la teoría 
del lenguaje primero). De acuerdo con Donald, algunas formas de culturas, ba-
sadas en habilidades miméticas, deben haber precedido al lenguaje y habilitado 
su evolución (la teoría de la cultura primero).32
La teoría de Donald conlleva consecuencias profundas. Primero, afirma 
que la mente humana está íntimamente conectada con la sociedad en la cual 
florece. Uno puede decir incluso que es cocreada por la comunidad. Las prác-
ticas comunales son constitutivas de la mente humana, tanto en su dimensión 
filogenética como en la ontogenética. Segundo, el lenguaje no es un individuo, 
sino un fenómeno de nivel de red: su evolución se asemeja a la evolución de 
un ecosistema en vez de un único organismo. Tercero, se sigue que “es poco 
probable que los neurocientíficos cognitivos encuentren un dispositivo inna-
to de adquisición del lenguaje, y deben redirigir sus investigaciones hacia los 
potentes sistemas de procesamiento analógico de los cuales el lenguaje puede 
emerger en la interacción grupal”.33
El surgimiento del lenguaje consolida aún más las prácticas humanas de se-
guimiento de reglas. Con el tiempo, también habilita la formulación lingüística 
de reglas de conducta y abre el camino para discutirlos críticamente. Por esta 
vía, la humanidad alcanzó la etapa en la cual hizo posible el desarrollo de siste-
mas morales y jurídicos, propiamente dichos.
4 LA ILUSORIA NATURALEZA DEL DERECHO
Para relacionar la imagen esbozada arriba sobre el surgimiento de la cultura 
con los problemas de la ontología jurídica, permítasenos introducir una dis-
tinción heurística entre reglas rudimentarias y reglas abstractas. Al calificar la 
distinción como heurística deseo enfatizar que esta puede representar solo una 
visión simplificada de las prácticas del seguimiento de reglas Al mismo tiempo, 
creo que es compatible con el escenario evolutivo delineado arriba, y también es 
altamente instructiva: incluso si una ontología compleja de las reglas requiriera 
distinciones conceptuales mucho más complejas, las categorías de las reglas ru-
dimentarias y abstractas capturan algunos aspectos cruciales de cualquier onto-
logía bien desarrollada.34
31 Cf. Donald 2005: 283-300.
32 Cf. Donald 2005: 283-300
33 Donald 2005: 294.
34 Esta distinción es desarrollada y defendida en Brożek 2013b: capítulo 2.
103
journal for constitutional theory and philosophy of law
(2023) 51
Ciencia cognitiva y naturaleza del derecho
Las reglas rudimentarias pueden ser caracterizadas como: (a) independien-
tes del lenguaje (un lenguaje completamente desarrollado, i.e. un sistema con-
sistente en un vocabulario bien definido y en reglas gramaticales es dependiente 
tanto evolucionaria como lógicamente, de la existencia de reglas rudimenta-
rias); (b) son simples y concretas (las reglas rudimentarias pertenecen a formas 
relativamente simples y concretas de comportamiento); (c) son normativamen-
te unificadas (las reglas rudimentarias no pueden dividirse en clases – en el 
nivel rudimentario no hay reglas matemáticas, lingüísticas, morales o jurídi-
cas); (d) son multiaspecto y no modales (las reglas rudimentarias pertenecen a 
alguna forma de comportamiento como un todo, ellas dicen que debe hacerse 
dadas ciertas circunstancias y no involucran operadores deónticos tales como 
obligatorio, prohibido o permitido).
Las reglas abstractas, por otro lado, dependen de la existencia de las reglas 
rudimentarias. Sin la forma rudimentaria de seguimiento de reglas sería difícil 
imaginar cómo emergieron las reglas abstractas: hubiera sido una especie de 
milagro. Si no es por las reglas rudimentarias, nuestros sistemas normativos 
tales como el lenguaje, la moralidad o el derecho se convertirían en un enigma 
evolutivo, y un logro tan único y cualitativamente diferente de la cultura de 
otros primates que cualquier intento de explicarlos estaría destinado al fracaso. 
Las reglas abstractas fueron desarrolladas a través de formulación lingüística 
y reflexión sobre las formas rudimentarias de seguimiento de reglas. En con-
traste con las reglas rudimentarias, ellas: (a) dependen del lenguaje (deben ser 
formuladas en lenguaje); (b) puede ser complejas (i.e. puede referir a complejos 
patrones de comportamiento) y generales (i.e. pueden referir a acciones defini-
das con generalidad, no concretamente); (c) son normativamente diferenciadas 
(i.e. ellas puede dividirse en clases: lingüísticas, morales, jurídicas); y (d) son 
aspectualizadas y modalizadas (i.e. pueden pertenecer sólo a ciertos aspectos de 
acciones, y ser expresadas con el uso de operadores deónticos).
Las reglas abstractas son por ende el resultado de teorizar algunos aspectos 
de las reglas rudimentarias. Para ilustrar este punto, consideremos el lenguaje. 
Lo que uno encuentra en diccionarios y libros de gramática es una imagen del 
lenguaje como un sistema aislado de reglas sintácticas y semánticas; pero esto 
es ya una consecuencia de desarrollar un abordaje teorético sobre el lenguaje. 
Pidiendo prestada una frase de los realistas jurídicos estadounidenses: el len-
guaje en los libros es un sistema de reglas que se comporta bien, mientras que el 
lenguaje en acción está constituido por una multiplicidad de patrones de com-
portamiento en los que aspectos lingüísticos, morales, matemáticos y de otras 
clases están íntimamente conectados entre sí. Cuando uno describe a alguien 
como culpable, uno puede decir que sigue una regla rudimentaria (uno hace lo 
que hace en circunstancias similares de su comunidad), pero desde una pers-
pectiva más teórica puede concebirse como observando simultáneamente una 
104
(2023) 51
journal for constitutional theory and philosophy of law
Bartosz Brożek
regla lingüística, moral, prudencial o jurídica. Crucialmente, la misma amalga-
ma de las reglas rudimentarias puede dar lugar a reconstrucciones teóricas dife-
rentes en el nivel de las reglas abstractas, y por ende uno puede tener abordajes 
divergentes del lenguaje, la prudencia o la moralidad.
Aun así, las reglas abstractas —i.e. los sistemas normativos que formulamos 
en lenguaje, discutimos, criticamos y cambiamos— en efecto, influencian nues-
tro comportamiento. La influencia es directa, cuando uno conscientemente 
aplica una regla abstracta, o (más frecuentemente) indirecta, cuando los pa-
trones de comportamiento prescritos por algunas reglas abstractas se vuelven 
fibras en la amalgama de las reglas rudimentarias. En otras palabras, las reglas 
abstractas pueden informar las prácticas rudimentarias de seguimiento de re-
glas. Este bucle de retroalimentación –la dependencia de las reglas abstractas en 
la existencia de reglas rudimentarias y la influencia que aquellas tienen sobre las 
prácticas rudimentarias de seguimiento de reglas– destaca el carácter heurísti-
co de la distinción entre reglas rudimentarias y abstractas: es una herramienta 
analítica útil para capturar algunos aspectos importantes del seguimiento de 
reglas, pero no nos ofrece una imagen completamente adecuada de la compleji-
dad del fenómeno en cuestión.
Estas observaciones tienen una relevancia directa con el problema de la na-
turaleza del derecho, o al menos eso sostengo. La naturaleza del derecho —o la 
correcta definición del derecho— ha sido desde hace mucho el santo grial de las 
personas dedicadas a la filosofía del derecho. Es sintomático, sin embargo, que 
hayan fracasado los repetidos esfuerzos por desarrollar una teoría del derecho 
comúnmente aceptada. Ya a inicios del siglo XX, Leon Petrażycki escribió:
El hecho de que hasta el día de hoy se haya probado imposible definir el derecho, 
aun a pesar del mucho esfuerzo dedicado a la tarea y que fueron propuestos –con el 
pasar del tiempo- innumerables intentos más o menos ingeniosos y fundamentales de 
caracterizar la esencia del derecho, ha conducido recientemente a dudas sobre si del 
todo la tarea puede ser completada, y a aceptar definiciones que son claramente defi-
cientes, así como evitar la cuestión sobre la esencia del derecho, para lograr así alguna 
paz (Petrażycki 1959: 25).
50 años después, H. L. A. Hart hizo una observación casi idéntica:
Pocas preguntas relacionadas con la sociedad humana han sido formuladas con tan-
ta persistencia y respondidas por pensadores serios de formas tan diversas, extrañas 
e incluso paradójicas como la pregunta ‘¿qué es el Derecho?’. Incluso si confinamos 
nuestra atención a la teoría del derecho de los últimos 50 años y descuidamos la espe-
culación clásica y medieval sobre la ‘naturaleza’ del derecho, nos encontraremos en 
una situación sin paralelos en ningún otro asunto sistemáticamente estudiado en una 
disciplina académica separada (Hart 2012: 1).
Las observaciones de Petrażycki y Hart se mantienen vigentes hasta hoy. 
Este hecho sorprendente puede llevar a la conclusión de que quizás hay ‘algo 
malo’ en el proyecto mismo de brindar una caracterización de la esencia del 
105
journal for constitutional theory and philosophy of law
(2023) 51
Ciencia cognitiva y naturaleza del derecho
derecho o en la búsqueda de la naturaleza del derecho. Los fracasos recurrentes 
en el proyecto sugieren que puede que se base en supuestos falsos.
Creo que los análisis presentados en las secciones precedentes arrojan algu-
na luz sobre este problema. Toda concepción del derecho es un abordaje teórico 
de aspectos seleccionados de las reglas rudimentarias. En otras palabras, toda 
concepción del derecho es una reconstrucción de las prácticas sociales reales, 
y las prácticas en sí mismas no caen bajo ninguna categoría natural preteórica. 
Las reglas rudimentarias no son jurídicas, morales o prudenciales: sólo son pa-
trones de conducta que son seguidos por una comunidad dada. Solo tras teori-
zar en algunos de sus aspectos puede uno construir un sistema jurídico o moral. 
Esta imagen se complica aún más debido al hecho de que nuestros esfuerzos 
teóricos, i.e. los sistemas normativos que desarrollamos, así como los abordajes 
del derecho o la moralidad que construimos, conllevan cierta influencia sobre 
las prácticas rudimentarias del seguimiento de reglas. Para una persona educa-
da del medioevo, el derecho era algo diferente que, para un positivista de inicios 
del siglo XX no sólo porque ellos adoptaron concepciones distintas del derecho, 
sino también debido al hecho de que esas concepciones (por un lado, la teoría 
del derecho natural, por el otro, el positivismo jurídico) influyeron [informed] 
en gran medida las prácticas sociales relevantes.
Así, si el escenario evolutivo sobre el surgimiento de la cultura que he bos-
quejado arriba es al menos a grandes rasgos correcto, uno no debería esperar 
capturar la naturaleza real del derecho. en ninguna definición. Esta conclusión 
es verdadera sin consideración a si uno entiende la naturaleza del derecho como 
refiriéndose a algo ahí afuera, i.e. algún elemento de la estructura de la realidad 
(como en el caso de las teorías del derecho natural sustantivas), o pertenecien-
do a condiciones necesarias y/o suficientes para aplicar el predicado derecho 
(como en el caso de algunas teorías contemporáneas positivistas del derecho). 
No existe una naturaleza del derecho en el sentido sustantivo de la palabra, ya 
que el derecho es el resultado de nuestros esfuerzos conjuntos prácticos y teóri-
cos (reflexivos). Por la misma razón, no hay una naturaleza del derecho enten-
dida como ciertos criterios para distinguir el derecho de otros fenómenos. Por 
supuesto, tales criterios pueden ser propuestos, pero uno no debe esperar en-
contrarlos fijados de una vez y para siempre o sin una alternativa sostenible. En 
cierta forma, en la reconstrucción teórica del derecho uno nunca puede adoptar 
un punto de vista totalmente externo: al desarrollar una teoría del derecho uno 
eo ipso participa en el discurso jurídico (y al menos potencialmente) influencia 
las formas de las prácticas sociales relevantes, aun si la influencia es muy limita-
da. Esto no significa, por supuesto, que nuestra empresa teórica en busca de la 
naturaleza del derecho sea completamente fútil: al menos ella informa nuestro 
entendimiento del derecho, y ayuda a mejorar, nuestra práctica jurídica. Pero 
esto no cambia el hecho de que no hay una imagen verdadera del derecho, así 
como no hay una imagen verdadera de la moralidad o el lenguaje.
106
(2023) 51
journal for constitutional theory and philosophy of law
Bartosz Brożek
Por supuesto, se puede insistir en que la concepción que he esbozado más 
arriba no muestra la inutilidad de buscar la naturaleza del derecho, sino que -un 
tanto paradójicamente- es en sí misma una exposición de lo que es la naturaleza 
del derecho. No tengo ningún problema en aceptar tal conclusión, siempre que 
se admita que la naturaleza del derecho no puede establecerse ni proporcionan-
do una definición esencial ni indicando condiciones (suficientes y/o necesarias) 
para la aplicación del concepto de derecho. Desde este punto de vista, desve-
lar la naturaleza del derecho se reduce a desarrollar una teoría que describa el 
mecanismo que subyace a las prácticas sociales jurídicas de seguir las normas 
jurídicas, reflexionar sobre estas prácticas y teorizar sobre los aspectos jurídi-
camente relevantes de la experiencia. En este sentido, cualquier teoría científi-
ca que proporcione una explicación mecanicista trata sobre la “naturaleza” de 
algo: la mente, el lenguaje, la visión, las emociones, la inflación o la intuición. Si 
uno desea seguir este lenguaje, sin duda puede hacerlo, pero, personalmente, lo 
encuentro algo engañoso.
Bibliografía
Bennett, M.R. & Hacker, P.M.S. (2003). Philosophical 
Foundations of Neuroscience, Wiley Blackwell.
Brożek, B. (2011). Philosophy in Neuroscience, en B. 
Brożek, J. Mączka, W. Grygiel (eds.), Philosophy 
in Science. Methods and Applications, Copernicus 
Copernicus Center Press.
Brożek, B. (2013a). Philosophy and Neuroscience. 
Three Modes of Interaction, en J. Stelmach, B. 
Brożek, Ł. Kurek (eds.), Philosophy in Neuro-
science, Copernicus Center Press.
Brożek, B. (2013b). Rule-following. From Imitation to 
the Normative Mind, Copernicus Center Press.
Brożek, B. (2015a). Explanation and Understanding, 
en B. Brożek, J. Stelmach, M. Hohol (eds.), The 
Concept of Explanation, Copernicus Center Press.
Brożek, B. (2015b). Imitation and the Emergence of 
Normative Orders, en B. Brożek, J. Stelmach, Ł. 
Kurek (eds.), The Emergence of Normative Orders, 
Copernicus Center Press.
Brożek, B., Hage, J., & Vincent, N. (eds.) (2021). Law 
and Mind. A Survey of Law and the Cognitive 
Sciences, Cambridge University Press.
Brożek, B. (s.f.) Models of Naturalizing Jurisprudence, 
en manuscrito.
Di Pellegrino, G., Fadiga, L., Fogassi, L., Gallese, V. 
& Rizolatti, G. (1991). Understanding Motor 
Events: A Neuropsychological Study, Experimen-
tal Brain Research 91(1), 176–180.
Donald, M. (2005). Imitation and Mimesis, en S. 
Hurley, N. Chater, (eds.), Perspectives on Imita-
tion vol. 2, (pp. 283-300). MIT Press.
Garland, B. (2004). Neuroscience and the Law. Dana 
Press.
Hickok, G. (2014). The Myth of Mirror Neurons, 
W.W. Norton & Company.
Hurley, S. & Chater, N. (2005). Introduction: The 
Importance of Imitation, en S. Hurley, N. Chater 
(eds.), Perspectives on Imitation vol. 2 (pp. 1-52). 
MIT Press.
Iacoboni, M. (2005). Understanding Others: Imita-
tion, Language and Empathy, en S. Hurley, N. 
Chater (eds.), Perspectives on Imitation vol. 1 (pp. 
77-100). MIT Press.
Jones, O.D., Schall, J.D. & Shen, F.X. (2014). Law and 
Neuroscience, Vanderbilt University Law School, 
Public Law and Legal Theory Working Paper, 12-
14.
Meltzof, A.N. (2005). Imitation and Other Minds: 
The “Like me” Hypothesis, en S. Hurley, N. 
Chater (eds.), Perspectives on Imitation vol. 2 (pp. 
55-77) MIT Press.
Pardo, M. & Patterson, D. (2013). Minds, Brain, and 
Law. Oxford University Press.
Popper, K. (1966). The Open Society and Its Enemies, 
II, Princeton University Press.
Rizolatti, G. (2005). The Mirror Neuron System and 
Imitation, en S. Hurley, N. Chater (eds.), Per-
spectives on Imitation vol. 1 (pp. 55-76). MIT Press.
Rowlands, M. (2010). The New Science of the Mind, 
MIT Press.
Silk, J. (2009). Forum, en M. Tomasello, Why we coo-
perate, (pp. 111-124). MIT Press.
Spranger, T. M. (ed.) (2012). International Neurolaw, 
Springer.
Tomasello, M. (1999). Cultural Origins of Human 
Cognition, Harvard University Press.
Tomasello, M. (2009). Why we cooperate, MIT Press.
Załuski, W. (2009). Evolutionary Theory and Legal 
Philosophy, Edward Elgar Publishing.
107
journal for constitutional theory and philosophy of law
(2023) 51
revus (2023) 51, 107–129
Johan Hermansson*
Structuring concepts of legal personhood
On legal personhood as a cluster property
Legal persons have traditionally been understood as entities with legal rights and/or du-
ties. This traditional concept of legal personhood has been challenged during the last 
decades: in legal practice through the global development in law where previous legal 
non-persons such as different non-human natural entities, non-human animals, fetuses 
and artificial intelligences have been ascribed, or been proposed to be ascribed, legal 
rights and/or status as legal persons; in legal theory through, inter alia, Visa Kurki’s Bun-
dle Theory of Legal Personhood. The purpose of this paper is to critically assess the 
concept of legal personhood proposed in the Bundle Theory. After discussing the defini-
tional structured traditional concepts of legal personhood and Kurki’s argument against 
them, I will focus on the first main tenet of the Bundle Theory: that legal personhood is a 
prototype structured cluster property which consists of incidents which are separate but 
interconnected. I will argue that such a concept is untenable since the legal personhood 
of legal persons in positive law is not structured in this way. Lastly, I will suggest the use 
of a dual structured concept of legal personhood that maintains the benefits of previous 
concepts but avoids their major deficiencies.
Keywords: legal person, legal personhood, the bundle theory of legal personhood, Kurki 
(Visa), conceptual structure
1 INTRODUCTION
Legal persons and legal personhood1 have received renewed attention within 
legal science during the first decades of the new millennia. Legal persons have 
traditionally been understood simply as entities with, or sometimes consisting 
of, legal rights and/or duties.2 These traditional concepts have been challenged 
during the last decades through the global development in positive law extend-
ing legal personhood to new kinds of entities: to previous legal non-persons 
who are now ascribed legal rights and/or status as legal persons.
The most frequent of these new kinds of entities in positive legal provisions 
so far are the different examples of non-human nature ascribed legal personhood 
* johan.hermansson@jur.lu.se | Department of Law, Lund University (Sweden). I would like 
to thank Matilda Bergström, Christian Dahlman, Eduardo Gill-Pedro, Ellika Sevelin, Lena 
Wahlberg and all the participants at my first seminar at the Faculty of Law at Lund University 
for comments on earlier drafts of this paper. I would also like to thank the two anonymous 
reviewers for their percipient observations and constructive critique.
1 With the term legal personhood, I refer to a certain set of legal positions and with legal person 
to the entity that has the capacity for, holds or consists in these positions. In English-language 
legal theory the terms are sometimes used interchangeably.
2 This is further developed in section 3.1 below.
108
(2023) 51
journal for constitutional theory and philosophy of law
Johan Hermansson
in different legal orders across the world.3 Although these provisions sometimes 
differ considerably from one another they are often labeled under the catch-all 
phrase rights of nature.4 Some of the more studied examples are the rights of 
nature under the 2008 constitution of Ecuador5 and the granting of status as 
legal persons or legal subjects to rivers in New Zealand6, India7 and Colombia.8 
Less successful, at least in practice, are the attempts to ascribe legal rights or 
status as legal person to non-human animals.9 Monkeys and apes have, however, 
been recognized as legal persons, or at least legal subjects (sujetos de derecho), 
in judgments from both Argentinian and Ecuadorian courts.10 The possible le-
gal personhood of artificial intelligence has also caused recent legal disputes, for 
example regarding the patent applications tried before the Legal Board of Appeal 
of the European Patent Office naming the AI DABUS as inventor.11
The legal doctrine treating, as well as anticipating, the development sketched 
in the previous paragraph is already vast.12 What unites most legal scholars 
writing on the subject, including legal philosophers, is, however, their adher-
ence to the traditional concepts of legal personhood, according to which legal 
persons are entities with legal rights and/or duties. This equation of legal per-
3 An overview of legal initiatives can for example be found in the Eco Jurisprudence Monitor. 
This database has so far registered around 80 approved legislative provisions concerning the 
rights of nature at municipal, national and constitutional level and around 80 cases concern-
ing rights of nature approved in courts. The database is largely based on the careful mapping 
of rights of nature initiatives carried out by Putzer e.al., see Putzer et al. 2022.
4 For a recent overview from a political science perspective, see Tanasescu 2022. 
5 Constitución de la República del Ecuador 2008, in particular articles 71 to 74.
6 Te Awa Tupua (Whanganui River Claims Settlement) Act 2017 section 14(1). It should be 
noted that, according to the actual phrasing of the provision, it is not the Whanganui River 
in itself that is constituted as a legal person, but an entity called Te Awa Tupua. This entity 
incorporates all the rivers physical and meta-physical elements, see section 12.
7 The very first example being the judgment from the High Court of Uttarakhand declaring riv-
ers Ganga and Yamuna as legal persons. See Mohd Salim v State of Uttarakhand & others, WP 
(PIL) 126/2014 (High Court of Uttarakhand) 2017 p. 19. It should be noted that the judgment 
have been stayed and is awaiting the ruling of the Supreme Court of India, see O’Donnell 
2018: 143.
8 Constitutional Court of Colombia, judgment T-622/16.
9 Attempts to grant non-human animals status as legal persons through litigation are for ex-
ample being carried out by the U.S.-based Non-Human Rights Project, compare Kurki 2021a: 
47–48.
10 Final Judgment No. 253-20-JH/22 from the Constitutional Court of Ecuador. The Argen-
tinian cases, recognizing the orangutang Sandra and the chimpanzee Cecilia as sujetos de 
derecho, are refereed in Kurki 2019: 198–199.
11 Case no. J 0009/20 (2021-12-21) Legal Board of Appeal of the European Patent Office. The 
application was rejected on the basis that DABUS was not a “person with legal capacity”, see 
section 4.3.1 of the judgment.
12 For example, a search on “rights of nature” in the LUBsearch research database of Lund uni-
versity libraries alone on September 22, 2023 resulted in 2 154 results.
109
journal for constitutional theory and philosophy of law
(2023) 51
Structuring concepts of legal personhood
On legal personhood as a cluster property
sonhood with a general holding of rights and/or duties in legal theory have 
largely remained unquestioned throughout the decades.
Finnish legal philosopher Visa Kurki has recently challenged the traditional 
legal personhood concepts in his book, A Theory of Legal Personhood.13 In the 
book Kurki introduces his own theory, the Bundle Theory of Legal Personhood 
(the Bundle Theory), according to which no singular legal positions constitute 
necessary or sufficient parts of legal personhood and no exact border between 
legal personhood and non-legal personhood exists.14 Since its release in 2019 
the book has been recognized as an important contribution to legal theory by 
reviewers,15 and is frequently quoted in new works on the subject.16 However, 
many works referring to the Bundle Theory so far do so briefly, and/or without 
fully adopting its original concept of legal personhood.17 Critique of the Bundle 
Theory has also so far been scarce, with the notable exception of a series of ar-
ticles on the theory published as a part of a symposium in the 44th issue of the 
international philosophy of law journal Revus.18
I find the Bundle Theory to be an interesting contribution to the develop-
ment of legal theory in general and theory on legal personhood in particular. I 
will show, however, that the definition of legal personhood as a cluster property 
in the first main tenet of the theory, which Kurki himself have described as 
the most original part of the theory,19 is untenable. I will both present relevant 
counterarguments to the concept of legal personhood proposed in the Bundle 
Theory and sketch a tenable alternative.
The paper is structured as follows: I will first elaborate on how I understand 
concepts as Fregean senses with different possible structures, before addressing 
the traditional concepts of legal personhood and Kurki’s argument for refuting 
them: the extension, or discrepancy, problem. In short, the extension problem 
relates to the fact that entities that are not considered legal persons in legal doc-
trine and positive law still holds certain legal positions. This creates a discrep-
ancy between the extension of the legal personhood concepts of traditional the-
ory (all entities with legal positions are legal persons) and the extension of the 
legal personhood concepts in positive law (some entities with legal positions 
are not legal persons). I conclude the section by arguing that this argument for 
13 Kurki 2019.
14 Kurki 2019: 93–94.
15 See for example Afrouzi 2020:1354–1358. Pratama & Rizkiyah 2022: 157–166.
16 Google Scholar lists 204 texts in which the book is quoted as of September 22, 2023.
17 Many scholars tend to refer to the theory for alternative views or for an overview of the litera-
ture on the topic, see for example Worthington & Spender 2021: 3. Of course, there are also 
examples of scholars applying the theory to a certain extent, see for example Mocanu 2022.
18 Stancioli 2021, Fasel 2021, Naffine 2021, Siltala 2021, Aalto-Heinilä & Karhu 2021, Banaś 
2021, Kurki 2021b.
19 Kurki 2021b: 185 (online § 43).
110
(2023) 51
journal for constitutional theory and philosophy of law
Johan Hermansson
a revised legal personhood concept is tenable but to a large extent dependent 
on marginal examples. In the next section I will focus on the legal personhood 
concept of the Bundle Theory and its first tenet, the claim that legal personhood 
is a prototype structured cluster property. I will argue that this is not a tenable 
theoretical concept of legal personhood since the legal personhood concepts 
of positive law necessarily includes certain legal positions, which cannot be 
the case if legal personhood would be a cluster property. In the last section, I 
suggest an alternative structure for a tenable concept of legal personhood. This 
concept of legal personhood combines the certainty of the definitional struc-
tured concepts of traditional theory with the flexibility and richness of the pro-
totype structured concept of the Bundle Theory into a dual structured concept 
that more accurately captures the legal personhood concepts of positive law.
2 THREE WAYS OF STRUCTURING (LEGAL) CONCEPTS
In order to clarify my critique of the concepts of legal personhood that are 
the topic of this paper some points regarding conceptual theory should firstly 
be made. Since Frege’s distinction between sign, sense, and referent,20 it has be-
come common to separate the term, concept, and referent of a word in ana-
lytical philosophy. According to this understanding, that I will adopt, a word 
(term) expresses a certain meaning in a certain context (concept) which refers 
to a certain object (referent).21
I will also distinguish between the extension and intension of a concept. By 
the extension of a concept, I refer to the group or class of entities described by 
the concept, and with the intension of a concept I refer to the conditions an 
entity must satisfy to fall within this class of entities.22
I will finally assume that the intension of concepts can be structured in differ-
ent ways. There are many theories on the structure of concepts, but for my pre-
sent argument it is sufficient to mention three of these. According to the classi-
cal theory of concepts, the intension of concepts has a definitional structure: they 
encode necessary and sufficient conditions for their application.23 According to 
its main rival, the prototype theory of concepts, the intension of concepts has a 
prototype structure: their application is dependent on the holding of a sufficient 
number of statistically prominent (typical) components, where some may be 
more significant than others.24 Finally, according to what might be called a dual 
20 Frege 1948: 27.
21 Margolis & Laurence 1999: 5–7; compare Zalta 2023.
22 Compare ‘Extension/Intension’ in Blackburn 2016. I do not, however, use referent and exten-
sion as synonymous terms, which sometimes is the case.
23 Margolis & Laurence 1999: 8–10.
24 Margolis & Laurence 1999: 27.
111
journal for constitutional theory and philosophy of law
(2023) 51
Structuring concepts of legal personhood
On legal personhood as a cluster property
theory of concepts, concepts have a definitional structured core of necessary and 
sufficient conditions and an additional prototype structured bundle of statisti-
cally prominent components that alone are neither necessary nor sufficient.25
In the following, it will be evident that the legal personhood concepts of tra-
ditional theory have a definitional structure since they set up necessary and suf-
ficient conditions for their application: the holding of rights and/or duties, or 
capacity for this. Kurki’s argument against the concepts of traditional theory, as I 
would rephrase it, is that the extension of these theoretical concepts differs from 
the extension of its referent, the actual concepts of legal personhood in posi-
tive law.26 Kurki’s alternative concept of legal personhood avoids this problem 
by proposing a prototype structured concept, where the holding of no single 
legal positions alone constitute necessary or sufficient conditions. This, however, 
causes other problems and more importantly does not seem to align to how legal 
personhood actually is structured in positive law. Instead, it seems likely that 
there exist necessary and together sufficient conditions for legal personhood in 
positive law, which, however, are different than the ones assumed in traditional 
legal theory, in addition to legal positions that are commonly but not necessarily 
held by legal persons. In the last section I present an alternative dual structured 
legal personhood concept that utilizes these observations and whose extension 
better aligns to the extension of the legal personhood concepts of positive law.
3 THE TRADITIONAL CONCEPTS OF LEGAL 
PERSONHOOD
3.1 The orthodox view of the traditional concepts
Legal persons are commonly understood as entities with legal rights and/
or duties, as already mentioned in the introduction. Kurki dedicates the first 
third of his book to this group of traditional concepts of legal personhood, 
their history, and their deficiencies.27 He calls them the orthodox view of legal 
personhood,28 a term I will also use in this section to refer to the concepts of 
legal personhood in traditional theory.
25 Compare Margolis & Laurence 1999: 42–43.
26 I think that one, in accordance with Frändberg, carefully should distinguish between legal con-
cepts with a law-stating function (stating the material legal content) and a juridical-operative 
function (concept for juridical handling of legal content), see Frändberg 2009: 2. The confusion 
on these two kinds of legal concepts can sometimes lead to confusion when the latter wrongly 
is used as an authoritative source in legal argumentation, compare Lindahl 1985: 37–39.
27 Kurki 2019: 29.
28 See for example Kurki 2019: vii.
112
(2023) 51
journal for constitutional theory and philosophy of law
Johan Hermansson
After explaining how understanding legal personhood as the holding of 
rights and/or the bearing of duties became the dominant view in Western legal 
theory Kurki distinguish five regularly occurring formulations of the orthodox 
view:29
– The definition of a legal person as an entity that holds at least one right or 
bear at least one duty is called the Rights-or-Duties position.
– The definition of a legal person as an entity that holds at least one right 
and bears at least one duty is called the Rights-and-Duties position.
– The definition of a legal person as an entity with the legal capacity to hold 
rights and bear duties is called the Capacity-for-Rights position.
– The definition of a legal person as an entity with the capacity to be a party 
to legal relations is called the Capacity-for-Legal-Relations view.
– Lastly, a fifth definition is mentioned but not named, the definition of 
legal persons as simply a bundle of rights and duties.30
Many existing definitions of legal persons, and hence different concepts of 
legal personhood, fit quite neatly into this systematization of the orthodox view. 
Some definitions of legal persons, however, also explicitly mention legal per-
sons’ capacity for standing, in addition to rights and duties.31 It should also be 
noted that there exists a fundamental metaphysical division in legal person-
hood theory which Kurki does not address in this systematization: Some schol-
ars understands legal persons as entities existing in the physical world holding 
legal positions and other scholars understand legal persons as purely legal enti-
ties consisting of legal positions and not existing (independently) in the physical 
world.32 The common denominator for both positions is, however, the equation 
of legal personhood with rights and duties, independent of which rights or du-
ties that a legal person is holding, or consisting in. Because of the definitional 
structure (necessary and sufficient conditions) given to their respective versions 
of the concept of legal personhood, the adherents of the orthodox view are thus 
logically forced to conclude that an entity has legal personhood (and conse-
29 Kurki 2019: 55.
30 Kurki 2019: 55–56. It should be noted that the Capacity-for-Rights position, despite its name, 
includes both the capacity to hold rights and to bear duties.
31 This is at least common in Swedish legal theory and in provision defining the legal person-
hood of artificial legal persons. See for example the Swedish Foundations Act (1994:1220) 
chapter 1 section 4 that states that a foundation can acquire rights, take on obligations and 
bring actions (föra talan) before courts and other authorities. Compare for example Agell et 
al 2010: 51–52; Grauers 2017: 47. It should be noted that legal personhood still is defined as 
a capacity for rights and duties in general, although a capacity for standing also is included. 
These versions of the orthodox view thus differ from the alternative legal personhood concept 
proposed in section 5 since the latter connects legal personhood to capacity for specific rights 
and duties, including standing.
32 The most well-known proponent of the latter position is probably Kelsen. Kelsen 2009: 173–
174. Kurki addresses this position in a later section of the book. Kurki 2019: 133, 137–138.
113
journal for constitutional theory and philosophy of law
(2023) 51
Structuring concepts of legal personhood
On legal personhood as a cluster property
quently also is a legal person) if this entity is deemed to have a right, a right and 
a duty, capacity for rights, etc.
3.2 The extension/discrepancy problem
Kurki’s main argument for revising the legal personhood concepts of the or-
thodox view is what in later articles has been called the extension problem of the 
orthodox view,33 or the discrepancy argument:34 the orthodox view comes into 
conflict with commonly held beliefs about the extension of the concept of legal per-
sonhood when combined with modern theories of rights.35 This needs elaboration.
The intension of the concept of legal person according to the orthodox view 
could be summarized as an entity with rights and/or duties, or the capacity to 
have rights and/or duties, as described in the previous subsection. In addition 
to these intensional beliefs Kurki identifies a number of extensional beliefs com-
monly held by jurists in Western legal systems: beliefs regarding which entities 
that are commonly perceived to fall within the class of legal persons. A number 
of these extensional beliefs are discussed in the book, but the more central be-
liefs for the extension problem are summarized in two convictions:
(i) Human beings who are born, currently alive and sentient are legal per-
sons.
(ii) Animals and fetuses are not legal persons, and slaves were not legal per-
sons.36
These convictions indeed seem to be generally held by jurists, at least they 
are regularly mentioned in connection with the orthodox view.37 They also 
seem to be generally compatible with positive legal norms (with the possible 
exception of sentience). That is, in Western legal systems humans are generally 
ascribed the status of legal person from birth to death, independent of their 
other abilities (and thus also legal personhood),38 while animals and fetuses are 
not ascribed the status of legal person.39 It is therefore possible to rephrase the 
33 Kurki 2021a: 51.
34 Aalto-Heinilä & Karhu 2021: 154 (online §§ 18–20).
35 Kurki 2019: 55 compare 15–19.
36 Kurki 2019: 14 compare 8–10 and 55.
37 For some examples from Swedish legal theory, see Håstad et al. 2016: 50; Ramberg and Malm-
ström 2016: 54)
38 This is sometimes explicitly stated in statues, sometimes not. Furthermore, the exact criteria 
for ascription of legal personhood at birth and its disappearance at death varies between legal 
systems. For a comparative overview, see Chloros, Rheinstein & Glendon 2007, 2–2 and 2–8.
39 It should be noted, however, that fetuses often are considered to be able to have limited legal 
rights on the condition that they are later born alive, see Chloros, Rheinstein & Glendon 
2007: 2-3, 2-4 and 2-5. That animals are not legal person is usually neither stated explicitly. 
That animals are not things, and indirectly neither persons, is, however, explicitly stated in 
several civil codes in Europe: section 285 a of the Civil Code of Austria; Article 3.38 of the 
114
(2023) 51
journal for constitutional theory and philosophy of law
Johan Hermansson
argument based on positive legal provisions: the extension of the concepts of 
legal personhood in traditional legal theory differs from the extension of the 
concepts of legal personhood in positive law in Western legal systems when 
combined with modern theories of rights.
3.3 The orthodox view and theories of rights
As long as humans fulfilling the above-mentioned criteria can have legal 
rights and duties, and animals and fetuses cannot, there is no discrepancy be-
tween the extension of the concepts of the orthodox view and the extension of 
the concepts in positive law. How legal rights are understood is thus fundamen-
tal for Kurki´s argument.
The Hohfeldian analytic structure of rights, that someone’s right necessarily 
correlates to someone else’s duty,40 is widely accepted within legal philosophy.41 
The function of rights, what it means for a right holder to have a right, is, howev-
er, heavily debated.42 The two main competing theories on how the function of 
rights should be understood are the interest theory and the will theory.43 Interest 
theorists generally hold that a duty borne by X constitutes a right for Y if Y’s in-
terests are typically served by the performance of the duty. Will theorists on the 
other hand generally hold that Y’s holding of a right consists in Y’s control over 
X’s duty: whether Y can, for instance, choose to demand the enforcement of X’s 
duty or not.44 So, while interest theorists tend to see rights as legally protected 
interests, will theorists tend to see rights as legally protected choices. Interest 
theories, furthermore, tend to be quite extensive and ascribe rights to both chil-
dren, animals, and fetuses, while will theories tend to be more restrictive and 
only ascribe rights to adult human beings with sufficient mental abilities.
Belgian Civil Code; section 494 of the Czech Civil Code; Section 90 a of the German Civil 
Code (BGB); Article 287 of Moldovan Civil Code; Book 3 Article 2a of the Dutch Civil Code; 
Article 333 of the Spanish Civil Code; Article 511-1(3) of the Civil Code of Catalonia; Article 
641a of the Swiss Civil Code. For an overview of some of these provisions, see Kempers 2021.
40 See generally Hohfeld 1913: 30 and onwards.
41 See Wenar 2021. For an overview on how Hohfeld’s theses has been developed in legal theory, 
see Arriagada Cáceres 2018.
42 For a general overview, see Wenar 2021: section 2. It should be noted that these theories not 
solely are used for explaining the function of rights as positive norms in the Hohfeldian ana-
lytic structure but also, inter alia, as reasons for rules, both as outside or extra-legal reasons 
and as inside or intra-legal reasons (such as rights often are understood in constitutional law). 
See Kennedy 2002: 185–187.
43 It should be noted that these consists of a great number of similar theories on rights, that, 
however, often are grouped together for pedagogical purposes.
44 This is Kurki’s own summary, see Kurki 2019: 61. A will-interest right could naturally also 
involve the choice to unilaterally change Y’s own position, for example through a gift. This 
would, however, also change the duties of X in relation to Y if X would be liable to such a 
change. The right would thus still be partially constituted by Y’s control over X’s duty.
115
journal for constitutional theory and philosophy of law
(2023) 51
Structuring concepts of legal personhood
On legal personhood as a cluster property
This brings us back to the core of the extension problem. The extension of 
the concept of legal personhood is either too wide (animals and fetuses are le-
gal persons since they have interest-based rights), or too narrow (children and 
adult humans with insufficient mental abilities are not legal persons since they 
lack will-based rights) in relation to extensional beliefs or positive law when the 
different concepts of the orthodox view are combined with modern right theo-
ries such as the interest or will theory.45
3.4 Assessing the extension problem
If the different concepts of legal personhood of the orthodox view fail to ac-
curately capture the extension of legal personhood in positive law (or common 
extensional beliefs) in the way described above this would be a strong argument 
for abandoning the traditional concepts of legal personhood. To summarize my 
view, I consider the argument tenable but weak, since it relies on rather mar-
ginal examples when some versions of the orthodox view are combined with 
some versions of the rights theories.
Kurki´s counterexample against a tenable combination of the interest the-
ory and the Rights-and-Duties position is that slaves in the antebellum U.S.A. 
bore duties while at the same time holding some limited interest-based rights 
and that infants are considered legal persons, and yet they bear no duties.46 
His counterexample against a tenable combination of a soft version of the will 
theory (according to which rights can be held through representants)47 and the 
Rights-or-Duties and Rights-and-Duties positions is that this relation resem-
bles a trustee rather than an agent relation, allowing for animals being the ben-
eficiaries of pet trusts48 to be considered legal persons (this is also an argument 
against the combination of the interest theory and Rights-and-Duties position). 
Many versions of the will theory would also ascribe rights to slaves in the ante-
bellum U.S.A. since they had the right to appeal criminal convictions.49
If one accepts that rights and duties can be carried out by representants, 
which after all is the case for all legal persons but none (or few) legal non-per-
sons in positive law, Kurki’s extension problem is dependent on two relatively 
marginal examples: U.S. antebellum slaves and U.S. pet trusts. It could be ques-
tioned if these examples are relevant for explaining contemporary legal person-
hood in Western legal orders in general. It thus appears as if the extension prob-
lem is not so problematic for some versions of the orthodox view combined 
45 Compare Kurki 2019: 61.
46 Kurki 2019: 71.
47 Hart 1982: 184 footnote 86.
48 A trust is commonly defined as “an equitable estate committed to the care of a fiduciary 
(trustee) for a beneficiary”, see ‘Trust’ in Garner 2011.
49 Kurki 2019: 67–68.
116
(2023) 51
journal for constitutional theory and philosophy of law
Johan Hermansson
with some versions of the rights theories as it appeared prima facie, at least if 
one is satisfied with a legal personhood concept that captures most but not all 
instances of what is considered legal personhood in positive law.
As stressed above, I do not find this argument untenable in itself: it is hard, 
probably impossible, to explain the extension of the concepts of legal person-
hood in positive law when understanding the holding of legal rights and/or du-
ties as necessary and sufficient conditions for legal personhood. For practical 
reasons legal non-persons are sometimes ascribed rights and duties.
In Sweden, limited liability corporations (aktiebolag) have been deemed to 
be able to hold rights and bear duties under certain circumstances before their 
registration, despite the Swedish Companies Act clearly stating this not being 
possible.50 In a similar manner limited liability corporations can exercise rights 
and bear duties under certain circumstances also after being dissolved, despite 
the main rule stating that the companies lose their capacity for rights and duties 
as well as standing after being dissolved.51 Thus traditional non-legal persons, 
not yet registered and dissolved corporations, are able to hold both interest and 
will based rights as well as duties in accordance with Swedish positive law.52 
Another Swedish example of non-legal persons previously holding rights and 
bearing duties are investment funds not considered legal persons but still hav-
ing rights and duties under Swedish tax law.53
The above-mentioned examples do not make the extension argument strong-
er. On the contrary, it underlines that this argument, at least presently, relies on 
marginal cases, which by some might be regarded as exceptions or anomalies. It 
is also clear, however, that non-legal persons sometimes are ascribed both rights 
and duties. One thus has a good reason to separate the general holding of legal 
positions from legal personhood, as suggested by Kurki. And the legal person-
hood concepts of traditional theory consequently work less well for explaining 
the exact extension of the legal personhood concepts in positive law.
50 See Lehrberg 2016: 73–75, compare NJA 1984 s 495 and chapter 2 section 25 in the Swedish 
companies act (2005:551). The Swedish legal philosopher Karl Olivecrona early observed the 
paradox that corporations need legal capacity to acquire legal capacity according to the law, 
see Olivecrona 1928: 136.
51 This follows from a judgment of the Swedish supreme court, see NJA 1979 s 700, and has been 
confirmed in later case law. For an overview, see Lehrberg 2016: 394 and subsequent pages. 
See also generally Lehrberg 1990.
52 This is not only the case in Sweden, for example similar constructions of preliminary compa-
nies in Germany and Romania award partial legal capacity to not yet formed companies, this 
is mentioned in Mocanu 2022: 7.
53 Two kinds of special investment funds (värdepappersfonder and specialfonder) are not con-
sidered legal persons in accordance with chapter 4 section 1 and chapter 12 section 1 in the 
statues regulating them (2004:46 and 2013:561). Despite this they previously had rights and 
obligations as subjects of tax law (skattesubjekt), this is stated explicitly in the preparatory 
works, see prop. 2011/21:1 p. 400.
117
journal for constitutional theory and philosophy of law
(2023) 51
Structuring concepts of legal personhood
On legal personhood as a cluster property
4 LEGAL PERSONHOOD AS A CLUSTER PROPERTY
4.2 The legal personhood concept of the Bundle Theory and 
its incidents
In this section I will examine Kurki’s proposed solution to the extension 
problem: the alternative legal personhood concept of the Bundle Theory. 
According to the first main tenet of the theory, which is central for his concept 
of legal personhood, the legal personhood of X is a cluster property and consists 
of incidents which are separate but interconnected.54 Somewhat simplified, legal 
personhood is understood not as a general holding of legal positions but as a 
bundle of legal position that varies from person to person and from one legal 
context to another. But to fully grasp his legal personhood concept it is neces-
sary to understand what Kurki means with both cluster property and incidents.
I will start with cluster property. Kurki is not the first to define legal per-
sonhood as a bundle of legal positions. The previous understandings of legal 
personhood as a bundle have, however, rested on the definitional structured 
concepts of legal personhood of traditional theory according to which the hold-
ing of rights and/or duties are necessary and sufficient criteria for legal person-
hood.55 In the Bundle Theory on the other hand, the claim is made that legal 
personhood is a bundle in the form of a cluster property in accordance with 
what Kurki calls the standard sense of the concept: “a property whose extension 
is determined based on a weighted list of criteria, none of which alone is neces-
sary or sufficient.”56
In other words (that Kurki does not employ himself), the legal personhood 
concept of the Bundle Theory is given a prototype structure: its application is 
dependent on the holding of a sufficient number of typical components (inci-
dents of legal personhood) of which some may be more significant than others, 
but none alone is necessary or sufficient. This definition of legal personhood as 
a cluster property is, as mentioned in the introduction, the truly novel inven-
tion of the Bundle Theory. It also has two major implications: (1) that none of 
the incidents of legal personhood identified by the theory alone are necessary 
or sufficient for an entity to have legal personhood and (2) that no exact border 
exist between legal personhood and non-legal personhood.
54 Kurki 2019: 5.
55 Kurki mentions Ngaire Naffine, Richard Tur and Jens David Ohlin as examples of this, Kurki 
2019: 93–94. Similar accounts for the relativity of legal personhood is evident also in Kelsen’s 
Pure Theory of Law, even if he uses the term complex instead, see Kelsen 2002: 173.
56 Kurki 2019: 93.
118
(2023) 51
journal for constitutional theory and philosophy of law
Johan Hermansson
To assess whether the claim that legal personhood is a cluster property is 
tenable one next has to examine the incidents57 of legal personhood identified 
in the theory. As I understand the term in the context of the Bundle Theory, it 
denotes a particular group of legal positions, rights and/or duties, that can be 
held by legal persons or legal non-persons ascribed limited legal personhood. 
It is not necessary for my argument to describe all of the different incidents in 
detail, but it should be relatively apparent from their names and groupings ap-
proximately what kind of legal positions they involve. See Table 1 for an over-
view of the incidents of legal personhood according to the Bundle Theory. I will 
address three of the incidents in detail shortly.
Table 1: The incidents of legal personhood according to the Bundle Theory58
Passive incidents
Substantive incidents Remedy incidents
- Fundamental protections: protection of life, 
liberty, and bodily integrity
- Standing
- Capacity to be the beneficiary of or party to 
special rights59
- Victim status in 
criminal law
- Capacity to own property - Capacity to undergo 
legal harms- Insusceptibility to being owned
Active incidents
Substantive incidents Remedy incidents
- Legal competences - Onerous legal 
personhood (legal 
responsibility)
57 The term incident is borrowed from Anthony Honoré’s division of ideal liberal ownership into 
eleven standard incidents. Honoré 1961: 107. In the Oxford Dictionary of English an incident 
is defined as “a privilege, burden or right attaching to an office, estate, or other holding”, see 
‘Incident’ in Oxford Dictionary of English, 3 ed., online-version. Oxford University Press, 2015.
58 Compare Kurki 2019: 95–96.
59 Both formulations are used in the book.
119
journal for constitutional theory and philosophy of law
(2023) 51
Structuring concepts of legal personhood
On legal personhood as a cluster property
Since legal personhood is defined as a cluster property, none of the incidents 
of legal personhood identified by the Bundle Theory can alone be necessary 
or sufficient for legal personhood. This implies (1) that holding a single one 
of these incidents is not sufficient for having legal personhood, which seems 
plausible. And (2) that there are, or at least could be, legal persons without each 
one of these different incidents since none of them are necessary parts of legal 
personhood. This second implication I find questionable. While some of these 
incidents are currently held by some legal persons in positive law but not by 
other legal persons, this does not seem to be the case for others of the inci-
dents. I will now examine three of these, potentially necessary incidents of legal 
personhood, to assess whether the concept of legal personhood of the Bundle 
Theory is a tenable description of legal personhood in positive law.
4.2 Standing as a non-necessary and non-sufficient incident of 
legal personhood
I will start with the passive remedy incident of standing as an example. Being 
a passive incident, this incident can be held by both passive and active legal per-
sons: both by persons exercising their legal personhood through a representa-
tive (such as infants and small children) and persons exercising it themselves 
(adults of sound mind and to a varying extent older children).60
The standing incident of legal personhood involves several different aspects 
of an entity’s capacity to be part of legal proceedings and enforce its rights 
through legal proceedings. Kurki makes a distinction both between the invested 
aspect of standing and the competence-related aspect of standing, and between 
abstract standing and standing in casu. This needs some clarification.
The invested aspect of standing concerns whether an entitlement of an en-
tity X is recognized by the legal system as enforceable in court. If not recog-
nized, the claims of X are either (1) unenforceable in courts or (2) enforced 
only when they happen to coincide with someone else’s enforceable rights. The 
competence-related aspect of standing concerns X’s legal competence to pursue 
the case in court, to initiate legal proceedings by itself.61
For example, in Swedish administrative law a decision by a public author-
ity, as the main rule, can be appealed if it affects someone’s situation in a not 
insignificant way. The decision can furthermore be appealed by those that the 
decision concerns, if the decision is to their disadvantage.62 The extension of 
the concepts of “someone’s” and “those that the decision concerns” thus delimit 
the number of entities that holds an invested aspect of standing in the decision. 
Other entities could of course be both affected and concerned by the decision, 
60 Kurki 2019: 95.
61 Kurki 2019: 108.
62 See section 41 and 42 of the Swedish Administrative Procedure Act (2017:900).
120
(2023) 51
journal for constitutional theory and philosophy of law
Johan Hermansson
but their entitlements are not independently recognized by the legal system. In 
addition, not all of those with an invested aspect of standing have a competence-
related aspect of standing: an adult human can normally initiate the appellation 
procedure herself, thus having the competence-related aspect of standing, while 
an infant normally would have to do this through a representative, thus lacking 
the competence-related aspect of standing.
Abstract standing on the other hand concerns whether an entity X has 
any kind of standing (invested or competence-related) at all in a legal system, 
whether an entity has the possibility to be a party to legal proceedings at all, 
while standing in casu refers to X’ standing in a particular case.63
Although other aspects of standing are mentioned, no exhaustive account of 
the different legal positions included in the standing incident of legal person-
hood is given.64 It is quite clear, however, that the aim has been to include most 
of the different legal positions that can be held by an entity in legal proceedings 
when recognized as a legal person in the particular legal system.
It seems plausible that this incident alone is not sufficient for legal person-
hood: Entities considered legal non-persons in positive law are occasionally 
recognized with different aspects of standing. Kurki himself mentions the oc-
casional considering of the merits of cases with animal plaintiffs in the U.S.65 
While I am personally unaware of cases where non-legal persons in Swedish 
positive law have been recognized as a party in legal proceedings, claims of 
non-legal persons are occasionally upheld or enforced with reference to their 
specific interests, giving them something resembling what the Bundle Theory 
describes as the invested aspect of standing. One recent example of this is the 
different decisions regarding the now deceased accused murderer of former 
Swedish prime minister Olof Palme, the so-called Skandia-man.66
There are apparently occasional cases of legal non-persons with standing, 
but it is hard to find examples of legal persons entirely without standing in posi-
tive law. Kurki himself gives no examples of this when discussing the incident 
of standing. Furthermore, the examples of legal persons in accordance with 
the extensional beliefs he builds his theory on, human beings who are born, 
currently alive and sentient and artificial persons such as corporations, are all 
63 Kurki 2019: 109.
64 Kurki 2019: 110. This was, according to Kurki, motivated by the fact that not all features of 
standing had much bearing on the questions that his investigation sought to understand.
65 Kurki 2019: 107.
66 See especially the decision from the ombudsman of the Swedish parliament (JO) regarding 
the Skandia-man’s right to be presumed innocent, JO Beslut dnr 6673-2020. See also the ad-
ministrative court of appeals in Stockholm’s decision on the 22 of December 2020 in cases 
number 5005-20, 5077-20 and 5079-20. The Skandia-man also figures in two preliminary 
investigations regarding defamation, of which one is now closed, see decision AM-86565-20 
from the Swedish Prosecution Authority (Åklagarmyndigheten).
121
journal for constitutional theory and philosophy of law
(2023) 51
Structuring concepts of legal personhood
On legal personhood as a cluster property
acknowledged some, although not all, of the aspects of standing. Indeed, ac-
cording to positive law as far as I know and according to widely held exten-
sional beliefs, standing seems to be a necessary, if not alone sufficient, condition 
for legal personhood.67 As mentioned above, there cannot exist any necessary 
conditions for being a legal person if legal personhood is a cluster property in 
the way claimed in the Bundle Theory. It thus seems like the legal personhood 
concept of the Bundle Theory does not capture the legal personhood concepts 
of positive law very accurately.
4.3 Beneficiary of special rights and ownership of property 
as non-necessary or non-sufficient incidents of legal 
personhood
The same objection as above can be raised in relation to other incidents of 
legal personhood identified in the Bundle Theory. For example, the incident 
capacity to be the beneficiary of, or the party to, special rights and the incident 
capacity to own property.
Special rights are described as rights that follow from the exercise of legal 
competences and that are limited to the parties who partake in a special agree-
ment or transaction, alone or through a representant: claims held against spe-
cific entities rather than against the world in general.68 What the capacity to own 
property concerns is rather evident, one or more of the different aspects of own-
ing property: the rights to possess, use, manage, destroy, transfer property, etc.69
Similar to the incident of standing, one can find examples of aspects of these 
incidents being held by entities considered legal non-persons in positive law. As 
I mentioned when discussing the extension problem of legal personhood, pet 
trusts in the U.S. can be said to ascribe pets who are beneficiaries of these trusts 
both some kind of special rights against the trustees as well as some aspects of 
ownership. In Swedish law, limited liability companies also hold these incidents 
to a limited extent before and after their registration. Likely other examples 
could be found, allowing one to conclude that holding these incidents alone 
are not sufficient for having legal personhood. Again, however, it is hard to find 
examples of legal persons in positive law not holding some aspects of these 
incidents. In general, at least in Swedish law, both all natural and all artificial 
persons have the capacity to hold special rights and to own property. Kurki dis-
67 It could of course be argued that standing is a commonly reoccurring but not conceptually 
necessary component of legal personhood, I will address this later.
68 Kurki 2019: 102.
69 Kurki never specificizes which aspects of owning that is sufficient or necessary for an entity 
to have the incident capacity to own. Kurki 2019: 103–106. Although this is not mentioned 
explicitly, this is probably due to the cluster property structure that Kurki seems to assume 
ownership has. I suspect, however, that the incident of transmissibility (capacity to transfer 
ownership) is a necessary incident of ownership.
122
(2023) 51
journal for constitutional theory and philosophy of law
Johan Hermansson
cusses a potential full legal personhood of animals that does not involve own-
ership, but instead fundamental protection accompanied by penal law regula-
tions and some aspects of standing.70 Something resembling this have actually 
already been applied in cases from South America regarding apes and monkeys, 
but these are still rare cases. You can furthermore question whether these pri-
mates should be considered legal persons, they are at least not described as such 
in the verdicts.71 While new kinds of legal personhood could be constructed in 
the future, current legal persons seem to have capacity for special rights as well 
as capacity to own property. To hold the incidents of being the beneficiary of 
special rights and the capacity to own thus also (currently) seem to be neces-
sary, although not sufficient, conditions for having legal personhood.
4.4 Three possible arguments in defense of a cluster property 
legal personhood concept
To summarize my assessment so far, the cluster property legal personhood 
concept proposed in the Bundle Theory, implicating that no incidents of legal 
personhood alone are necessary or sufficient, is hard to defend; some of the 
incidents of legal personhood indeed seems to be necessary since they are held 
by all legal persons in positive law. It is unclear what kind of legal person in 
positive law, if any, that satisfies the variable in the first tenet of the theory (that 
the legal personhood of x is a cluster property and consists of incidents which 
are separate but interconnected). Before concluding this part of the paper, I will 
discuss three possible arguments in defense of constructing a cluster property 
legal personhood concept.
One could argue that emphasizing the entities that are deemed as legal per-
sons in accordance with positive law or widely held extensional beliefs is not 
necessary, and that all entities with some sort of legal position could be consid-
ered as a kind of legal person. With such an understanding of legal persons it 
would be possible to identify legal persons in positive law with legal positions 
but without any aspects of the incident of standing. This would consequent-
ly support the claim that legal personhood should be understood as a cluster 
property. While it is possible to revise the theory in this way, it is quite clear that 
this is not Kurki’s intention in the Bundle Theory. He makes a clear distinction 
between the paradigmatic examples of legal persons or legal personhood “tout 
court”, entities being endowed with significant (but unspecified) numbers of 
incidents of legal personhood, and something called legal subjects and legal sub-
jecthood, referring to the status of an entity within a specific field of law or with 
70 Kurki 2019: 121.
71 See footnote 10 above for references. To my knowledge three successful cases exist. The ex-
pression sujetos de derecho rather than personas naturales or personas jurídicas, the usual 
equivalents of legal person, is used by the courts.
123
journal for constitutional theory and philosophy of law
(2023) 51
Structuring concepts of legal personhood
On legal personhood as a cluster property
regard to a specific legal institution. In accordance with this distinction, ani-
mals can be seen as subjects of animal welfare law, being legally protected from 
certain forms of maltreatment through animal welfare acts, without being legal 
persons in general or legal subjects in other legal fields.72 Even if legal subject-
hood might be understood as a cluster property, without necessary or sufficient 
conditions, support for defining legal personhood as a cluster concept is harder 
to find in positive law. And more importantly, if legal personhood was defined 
in this way, it would fall back into the equation in traditional theory of legal per-
sonhood with the general holding of rights and/or duties, precisely what Kurki 
wants to avoid with his alternative concept.
A second possible argument for a cluster property concept of legal person-
hood could be to argue that while it indeed might be so that all legal persons cur-
rently have capacity for specific legal positions, or incidents, this does not have 
to be the case in the future. No explicit claim in this direction is made, but Kurki 
mentions it indirectly when giving the above-mentioned example of the poten-
tial legal personhood of animals not including capacity for owning property.73 
One could firstly object that not much is gained by adopting a theory to poten-
tial examples if this forces the theory to obscure how legal personhood currently 
works. Secondly, one could object that even if future legal personhood might 
be structured differently, it seems rather implausible that there would exist a le-
gal person entirely without the capacity for any of the legal positions included 
in the standing incident of legal personhood. This would be a legal person not 
only lacking the competence to initiate legal proceedings (competence-related 
aspect of standing) and being the party in a legal proceeding overall (abstract 
standing), but also completely lacking rights enforceable in courts or just having 
rights enforced only when they happen to coincide with someone else’s enforce-
able rights (the invested aspect of standing). It is rather hard to see the purpose 
of creating a legal person with certain legal positions (such as rights) but not giv-
ing this person or any other persons any legal possibilities to uphold them. This 
is radically different from the legal personhood of existing legal persons.
A third possible argument for defining legal personhood as a cluster prop-
erty could be that even if all current legal persons hold certain legal position 
and that future legal persons probably would as well, legal personhood is still 
theoretically best understood as a cluster property within legal science. But 
even if this would be supported by the extension of the concept in positive law 
in the future (some legal concepts are after all partly structured as prototypes)74, 
72 Kurki 2019: 121–124.
73 Kurki 2019: 121.
74 An example of this is the concept of employee in Swedish private law. This concept is not 
defined in any statue, but instead an overall assessment of criteria developed in case law and 
preparatory works is made. These criteria include factors such as whether the person has a 
personal duty to perform work according to a contract, whether the work tasks are predeter-
124
(2023) 51
journal for constitutional theory and philosophy of law
Johan Hermansson
such a concept would still be problematic. This is evident if one considers the 
general deficiencies that necessarily follows when concepts are structured as 
prototypes.75 Most importantly for the present argument prototype structured 
concepts cause problems with extension determination and accommodating 
analytical inferences: one can neither determine exactly which or how many 
components that are sufficient for determining that an entity falls within the 
extension of the concept, nor can inferences be drawn from the fact that an en-
tity falls within or without the extension of the concept, or possess some of the 
common components of the concept.76 Thus, if one defines legal personhood 
as a prototype structured cluster property one cannot determine which entities 
that have or does not have legal personhood nor can one conclude that a legal 
person holds certain legal positions since it is a legal person. These implications 
cannot be avoided since they follow from how the concept is structured. If nec-
essary or sufficient conditions where provided the concept would have a defini-
tional structure instead. I see few, if any, theoretical advantages of adopting such 
a descriptive77 concept of legal personhood if not supported in positive law.
5 STRUCTURING CONCEPTS OF LEGAL 
PERSONHOOD
5.1 A dual structured concept of legal personhood
I have so far concluded that the concepts of legal personhood of traditional 
theory (the orthodox view) are problematic, since the extension of these con-
cepts differs from the extension of legal personhood concepts in positive law, 
its referent. The alternative concept proposed by Kurki in the Bundle Theory 
likewise seems to be empirically untenable, since some of the incidents of legal 
personhood that should be non-necessary in fact are held by all legal persons in 
positive law. This does not mean that the concepts of legal personhood assessed 
so far are entirely mistaken. Legal persons do have rights and duties, even if 
legal non-persons also do occasionally. And these rights and duties do vary, but 
not to the extreme extent suggested in the Bundle Theory. I will in this final sec-
mined, whether the person is prevented from performing similar work for other persons, and 
so on. However, none of these criteria are alone necessary or sufficient for a person being an 
employee, and some of them are considered as more important than others for the outcome 
of the assessment. For an overview in English, see generally Rönnmar 2004.
75 For an overview, see Margolis & Laurence, 1999, pp. 32–43.
76 A similar objection is raised by Mocanu when discussing the application of the Bundle Theo-
ry. See Mocanu 2022: 8–9.
77 It is of course possible that Kurki´s intention with constructing his legal personhood concept 
is not descriptive but normative, this is, however, not what is implied by his argumentation.
125
journal for constitutional theory and philosophy of law
(2023) 51
Structuring concepts of legal personhood
On legal personhood as a cluster property
tion suggest how a concept of legal personhood that incorporates the benefits 
but avoids the deficiencies of these concepts can be structured.
Instead of constructing a concept with a traditional definitional structure, 
like the legal personhood concept of traditional theory, or a concept with a pro-
totype structure, like the legal persons concept of the Bundle Theory, I suggest a 
dual structured concept of legal personhood. This concept has a core of neces-
sary legal positions, although different from the ones traditionally assumed, and 
an additional bundle of typical but non-necessary legal positions. And while 
this might have to be determined after further studies, it seems that the neces-
sary legal positions of this core together also are sufficient for legal personhood 
in contemporary positive legal orders.
If one analyzes the entities referred to as legal persons (and equivalent 
terms) in law, that is, all the entities acknowledged the status of legal person in 
positive law in all the different legal systems, one will find a common core con-
sisting in capacity for a certain number of legal positions: a group of legal posi-
tions that are a common denominator for all entities with this status.78 These 
positions are potentially held by, or included in, all kinds of legal persons, from 
classic instances such as humans that are born and currently alive, corpora-
tions and public legal persons (states and municipalities) to newer instances 
such as the rivers79 declared as legal persons mentioned in the introduction. 
These legal positions are the legal positions required for an entity to function as 
a legal person in property law, understood in a very wide sense. This includes 
at least legal positions pertaining to the (relatively) free acquisition, transfer 
and holding of property through agreements (contracts) and other kinds of 
transactions, liability for violations of such agreements, liability for other kinds 
of damages to the property of others, and the procedural capacities required 
for protecting and upholding the rights and duties following from these po-
sitions. The exact scope of this common core of legal personhood will vary 
between different legal orders, and potentially also among legal persons in the 
same legal order, but a common core could probably be identified that roughly 
corresponds to the above-mentioned legal positions pertaining to property law 
understood in this very wide sense.80
78 It is possible that these legal positions are given to these entities since they are held as legal 
persons, but also that they are held legal persons since they have these legal positions. I believe 
the former to be correct rather than the latter.
79 It should be noted that it, so far, only are human individuals that are exercising the rights and 
fulfilling the obligations of these legal persons as their guardians. No ‘behaviour’ of the rivers 
(floods or droughts etc.) is imputed to the legal person bearing their name.
80 This observation is in fact rather common in older legal theory, especially regarding artificial 
persons. Savigny considered artificial persons as an “artificially created subject having legal 
capacity under property law”. Translated and referred in Beran 2020: 68. The connection be-
tween commodity ownership and the legal subject is also an important aspect of Pashukanis 
commodity exchange theory of law, see Milanovic´s introduction in Pashukanis 2002: 14–15.
126
(2023) 51
journal for constitutional theory and philosophy of law
Johan Hermansson
In addition to this core of necessary and sufficient positions legal person-
hood can, but does not have to, be composed of other kinds of legal positions, 
some more statistically common than others. Such positions are for example 
fundamental legal protection, victim status in criminal law and criminal-law re-
sponsibility, to borrow some examples from the Bundle Theory. Legal positions 
pertaining to family law institutions such as marriage or parenthood could also 
be mentioned. The proposed legal personhood concept can thus also come in 
thicker and thinner bundles of legal position. These bundles of characteristic 
legal positions attaching to the core of legal personhood indeed seems to have 
a prototype-like structure, with none of these additional positions alone being 
necessary conditions of legal personhood but some of them being more typical 
and important than others.
5.2 Benefits of the proposed concept
The proposed concept of legal personhood would solve the extension prob-
lem of the concepts of traditional theory since it can account for the extension 
of the concept according to both positive law and widely held beliefs when 
combined with the modern right theories. It can explain why fetuses and ani-
mals do not have legal personhood and why born and alive humans do: neither 
fetuses nor animals can own property, enter into agreements, be parties to legal 
proceedings, and so on, while all humans can, although some through repre-
sentatives. It can also explain more marginal cases of legal personhood such 
as the ascription of legal personhood to idols in India or to ships, both have 
the capacity to own property through representatives,81 as well as the (alleged) 
ascription of legal personhood to non-human nature and animals mentioned 
in the introduction. It can furthermore explain why slaves did not have legal 
personhood: they did not possess these legal positions, and if they did only to a 
limited extent. The same applies for other marginal examples discussed above, 
such as limited liability companies before registration and after dissolvement. 
These cannot hold these positions in general, only for the limited purposes 
deemed necessary.
Since this concept, with its definitional structured core, allows for deter-
mination of the proper extension of legal personhood and legal inferences, ju-
rists could with support of such a concept explain the consequences of being 
a legal person and that legal personhood necessarily entails some fundamen-
tal legal positions. The additional legal positions of the prototype structured 
bundle would serve as a basis for further investigation or examination. The 
81 For an overview on legal personhood of artificial persons and ships in general and Indian 
idols in particular see the lengthy reasoning of the Supreme Court of India, which clearly 
connects (artificial) legal personality to holding of property, in the 2019 Supreme Court ver-
dict on Ayodhya dispute, M. Siddiq v. Mahant Suresh Das, Judgment of the Supreme Court of 
India dated 09.11.2019 in Civil Appeals No. 10866-10867 of 2010, paragraphs 86 to 124.
127
journal for constitutional theory and philosophy of law
(2023) 51
Structuring concepts of legal personhood
On legal personhood as a cluster property
division in a definitional structured core and a prototype structured bundle 
would also be helpful in illuminating why or why not juristic persons should be 
acknowledged certain legal positions natural persons have, and the other way 
around. Finally, the proposed concept is neutral to the interest and will theo-
ries of rights: it allows for interest theorists to recognize rights for any entity 
they deem morally relevant without necessarily concluding that these entities 
have legal personhood; the will theorists could likewise continue to refuse to 
recognize the rights of infants as long as they accept that they function as legal 
persons through representatives with the relevant legal positions. An equation 
of legal personhood with a specific set of legal positions rather than legal posi-
tions in general thus makes the debate between will and interest theorists less 
relevant for legal personhood theory.
To conclude my argument: There are good reasons to separate the general 
holding of legal positions and legal personhood, as argued for in the Bundle 
Theory. The extension of the concepts of legal personhood in traditional theory 
differs from the extension of the concepts of legal personhood in positive law. 
But there are also good reasons not to adopt the alternative concept of legal 
personhood of the Bundle Theory. Some legal positions, such as standing, are 
held by all legal persons in positive law and thus likely necessary, contrary to 
what would be the case if legal personhood was a cluster property consisting 
of non-necessary and alone non-sufficient incidents. And even if legal person-
hood in positive law would change in the future, prototype structured concepts 
are still problematic since they do not allow for clear extension determination 
or analytical inferences. Unless support for the claim that legal personhood in 
positive law actually is structured as a cluster property, I think such concepts 
should be avoided. Both the disadvantages of the rigid definitional structured 
legal personhood concepts of traditional theory and of the overly flexible pro-
totype structured legal personhood concept of the Bundle Theory can be avert-
ed through a dual structured legal personhood concept. The concept of legal 
personhood proposed in this final section of the paper combines the clarity 
of definitional structured concepts with the flexibility of prototype structured 
concepts. The proposed concept is in need of further development, but it is my 
hope that this paper might benefit the development of legal personhood theory 
and thus the understanding of one of the fundamental structures of law.
Bibliography
Aalto-Heinilä, M., & Karhu, J. (2021). Animals, 
slaves, and beyond. Revus. Journal for Consti-
tutional Theory and Philosophy of Law, 44, 147–
162. https://doi.org/10.4000/revus.7004
Agell, A., Malmström, Å., Ramberg, C., & Sigeman, 
T. (2010). Civilrätt. Liber.
Arriagada Cáceres, M. B. (2018). Fundamentality, 
interdefinability, and circularity. Revus. Journal 
for Constitutional Theory and Philosophy of Law, 
35, 45–84. https://doi.org/10.4000/revus.4173
Banaś, P. (2021). Why cannot anything be a legal per-
son? A critique of Visa Kurki’s Theory of Legal 
Personhood. Revus. Journal for Constitutional 
Theory and Philosophy of Law, 44, 163–171. https:// 
128
(2023) 51
journal for constitutional theory and philosophy of law
Johan Hermansson
doi.org/10.4000/revus.7335
Beran, K. (2020). The Concept of Juristic Person. 
Wolters Kluwer Czech Republic.
Blackburn, S. (2016). Extension/intension. In A 
Dictionary of Philosophy. Oxford University Press. 
https://www.oxfordreference.com/view/10.1093/ 
acref/9780198735304.001.0001/acref-9780198 
735304-e-1203
Chloros, A., Rheinstein, M., & Glendon, M. A. 
(2007). International encyclopedia of comparative 
law. Vol. 4, Persons and family. Mohr.
Ebrahimi Afrouzi, A. (2020). Visa A.J. Kurki, A 
Theory of Legal Personhood, Oxford: Oxford 
University Press, 2019, 240 pp, hb £70.00. The 
Modern Law Review, 83(6), 1354–1358. https://
doi.org/10.1111/1468-2230.12531
Fasel, R. N. (2021). Shaving Ockham. A review of 
Visa Kurki’s ‘A Theory of Legal Personhood’ 
Revus. Journal for Constitutional Theory and 
Philosophy of Law, 44, 113–126. https://doi.org/ 
10.4000/revus.6921
Frege, G. (1948). Sense and Reference. The Philo-
sophical Review, 57(3), 209–230. https://doi.org/ 
10.2307/2181485
Garner, B. A. (2011). Garner’s dictionary of legal us-
age. Oxford University Press.
Grauers, P. H. (2017). Person och avtal: En kortfattad 
inledning till person- och avtalsrätten. Liber.
Hage, J. C., & Pfordten, D. von der (Eds.). (2009). 
Concepts in law. Springer.
Hart, H. L. A. (1982). Legal Rights. In H. L. A. 
Hart (Ed.), Essays on Bentham: Jurisprudence 
and Political Philosophy (pp. 162–193). Oxford 
University Press. https://doi.org/10.1093/acprof:
oso/9780198254683.003.0008
Håstad, T., Björkdahl, E. P., Brattström, M., & Zacka-
riasson, L. (2016). Civilrättens grunder. Iustus.
Hohfeld, W. N. (1913). Some Fundamental Legal 
Conceptions as Applied in Judicial Reasoning. 
The Yale Law Journal, 23(1), 16–59. https://doi.
org/10.2307/785533
Kelsen, H. (2002). Pure theory of law (M. Knight, 
Trans.). Lawbook Exchange.
Kempers, E. B. (2021). Neither Persons nor Things: 
The Changing Status of Animals in Private Law. 
European Review of Private Law, 29(1), 39–70.
Kurki, V. A. J. (2019). A Theory of legal personhood. 
Oxford University Press.
Kurki, V. (2021a). Legal Personhood and Animal 
Rights. Journal of Animal Ethics, 11(1), 47–62.
Kurki, V. A. (2021b). On legal personhood: Re-
joinders, reflections and restatements. Revus. 
Journal for Constitutional Theory and Philosophy 
of Law, 44, 173–194. https://doi.org/10.4000/re 
vus.7425
Lehrberg, B. (1990). I aktiebolagens skymningsland: 
En studie rörande upplösta aktiebolags rättskapac-
itet, med beaktande även av rättsläget vid hand-
lande för ett nybildat, ännu ej registrerat aktie-
bolag. Iustus.
Lehrberg, B. (2016). Aktiebolagsrätt. Iusté.
Lindahl, L. (1985). Definitioner, begreppsanalys och 
mellanbegrepp i juridiken. (Definitions, Con-
ceptual Analysis and Intermediate Concepts in 
the Law.). In Rationalitet och empiri i rättsveten-
skapen (pp. 37–62). Juristförlaget vid Stockholms 
universitet.
Margolis, E., & Laurence, S. (1999). Concepts: Core 
readings. MIT Press.
Mocanu, D. M. (2022). Gradient Legal Personhood 
for AI Systems—Painting Continental Legal 
Shapes Made to Fit Analytical Molds. Frontiers in 
Robotics and AI, 8, Article 788179.
Naffine, N. (2021). Hidden presuppositions and the 
problem of paradigm persons. Visa AJ Kurki’s ‘A 
Theory of Legal Personhood’. Revus. Journal for 
Constitutional Theory and Philosophy of Law, 44, 
127–135. https://doi.org/10.4000/revus.6953
O’Donnell, E. (2018). At the intersection of the sacred 
and the legal: Rights for nature in Uttara khand, 
India. Journal of Environmental Law, 30, 135–
144. https://doi.org/10.1093/jel/eqx026/4364852/ 
At-the-Intersection-of-the-Sacred-and-the-Legal
Olivecrona, K. (1928). Studier över begreppet jurid-
isk person i romersk och modern rätt. Appelbergs 
boktryckeri.
Pashukanis, E. B. (2002). The general theory of law 
and Marxism (B. Einhom, Trans.). Transaction 
Publishers.
Pratama, M. A., & Rizkiyah, E. L. (2022). A Theory of 
Legal Personhood by Visa A.J. Kurki. Indonesian 
Journal of Law and Society, 3(1), 157–166. DOI: 
10.19184/ijls.v3i1.28220
Putzer, A., Lambooy, T., Jeurissen, R., & Kim, E. 
(2022). Putting the rights of nature on the map. A 
quantitative analysis of rights of nature initiatives 
across the world. Journal of Maps, 18(1), 89–96. 
https://doi.org/10.1080/17445647.2022.2079432
Ramberg, C., & Malmström, Å. (2016). Malmströms 
civilrätt. Liber.
Siltala, R. (2021). Earth, Wind, and Fire, and other 
dilemmas in A Theory of Legal Personhood - a 
vindication of legal conventionalism. Revus. 
Journal for Constitutional Theory and Philosophy 
of Law, 44, 137–146. https://doi.org/10.4000/re 
vus.6974
Stancioli, B. (2021). A bundle of rights and Pa-
chamama: Visa Kurki’s theory of legal person-
hood. Revus. Journal for Constitutional Theory 
and Philosophy of Law, 44, 103–112. https://doi.
org/10.4000/revus.6893
Tanasescu, M. (2022). Understanding the Rights 
of Nature: A Critical Introduction (1st ed., Vol. 
6). transcript Verlag. https://doi.org/10.14361/ 
9783839454312
Wenar, L. (2021). Rights. In E. N. Zalta (Ed.), The 
129
journal for constitutional theory and philosophy of law
(2023) 51
Structuring concepts of legal personhood
On legal personhood as a cluster property
Stanford Encyclopedia of Philosophy (Spring 2021). 
Metaphysics Research Lab, Stanford University. 
https://plato.stanford.edu/archives/spr2021/ 
entries/rights/
Worthington, M., & Spender, P. (2021). Constructing 
legal personhood: Corporate law’s legacy. Griffith 
Law Review, 30(3), 348–373. https://doi.org/10.1
080/10383441.2021.2003742
Zalta, E. N. (2023). Gottlob Frege. In E. N. Zalta & 
U. Nodelman (Eds.), The Stanford Encyclopedia 
of Philosophy (Summer 2023). Metaphysics Re-
search Lab, Stanford University. https://plato.stan 
ford.edu/archives/sum2023/entries/frege/

131
journal for constitutional theory and philosophy of law
(2023) 51
revus (2023) 51, 131–150
Ana Lúcia Costa*
Vulnerability, sexual violence and the law
A conceptual analysis
This article examines the concept of vulnerability and is envisioned as a reply to the work 
of Susanna Pozzolo. Although agreeing with the author on the usefulness of vulnerability 
as a heuristic instrument to make visible the flaws on the laws on sexual violence, I differ 
from her in rejecting to consider the body as a source of women’s vulnerability to sexual 
violence and in considering the non-mutually exclusive relation between vulnerability 
and autonomy relevant in what regards sexual violence. I further discuss the polysemic 
nature of vulnerability and the difficulties with the use of the concept in cases of sexual 
violence, which arise from that nature. I finish with a note of hope by considering the 
feminist political struggle over the meaning of vulnerability.
Keywords: vulnerability, sexual violence, women, autonomy, consent, female stereotypes
1 INTRODUCTION
Echoing calls for the use of vulnerability as an epistemological framework 
for understanding sexual violence against women,1 Susanna Pozzolo defends 
the idea of vulnerability as a useful heuristic instrument for its capacity to make 
visible the flaws in the laws on sexual violence.2
Such a position, however, does not underestimate the difficulties arising 
from the use of this concept. Pozzolo rightly notes the discriminatory effects 
and paternalistic measures usually associated with it.3 The solution proposed 
is twofold.4 On the one hand, the author advocates an understanding of vul-
nerability contextually rather than corporeally.5 In her view, in determining if 
someone is vulnerable, one should look at “the (cultural and/or legal) context” 
* analucia.correiadacosta@gmail.com | Margarita Salas Postdoctoral Fellow, Law and Philoso-
phy Research Group, Law Department, Pompeu Fabra University, Barcelona, Spain.
1 Bergoffen 2003, 2009, 2011; Code 2009, Gilson 2016.
2 Pozzolo 2020: 16.
3 Pozzolo 2020: 2, 9, 11.
4 Pozzolo 2020: 2.
5 As expressly stated, Pozzolo bases her proposal in Catriona Mackenzie’s classification of the 
sources of vulnerability, which she reinterprets (Pozzolo 2020: 3). As such, whereas Mack-
enzie proposes a taxonomy that distinguishes between inherent, situational, and pathogenic 
vulnerability, Pozzolo speaks of corporeal, contextual, and pathogenic vulnerability. As I see 
it, there is at least one important difference. I am referring to inherent and corporeal vul-
nerability, as inherent vulnerability refers to much more than just the body, such as human 
beings’ social, affective, and political nature which makes us “emotionally and psychological 
vulnerable”, in addition to making us vulnerable to “exploitation, manipulation, political vio-
lence, and rights abuses.” (Mackenzie et al. 2014: 1).
132
(2023) 51
journal for constitutional theory and philosophy of law
Ana Lúcia Costa
in virtue of which vulnerability arises, “instead of looking for something ‘weak’ 
in the person herself.”6 On the other side, Pozzolo highlights the importance of 
understanding vulnerability and autonomy not as mutually exclusive but rather 
as terms whose oppositional relationship is a matter of degree: although it is 
true that the more vulnerable someone is the less autonomy and freedom of 
choice she has, the mere presence of vulnerability does not determine the total 
absence of autonomy and vice versa.7
Reading Pozzolo together with Catriona Mackenzie’s work on vulnerability 
and relational autonomy (which Pozzolo expressly refers to8), I interpret the 
reasoning behind her twofold proposal as follows. First, in addition to a more 
accurate description of the reality of (gender) oppression, the concept of vul-
nerability is aimed at capturing and disconnecting the source of vulnerability 
from the vulnerable person and locating it at the level of context, society, and 
institutions, offering a suitable solution for the discriminatory effects vulner-
ability often gives rise to. Second, understanding vulnerability and autonomy 
not as mutually exclusive is aimed at avoiding paternalistic measures that as-
sume vulnerable people’s incapacity to give true consent and protect their own 
interests in virtue of their lack of autonomy.
Whereas I agree with Pozzolo’s proposal at a general level, I see problems in 
its application to sexual violence against women. Problems that have to do with 
the specific nature of the topic at hand and the specific type of state responses to 
it. And so, problems that need to be properly identified and dealt with if vulner-
ability is to constitute a “useful heuristic instrument that allows us to identify 
the flaws in the laws [on sexual violence]”.9
My argument is that those problems are inextricably related with the con-
cept of vulnerability. On the one hand, the concept needs to capture and refer to 
the relevant circumstances of sexual violence against women, the circumstances 
that make women vulnerable to this kind of aggression. On the other, it is fun-
damental to identify the different uses of the concept at work in the legal system 
that rather than protecting women, make them more vulnerable. As a result, 
the methodology adopted is conceptual analysis. Following Kenneth Himma’s 
definition of conceptual analysis, my aim is to “locate [the concept of vulner-
ability] among a general conceptual framework that [can guide] both our lin-
guistic practices regarding the relevant concept-words and our legal practices 
themselves.”10
6 Pozzolo 2020: 14, own translation.
7 Pozzolo 2020: 5.
8 Pozzolo 2020: 3, 6, 9.
9 Pozzolo 2020: 16, own translation.
10 Himma 2015: 67.
133
journal for constitutional theory and philosophy of law
(2023) 51
Vulnerability, sexual violence and the law
A conceptual analysis
My argument will unfold as follows. Section 2 addresses the relevance of 
the body for a proper understanding of women’s vulnerability in the context of 
sexual aggression. Section 3 delves into the relationship between vulnerability 
and autonomy, arguing that, contrary to other contexts, what is needed regard-
ing sexual violence is a greater focus on the oppositional relationship between 
vulnerability and autonomy. Section 4 examines the polysemic nature of vul-
nerability, claiming that such nature is what lies at the root of the discrimi-
natory potential of vulnerability in what regards the laws on sexual violence. 
Subsection 4.1 analyses the binary structure of female stereotypes, divided into 
opposing representations of “good” and “bad” women. Subsection 4.2 asserts 
the correspondence between the “good woman” stereotype and the ideal victim 
of rape, whereas subsection 4.3 contends that women whose behaviour does 
not conform with the “good woman” stereotype are framed as bad women and, 
consequently, become “bad victims”, victims whose sexual aggression is often 
denied by the judicial system. Subsection 4.4 concludes by affirming that it is 
not the idea of women’s vulnerability but rather the perception of women as ac-
tive agents, perfectly capable of resistance to sexual aggression, that leads to the 
effacement of countless cases of sexual violence by law. The paper finishes with 
a note of hope by considering, in its concluding remarks, the feminist political 
struggle over the meaning of vulnerability.
2 THE BODY
In her analysis of the “Manada” case,11 Pozzolo excludes the general differ-
ence in strength between men and women as a source of women’s vulnerability 
to sexual violence.12 As said before, this exclusion is related to an understand-
ing of vulnerability in terms of context rather than corporeality. Hence, regard-
ing the “Manada” case, the author states: “what can be seen, at a closer look, is 
situational vulnerability, that is, vulnerability caused by the particular context 
– without a necessary connection with the body (…).”13 As I interpret Pozzolo, 
this exclusion is, as previously mentioned, related with the intent of avoiding 
discriminatory effects against people considered vulnerable: “[to] qualify some-
one as vulnerable becomes potentially discriminatory due to the insistence in 
11 The case that became known as “la Manada” refers to a rape of an 18-year-old woman by five 
men on the 7th of July 2016 during the San Fermín Festival, that takes place annually in Pamplo-
na, Spain. The attackers, that included a member of the Civil Guard and another of the Span-
ish Army, made several videos of the attack, which took place in a vestibule of an apartment 
building. The case drew intensive public attention and was at the center of an exhaustive public 
debate concerning the definition of rape and its differentiation from sexual abuse. It eventually 
led to an important change in the law, that ended the distinction between both crimes.
12 Pozzolo 2020: 12.
13 Pozzolo 2020: 13, own translation.
134
(2023) 51
journal for constitutional theory and philosophy of law
Ana Lúcia Costa
anchoring vulnerability in her being – her lack of [something]”.14 In my under-
standing, however, such effects emerge precisely out of the lack of consideration 
of the corporeal element or, more accurately, the general difference in strength 
between men and women. Let me explain.
In the Manada case, the first instance court15 ruled it to be a case of sexual 
abuse and not rape.16 The reason was the lack of violence and intimidation on 
the part of the aggressors and the lack of resistance on the part of the victim 
from which (vitiated) consent17 was inferred.18 Given that this is a case of a 
sexual aggression by five men, the difference in strength between the victim 
and the aggressors seems to be of essence in the consideration of the reason and 
meaning of the lack of resistance on the part of the victim. The same can, in my 
view, be said of cases where there is only one aggressor with a superior physi-
cal strength. The probability of success when fighting someone with a superior 
strength, as well as the level of intimidation, is definitely not the same as when 
fighting someone of equal or inferior physical strength. Hence, the likeliness 
of the perception on the part of the victim that resistance is both pointless and 
dangerous. Not to consider this factor can be seen as a form of discrimination 
against women masked as gender neutrality. A preference for formal equality to 
the detriment of substantial equality.
I would like to make clear, however, that I am not defending that the only 
source of women’s vulnerability in cases of sexual aggression is the body, wom-
en’s body – neither at a concrete case level, nor at an abstract/group one. There 
is no doubt in my mind over the fact that the source needs to be located funda-
mentally at the level of the patriarchal social relations of gender and sexuality, 
and that this, therefore, is a type of situational or, more accurately, pathogenic 
14 Pozzolo 2020: 8, own translation.
15 Decision No. 000038/2018 of the Second Section of the Audiencia Provincial de Navarra.
16 A clarification concerning terminology is in order here. At the time of the ruling, Spanish law 
differentiated between the crime of sexual aggression and the crime of sexual abuse, the for-
mer but not the latter being a synonym of rape. I will use the word “rape” to refer to what the 
Spanish law used to name sexual aggression (“agresión sexual”) and use “sexual aggression” as 
a general term to refer to sexual violence, sexual attack and sexual assault, and so, as including 
both rape and sexual abuse.
17 At that moment, Spanish law differentiated between “lack of consent” and “vitiated consent”. 
Whereas lack of consent was a constitutive element of the crime of rape, vitiated consent was 
constitutive of the crime of sexual abuse. The decision under analysis defines vitiated consent 
in opposition to free consent and as synonym of consent that is “coerced or pressured by [the] 
situation.” (SAP NA 38/2018: 99, own translation).
18 At the time when the events took place, Spanish law defined rape as involving violence or 
intimidation. Even though the first instance ruling on the “Manada” case expressly states that 
the legal qualification of the facts as rape depends exclusively on the conduct of the aggressor 
and not the victim’s (SAP NA 38/2018: 98), the decision also refers to the requirement of re-
sistance on the part of the victim not having “to be desperate, but real, true, determined, con-
tinued and that unequivocally expresses a will contrary to the sexual act.” (SAP NA 38/2018: 
95, own translation).
135
journal for constitutional theory and philosophy of law
(2023) 51
Vulnerability, sexual violence and the law
A conceptual analysis
vulnerability: a subset of situational vulnerability that arises out of “morally 
dysfunctional or abusive interpersonal and social relationships and sociopoliti-
cal oppression or injustice.”19 Nonetheless, the difference in strength between a 
man and a woman cannot be discarded as irrelevant to the concrete situation 
of vulnerability – and thus diminished autonomy – a female victim is in at the 
time of a sexual aggression. This is particularly so when the law directly or in-
directly establishes the requirement of resistance on the part of the victim to 
frame the aggression as rape. Doing so constitutes, in my view, a form of patho-
genic vulnerability.
Concerning the relevance of the corporeal factor in cases of sexual aggres-
sion against women and the type of vulnerability at stake in this kind of case, 
I would like to call attention to the relationship between inherent and situa-
tional vulnerability. When proposing a taxonomy of sources of vulnerability, 
Mackenzie, Rogers, and Dodds highlight the fact that those two sources “are 
not categorically distinct”.20 As Mackenzie further elaborates:
The distinction between inherent and situational vulnerability is not categorical, as 
these two sources of vulnerability can be causally interconnected. Situational vulne-
rability can give rise to inherent vulnerability (…). And some kinds of inherent vul-
nerability will render people more liable to situational vulnerability (…). (Mackenzie 
2014: 39)
Let me apply this idea to sexual violence against women. Were it not for the 
patriarchal social relations of gender, the difference in strength between men 
and women would not, by itself, make women vulnerable to sexual aggression, 
at a general, group level. The same observation can be made in relation to con-
crete cases of sexual attack: were it not for that specific context, the difference in 
strength would not, by itself, make women vulnerable. Yet, given those relations 
and specific contexts of sexual aggression, the difference in strength between 
victim and attacker in concrete situations needs to be taken into account, as such 
difference is causally interconnected with the context in making women vulner-
able to sexual aggression and in explaining their lack of resistance to the attack.
3 VULNERABILITY AND AUTONOMY
As rightly noted by Susanna Pozzolo, the classification of a group as vulner-
able frequently motivates paternalistic measures towards those groups.21 As I 
interpret her, this is one of the reasons why the author defends that vulnerabil-
ity and autonomy should not be understood as mutually exclusive.22 Her rea-
19 Mackenzie et al. 2014: 9.
20 Mackenzie et al. 2014: 8.
21 Pozzolo 2020: 9.
22 Pozzolo 2020: 5, 6.
136
(2023) 51
journal for constitutional theory and philosophy of law
Ana Lúcia Costa
soning seems to follow Catriona Mackenzie’s, who Pozzolo quotes when refer-
ring to the relationship between vulnerability and paternalism.23 As Mackenzie 
explains,
(…) notions of vulnerability and protection can be, and historically have been, used 
to justify coercive or objectionably paternalistic social relations, policies, and instituti-
ons, which often function to compound rather than ameliorate the vulnerability of the 
persons or groups they are designed to assist. (Mackenzie 2014: 35)
In her view, the reason lies in the conceptualization of vulnerability and au-
tonomy in oppositional terms – by which Mackenzie means mutually exclu-
sive terms –24 as, in her understanding, such a conceptualization implies that 
“there is a tension between responding to human vulnerability and promoting 
autonomy”.25 I.e., in the author’s view, protecting vulnerable people from harm 
and protecting their autonomy are mutually exclusive.
Although I agree with Mackenzie and Pozzolo on a general level, I fail to see 
the relevance of this remark regarding sexual violence against women and the 
specificity of the law’s flaws in this context. As I see it, there are two important 
differences between the cases on which Pozzolo bases her reasoning and cases 
of sexual violence against women. The first is related to what is disputed, that is, 
what is at stake in these two different kinds of cases, and the second concerns 
the (problematic) measures typically applied by the state in both types of cases.
The cases brought up by Pozzolo are cases in which what is at stake is the 
capacity of an individual or group, and particularly the capacity to consent. As 
the author tells us, in those cases:
[…] vulnerability indicates precisely a lack of power – a deficiency in control. It indi-
cates, ultimately, an insufficiency at the level of autonomy, of freedom of election and 
action. More precisely, I would say that [the idea of vulnerability] indicates a reduction 
at the decisional level, because, when a person is classified as vulnerable, that means 
23 Pozzolo 2020: 9.
24 Mackenzie does not hesitate to acknowledge the oppositional relation of degree between vul-
nerability and autonomy when affirming that “some social relationships and environments 
provide hostile conditions for autonomy. Environments characterized by corrosive disadvan-
tage (social, political, economic, educational) or social relationships characterized by abuse, 
coercion, violence, or disrespect [– environments and social relationships, therefore, that cor-
respond to situations of pathogenic vulnerability –] may seriously thwart the development of 
many of the skills and competences required for self-determination or may constrain their 
exercise.” (Mackenzie 2014)»abstract»:»This chapter discusses a number of issues related to 
measuring poverty over time. It highlights some of the key normative decisions that have to 
be taken, in particular, the role of compensation over time (whether poverty spells can be 
compensated for by non-poverty spells In addition, when writing with Wendy Rogers and Su-
san Doods, Mackenzie expressly refers to the undermining of autonomy or the exacerbation 
“of the sense of powerlessness engendered by vulnerability in general” as one of pathogenic 
vulnerability’s key features (Mackenzie et al. 2014: 9).
25 Mackenzie 2014: 34.
137
journal for constitutional theory and philosophy of law
(2023) 51
Vulnerability, sexual violence and the law
A conceptual analysis
that she has a characteristic that situates her at an inferior level than the average [hu-
man being]. (Pozzolo 2020: 3-4, own translation)
Pozzolo is not alone. The literature on vulnerability largely emerged from 
contexts in which the capacity to consent is at stake. Bioethics and disabilities 
studies are probably the best examples,26 the latter being expressly referred to by 
Pozzolo.27 There the idea of vulnerability was meant to signal a special obliga-
tion from the state to protect those who are not able to make choices on their 
own, thus requiring a special state intervention aimed at protecting them from 
harm – a harm caused by the choices vulnerable people supposedly have no 
capacity to make.28
But the capacity to consent is not at all what is usually at stake in cases of 
female rape. In cases where a woman claims to have been raped, what is often 
put into question is not whether or not she has the capacity to consent, to make 
a choice, but rather consent itself, that is, the content of her choice: if she indeed 
agreed to the sexual interaction or not. This is important because of the type of 
problems one encounters in relation to state intervention in cases of rape: the 
issue here is not the paternalistic measures through which the state impedes 
the vulnerable person to make choices or invalidate them, but rather the ample 
denial that there was a lack of consent. Therefore, in cases of sexual violence the 
problem is not paternalism, but exactly the opposite: a lack of protection.
Now, if the understanding of vulnerability and autonomy not as mutually 
exclusive but rather in terms of an inverse relationship of degree is an impor-
tant instrument against paternalism, and paternalism is not the problem in 
cases of sexual violence, then, although true and generally very important, 
such an understanding does not illuminate or contribute to the correction of 
law’s flaws in these types of cases. In fact, in cases of sexual violence what is 
needed, instead, is a greater focus on the oppositional relationship between 
vulnerability and autonomy, in addition to a greater awareness that cases of 
sexual aggression are frequently ones where the degree of vulnerability is clos-
er to its highest and, thus, the degree of autonomy closer to its lowest. Only 
this focus and awareness can help law enforcers to reconceptualize the lack of 
resistance on the part of a victim of sexual aggression: not as consent, as is too 
often the case, but rather as a lack thereof.
Let me apply this reasoning to the Manada case. The case was not framed 
as rape because of the absence of violence on the part of the aggressors. Such a 
requirement translates into one of resistance on the part of the victim – since 
often violence is exerted as an instrument against the resistance of the victim 
to a sexual interaction. This latter requirement assumes that resistance was a 
26 Mackenzie et al, 2014:2.
27 Pozzolo 2020: 21-23, notes 5, 6, 7, 12.
28 Rogers 2014: 64-66.
138
(2023) 51
journal for constitutional theory and philosophy of law
Ana Lúcia Costa
possibility, disregarding the situation of reduced alternatives the victim had at 
the time of the attack. It disregards, therefore, the victim’s situation of extreme 
vulnerability – vulnerability defined as the condition of “increased risk of harm 
or […] reduced capacity or power to protect one’s interests” –29 and its oppo-
sitional relationship with autonomy. Due to the difference in strength between 
the attackers and the victim, and so the possible risk to her physical integrity 
and life, resistance was not a real possibility, and so, from its absence, consent 
cannot be inferred.
Summarising my position on the topic, I would like to say that I am in total 
agreement with Pozzolo when she says that the legal requirement of a violent 
reaction on the part of the victim for the typification of the facts of the Manada 
case as rape configures pathogenic vulnerability: “here it is obvious that it is the 
law that makes women vulnerable.”30 In my view, however, the source of that 
specific cause of vulnerability is not only the fact that such requirement has a 
masculine character disguised as neutral.31 My suggestion is that, in addition, 
(1) this requirement is discriminatory because it ignores the frequent difference 
in strength between victim and attacker(s), (2) equals a demand that victims 
endure further danger to their physical integrity and life, and/or (3) translates 
into a lack of recognition that in cases such as this, because of that danger, re-
sistance is not a real possibility for the victim. It ultimately amounts to a lack 
of acknowledgement of the situation the victim is in with a constrained set of 
options, as it does not shelter the oppositional view on the relationship between 
vulnerability and autonomy, thus, not recognizing the lack of autonomy that 
such a situation of extreme vulnerability implies.
4 THE POLYSEMIC NATURE OF VULNERABILITY
While defending vulnerability as a useful heuristic instrument to make the 
flaws in the laws on sexual violence visible,32 Susanna Pozzolo does not un-
derestimate its discriminatory effects. Nonetheless, in my understanding, she 
does not fully account for the discriminatory potential of the concept, thus 
missing out on some important reasons for the law’s overall lack of protection 
of women in this context. The aim of the present section is to elaborate on 
some fundamental challenges the idea of vulnerability poses in the context of 
sexual violence against women. It is my belief that a proper understanding of 
those challenges might illuminate and neutralize the discriminatory potential 
of this concept.
29 Mackenzie 2014: 33-34.
30 Pozzolo 2020: 15.
31 Pozzolo 2020: 15.
32 Pozzolo 2020: 16.
139
journal for constitutional theory and philosophy of law
(2023) 51
Vulnerability, sexual violence and the law
A conceptual analysis
While Pozzolo associates the discriminatory potential of vulnerability with 
the idea that the source of vulnerability is a woman’s body, I believe such potential 
is grounded in vulnerability’s polysemy. In understanding this argument, Erinn 
Gilson provides us with a good starting point: “a particular understanding of 
vulnerability dominates the sociocultural imaginary of the industrialised, capi-
talist Western parts of the world”.33 Such understanding associates vulnerabil-
ity with “weakness, dependency, incapacity, incapability, and powerlessness”,34 
and, in turn, all those ideas with women. The result is the femininisation of 
vulnerability: vulnerability is a “feminised concept”.35 From a feminist point of 
view, the problem with this is the “vulnerabilisation” of women or, more ac-
curately, the invocation and perpetuation of a female stereotype that represents 
women’s nature as vulnerable in a specific sense. A stereotype, which, in turn, is 
strongly associated with women’s discrimination and repression.
This idea strongly resonates with the debate on victimhood and agency 
fought over in the 1980s and 1990s within feminism. At the time, it was argued 
that the dominant feminist focus on sexual violence against women depicted 
“women as shaped by pervasive male sexual coercion [and that] was to tell a 
partial, and potentially injurious, story.”36 As Kathryn Abrams explains, the fo-
cus on sexual violence was said to assign women a “victim status” that encour-
aged a “wounded passivity on the part of women and a repressive regulatory 
urge on the part of state authorities.”37 Instead, those critical voices claimed for 
a greater emphasis on women’s agency, assuming a mutually exclusive relation-
ship between victimhood and agency.
I would like to bring the idea that depicting women as victims is potentially 
injurious into the debate on the dangers of using vulnerability as a framework 
through which to analyse rape and complement it with the suggestion that vul-
nerability mobilises female stereotypes. Here is my proposal: female stereotypes 
lie at the heart of women’s vulnerability to sexual violence, and that is the case 
in at least two senses. First, they are directly connected with the acts of sexual 
violence perpetrated by men against women; and second, they are central to the 
law’s lack of protection of women in cases of rape. My focus here is on the latter 
connection. However, going into the specifics of the law’s flaws requires a previ-
ous digression into the content of women’s stereotypes.
33 Gilson 2016: 74.
34 Gilson 2016: 74.
35 Gilson 2016: 71.
36 Abrams 1995: 305.
37 Abrams 1995: 305.
140
(2023) 51
journal for constitutional theory and philosophy of law
Ana Lúcia Costa
4.1 Female stereotypes
Stereotypes about women have historically been organized around a binary 
opposition, which divides women into two types. There are different designa-
tions for them: the good and the bad woman, the virtuous and the fallen, and 
the Madonna and the whore, just to mention some of the most popular ones. 
In my view, this dualism is based on three criteria, the first being sexuality, the 
second being morality, and the third being what some call potency or power. 
Accordingly, women have been perceived as either 1) desexualized, good, and 
powerless, or 2) sexualized, evil, and powerful.
The extensive work that has been done over the last decades on stereotypes 
might help us to better understand this. Stereotypes seem to work in the same way 
our perception of other people does. We perceive people along two dimensions. 
Some call it intellectual and social,38 others competence and warmth,39 and fur-
ther there are those that refer to it as power/potency and morality.40 Concerning 
stereotypes of out-groups, Susan Fiske and her colleagues have shown a preva-
lent inverse relationship between those two dimensions: where there is a positive 
regard of one of the dimensions, there is a negative assessment of the other.41 In 
terms of gender, this is exactly what happens. But when it comes to women, a 
further dimension enters the equation: sexuality.42 And when it comes to sexu-
ality, women are again divided into opposing types, some with favourable con-
notations and others with highly negative ones.43 It is important to note how the 
morality and sexuality dimensions seem to vary in the same direction: a sexually 
virtuous woman is usually perceived as morally virtuous too.44
Allow me to make an attempt to synthesise the content of the good and the 
bad woman stereotypes. The good woman is chaste. She is sexually innocent, 
pure, and passive.45 The bad woman, instead, is sexually active: a seductor, a 
provocateur, and a temptress.46 But, as previously said, the opposition in ste-
reotypes of women is not restricted to their sexuality – even if, at least in the 
case of rape, sexuality can be seen as a starting point in the framing of women 
into one or the other stereotype. Morality is also of essence in such opposition. 
Women are, thus, divided into angels and devils, Marys and Eves. Historically, 
it can be observed as, from modernity onwards, the figure of the wife/mother 
38 Rosenberg et al. 1968.
39 Fiske et al. 2002.
40 Jasper et al. 2020.
41 Fiske et al. 2002.
42 Clifton et al. 1976; Noseworthy & Loth 1984; Six & Eckes 1991.
43 Six & Eckes 1991; Vonk & Ashmore 2003.
44 For this reason Dewall et al. (2005) propose to unite morality and sexuality into one single 
category which they designate as virtue.
45 Cott 1978; Trudgill 1976: 56-57; Welter, 1966: 154-158.
46 Higgins 1976.
141
journal for constitutional theory and philosophy of law
(2023) 51
Vulnerability, sexual violence and the law
A conceptual analysis
has epitomised the former side of this dichotomy, with married women being 
represented and perceived as kind and completely selfless, entirely devoted to 
their children and their husbands, whom they often saved from immorality.47 
They were supposedly connected with God, religion, and charity.48 In opposi-
tion, promiscuous women were perceived as endowed with evilness and selfish-
ness, capable of deceit, revenge, and whatever necessary to achieve their aims.49 
This brings us to the final element in the opposition between the good and the 
bad woman stereotypes: what I have referred to as power/potency. Whereas 
bad women have been associated with all the necessary capacities for evilness,50 
good women have often been perceived as naïve, incompetent, and even idiots, 
uncapable, therefore, of protecting themselves or their interests.51
The idea of vulnerability is strongly associated with this latter characteristic 
of the good woman. But not exclusively so. All the three elements in the di-
chotomy contribute to an opposing placement of the good and the bad woman 
in terms of vulnerability to danger: whereas innocent and respectable women 
are seen as vulnerable to sexual attack and ill-intended male seduction, promis-
cuous women are depicted as active seductresses who bring destruction and 
all kinds of harms upon themselves and everyone that surrounds them. Bad 
women, are thus, not perceived as vulnerable to danger but rather as a source 
of danger. Glick and Fiske, for instance, have shown how sexually “bad” women 
are perceived as threatening to men because of the idea that they can use men’s 
attraction to them to manipulate them.52 Such an idea is of essence in the con-
text of rape and the judicial system’s failure to acknowledge and punish it. The 
idea of the “ideal victim” might help us to see why.
4.2 The ideal victim
Nils Christie has famously defined the ideal victim as “a person or a category 
of individuals who – when hit by crime – most readily are given the complete 
and legitimate status of being a victim.”53 According to the author, being framed 
as an ideal victim requires that the victim, the circumstances of the crime, and 
the offender be endowed with certain attributes: 1) the victim must be weak, 2) 
she must have been carrying out a respectable project at the time of the crime, 
3) she must have been in a place where she could not possibly be blamed for 
being, 4) the offender must be big and bad, and also 5) unknown and in no 
47 Trudgill 1976: 70-71, 76-77, 78-80.
48 Welter 1966: 152-154.
49 Higgins 1976; Lombroso & Ferrero 2004: 183.
50 Lombroso & Ferrero 2004: 83.
51 Trudgill 1976: 66.
52 Glick & Fiske 1996: 509.
53 Christie 1986: 18.
142
(2023) 51
journal for constitutional theory and philosophy of law
Ana Lúcia Costa
personal relationship to the victim.54 There is also a sixth attribute that I will 
keep for now in order to make the following point: in relation to sexual violence 
against women, the ideal victim is the good (vulnerable) woman. Christie con-
firms this. According to him, in rape cases, the ideal scenario:
is the young virgin on her way home from visiting sick relatives, severely beaten or 
threatened before she gives in. … [She] is weak compared to the unrelated offender, as 
well as having put a reasonable energy into protecting herself … against becoming a 
victim. (Christie 1986: 19)
As it can be noted, then, the ideal victim of rape needs to be more than weak: 
she needs to be chaste. The ideal victim therefore, is sexually innocent, pure, and 
passive. This means that, in the context of rape, the second and third character-
istics of the ideal victim are translated into the requirements that, at the time of 
the crime, the victim is not involved in any potentially sex-related activity or at 
a place prone to or associated with such types of activities. In my view, however, 
chastity adds further conditions to the ideal victim of rape, conditions that go 
beyond the time and the place of the crime. What is required is not only a chaste 
behaviour at the time of the crime, but instead a chaste character. This appears to 
be confirmed by Christie’s examples of the non-ideal victim in rape cases.
So, who is the non-ideal victim? Christie refers to “the experienced lady on 
her way home from a restaurant, not to talk about the prostitute who attempts 
to activate the police in a rape case.”55 I would briefly translate it as the bad 
woman: the non-ideal victim is framed within the stereotype of the bad woman. 
In fact, due to the binary character of the stereotypes of the good and bad wom-
en, it can succinctly be said that anyone who does not completely fit the good 
woman stereotype is framed as a bad woman.
After this brief digression into the content of female stereotypes and the 
concept of the ideal victim of rape, I will now come back to the idea that moti-
vated it: the dangers of the use of vulnerability in relation to rape.
4.3 Bad women, non-ideal victims, and the denial of rape
The problem with vulnerability is that it has always been associated with 
a particular stereotype of woman – the good woman –, which, in turn, corre-
sponds to the ideal victim of rape. What is problematic about this is that, in most 
cases, real victims do not match the stereotype of the good (vulnerable) woman 
with regard to chastity. As a result, they are framed as “bad women” and as “non-
ideal victims”, and this has disastrous consequences in terms of victim blaming. 
Bad women are generally perceived as responsible for the rape, and that is far 
from being uncommon. In fact, victim blaming is a huge problem with regard 
to complaints of sexual aggression: “[u]nlike many other interpersonal crimes 
54 Christie 1986: 19.
55 Christie 1986: 19.
143
journal for constitutional theory and philosophy of law
(2023) 51
Vulnerability, sexual violence and the law
A conceptual analysis
such as robberies or muggings, victims of sexual assault are particularly vul-
nerable to being blamed for their attack”.56 As is extensively documented,57 the 
main reason for victim blaming is the violation of the good/vulnerable woman 
stereotype, and, hence, the framing of the victim as the bad woman.
A brief look into the factors that influence victim blaming in rape cases 
should be enough to prove this point. The very first is “social respectability,” or 
the victim’s sexual behaviour.58 As a result:
both legal practitioners and laypersons attribute blame to rape victims on the basis of 
extralegal factors such as clothing (Johnson, 1995; Vali & Rizzo, 1991), alcohol con-
sumption (Corcoran & Thomas, 1991; Scronce & Corcoran, 1995), and whether the 
victim has had multiple sex partners in the past (Marx & Gross, 1995). (Abrams et al. 
2003: 113.)
Another key example of the importance of sexual respectability on the at-
tribution of blame to the victim is the higher rate of victim blaming in acquaint-
ance rape. As Abrams et al. tell us,59 the reason for this is that acquaintance rape 
victims are perceived “as having behaved in a manner that is inappropriate for a 
woman”. Behind this idea lies the stereotype of the good woman and its binary 
character. Someone who does not behave as a good woman is framed as a bad 
woman. An important characteristic of the binary opposition between good 
and bad women is that the first is perceived as the norm in its descriptive and 
normative senses – how women normally act and how they should act – and 
the second as a deviation from it. “[A] woman who invites a relationship with 
a man” is violating that norm and is, therefore, perceived as “responsible for 
anything unfortunate that may happen to her.”60
It should be noted, however, that what is at stake is not merely a matter of 
victim blaming, at least not explicitly. Adherence to female stereotypes leads to 
more restrictive rape definitions, which translates into the denial of many actual 
rapes.61 The central piece here is, of course, consent. Martha Burt and Rochelle 
Albin make this connection clear:
women with certain reputations and identities … are stereotypically assumed to con-
sent more readily, to more men, in more situations. Having assumed a generalized 
propensity to consent and attached it to whole classes of women, this line of reaso-
ning then particularizes the argument to this woman (victim) in this situation (alleged 
rape) and infers consent to this man (alleged assailant). Therefore, following this rea-
soning, this situation is not a rape. (Burt & Albin 1981: 214)
56 Gravelin, Biernat, & Bucher 2019: 2.
57 Pedersen & Strömwall 2013; Persson et al. 2018; Viki & Abrams 2002; Yamawaki et al. 2007.
58 Krahé 1991: 282.
59 Abrams et al. 2003: 121.
60 Abrams et al. 2003: 121.
61 Burt & Albin 1981: 213.
144
(2023) 51
journal for constitutional theory and philosophy of law
Ana Lúcia Costa
Accordingly, my point is that unchaste sexual behaviour leads to the framing 
of the victim as a bad woman and such framing, in turn, translates not only into 
the idea that the victim is the one to blame for the assault, but further into the 
idea that there was no assault at all. The reason is that bad women are perceived 
as always “ready to go”, which is seen as incompatible with the absence of con-
sent. The result is the denial of rape at all.
To this, I would like to add a last element in the stereotype of the bad wom-
an, which I believe is at work in the pervasive denial of rape by the judicial 
system. It concerns the widespread suspicion that victims are lying about the 
occurrence of a sexual aggression. Such suspicion, I propose, is related to the 
bad woman stereotype, which depicts women as evil and vengeful, and which 
relates to the widespread rape myth that “women ‘cry rape’ only when they have 
been jilted or have something to cover up”62 and that women are prone to mak-
ing false allegations out of spite, revenge, or fantasy.
4.4 Vulnerability and the law
At the start of Section 4 I defended that the discriminatory potential of vul-
nerability goes beyond the idea that the source of women’s vulnerability to sexual 
violence is our body, an idea I interpret Susanna Pozzolo’s argument to presume. 
I pointed out the polysemic nature of vulnerability to call attention to a particu-
larly dominant understanding of vulnerability that associates it with “weakness, 
dependency, incapacity, incapability, and powerlessness” and, in turn, all those 
ideas with women.63 I then suggested that understanding mobilises female ste-
reotypes that lie at the heart of the law’s failure in acknowledging and punishing 
sexual aggression against women. I was referring to the dichotomy of the good/
bad woman which I then attempted to show is incorporated by laws and judi-
cial practices on sexual violence: the good woman becomes the norm with which 
every victim needs to comply in order to become an “ideal victim” and to have the 
sexual violence against her acknowledged and punished by the system. As a re-
sult, a lack of conformity on the part of victim leads to her framing as bad woman 
and the consequent unrecognition of the sexual assault by the judicial system.
As also stated before, vulnerability seems to constitute a new chapter in the 
feminist debate on victimhood and agency that started in the 1980s and 1990s. 
As Rebecca Stringer has summarised, many have accused feminist rape law re-
form efforts at reinscribing “patriarchal constructions of femininity as embod-
ied vulnerability, perpetuating a sexist linking of femininity with victimhood 
rather than agency.”64 At first sight, there seems to be something right with this 
idea. By posing the good woman stereotype as the norm with which victims of 
62 Burt 1980: 217.
63 Gilson 2016: 74.
64 Stringer 2013: 148-149.
145
journal for constitutional theory and philosophy of law
(2023) 51
Vulnerability, sexual violence and the law
A conceptual analysis
sexual violence need to conform with, law does seem to adopt a language that, 
as Sharon Marcus puts it, “solicits women to position ourselves as endangered, 
violable, and fearful”65 and excludes “women’s will, agency, and capacity for 
violence”.66 If that would be true, then, the feminist use of vulnerability in the 
context of sexual violence - even if with a very different meaning – would likely 
serve to reinforce and perpetuate patriarchal constructions of femininity that 
hinder the law’s capacity to acknowledge and punish rape, due to the polysemic 
nature of vulnerability.
Yet, at a closer look, what can be observed is exactly the opposite. As Stringer 
notes, “rape law typically figures femininity not as embodied vulnerability but 
as responsible agency”.67 Rather than an expectation and demand of women’s 
passivity, “modern rape law has typically represented women as capable of re-
sisting rape, invoking images of women as agents in order to deny that sexual 
victimization has taken place.”68 In other words, the resistance requirement is 
evidence of the law’s perception of women as active agents,69 and this – and not 
the idea of women’s vulnerability – is precisely what leads to the effacement of 
countless cases of sexual violence.
Furthermore, there is a second way in which the agency of women is pre-
sumed in cases of sexual violence. Stringer refers to it as “bad agency”. As she 
explains, “[i]n addition to the good agency ascribed to the ideal victim by the 
resistance requirement, there is the bad agency ascribed”70 to non-ideal victims. 
Here I will slightly depart from Stringer’s view to suggest that “bad agency” is 
ascribed to victims of sexual violence framed as “bad women”. It is not vulner-
ability but agency that stands at the core of victim blaming. Moreover, it is spe-
cifically sexual agency that lies behind the stereotypical assumption of consent 
by women with certain (unchaste) reputations and identities. Finally, it is agen-
cy once again and in no way vulnerability that grounds the commonplace idea 
that women are prone to make false allegations out of spite, revenge, or fantasy.
As a result, one might conclude that if the legal system is too often “a noto-
rious site of incredulity” 71 – to use Lorraine Code’s brilliant expression – with 
regards to women’s claims of sexual violence, this is certainly not related with 
vulnerability or, more accurately, a construction of femininity as vulnerable in 
the sense of weakness, passivity, incapability, and powerlessness. Since, when 
understood in this sense, vulnerability is too often imagined as opposite to 
agency, it is possible to say that, in fact, just the contrary is true. What lies at the 
65 Marcus 2013: 390.
66 Marcus 2013: 393.
67 Stringer 2013: 149.
68 Stringer 2013: 149.
69 Stringer 2013: 153.
70 Stringer 2013: 161.
71 Code 2009: 336.
146
(2023) 51
journal for constitutional theory and philosophy of law
Ana Lúcia Costa
heart of the law’s flaws regarding sexual violence against women is the demand 
and assumption of our agency.
5 CONCLUSION
Vulnerability is a challenging concept. Its polysemic nature presents us with 
important difficulties arising from its “ambiguities, complexities, and tensions”.72 
Understood in one sense, vulnerability captures the condition of “contingent 
susceptibility of particular persons or groups to specific kinds of harm or threat 
by others”,73 thus, helping to shed light on the state’s special moral and justice 
obligations towards those considered vulnerable. Understood in another sense, 
vulnerability mobilises female stereotypes that are at the origin of the state’s fail-
ure in acknowledging and punishing sexual violence against women. As such, 
vulnerability is both potentially apt to illuminate women’s sexual oppression 
and perpetuate such a condition, both resistant and complicit, both emancipa-
tory and oppressive.
Given such ambiguity, doubts arise concerning the usefulness of vulnerabil-
ity as an epistemological framework for understanding sexual violence against 
women as endemic to male-female relations. Despite its vexing nature, I side with 
those who defend vulnerability as a lens through which to (re)conceptualize rape 
and sexual aggression more broadly. As Susanna Pozzolo, I too argue in favour 
of vulnerability’s value as a heuristic instrument capable of rendering intelligible 
the law’s flaws in this regard. The reason is my firm commitment to the belief in 
the transformative power of language: “(…) behind every word there is a history, 
just as behind every history there is a struggle to fix or change the meaning of 
words.”74 As Joan Scott tells us, “words, like the ideas and things they are meant to 
signify, have a history.”75 That history is political, as the meaning of words shapes 
and is shaped by power. For all this, language is a privileged locus of political ac-
tion and resistance.76 I would like to bring this insight concerning the changing 
and political nature of language to the present reflection on vulnerability.
Nils Christie added to the already mentioned five conditions to be an ideal 
victim, a sixth one: be “powerful enough to make your case known and suc-
cessfully claim the status of an ideal victim.”77 He arrived at this condition pre-
cisely by looking at the feminist fights and achievements on rape. Writing in the 
1980s, he argued:
72 Gilson 2016: 73.
73 Mackenzie et al. 2014: 6.
74 Preciado 2009: 14, own translation.
75 Scott 1986: 1035.
76 Preciado 2009: 16.
77 Christie 1986: 21.
147
journal for constitutional theory and philosophy of law
(2023) 51
Vulnerability, sexual violence and the law
A conceptual analysis
Wives are not “ideal victims.” Not yet. But they are approaching that status. They are 
more ideal today than yesterday. (…) They are also closer to a position where they can 
claim that their definition of the situation is the valid one. They can make the political 
claim of being real victims. (Christie 1986: 20)
When Christie talked of a position of power, he was exclusively referring to 
material conditions. However, in achieving a position of power, in being able to 
claim to be a real victim, the struggle over words and their meaning is of abso-
lute essence. As Linda Alcoff puts it, language is “part of what gives people ideas 
of what they can do and of what they have just experienced.”78 Marital rape is 
a prime example of this, as before feminists started talking about rape in mar-
riage, not only was this rarely acknowledged by law, but the victims themselves 
did not recognize either the rape or their own victimization.79 It was by naming 
marital rape that both marriage and rape were redefined. Such a redefinition 
occurred within a wider process of meaning mutability in which new ideas be-
came possible and others came to be contested. What made this process pos-
sible was the political struggle over the meaning of words.
In more general terms, Rebecca Stringer refers to “feminist efforts to reform 
rape law” as a “political work that endeavours to counter the linguistic, cultur-
al, and legal effacement of particular forms of suffering, through the invention 
of new idioms that give suffering visibility.”80 Not only do I completely agree 
with her as, for me, her words shed light on what is often a vexed point in the 
conceptualisation of vulnerability: its relation to resistance. As Judith Butler, 
Zeynep Gambetti, and Leticia Sabsay point out, several popular and theoreti-
cal discourses assume vulnerability and resistance to be mutually opposite.81 
Yet, as the feminist work on sexual violence has been witness to vulnerability, 
and more specifically the consciousness of group vulnerability to certain harms, 
it “is one of the conditions of the very possibility of resistance”.82 The political 
struggle over the meaning of words is a crucial mode of resistance that both 
emerges out of and allows for the visibility of a group’s vulnerability.
As I see it, the feminist struggle over the meaning of the word vulnerabil-
ity in relation to sexual violence against women is on the right path. Both the 
theoretical discussions on the matter and the public mobilisations against deci-
sions and trials such as the “Manada” case are the best evidence of just that, as is 
the final decision on this case.83 Such a struggle, however, cannot remain blind 
78 Alcoff 2018: 3.
79 For a detailed account of the impact of the feminist movement on the marital rape exemption 
in US Law, see Ryan 1995.
80 Stringer 2013: 148.
81 Butler et al. 2016: 1.
82 Butler et al. 2016: 1.
83 Spain’s Supreme Court found the five men involved in the case guilty of gang rape, overruling 
previous decisions that had convicted them of the offence of sexual abuse. In its ruling, the 
148
(2023) 51
journal for constitutional theory and philosophy of law
Ana Lúcia Costa
to the semantic difficulties vulnerability is associated with. It needs to be very 
aware of them, to understand their workings, and to be ready to identify them 
when they are at stake both in concrete cases and general doctrines. For the 
rest, the most important change has already started to happen. As Paul Beatriz 
Preciado tells us in relation to the transformation in the meaning of the word 
queer, what changed was the uttering subject.84 In relation to vulnerability and 
sexual violence against women, that subject is no longer (exclusively) the sexist 
male judicial operator, but feminists who are transforming the meaning of vul-
nerability and, with it, its role: from an instrument of social oppression to one of 
political resistance and social transformation.
court considered that there was a genuine scenario of intimidation in which the victim never 
consented to the sexual acts performed by the accused (STS 344/2019).
84 Preciado 2009: 16, own translation.
Bibliography:
Abrams, D., Viki, G. T., Masser, B., & Bohner, G. 
(2003). Perceptions of Stranger and Acquaintance 
Rape: The Role of Benevolent and Hostile Sexism 
in Victim Blame and Rape Proclivity. Journal of 
Personality and Social Psychology, 84(1), 111–
123. https://doi.org/10.1037/0022-3514.84.1.111
Abrams, K. (1995). Sex Wars Redux: Agency and 
Coercion in Feminist Legal Theory. Columbia 
Law Review, 95(2), 304–376. https://doi.org/10.23 
07/1123232
Alcoff, L. M. (2018). Rape and Resistance: 
Understanding the Complexities of Sexual Viola-
tion. Polity Press. ISBN 9780745691916
Bergoffen, D. (2003). February 22, 2001: Toward a 
Politics of the Vulnerable Body. Hypatia, 18(1), 
116–134. https://doi.org/10.1111/j.1527-2001.20 
03.tb00782.x
Bergoffen, D. (2009). Exploiting the Dignity of the 
Vulnerable Body: Rape as a Weapon of War. 
Philosophical Papers, 38(3), 307–325. https://doi.
org/10.1163/9789004259782_007
Bergoffen, D. (2011). Contesting the Politics of 
Genocidal Rape: Affirming the Dignity of the 
Vulnerable Body. Routledge. https://doi.org/ 
10.4324/9780203340134
Burt, M. R., & Albin, R. S. (1981). Rape Myths, 
Rape Definitions, and Probability of Conviction. 
Journal of Applied Social Psychology, 11(3), 212–
230. https://doi.org/10.1111/j.1559-1816.1981.
tb00739.x
Burt R., M. (1980). Cultural Myths and Supports for 
Rape. Journal of Personality and Social Psychology, 
38(2), 217–230. https://doi.org/10.1037/0022-35 
14.38.2.217
Butler, J., Gambetti, Z., & Sabsay, L. (2016). 
Introduction. In J. Butler, Z. Gambetti, & L. 
Sabsay (Eds.), Vulnerability in Resistance (pp. 
1–11). Duke University Press. https://doi.org/ 
10.2307/j.ctv11vc78r.5 9780822362791
Christie, N. (1986). The Ideal Victim. In E. A. 
Fattah (Ed.), From Crime Policy to Victim Policy: 
Reorienting the Justice System (pp. 17–31). 
Macmillan Press. https://doi.org/10.1007/978-1-
349-08305-3
Clifton, A. K., McGrath, D., & Wick, B. (1976). 
Stereotypes of Woman: A Single Category? Sex 
Roles, 2(2), 135-148. https://doi.org/10.1007/BF00 
287246
Code, L. (2009). A New Epistemology of Rape? 
Philosophical Papers, 38(3), 327-345. https://doi.
org/10.1080/05568640903420897
Cott, N. F. (1978). Passionless: An Interpretation 
of Victorian Sexual Ideology, 1790-1850. Signs, 
4(2), 219–236. https://doi.org/10.1086/493603
Fiske, S. T., Cuddy, A. J. C., Glick, P., & Xu, J. (2002). 
A Model of (Often Mixed) Stereotype Content: 
Competence and Warmth Respectively Follow 
From Perceived Status and Competition. Journal 
of Personality and Social Psychology, 82(6), 878-
902. https://doi.org/10.1037/0022-3514.82.6.878
Gilson, E. C. (2016). Vulnerability and Victimization: 
Rethinking Key Concepts in Feminist Discourses 
on Sexual Violence. Signs, 42(1), 71–98. https://
doi.org/10.1086/686753
Glick, P., & Fiske, S. T. (1996). The Ambivalent 
Sexism Inventory: Differentiating Hostile and 
Benevolent Sexism. Journal of Personality and 
Social Psychology, 70(3), 491-512. https://doi.
org/10.1037/0022-3514.70.3.491
Gravelin, C. R., Biernat, M., & Bucher, C. E. (2019). 
Blaming the Victim of Acquaintance Rape: 
Individual, Situational, and Sociocultural factors. 
In Frontiers in Psychology, 9 (Jan), 1–22). https://
doi.org/10.3389/fpsyg.2018.02422
149
journal for constitutional theory and philosophy of law
(2023) 51
Vulnerability, sexual violence and the law
A conceptual analysis
Higgins, J. M. (1976). The Myth of Eve: The 
Temptress. Journal of the American Academy of 
Religion, 44(4), 639–647. https://doi.org/10.1093/
jaarel/XLIV.4.639
Himma, K. E. (2015). Conceptual Jurisprudence: 
An Introduction to Conceptual Analysis and 
Methodology in Legal Theory. Revus, 26, 65-92. 
https://doi.org/10.4000/revus.3351
Jasper, J. M., Young, M. P., & Zuern, E. (2020). Public 
Characters. Oxford University Press. https://doi.
org/10.1093/oso/9780190050047.001.0001
Krahé, B. (1991). Social Psychological Issues in 
the Study of Rape. European Review of Social 
Psychology, 2(1), 279–309. https://doi.org/10.10 
80/14792779143000097
Lombroso, C., & Ferrero, G. (2004). Criminal 
Woman, the Prostitute, and the Normal Woman. 
Duke University Press. https://doi.org/10.1215/ 
9780822385592
Mackenzie, C. (2014). The Importance of Relational 
Autonomy and Capabilities for an Ethics of 
Vulnerability. In C. Mackenzie, W. Rogers, & S. 
Dodds (Eds.), Vulnerability: New Essays in Ethics 
and and Feminist Philosophy (pp. 33–59). Oxford 
University Press. https://doi.org/10.1093/acprof:
oso/9780199316649.003.0002
Mackenzie, C., Rogers, W., & Dodds, S. (2014). 
Introduction: What Is Vulnerability, and 
Why Does It Matter for Moral Theory? In C. 
Mackenzie, W. Rogers, & S. Dodds (Eds.), 
Vulnerability: New Essays in Ethics and Feminist 
Philosophy (pp. 1–29). Oxford University Press. 
https://doi.org/10.1093/acprof:oso/97801993166 
49.003.0001
Marcus, S. (2013). Fighting Bodies, Fighting Words: 
A Theory and Politics of Rape Prevention. In J. 
Butler & J. Scott (eds.), Feminists Theorize the 
Political (pp. 385-204). Routledge. https://doi.
org/10.4324/9780203723999
Nathan Dewall, C., William Altermatt, T., & 
Thompson, H. (2005). Understanding the 
Structure of Stereotypes of Women: Virtue and 
Agency as Dimensions Distinguishing Female 
Subgroups. Psychology of Women Quarterly, 
29(4), 396–405. https://doi.org/10.1111/j.1471-64 
02.2005.00239.x
Noseworthy, C. M.; Loth, A. J. (1984). The Cognitive 
Organization of Gender Stereotypes. Personality 
and Social Psychology Bulletin, 10(3), 474–481. 
https://doi.org/10.1177/0146167284103016
Pedersen, S. H., & Strömwall, L. A. (2013). Victim 
Blame, Sexism and Just-World Beliefs: A Cross-
Cultural Comparison. Psychiatry, Psychology and 
Law, 20(6), 932-941. https://doi.org/10.1080/132
18719.2013.770715
Persson, S., Dhingra, K., & Grogan, S. (2018). 
Attributions of Victim Blame in Stranger and 
Acquaintance Rape: A Quantitative Study. 
Journal of Clinical Nursing, 27(13–14), 2640–
2649. https://doi.org/10.1111/jocn.14351
Pozzolo, S. (2020). ¿Vulnerabilidad Personal o 
Contextual? Aproximaciones al Análisis de 
Derecho en Perspectiva de Género. Isonomía 
- Revista de Teoría y Filosofía Del Derecho, 51, 
1–28. https://doi.org/10.5347/isonomia.v0i51.226
Preciado, P. B. (2009). Historia de una Palabra: 
Queer. Parole de Queer, 1, 14–17. Available at htt-
ps://es.scribd.com/fullscreen/79992238?access_
key=key-2l64jqncgcgodxmcd3jr
Rogers, W. (2014). Vulnerability and Bioethics. In 
C. Mackenzie, W. Rogers, & S. Dodds (Eds.), 
Vulnerability: New Essays in Ethics and Feminist 
Philosophy (pp. 60–87). Oxford University Press. 
https://doi.org/10.1093/acprof:oso/978019931 
6649.003.0003
Rosenberg, S., Nelson, C., & Vivekananthan, P.rogers 
S. (1968). A Multidimension Approach to the 
Structure of Personality Impressions. Journal of 
Personality and Social Psychology, 9(4), 283-294. 
https://doi.org/10.1037/h0026086
Ryan, R. M. (1995). The Sex Right: A Legal 
History of the Marital Rape Exemption. Law 
& Social Inquiry, 20(4), 941–1001. https://doi.
org/10.1111/j.1747-4469.1995.tb00697
Scott, J. W. (1986). Gender : A Useful Category of 
Historical Analysis. The American Historical 
Review, 91(5), 1053–1075. https://doi.org/10.230 
7/1864376
Six, B., & Eckes, T. (1991). A Closer Look at the 
Complex Structure of Gender Stereotypes. Sex 
Roles, 24(1–2). https://doi.org/10.1007/BF0028 
8703
Stringer, R. (2013). Vulnerability After Wounding: 
Feminism, Rape Law, and the Differend. Sub-
Stance, 42(3), 148–168. https://doi.org/10.1353/
sub.2013.0031
Trudgill, E. (1976). Madonnas and Magdalens: 
The Origins and Development of Victorian 
Sexual Attitudes (First Edit). Heinemann. ISBN 
9780434794621
Viki, G. T., & Abrams, D. (2002). But She Was 
Unfaithful: Benevolent Sexism and Reactions to 
Rape Victims Who Violate Traditional Gender 
Role Expectations. Sex Roles, 47(5–6), 289-293. 
https://doi.org/10.1023/A:1021342912248
Vonk, R., & Ashmore, R. D. (2003). Thinking 
About Gender Types: Cognitive Organization 
of Female and Male Types. British Journal of 
Social Psychology, 42(2), 257–280. https://doi.
org/10.1348/014466603322127247
Welter, B. (1966). The Cult of True Womanhood: 
1820-1860. American Quarterly, 18(2), 151–174. 
https://doi.org/10.2307/2711179
Yamawaki, N., Darby, R., & Queiroz, A. (2007). The 
Moderating Role of Ambivalent Aexism: The 
Influence of Power Status on Perception of Rape 
150
(2023) 51
journal for constitutional theory and philosophy of law
Ana Lúcia Costa
Victim and Rapist. Journal of Social Psychology, 
147(1), 41–56. https://doi.org/10.3200/SOCP.147. 
1.41-56
Spanish caselaw
Sección Segunda de la Audiencia Provincial de 
Navarra, Decision nº 38, 20th March 2018. 
Available at: https://www.diariodesevilla.es/2018/ 
04/26/SENTENCIA_SUMARIO_426_DE_201 
6_cen.pdf?hash=e20ec174f9f8f973713214dc7e6
703d33861d342
Tribunal Supremo, Decision nº 344, 4th July 2019. 
Available at: https://www.diariodesevilla.es/2019/ 
07/05/Sentencia_Sala_Segunda_caso_La_Mana 
da_JG.pdf?hash=c2bc31bd7c438c8d4dc829a6b2
d4e4c79cb3e064
151
journal for constitutional theory and philosophy of law
(2023) 51
revus (2023) 51, 151–167
Kenneth Einar Himma*
In defence of the Coercion Thesis and 
modest conceptual methodology
A reply to critics
This is the author’s reply to a book symposium on law and coercion. Himma’s book – 
Coercion and the Nature of Law (OUP 2020) – had two principal objectives. The first was 
to show it is a necessary condition for a normative system to count as one of law that it 
includes prohibitions on violence and theft that are backed by detriment that counts as 
a sanction in virtue of being reasonably contrived to deter enough noncompliance to 
enable the system to minimally achieve its function of keeping the peace (the Coercion 
Thesis). The second was to defend and illustrate the methodology of modest conceptual 
analysis as defined by Frank Jackson. The book was critically discussed by Brian Bix, 
Thomas Bustamante, Frank Jackson, Paolo Di Lucia and Lorenzo Passerini Glazel, Anna 
Pintore, Pablo Rapetti and Kara Woodbury-Smith in the 45th issue of Revus.
Keywords: law, coercion, sanction, modest conceptual analysis, normativity, law’s 
function
The book had two principal objectives. The first was to show it is a necessary 
condition for a normative system to count as one of law that it includes prohibi-
tions on violence and theft that are backed by detriment that counts as a sanc-
tion in virtue of being reasonably contrived to deter enough noncompliance to 
enable the system to minimally achieve its function of keeping the peace (the 
Coercion Thesis). The second was to defend and illustrate the methodology of 
modest conceptual analysis (MCA). MCA explicates the nature of a kind as it 
is determined by our semantic conventions for using the corresponding term 
as qualified in hard cases by certain shared philosophical assumptions about its 
nature – i.e. as it is determined by our conceptual practices. Immodest concep-
tual analysis (ICA) explicates the nature of a kind as it is determined by objec-
tive considerations that are utterly independent of anything we do with words 
– i.e., as it is determined independently of our conceptual practices.
The book adopts MCA because it is the only epistemically viable methodol-
ogy for conceptual analysis; it seems clear to me that no ordinary human being 
has reliable epistemic access to the nature of a kind as it is determined inde-
pendently of our conceptual practices by purely objective considerations. To 
do ICA, one would need something akin to a God’s-eye view of the nature of 
kinds that none of us can plausibly claim to have. We have, after all, little reason 
* khimma@gmail.com | Faculty of Law, University of Zagreb (Croatia). I want to thank the edi-
tors of Revus for sponsoring this symposium on Coercion and the Nature of Law (OUP 2020). 
I am deeply indebted to each of the philosophers who honoured me with their thoughtful 
reviews of the book. I apologize to all for my delay in replying.
152
(2023) 51
journal for constitutional theory and philosophy of law
Kenneth Einar Himma
to believe our basic perceptions of the objective material world mirror what it is 
actually like — though they clearly enable us to thrive in this world. But we have 
even less reason to believe that we have insight into the nature of a kind as it is 
determined independently of anything we do with words. I adopt MCA, then, 
because it is, as far as I can tell, the only game in town.
It bears noting that each of the reviewers accepts some version of the claim 
that law is, by nature, coercive in the sense it is a necessary condition for a sys-
tem to count as one of law that some of its norms are backed by the threat of 
a sanction. And I surmise some would go further, as I would, and argue that 
it is a necessary condition for a mandatory norm to count as one of law that 
it is backed by a sanction. In this respect, each of the reviews vindicates the 
Coercion Thesis.
1 REPLY TO BRIAN BIX
Bix’s review begins with an exposition of the book’s methodology and argu-
ments and then criticizes it for not engaging more directly with the various crit-
icisms of the project of conceptual analysis.1 As Bix (2019: 32) describes Brian 
Leiter’s criticisms, conceptual analysis is “an outdated ‘armchair’ view to do 
philosophy, an approach that other philosophical disciplines have abandoned”. 
Leiter and others argue that conceptual analysis should not be done because, 
among other reasons, it is “uninteresting” and has no normative implications 
that would enable us to improve our legal practices.
I don’t know whether Bix is endorsing these criticisms. Either way, however, 
they are important enough to warrant a response. To begin, it is not true that 
philosophers have abandoned conceptual analysis. Philosophers specializing in 
metaphysics (which concerns, among other things, the nature of time, space, 
free will, causation, etc.) or metaethics (which concerns, among other things, 
the nature of good, right, wrong, morality, etc.) continue to pursue conceptual 
analysis. Visit any accredited philosophy department and you will find someone 
competent with metaphysics and someone competent with metaethics.
Indeed, and on the contrary, the project of conceptual analysis is enjoying a 
philosophical resurgence — and one, moreover, that highlights the ascendency 
of MCA. The projects of experimental philosophy and conceptual engineering 
are becoming increasingly influential: experimental methods are deployed to 
ascertain the intuitions of competent speakers pertaining to hard cases so as to 
better understand our conceptual practices. Similarly, conceptual engineering 
is concerned with understanding and evaluating our conceptual practices so as 
to revise them to eliminate perceived problems.
1 I address these criticisms in Section 4 of Chapter 1 of Himma 2019.
153
journal for constitutional theory and philosophy of law
(2023) 51
In defence of the Coercion Thesis and modest conceptual methodology
A reply to critics
Bix also worries about my apparent lack of engagement with John Finnis’s 
substantive views, but I do not engage his natural law theory for two reasons. 
First, with one crucial exception discussed below, the arguments I make in the 
book are non-partisan in the sense they do not presuppose the truth of legal 
positivism. Second, and more importantly, Finnis (1996: 203, 204) makes clear 
he accepts positivism’s separability thesis:
‘There is no necessary or conceptual connection between positive law and morali-
ty.’ True, for there are immoral positive laws; ‘there are two broad categories (with 
many sub-classes) of unjust laws…’. And a conceptual distinction or disconnection 
is effortlessly established by the move made in the Summa, of taking human positive 
law as a subject for consideration in its own right (and its own name), a topic readily 
identifiable and identified prior to any question about its relation to morality…. ‘The 
identification of the existence and content of law does not require resort to any moral 
argument.’ True, for how else could one identify wicked laws such as Israel’s prophet 
denounced in words so often quoted by Aquinas: ‘Woe to those who make unfair laws 
[leges iniquas] who draw up instruments imposing injustice [iniustitiam], and who 
give judgments oppressing the poor’?
As I argue in Morality and the Nature of Law,2 Finnis’s natural law theory is 
intended to explicate an evaluative usage of law – i.e., law in its ideal sense or 
law as it should be, and he believes that positivism’s separability thesis is true of 
the descriptive usage of law because that thesis cannot be plausibly disputed: it 
is simply undeniable, on the descriptive usage we construct with our conceptual 
practices, that there can be wicked law.
My concern in the book was to explicate the descriptive usage, which is the 
only one reported in mainstream dictionaries.3 While law in its ideal sense 
might have a moral purpose (e.g., to do justice, as determined by substantive 
natural law theory), there is no reason to think law in its descriptive sense has 
such a purpose; and the idea that the basic purpose of law is the purely descrip-
tive one of maintaining order has an impressive pedigree that goes back as far as 
Locke, Hobbes, Blackstone, Aquinas. There is, of course, obvious moral value in 
keeping the peace (war is bad). However, we are motivated, first and foremost, 
by the prudential value of keeping the peace; if there is no objective morality, we 
would still be motivated to create some sort of normative system contrived to 
keep the peace. Few people would be willing to try to live together without the 
coercive mechanisms of law.
2 Himma 2019.
3 Dictionaries are compiled by lexicographers who ground their definitions in scientifically 
rigorous polls of competent speakers that are designed to identify our semantic conventions 
regarding the relevant terms. They have epistemic authority over philosophers with respect to 
the content of our semantic conventions in virtue of having formal training in the empirical 
methods of the social sciences — which we philosophers lack. Our expertise is relevant in ap-
plying philosophical methods to extract the underlying assumptions (and their entailments) 
that qualify these conventions in hard cases.
154
(2023) 51
journal for constitutional theory and philosophy of law
Kenneth Einar Himma
As for the worry that it is not desirable to focus on conceptual analysis 
without worrying about the moral evaluation of law, my voice is certainly not 
needed to contribute to the moral evaluation of law. There are, after all, vastly 
more theorists evaluating our legal practices than there are theorists addressing 
conceptual problems. And I would be stunned if I have anything new or worth-
while to contribute to that discussion.
2 REPLY TO THOMAS BUSTAMANTE
Bustamante is a Dworkinian who, like Dworkin, accepts the Coercion 
Thesis. However, he believes that law’s coercive nature cannot be satisfactori-
ly grounded in MCA. As he expresses the matter, “other methodologies, like 
Dworkin’s interpretivism (if interpreted as endorsing an inferentialist theory of 
meaning), can provide a more plausible account of the coercive nature of law”.4
Three observations might be helpful here. First, Dworkin accepts the 
Coercion Thesis but gives no argument for it; he simply builds it into his analy-
sis of the concept of law, which holds that the law of a community includes not 
only those rules enacted by the relevant authoritative body but also those moral 
principles that justify the use of coercive enforcement mechanisms. Second, 
there is nothing in any of my arguments that Dworkin is committed to reject-
ing. Indeed, he could adopt any of them without even prima facie inconsistency. 
Third, and most importantly, Dworkin (1986: 103–104), like Finnis,5 makes 
clear he is concerned with explicating an evaluative – or “preinterpretive” – 
concept of law and that he believes positivism is consistent with his analysis:
We need not deny that the Nazi system was an example of law … because there is an 
available sense in which it plainly was law. But we have no difficulty in understanding 
someone who does say that Nazi law was not really law, or was law in a degenerate 
sense, or was less than fully law. For he is not then using ‘law’ in that sense; he is not 
making that sort of preinterpretive judgment but a skeptical interpretive judgment 
that Nazi law lacked features crucial to flourishing legal systems whose rules and pro-
cedures do justify coercion.
Though there are clearly some differences between Dworkin and positivists,6 
the misunderstanding that Dworkin’s mature theory of law in Law’s Empire is 
inconsistent with the basics of positivism continues to persist despite the unam-
biguous character of these remarks.
4 Bustamante 2021: 39.
5 See above for Finnis’s views on the matter.
6 Dworkin is critical of the view that the criteria of validity are necessarily exhausted by a rule 
of recognition that counts as conventional; this — and not the idea law is a social artifact — is 
the target of his semantic sting argument.
155
journal for constitutional theory and philosophy of law
(2023) 51
In defence of the Coercion Thesis and modest conceptual methodology
A reply to critics
Much of what Bustamante says in his review of my book is less concerned 
with my views than with mounting a defence of the Coercion Thesis grounded 
in the evaluative methodology that Dworkin deploys in defence of his interpre-
tivism. Indeed, and somewhat oddly, Bustamante never offers a reason to think 
Dworkin would reject the use of MCA in defending the claim that the Nazis 
had a legal system; and I see no other reasons to think Dworkin is committed 
to rejecting any of the substantive claims I make in the book in defence of the 
Coercion Thesis.
Bustamante argues that those arguments count as pragmatic in character 
because they are grounded in claims about our practices. In particular, he argues 
that Hart’s remark that The Concept of Law is “an essay in descriptive sociology” 
shows he intends his arguments as pragmatic. But a consideration of the full 
context of those remarks makes clear Hart (2012: vi) is concerned to give an 
analysis of what we do with words:
The lawyer will regard the book as an essay in analytic jurisprudence, for it is concer-
ned with the clarification of the general framework of legal thought, rather than with 
the criticism of law or legal thought. Moreover, at many points, I have raised questions 
which may well be said to be about the meanings of words.… Notwithstanding its con-
cern with analysis the book may also be regarded as an essay in descriptive sociology; 
for the suggestion that inquiries into the meanings of words merely throw light on 
words is false. Many important distinctions, which are not immediately obvious, betwe-
en types of social situations or relationships may best be brought to light by an examina-
tion of the standard uses of the relevant expressions and of the way in which these depend 
on a social context, itself left unstated (Emphasis added).
Although my argument relies on the claim law has a basic function, this 
does not constitute the strategy as pragmatic. It is an uncontentious conceptual 
truth that every artifact performs some task we want or need it to do; it is nearly 
universally accepted among artifact theorists that an artifact is, by nature, con-
trived to perform some basic function – i.e., one it needs to be able to perform 
in order to do anything else it is standardly used to do.7 If law counts as an arti-
fact, it is partly because it has a basic function.
It should be clear that both the Coercion Thesis and its supporting argu-
ments are conceptual in character. To begin, it is worth noting that Oxford 
English Dictionary defines law as “the system of rules which a particular country 
or community recognizes as regulating the actions of its members and which it 
may enforce by the imposition of penalties.” Further, Chapters 4 through 8 are 
all aimed at vindicating the conceptual truism that law is normative in the sense 
that it is a necessary condition for a system to count as one of law that its man-
datory norms give rise to reasons to comply. The point of the last two chapters 
discussing international law and the society of angels is to show that neither is a 
counterexample to the Coercion Thesis.
7 See, e.g., Preston 2022: sec. 2.3 and Burazin 2023: 1846.
156
(2023) 51
journal for constitutional theory and philosophy of law
Kenneth Einar Himma
Bustamante (2021: 43) accepts my conclusions but argues “they are based on 
empirical claims.” That is true, but inconsequential because those claims describe 
our conceptual practices and hence count as metaphysical in virtue of purport-
ing, as every conceptual claim does, to be necessarily true. There are no other 
empirical observations that could justify a necessary truth about the nature of a 
kind.8 It is a necessary truth on our conceptual practices that every bachelor is an 
unmarried man because — and only because — we converge in using the term 
bachelor, as a contingent empirical matter, to refer only to unmarried men.
3 REPLY TO PAOLO DI LUCIA AND LORENZO 
PASSERINI GLAZEL
Di Lucia and Glazel focus on what they deem to be “risks” of MCA, which, 
they argue, illegitimately grounds substantive conclusions about the nature of 
law in “the canons or ordinary usage and on the philosophical assumptions of 
an undefined and contingent linguistic and cultural community”.9
It is worth noting they do not try to show that ICA is epistemically viable. 
Instead, they describe the following four problems with MCA. The first is that, 
“among people who construct definitions for a living, there is no single, gener-
ally accepted cognitive model, even for such common concepts as ‘mother’”.10 
The second is that there are “divergent” or “alternative” usages of the term law.11 
The third is that “linguistic analysis [is] … not merely descriptive or explicative 
but critical [in the sense] … [o]rdinary language is rendered more rigorous and 
less flexible, or indeed entirely supplanted by, scientific language”.12 The fourth 
is that “it is unclear whom the adjective ‘ours’ refers to [in our conceptual prac-
tices] – except for the exclusion of the academic philosophical community”.13
 I begin with the fourth: the term “our” refers to the community of compe-
tent speakers of our language, which does not exclude members of the academ-
ic community; the intuitions they have about the nature of law are also obvi-
ously conditioned by their competence with the term law. The relevant point is 
simply that we philosophers are constrained in giving an account of the nature 
of law by our conventions for using the term. That seems indisputable because 
any theory of law inconsistent with the idea that law consists of norms can be 
8 They can, of course, help us to identify counterexamples to a claim purporting to be necessar-
ily true. But that is a different matter.
9 Di Lucia & Glazel 2021: 57.
10 Di Lucia & Glazel 2021: 62.
11 Di Lucia & Glazel 2021: 62.
12 Di Lucia & Glazel 2021: 63 (emphasis added).
13 Di Lucia & Glazel 2021: 63.
157
journal for constitutional theory and philosophy of law
(2023) 51
In defence of the Coercion Thesis and modest conceptual methodology
A reply to critics
rejected – and conclusively – for that reason. It would, for example, be obvi-
ously preposterous to claim that law is, by nature, composed of H2O molecules.
As to the third criticism that what they characterize as linguistic analysis is 
“critical” in the sense described above: that might be true, but that’s not incon-
sistent with anything I say because I do not claim to be doing that kind of lin-
guistic analysis. Philosophers are free, of course, to pick out aspects of our usage 
that are problematic and recommend changes. This is being done with great suc-
cess by those engaged in the project of conceptual engineering. One could group 
the two projects together, but I see no reason to do so. The projects of describing 
our conceptual practices and evaluating them might be related, but they are also 
clearly distinct, and it accomplishes nothing of value to suggest otherwise.
Further, it is false that the point of analysing our conceptual practices is nec-
essarily to replace ordinary language with more rigorous scientific language. 
The physical sciences have contributed to improving our conceptual practices, 
as they did in leading us to revise the term water to refer only to liquids com-
posed of H20 molecules. But that doesn’t change the fact that description/expli-
cation and evaluation/revision are different projects.
 As to the second criticism that there are different usages of “law”: that is 
true but irrelevant. There are usages that refer to the propositions that pur-
port to describe causal regularities in the universe (the so-called laws of nature) 
and to morality. But the Coercion Thesis is concerned by its own terms with 
only the usage that refers to what John Austin and Jeremy Bentham describe 
as “positive” or “posited law” and there is no reason to think my focus on that 
usage is problematic.
It is worth noting the descriptive usage with which my book is concerned is 
the only one recognized by mainstream dictionaries because that is the only one 
we use in ordinary contexts. When one travels to a foreign country and wants 
to know whether, say, the law prohibits use of cannabis, one is asking whether 
there is a legal prohibition that is enforced in that country, which is all that really 
matters as a prudential matter.
Indeed, the only competent speakers I have ever encountered who use the 
term in an evaluative sense are legal philosophers, such as Finnis and Dworkin. 
However, this technical evaluative use does not, as far as I can tell, enable us to 
do anything we cannot do more clearly by simply thinking in terms of the mor-
al principles governing legitimate authority. I would hypothesize that the reason 
the evaluative usage has not gained currency in ordinary usage is that it is just 
not needed to discuss anything we need to worry about when it comes to law.
Finally, as to the first criticism that there is no commonly accepted “cogni-
tive model” for ordinary concepts: that might be true, but it is clear that there 
is core content to the relevant usage of the term “law” that every speaker who 
158
(2023) 51
journal for constitutional theory and philosophy of law
Kenneth Einar Himma
counts as competent with the term would accept in virtue of being competent 
with that term — such as the claim that law consists of norms.
And it is not the business of descriptive conceptual analysis to produce 
cognitive models. “Cognitive model” as defined by the American Psychological 
Association Dictionary of Psychology, means “a theoretical view of thought and 
mental observations, which provides explanations for observed phenomena 
and makes predictions about an unknown future.” There is a trivial sense in 
which MCA is concerned with explaining an observed phenomena (i.e., our 
conceptual practices for using words) and making predictions about an un-
known future (i.e., that we will continue to use words in this way), but it distorts 
the idea of conceptual analysis beyond recognition to describe it as a “psycho-
logical theory” and it distorts the idea of a cognitive model as much to suggest 
it is concerned with our conventions for using words.
4 REPLY TO FRANK JACKSON
Jackson’s comments are sympathetic to both the Coercion Thesis and the 
arguments of the book. He says, for instance, “I am no philosopher of law, but 
[Himma] convinced me (a task made easier by the fact that I was anteced-
ently disposed to accept something like the Coercion Thesis)”.14 Although he 
observes that my views about MCA diverge in some particulars from his, he 
defends the methodology as I have articulated it: “Himma’s methodology in 
defence of the Coercion Thesis is just fine, or so I will argue”.15
Jackson has two principal worries. While I characterize the Coercion Thesis 
as a purely descriptive claim, he argues certain remarks I make express a norma-
tive claim — namely that “it would be good to make the possibility of punish-
ment for breaking its law a necessary condition for being a legal system, for it is 
part of the essential rationale for having legal systems”.16 He points to the follow-
ing remarks I made as evidence the Coercion Thesis expresses a normative claim:
All that stands between civilisation and the state of nature is an institutional norma-
tive system that backs mandatory norms prohibiting certain assaults on persons or 
property with the threat of severe detriment; this is why every existing system of our 
world … backs mandatory norms prohibiting acts likely to lead to breaches of the 
peace with the threat of incarceration – or worse (Himma 2020: 24).
 I was not intending to make a normative claim in that passage; however, my 
language invites that interpretation. This text occurs in a chapter of the book 
where I was making an empirical argument that the Coercion Thesis coheres 
14 Jackson 2021: 71.
15 Jackson 2021: 72.
16 Jackson 2021: 73.
159
journal for constitutional theory and philosophy of law
(2023) 51
In defence of the Coercion Thesis and modest conceptual methodology
A reply to critics
with existing legal practice because every system in our world that counts as 
one of law backs mandatory prohibitions on acts likely to lead to a breach of the 
peace with something that counts as a sanction. But that claim about existing le-
gal systems should be distinguished from the conceptual claim about the nature 
of law: the empirical claim is that without law we would likely fall into a state of 
nature; the conceptual claim is that the Coercion Thesis is true because we have 
adopted the term “law” to refer only to systems that are minimally efficacious in 
keeping the peace in virtue of backing those prohibitions with a sanction.
Jackson’s methodological concern is that I give MCA too large a role in 
structuring the world. On his view, the problem is that:
There is a sense in which we create the classifications and a sense in which we do not. 
The items to be found in our world are alike and unalike in their many ways, indepen-
dently of us, or so we realists insist… What we create are the classifications we choose, 
implicitly or explicitly, to employ in making sense of the world, and not whether some 
item falls under some classification…. Anthropologists decided that it would be good 
to classify people together in terms of whether or not they have a parent in common, 
co-opting the term “sibling” to do the job in language…. They created the term and 
the classification, but they did not thereby create any siblings. Parents do that.17
Jackson is correct, and I did not mean to suggest otherwise. As I typically put 
the matter, what we do with words structures the world of our experience by or-
ganizing kinds and their instances into an ontology. But those conceptual prac-
tices assume the things we name already exist in the world of our experience. 
Indeed, it is the fact that we notice they exist and are salient that explains why we 
adopt words to talk about them. Our conceptual practices structure the world, as 
I intend that idea, only in the limited sense that they create a framework that as-
signs words to the various kinds we experience so we can talk about those kinds.
Jackson worries that this claim is incompatible with ontological realism, but 
I disagree. It is, as far as I can tell, compatible not only with realism but also 
with antirealism. That I consistently speak in terms of the world of our experi-
ence or our world was intended to acknowledge that what we perceive might 
not be really true; but that latter claim is compatible with the realist view the 
world really exists and with the antirealist view that it doesn’t really exist. Those 
perceptual beliefs might be false because the world doesn’t exist, or they might 
be false because our perceptions systematically misrepresent what the world 
really looks like. The Coercion Thesis and supporting arguments are compat-
ible with both ontological views.
This is as it should be. A modest conceptual theory of law should not make 
any assumptions about whether the world as we perceive it exists in some im-
modest sense. Since our conceptual practices are what they are regardless of 
whether realism or antirealism is true, an account of those practices should be 
17 Jackson 2021: 73.
160
(2023) 51
journal for constitutional theory and philosophy of law
Kenneth Einar Himma
agnostic about the disputes between realism and antirealism. Our conceptual 
practices structure the world of our experience, and, for all we can be certain of, 
our beliefs about the mind-independent world might be systematically mistaken.
5 REPLY TO ANNA PINTORE
Pintore’s review (2021: 79) begins with a strikingly aggressive remark: “To 
dispel any doubt, I will say immediately that there are very few points in the 
book that I agree with, notwithstanding the fact that I believe, like its author, 
that coercion plays a crucial and ineluctable role in the legal world”.18 Ouch.
Unfortunately, Pintore’s review tends to meander, so it is not always easy 
to understand. However, her principal complaints seem to be that my analy-
sis: (1) “equat[es] legal coercion with authorization to impose sanctions,” (2) 
“oscilat[es] between law in general and law here and now,” (3) relies on a con-
cept of normativity that is either unclear or nonstandard, and (4) “reduce[s] the 
motives for law-abidance to a single one – the desire to avoid a sanction.”
I am baffled by these claims. I have no idea why, to begin, she thinks (1) is 
problematic, since she accepts the Coercion Thesis. It is an obvious truism that 
one can coerce only by threatening detriment. If the relevant norms are coercive in 
the relevant sense, it is because they are backed by the threat of a sanction. While 
Pintore points out law is not the only normative system that backs its prescrip-
tions with a sanction, there is nothing in the book inconsistent with that claim; 
the Coercion Thesis is, after all, limited by its own clear terms to systems of law.19
Similarly, I do not understand which elements of my argument Pintore be-
lieves “oscillate” in an objectionable way between law in general and particular 
instances of legal systems. If one wants to show, for instance, that international 
18 She is not shy about expressing her disdain for the concerns of the “Anglo-American” world 
of legal philosophers. She writes, “At the risk of seeming to be the proverbial bull in a china 
shop, I confess that I cannot get excited about the discussion of the normativity of law, which 
so absorbs the Anglo-American world and also Himma” (Pintore 2021: 87). Her tone certainly 
invites a reaction, but not because she comes across as a bull in a china shop.
19 She also points out (Pintore 2021: 82), first, that “not all sanctions are penal sanctions” and, 
second, not all penal sanctions … presuppose culpability”. As to the first point, there is noth-
ing in my explanation of a sanction that presupposes all sanctions are penal in the sense they 
are intended to punish. I claim only that detriment counts as a sanction if reasonably likely 
— and thus reasonably contrived — to deter noncompliance to the requisite extent. As to the 
second, there is nothing in the Coercion Thesis or my analysis of a sanction that entails all 
penal sanctions presuppose culpability. However, that claim is obviously true: even a law that 
imposes strict liability presupposes that the commission of some acts are inherently culpable 
because of the likelihood that they result in grievous harm. It is the law itself that defines stan-
dards of legal culpability. If there is an existing system of criminal law that does not require a 
showing of what it deems culpable intent on the part of a subject to justify imposing a sanc-
tion, I would like to see it.
161
journal for constitutional theory and philosophy of law
(2023) 51
In defence of the Coercion Thesis and modest conceptual methodology
A reply to critics
law is not a counterexample to the Coercion Thesis, as I argued in Chapter 9 of 
the book, one must discuss the particular system of international law governed 
by the U.N. Charter. I suppose that might count in some literal sense as “oscil-
lating,” but that is part of what must be done to show a claim is true as a matter 
of conceptual necessity.
Pintore’s concerns about the concept of normativity are just as confusing 
(“which normativity?” she asks in the title to her review). Much of the analysis 
in the central part of my book is concerned with defending the claim that law is 
conceptually normative in the sense that it is part of the nature of a legal system 
that at least some of its norms create reasons to comply. The book argues that if 
law is conceptually normative in that sense, the Coercion Thesis uniquely vin-
dicates its conceptual normativity; and that cannot be plausibly denied. Since 
there can be wicked systems that count as law, not every legal system creates 
moral reasons to comply. But if that is correct, then the only reasons a system 
of law could create as a matter of conceptual necessity are prudential and the 
Coercion Thesis provides the only remotely plausible explanation of how it can 
create such reasons as a matter of conceptual necessity.
As to her final concern, there is nothing in either the book or the Coercion 
Thesis that “reduce[s] the motives for law-abidance to a single one – the desire 
to avoid a sanction.” Indeed, I explicitly reject that claim in the very first chapter 
of the book: “but none of this implies it is a conceptual truth that people whose 
behavior complies with the law must be motivated, even in part, by a desire to 
avoid the sanctions judges are authorized to impose on subjects for non-compli-
ance.” What I claim — a few sentences later — is that the prospect of incurring 
a sanction is normatively relevant even for those who comply for other reasons:
While there are some legally prohibited acts from which I abstain, such as jaywalking 
in front of a police officer, only out of a desire to avoid the sanctions, the fact that I 
have never intentionally killed someone is not at all explained by the fact that murder 
is punishable by a long term of incarceration; however, if I am ever tempted to do so, I 
would certainly regard the fact that I would likely be sentenced to prison as normati-
vely relevant (Himma 2020: 19).
What law does, on the Coercion Thesis, is provide an incentive to comply in 
the form of a deterrent to noncompliance for those not antecedently inclined 
to comply.
6 REPLY TO PABLO RAPETTI
Pablo Rapetti’s review expresses a number of interesting concerns, but his 
two main worries seem to be that despite my remarks to the contrary (1) the 
analysis is positivist in character, and (2) my methodology is normative insofar 
as I claim the basic function of law is to keep the peace.
162
(2023) 51
journal for constitutional theory and philosophy of law
Kenneth Einar Himma
As to (1), there is only one place where I make an argument that presup-
poses a positivist view — i.e., in defence of the claim that only the Coercion 
Thesis can vindicate the truism that it is a constitutive property of a legal system 
that some of its norms create reasons to comply.20 In particular, Chapter 6 ar-
gues that if law is conceptually normative in that sense, those reasons must be 
prudential because a wicked legal norm obviously cannot create a moral reason 
to comply (which presupposes the falsity of natural law theories construed as 
anti-positivist). The relevant reasons would have to be prudential, as those are 
the only remaining reasons relevant in explaining law’s conceptual normativity.
As to (2), there is nowhere in the book that deploys a normative methodol-
ogy. Rapetti (2021: 98) grounds his belief to the contrary in my claim that law, 
like any other artifact, has a basic function, which is to keep the peace:
[Himma’s] account on law’s “conceptual function” raises the first suspicion [that he is 
making normative claims in addition to purely descriptive claims]…. Himma states 
that we do not value the regulation of behaviour for its own sake, that his argument 
borrows a device belonging to normative political philosophy, that the function of kee-
ping the peace is morally valuable, and that such a function grants the law a claim to 
moral legitimacy. Whereas these two latter statements can be taken as mere side com-
ments, the former two cannot, for they have heavy bearing on the argument he deplo-
ys to establish a conceptual connection between law and coercion, i.e. the Coercion 
Thesis”.
It is crucial to note here that the conceptual claim that a necessary condition 
for something to count as an artifact is that it is contrived to do certain things 
is a purely descriptive claim — and one that is uncontentious among artifact 
theorists.21 Indeed, even art has a basic function: its point, on the prevailing 
conceptual theory of art, is to induce an aesthetic experience.22 If, for instance, I 
throw some logs in a pile, the pile counts as artificial but not as artifactual, un-
less I have created it to perform some task or function.
I argued that (1) a normative system cannot count as one of law unless it 
is reasonably contrived to perform law’s basic function of keeping the peace, 
which is the function it must succeed in performing, at least minimally, to do 
anything else it can be used to do, and that (2) a normative system is not rea-
sonably contrived to keep the peace among rational self-interested beings like 
us who inhabit a world of acute material scarcity, where we must compete for 
everything we need or want, unless some of its norms are backed by the threat 
of a sanction.
20 As discussed above in my replies to Brian Bix and Thomas Bustamante, both Finnis and 
Dworkin concede that positivism’s separability thesis is true of the descriptive use of law. See 
above for quotes.
21 See note 7, above.
22 See, e.g., Beardsley 1982.
163
journal for constitutional theory and philosophy of law
(2023) 51
In defence of the Coercion Thesis and modest conceptual methodology
A reply to critics
Both of these claims are descriptive. The attribution of a basic function to 
something is wholly descriptive and does not entail that one should use it to per-
form that function. It is clear that guns have a basic function, namely, to propel 
a small metallic object at speeds sufficient to pierce bodies and other material 
objects;23 but that is obviously consistent with the altogether sensible view that 
the possession of assault weapons by ordinary citizens should be criminalized. 
While it is equally clear that nuclear bombs have a basic function (“to cause an 
explosion”24), it is obvious they should not be used except to deter, or defend 
against, a nuclear attack. These conceptual claims are all purely descriptive.
Rapetti also argues that I make no argument supporting the claim that the 
basic function of law is to keep the peace, but that is the point of invoking the 
state-of-nature device. While Hobbes might have been too pessimistic about 
how bad a state of nature would be, it would nonetheless be bad enough that the 
vast majority of us would oppose, and vehemently, just repealing the use of en-
forcement mechanisms. Even libertarians like Robert Nozick25 do not oppose 
the existence of the state; they simply claim its only legitimate function, which 
is a necessary one, is to keep the peace. I see no reason to take a poll to establish 
something so obvious. I think it fair to say that the evidentiary burden rests 
with Rapetti to make a case to the contrary if he believes this is false.
Finally, Rapetti (2021: 101) argues “the conclusion that – as a matter of con-
tingent social fact – we are inclined to do as the law commands if it backs its 
commands with threat of sanctions is drawn by Himma from the fact … that – 
as a matter of normative objective rationality – we should do so”, which Rapetti 
takes to be normative in character.
There are two problems with this concern: first, I never claim “we are in-
clined to do as the law commands if … back[ed] … with threat of sanctions.” As 
I indicate in my response to Pintore, I acknowledge in the book that most peo-
ple do not refrain from murder only— or even primarily — because they do not 
want to incur a sanction. What I claim — and this is not only a purely descrip-
tive claim but one that strikes me as obvious — is that rational self-interested 
beings like us who live in a world of acute material scarcity like ours are, all 
else being equal, are more likely to comply with norms prohibiting violence and 
theft if backed by a sanction than if not backed by a sanction. And that claim/
prediction, like the others discussed above, strikes me as obvious enough not to 
need empirical confirmation in the form of a poll.
Second, the claim that we are more likely to comply with a norm if backed 
by a sanction than if not backed by a sanction because we converge in believing 
23 As Oxford English Dictionary defines the term gun, it means a “weapon incorporating a metal 
tube from which bullets, shells, or other missiles are propelled by explosive force.”
24 Oxford English Dictionary.
25 Nozick 1974.
164
(2023) 51
journal for constitutional theory and philosophy of law
Kenneth Einar Himma
that we should treat that as an additional reason to comply, has no normative 
implications. The claim that we converge in believing objective norms of practi-
cal reasoning requires us to treat a sanction as an additional reason to comply 
is also a purely descriptive claim with no normative implications. I am merely 
describing what I take to be our shared beliefs about the matter; I am not mak-
ing claims about what we should do because it might be good for us.
Accordingly, there is nothing in my defence of the purely descriptive 
Coercion Thesis that relies either on normative claims or presupposes a norma-
tive methodology. Indeed, this should surprise no one. One reason many theo-
rists believe conceptual analysis should not be done is that it has no normative 
implications. As Richard Posner (1996: 3) expresses the complaint:
I have nothing against philosophical speculation. But one would like it to have some 
pay-off; something ought to turn on the answer to the question ‘What is law?’ if the 
question is to be worth asking by people who could use their time in other socially 
valuable ways. Nothing does turn on it…. [T]he central task of analytic jurisprudence 
is, or at least ought to be, not to answer the question ‘What is law?’ but to show that it 
should not be asked, because it only confuses matters.
Obviously, I disagree, but not because I think conceptual analysis has nor-
mative implications. I disagree because I see no reason to believe that the only 
theories that are legitimately pursued are those with obvious normative impli-
cations. If that were true, then mathematicians should not have pursued, for 
instance, the theory of imaginary numbers26 or non-Euclidian geometries that 
had no such applications before Einstein invented his theories of relativity.27 It 
stuns me that Posner’s anti-intellectual views here have gained so much traction 
among legal theorists trained in philosophy.
7 REPLY TO KARA WOODBURY-SMITH
Woodbury-Smith expresses a number of concerns about the arguments in 
Coercion and the Nature of Law, one of which pertains to methodology while 
the others pertain to substantive claims of the book. As she expresses the meth-
odological concern:
Engagement with [the book], and with Himma’s other writings on legal normativity, 
is particularly difficult because … we mean different things when we think and write 
about the conceptual analysis of law as such. I think a conceptual theory of law ought 
to account for law at its most general, and I mean this in the Hartian sense that it can 
26 Imaginary numbers are formed by taking the square root of -1.
27 Non-Euclidean geometries deny Euclid’s parallel postulate, which holds that given a line and 
a point off that line, there is one and only one line passing through that point parallel to the 
given line.
165
journal for constitutional theory and philosophy of law
(2023) 51
In defence of the Coercion Thesis and modest conceptual methodology
A reply to critics
account for all instances of law anywhere – including Twin Earth, or Heaven. Himma, 
however, argues that such a project cannot be done.28
I don’t see anything in MCA that entails this. Though I regard MCA as the 
only viable approach to conceptual analysis, this is consistent with Woodbury-
Smith’s description of the project. Regardless of how the nature of a kind is 
determined, the locution nature of a kind, as discussed in Chapter 2, refers to 
those properties that constitute something as an instance of that kind; these 
properties account “for all instances of law anywhere.” The epistemological 
claim I make should thus be construed as asserting nothing stronger than that 
the only conditions explaining all instances of law to which we have epistemic 
access are those entailed by our conceptual practices.
Indeed, it is crucial to note in this connection that both Raz and Hart adopt 
a modest methodology. Since I have already reproduced Hart’s remarks on the 
matter in my reply to Pintore, I will confine myself here to what Raz has said:
The notion of law as designating a type of social institution is not, however, part of 
the scholarly apparatus of any learned discipline. It is not a concept introduced by 
academics to help with explaining some social phenomena. Rather it is a concept en-
trenched in our society’s self-understanding. It is a common concept in our society 
and one which is not a preserve of any specialized discipline …. It occupies a central 
role in our understanding of society, our own as well other societies. In large measure 
what we study when we study the nature of law is the nature of our self-understan-
ding …. It is part of our self-consciousness, of the way we conceive and understand 
our society…. That consciousness is part of what we study when we inquire into the 
nature of law.29
Raz is also clear in the last sentence of his discussion on the society of angels 
that his analysis is concerned with our conceptual practices. He says of the nor-
mative system of the angels, “If such a normative system has all the features of 
a legal system described above, then it would be recognized as one by all despite 
its lack of sanctions.”30
Woodbury-Smith’s second worry has to do with my view of what the con-
ceptual normativity of law amounts to. She argues:
Himma’s bar for what constitutes a successful account of legal normativity is higher 
than what a conceptual theory of law as such needs to offer. Being able to precisely 
state how law qua legal system provides us with objective reasons for complying with 
its mandatory norms is dependent not only on how we conceive of law as such (in the 
28 Woodbury-Smith 2021: 108 (emphasis added).
29 Raz 2009: 31.
30 Raz 2002: 159–160 (emphasis added). This not only begs the empirical question of how peo-
ple would react, but is also inconsistent with his view that someone counts as having practical 
authority only if “effective in imposing [her] will on many over whom [she] claims authority.” 
Raz 1994: 211.
166
(2023) 51
journal for constitutional theory and philosophy of law
Kenneth Einar Himma
Hartian sense), but also on understanding the ways in which humans can be motiva-
ted to comply with norms.31
On her view, all that is needed to explain the conceptual normativity of law 
is that it “claims to be a legitimate authority and, as such, claims to provide us 
with objective reasons for action”.32
I have two reactions. First, the idea that law claims legitimate authority as 
a matter of conceptual necessity is contentious and hence does not count as 
a truism. As I argued in Chapter 5 of Morality and the Nature of Law33 and 
elsewhere,34 a legal system is an abstract object metaphysically incapable of 
making claims. It makes no more sense to attribute communicative acts to a le-
gal system than it does, for instance, to attribute such acts to the abstract object 
to which the numeral “1” refers. And it is implausible to attribute that claim to 
individual legislators and judges. The only claim of authority that can plausibly 
be attributed to them on just the basis of their making and adjudicating rules 
is that they have authority that counts as legal in virtue of being conferred by 
law. Indeed, it is worth noting in this regard that judges and legislators com-
monly express their disagreement with a norm or holding by denying its moral 
legitimacy — and, lately, in an increasingly strident tone that contributes to the 
increasing polarization in the U.S. between left and right.35
Second, just claiming to give us reasons to comply is not enough to vindi-
cate the conceptual truism that legal systems are normative. There is nothing in 
the fact that some system claims to be morally legitimate that it is reasonably 
likely to motivate someone to comply because it should motivate her to comply; 
anyone, after all, can claim to have legitimate authority. To claim my account 
of law’s conceptual normativity is too strong is to deny that law is actually nor-
mative as a matter of conceptual necessity. Claiming to provide reasons might 
count as purportedly normative but it does not count as actually normative.
In closing, I want to thank everyone who contributed to the symposium 
on my book and to the editors of Revus for publishing it. While my reviewers 
and I disagree a great deal, I am grateful that each took the time to share their 
thoughts on my work. I learned much from their reviews.
31 Woodbury-Smith 2021: 109.
32 Woodbury-Smith 2021: 112.
33 Himma 2019.
34 See Himma 2001.
35 Antonin Scalia was especially aggressive in expressing his views about the legitimacy of deci-
sions he disagreed with. See, e.g., Stern 2015. But he is far from unique in couching his dis-
agreement in such moral terms.
167
journal for constitutional theory and philosophy of law
(2023) 51
In defence of the Coercion Thesis and modest conceptual methodology
A reply to critics
Bibliography
Beardsley, M. (1982). The Aesthetic Point of View. 
Cornell University Press.
Bix, B. H. (2021). Coercion, function, and the “why” 
of law. Comments on Kenneth Einar Himma’s 
Coercion and the Nature of Law. Revus. Journal 
for constitutional theory and philosophy of law, 45, 
31–38. https://doi.org/10.4000/revus.7532
Burazin, L. (2023). Law as an Artifact. In Mortimer 
Sellers & Stephan Kirste (eds.), Encyclopedia of 
the Philosophy of Law and Social Philosophy (pp. 
1845–1850). Springer. https://doi.org/10.1007/97 
8-94-007-6519-1
Bustamante, T. (2021). Should the Coercion Thesis 
really be defended as a conceptual claim? A 
Dworkinian rumination on Himma. Revus. 
Journal for constitutional theory and philosophy 
of law, 45, 39–56. https://doi.org/10.4000/revus 
.7565
Di Lucia, P. & Passerini Glazel, L. (2021). The first 
and the last word. Critical remarks on Himma’s 
philosophical enterprise. Revus. Journal for con-
stitutional theory and philosophy of law, 45, 57–
69. https://doi.org/10.4000/revus.7612
Dworkin, R. (1986). Law’s Empire. Harvard Uni-
versity Press.
Finnis, J. (1996). The Truth in Legal Positivism. In 
Robert P. George (ed.), The Autonomy of Law (pp. 
195–214). Clarendon Press.
Hart, H.L.A. (2012). The Concept of Law (3rd ed.). 
Oxford University Press.
Himma, K. E. (2001). Law’s Claim of Legitimate 
Authority. In Jules L. Coleman (ed.), Hart’s Post-
script: Essays on the Postscript to the Concept of 
Law. Oxford University Press.
Himma, K. E. (2019). Morality and the Nature of 
Law. Oxford University Press.
Himma, K. E. (2020). Coercion and the Nature of 
Law. Oxford University Press.
Jackson, F. (2021). Conceptual analysis and the 
Coercion Thesis. Revus. Journal for constitutional 
theory and philosophy of law, 45, 71–77. https://
doi.org/10.4000/revus.7594
Nozick, R. (1974). Anarchy, State, and Utopia. Basic 
Books.
Pintore, A. (2021). Which coercion, which method, 
which normativity? Observations on Kenneth E. 
Himma’s Coercion and the Nature of Law. Revus. 
Journal for constitutional theory and philosophy 
of law, 45, 79–92. https://doi.org/10.4000/revus 
.7647
Posner, R. (1996). Law and Legal Theory. Clarendon 
Press.
Preston, B. (2022). Artifact. In Edward N. Zalta & Uri 
Nodelman (eds.), The Stanford Encyclopedia of 
Philosophy. <https://plato.stanford.edu/archives/ 
win2022/entries/artifact/>.
Rapetti, P. A. (2021). Law’s function, descriptive 
conceptual analysis, and legal positivism. Some 
methodological qualms. Revus. Journal for con-
stitutional theory and philosophy of law, 45, 93–
105. https://doi.org/10.4000/revus.7665
Raz, J. (1994). Ethics in the Public Domain. Clarendon 
Press.
Raz, J. (2002). Practical Reason and Norms. Oxford 
University Press.
Raz, J. (2009). Can There Be a Theory of Law? In 
Joseph Raz, Between Authority and Interpretation 
(pp. 17–46). Oxford University Press.
Stern, M. (Feb. 24, 2015). Scalia on Supreme Court 
gay marriage ruling: invalid, illegitimate. Slate. 
https://slate.com/human-interest/2015/09/sca 
lia-on-supreme-court-gay-marriage-ruling-inva 
lid-illegitimate.html
Woodbury-Smith, K. (2021). Legal normativity in 
Kenneth Einar Himma’s Coercion and the Nature 
of Law. Revus. Journal for constitutional theory 
and philosophy of law, 45, 107–119. https://doi.
org/10.4000/revus.7699

169
journal for constitutional theory and philosophy of law
(2023) 51
SYNOPSES, KEYWORDS, AND BIOGRAPHICAL NOTES
Synopsis
David Duarte
Rights as formal combinations of normative variables
SLOV. | Pravice kot formalne kombinacije normativnih spremenljivk. Čeprav 
je v luči Hohfeldovih predpostavk njegova razpredelnica pravnih položajev 
povsem utemeljena, se zdi, da jo pestijo prav tiste hibe, ki jih je mogoče pripisati 
omenjenim predpostavkam. Natančneje gre za predpostavki, da norma ni nujni 
pogoj za pravni položaj in da je eno samo dejanje v korelativni vrsti zadosten 
pogoj za to, da se povzroči rezultat. S preprostim predlogom poposameznjenja 
ali individuacije norme avtor v tem članku ponudi povsem na normah osnova-
no razpredelnico pravnih položajev, v kateri je upoštevano tudi  hkratno delo-
vanje korelativnih igralcev (ne da bi bila ogrožena atomičnosti). Ker je pravni 
položaj zgolj rezultat neke kombinacije normativnih spremenljivk, ta pristop 
prinaša strogo formalno sestavo pravnih položajev na način, ki izpodbija vlogo, 
ki so jo doslej imele tradicionalne teorije pravic.
Ključne besede: Hohfeld (Wesley Newcomb), pravne norme, pravni 
položaji, korelacijske vrste, korelacijski igralci
ENG. | Hohfeld’s table of legal positions, though highly consistent under 
his own assumptions, seems to be vulnerable to the exact flaws assignable to 
those assumptions. Specifically, the assumptions that a norm is not a necessary 
condition of a legal position and that one single action in a correlativity line is 
sufficient to bring about the action’s result. With a simple proposal of norm in-
dividuation, this paper develops a totally norm-based table of legal positions in 
which co-action from correlated agents is also considered (without threatening 
atomicity). Since a legal position is just the result of a combination of norma-
tive variables, the present approach leads to a strictly formal composition of 
legal positions in a way that challenges the role played so far by the traditional 
theories of rights.
Keywords: Hohfeld (Wesley Newcomb), legal norms, legal positions, 
correlativity lines, correlated agents
Summary: 1 The legal normativist approach: Two basic presuppositions – 2 
Why the Hohfeldian strategy does not work: Too little and too much – 3 Norm 
individuation: A brief scheme – 3.1 A norm: Material elements – 3.2 A norm: 
Subjective components – 3.3 A complete norm: Adding up the elements – 4 
First order positions: Liberty and duty not to defeat the internal possibility – 
170
(2023) 51
journal for constitutional theory and philosophy of law
SYNOPSES, KEYWORDS, AND BIOGRAPHICAL NOTES
4.1 Duty not to defeat and duty not to: Adaptative content – 4.2 Liberties are 
protected: Weak and strong protections – 5 First order positions: Duty and 
claim-right – 5.1 Waiving: Weak and strong waiving and non-waivability – 6 
Second order positions: Power and duty not to defeat – 6.1 Without power: 
Disability and immunity are not legal positions – 7 A table of legal positions: 
Five atomic positions – 8 Sets of addressees: Who, how many and how – 9 
Correlativity: Assuming asymmetry and differentiating directedness – 10 The 
formality of rights: Combining normative variables – 10.1 Why theories of 
rights are almost useless: Facing epistemic insufficiency
David Duarte is an Associate Professor at Lisbon Legal Theory. Univer-
sity of Lisbon (Portugal) | E-mail: davidduarte@fd.ulisboa.pt
171
journal for constitutional theory and philosophy of law
(2023) 51
SYNOPSES, KEYWORDS, AND BIOGRAPHICAL NOTES
Synopsis
Miklós Könczöl
Parental proxy voting and political representation
SLOV. | Zastopniško glasovanje staršev in politično predstavništvo. V prispev-
ku je predstavljen predlog, da se staršem podelijo dodatni glasovi, ki jih lahko 
oddajo kot zastopniki v imenu svojih otrok. Proučene so utemeljitve glasovanja 
na voltivah po starševskem zastopanju s poudarkom na različnih razlagah poj-
ma "zastopnik". V prvem delu razprave je ocenjena trditev, da zastopniško gla-
sovanje staršev ne krši načel enake in neposredne volilne pravice. V nasprotju z 
zagovorniki zastopniškega glasovanja staršev avtor trdi, da starševskega glaso-
vanja v imenu otrok ni mogoče šteti zgolj kot izraz političnih preferenc otrok 
in da osebe, za katere velja, da same ne morejo sprejemati odločitev, ne morejo 
biti zastopane na takšen način. Zastopniško glasovanje staršev tem dejansko 
podeljuje dodatne glasovalne pravice, njihovim preferencam pa v postopkih 
odločanja daje dodatno težo. V drugem delu razprave so starši obravnavani kot 
mogoči zastopniki interesov otrok, pri čemer naj bi bili dodatni glasovi staršev 
namenjeni njihovi prevladi nad interesi starejših ali tistih, ki niso starši. Tako 
razumljeno zastopniško glasovanje staršev bi lahko podprli s trditvijo, da star-
ši bolje zastopajo interese svojih otrok kot povprečni volivci. Predlogi zasto-
pniškega glasovanja staršev se običajno sklicujejo na to, da imajo starši boljši 
dostop do informacij, da imajo z otroki skupne interese in/ali da so nesebični. 
Avtor pokaže, da so ti argumenti bodisi nepomembni bodisi vprašljivi in zato 
dejansko ne govorijo v prid uvedbi zastopniškega glasovanja staršev. Čeprav se 
to običajno prikazuje kot sredstvo, ki hkrati omogoča večjo demokratičnost in 
previdnost političnih odločitev, avtor opozarja, da to ne drži. In ker omenjenih 
ciljev ni mogoče doseči z eno samo institucijo, bo za njuno uresničitev treba 
poiskati več različnih metode.
Ključne besede: politično predstavništvo, glasovanje, volilna pravica, 
starostna diskriminacija, zastopniško glasovanje staršev
ENG. | This paper reviews the proposal to give parents extra votes that they 
can cast as proxies on behalf of their children. Justifications of parental proxy 
voting (PPV) are examined with a focus on various interpretations of the con-
cept of ‘proxy’. The first part of the paper assesses the notion that PPV does 
not violate the principles of equal and direct suffrage. Contrary to proponents 
of PPV, I argue that parents voting on behalf of their children cannot be con-
172
(2023) 51
journal for constitutional theory and philosophy of law
SYNOPSES, KEYWORDS, AND BIOGRAPHICAL NOTES
sidered as merely expressing children’s political preferences, and that persons 
who are taken to be unable to make a decision themselves cannot be repre-
sented in this way. Thus, PPV actually allocates extra voting rights to parents, 
giving additional weight to their preferences in decision-making. The second 
part turns to parents as possible proxies for children’s interests, with their extra 
votes being meant to overweigh those of the elderly or of non-parents. PPV 
thus understood could be supported by the claim that parents are better situated 
to represent their children’s interests than the average voter. Proposals of PPV 
usually refer to parents’ better access to information, their shared interests with 
their children, and/or their selflessness. These arguments are, however, either 
irrelevant or questionable, and do not therefore actually speak in favour of the 
introduction of PPV. In conclusion, while PPV is usually depicted as making 
political decisions simultaneously more democratic and more prudent, it does 
neither. Since these aims cannot be achieved through a single institution, differ-
ent methods to achieve each aim need to be explored.
Keywords: political representation, voting, suffrage, age discrimination, 
parental proxy voting
Summary: 1 Introduction – 2 The equal suffrage account – 2.1 Casting 
the vote – 2.2 Making the decision – 3 The unequal suffrage account – 3.1 
Information – 3.2. Altruism – 3.3 Shared interests – 4 Conclusion
Miklós Könczöl is a research fellow at the Centre for Social Sciences, In-
stitute for Legal Studies, and associate professor at Pázmány Péter Cath-
olic University, Faculty of Law, Budapest (Hungary) | E-mail: konczol.
miklos@jak.ppke.hu
173
journal for constitutional theory and philosophy of law
(2023) 51
SYNOPSES, KEYWORDS, AND BIOGRAPHICAL NOTES
Synopsis
Bartosz Brożek
Ciencia cognitiva y la naturaleza del derecho
SLOV. | Kognitivne znanosti in narava prava. Cilj članka je preučiti, kako bi 
odkritja kognitivnih znanosti lahko vplivala na ontologijo prava. Avtor najprej 
ovržbe argument, da je naša pojmovna shema - in z njo naša osnovna ontolo-
gija - apriorna v odnosu do katerega koli znanstvenega nauka. Nato na podlagi 
sodobnih evolucijskih scenarijev in nevroznanstvenih naukov skicira nastanek 
kulture. Končno pojasni, da pravilno razumevanje prava ni in ne more biti samo 
eno, iz česar sledi, da je prav mogoče razviti več konkurenčnih in vendar enako 
sprejemljivih ontologij prava.
Ključne besede: pravo, kognitivne znanosti, pravna onologija, imitacija, 
narava prava, ravnanje po pravilih
ENG. | Cognitive science and the nature of law. The goal of the paper is to con-
sider the possible impact of the discoveries in cognitive science on legal ontol-
ogy. I begin by rebutting an argument to the effect that our conceptual scheme 
– and hence our basic ontology – is a priori in relation to any scientific theory. 
Then I sketch a picture of the emergence of culture as found in the recent evolu-
tionary scenarios and neuroscientific theories. Against this background I argue 
that there is no – and there cannot be – one correct understanding of what the 
law is, which explains why it is possible to develop competing, but equally ac-
ceptable legal ontologies.
Keywords: law, cognitive science, legal ontology, imitation, the nature of 
law, rule-following
Summary: 1 Introducción – 2 La ontología jurídica se encuentra con la 
ciencia cognitive – 3 El surgimiento de la cultura – 4 La ilusoria naturaleza 
del derecho
Bartosz Brożek is Professor of Law and Philosophy, Jagiellonian Univer-
sity, Krakow (Poland) | Address: Department for the Philosophy of Law 
and Legal Ethics, Jagiellonian University, Bracka Str. 12, 31-005 Krakow, 
Poland | E-mail: bbroze@yahoo.com
174
(2023) 51
journal for constitutional theory and philosophy of law
SYNOPSES, KEYWORDS, AND BIOGRAPHICAL NOTES
Synopsis
Johan Hermansson
Structuring concepts of legal personhood
On legal personhood as a cluster property
SLOV. | Strukturirajoči pojmi pravne osebnosti. Pravne osebe se je običajno 
opredeljevalo kot subjekte s pravnimi pravicami in/ali dolžnostmi. To tradici-
onalno pojmovanje pravne osebnosti je bilo v zadnjih desetletjih postavljeno 
pod vprašaj. V pravni praksi je prišlo do tega z globalnim razvojem prava, ker 
so bile prejšnjim pravnim neosebam, kot so različne nečloveške naravne enti-
tete, nečloveške živali, zarodki in umetne inteligence, pripisane oz. Dodeljene 
pravne pravice in/ali status pravnih oseb (ali pa je bilo vsaj predlagano, da se 
jim jih dodeli). V pravnem nauku je bilo običajno pojmovanje postavljeno pod 
vprašaj med drugim s Kurkijevo teorijo pravne osebnosti kot svežnja pravic. 
Namen tega članka je podati kritično oceno pojma pravne osebnosti, ki izhaja 
iz omenjene teorije. Po obravnavi opredelitvenih strukturiranih tradicionalnih 
pojmov pravne osebnosti in Kurkijevega argumenta proti njim se avtor osredo-
toči na prvo glavno načelo teorije svežnjev, in sicer da je pravna osebnost pro-
totipno strukturirana lastnost, sestavljena iz več ločenih, a med seboj povezanih 
elementov. Avtor utrjuje, da je takšen pojem nevzdržen, saj pravna osebnost 
pravnih oseb v pozitivnem pravu ni strukturirana na takšen način. Končno pa 
avtor predlaga, da uporabljamo dvojno strukturirani pojem pravne osebnosti, 
ki ohranja prednosti prejšnjih pojmov, obenem pa se izogne njihovim glavnim 
pomanjkljivostim.
Ključne besede: pravna oseba, pravna osebnost, teorija svežnjev, Kurki 
(Visa), pojmovna struktura
ENG. | Legal persons have traditionally been understood as entities with 
legal rights and/or duties. This traditional concept of legal personhood has been 
challenged during the last decades: In legal practice through the global devel-
opment in law where previous legal non-persons such as different non-human 
natural entities, non-human animals, fetuses and artificial intelligences have 
been ascribed, or been proposed to be ascribed, legal rights and/or status as 
legal persons; in legal theory through, inter alia, Visa Kurki’s Bundle Theory of 
Legal Personhood. The purpose of this paper is to critically assess the concept 
of legal personhood proposed in the Bundle Theory. After discussing the defini-
tional structured traditional concepts of legal personhood and Kurki’s argument 
175
journal for constitutional theory and philosophy of law
(2023) 51
SYNOPSES, KEYWORDS, AND BIOGRAPHICAL NOTES
against them, I will focus on the first main tenet of the Bundle Theory: That 
legal personhood is a prototype structured cluster property which consists of 
incidents which are separate but interconnected. I will argue that such a concept 
is untenable since the legal personhood of legal persons in positive law is not 
structured in this way. Lastly, I will suggest the use of a dual structured concept 
of legal personhood that maintains the benefits of previous concepts but avoids 
their major deficiencies.
Keywords: legal person, legal personhood, the bundle theory of legal 
personhood, Kurki (Visa), conceptual structure
Summary: 1 Introduction – 2 Three ways of structuring (legal) concepts – 
3 The traditional concepts of legal personhood – 3.1 The orthodox view of 
the traditional concepts – 3.2 The extension/discrepancy problem – 3.3 The 
orthodox view and theories of rights – 3.4 Assessing the extension problem – 
4 Legal personhood as a cluster property – 4.2 The legal personhood concept 
of the Bundle Theory and its incidents – 4.2 Standing as a non-necessary 
and non-sufficient incident of legal personhood – 4.3 Beneficiary of special 
rights and ownership of property as non-necessary or non-sufficient 
incidents of legal personhood – 4.4 Three possible arguments in defense 
of a cluster property legal personhood concept – 5 Structuring concepts of 
legal personhood – 5.1 A dual structured concept of legal personhood – 5.2 
Benefits of the proposed concept
Johan Hermansson is a Doctoral Candidate in Jurisprudence at the at 
Department of Law, Lund University (Sweden) | Address: Lunds univer-
sitet, Juridiska fakulteten, Lilla Gråbrödersgatan 4–6, 221 00 Lund, Swe-
den w E-mail: johan.hermansson(at)jur.lu.se
176
(2023) 51
journal for constitutional theory and philosophy of law
SYNOPSES, KEYWORDS, AND BIOGRAPHICAL NOTES
Synopsis
Ana Lúcia Costa
Vulnerability, sexual violence and the law
A conceptual analysis
SLOV. | Ranljivost, spolno nasilje in pravo: pojmovna analiza. Ta članek pre-
učuje pojem ranljivosti in je zamišljen kot odgovor na delo Susanne Pozzolo. 
Čeprav se z avtorico strinjam o uporabnosti pojma ranljivosti kot hevristične-
ga instrumenta za izpostavljanje pomanjkljivosti v zakonih o spolnem nasilju, 
se moj pogled od njenega razlikuje v dveh točkah: v zavračanju razumevanja 
telesa kot vira ranljivosti žensk za spolno nasilje in pri upoštevanju vzajemno 
neizključujočega razmerja med ranljivostjo in avtonomijo, ki je pomembno v 
zvezi s spolnim nasiljem. Nadalje razpravljam o polisemični naravi ranljivosti 
in težavah pri uporabi tega pojma v primerih spolnega nasilja, ki izhajajo iz te 
narave. Zaključim z upanjem, ki ga navdihuje premišljevanje o feminističnem 
političnem boju za pomen ranljivosti. 
Ključne besede: ranljivost, spolno nasilje, ženske, avtonomija, soglasje, žen-
ski stereotipi
ENG. | This article examines the concept of vulnerability and is envisioned 
as a reply to the work of Susanna Pozzolo. Although agreeing with the author on 
the usefulness of vulnerability as a heuristic instrument to make visible the flaws 
on the laws on sexual violence, I differ from her in rejecting to consider the body 
as a source of women’s vulnerability to sexual violence and in considering the 
non-mutually exclusive relation between vulnerability and autonomy relevant in 
what regards sexual violence. I further discuss the polysemic nature of vulner-
ability and the difficulties with the use of the concept in cases of sexual violence, 
which arise from that nature. I finish with a note of hope by considering the 
feminist political struggle over the meaning of vulnerability.
Keywords: vulnerability, sexual violence, women, autonomy, consent, 
female stereotypes
Summary: 1 Introduction – 2 The body – 3 Vulnerability and autonomy – 
4 The polysemic nature of vulnerability – 4.1 Female stereotypes – 4.2 The 
ideal victim – 4.3 Bad women, non-ideal victims, and the denial of rape – 4.4 
Vulnerability and the law – 5 Conclusion
Ana Lúcia Costa is Margarita Salas Postdoctoral Fellow, Law and Phi-
losophy Research Group, Law Department, Pompeu Fabra University, 
Barcelona, Spain | E-mail: analucia.correiadacosta@gmail.com
177
journal for constitutional theory and philosophy of law
(2023) 51
SYNOPSES, KEYWORDS, AND BIOGRAPHICAL NOTES
Synopsis
Kenneth Einar Himma
In defence of the Coercion Thesis and modest conceptual 
methodology
A reply to critics
SLOV. | V bran tezi o prisili in zmerni pojmovni analizi. To je avtorjev odgo-
vor na Revusov simpozij o njegovi knjigi Coercion and the Nature of Law (OUP 
2020). Knjiga je imela dva glavna cilja. Prvi je bil pokazati, da je za obravnavo 
normativnega sistema kot pravnega sistema nujnen pogoj to, da sistem vključuje 
prepovedi nasilja in kraje, ki sta podkrepljeni s sankcijo, ki neupoštevanje sistema 
na razumen način odvrača v tolikšni meri, da temu omogoča minimalno izvrše-
vanje svoje naloge ohranjanja miru (to je t. i. Teza o prisili). Drugi cilj knjige sta 
bili obramba in ponazoritev metodologije zmerne pojmovne analize po oprede-
litvi Franka Jacksona. V simpoziju o knjigi, ki je bil objavljen v Revusu (2021) 45, 
so sodelovali Brian Bix, Thomas Bustamante, Frank Jackson, Paolo Di Lucia in 
Lorenzo Passerini Glazel, Anna Pintore, Pablo Rapetti in Kara Woodbury-Smith.
Ključne besede: pravo, prisila, sankcija, zmerna pojmovna analiza, 
normativnost, funkcija prava
ENG. | This is the author’s reply to a book symposium on law and coercion. 
Himma’s book – Coercion and the Nature of Law (OUP 2020) – had two prin-
cipal objectives. The first was to show it is a necessary condition for a normative 
system to count as one of law that it includes prohibitions on violence and theft 
that are backed by detriment that counts as a sanction in virtue of being reasona-
bly contrived to deter enough noncompliance to enable the system to minimally 
achieve its function of keeping the peace (the Coercion Thesis). The second was 
to defend and illustrate the methodology of modest conceptual analysis as de-
fined by Frank Jackson. The book was critically discussed by Brian Bix, Thomas 
Bustamante, Frank Jackson, Paolo Di Lucia and Lorenzo Passerini Glazel, Anna 
Pintore, Pablo Rapetti and Kara Woodbury-Smith in the 45th issue of Revus.
Keywords: law, coercion, sanction, modest conceptual analysis, normativity, 
law’s function
Summary: 1 Reply to Brian Bix – 2 Reply to Thomas Bustamante – 3 Reply to 
Paolo Di Lucia and Lorenzo Passerini Glazel – 4 Reply to Frank Jackson – 5 Reply 
to Anna Pintore – 6 Reply to Pablo Rapetti – 7 Reply to Kara Woodbury-Smith
Kenneth Einar Himma is a Continuing Guest Professor at the at Faculty 
of Law, University of Zagreb (Croatia) | Address: Trg Republike Hrvatske 
14, 10000 Zagreb, Croatia | E-mail: khimma@gmail.com