155UDK 902.01''634''>811.29
Documenta Praehistorica XXXIV (2007)
Indo-Europeanization – the seven dimensions
in the study of a never-ending process
Harald Haarmann
Institute of Archaeomythology (Sebastopol, USA), European Branch
harald.haarmann@pp.inet.fi
Introduction: A Complex Agenda
The history of Europe's peoples, cultures and langu-
ages is the history of Indo-Europeanization. This pro-cess started somewhere, some time, and it is still un-folding. In fact, the Indo-Europeanization of Europeand other regions of the world will not come to anend as long as there are non-Indo-European langu-ages and cultures that withstand the pressure of con-stant assimilation and acculturation.
As for Europe, the great majority of its recent popu-
lations speak Indo-European languages. Until theNeolithic, the proportions of non-Indo-European andIndo-European languages in Europe were the oppo-site of modern times, with Palaeo-European langu-ages of non-Indo-European affiliation dominatingthe linguistic landscape ( Haarmann 2002 ).In Western Europe, Basque in southwestern France
and northern Spain is the only surviving non-Indo-European language from antiquity ( Haarmann
1998a ). The Basque community has suffered a conti-
nual loss of its area of distribution and of the num-ber of speakers of Basque since the tenth centuryAD, under the pressure of Spanish and French.
In central Europe, Hungarian (of Finno-Ugric affilia-
tion within the Uralic language family) is like a non-Indo-European island amidst Indo-European speechcommunities. Northeastern Europe is home to a num-ber of Finno-Ugric languages. Of these, Finnish, Esto-nian and Saami are the best known. Among the Fin-no-Ugric minority languages in the European part ofRussia, processes of assimilation to the Russian-spea-ABSTRACT – This contribution focuses on the multifaceted process of Indo-Europeanization which
started out, in the Pontic-Caspian region, with the formation of a distinct ethno-cultural epicenter,the Proto-Indo-European complex. Since the late Neolithic, the Indo-Europeanization of Europe andparts of Asia produced various scenarios of contact and conflict. Altogether seven dimensions arehighlighted as essential for the study of the contacts which unfolded between Indo-Europeans andnon-Indo-European populations ( i.e., Uralians, Caucasians, ancient populations in southern and
central Europe). Selective aspects of cultural and linguistic fusion processes during the Neolithic andsubsequent periods are discussed, and the controversial term ‘migration’ is redefined.
IZVLE∞EK – Ta prispevek se osredoto≠a na ve≠fasetni proces indoevropeizacije, ki se je za≠ela na
Pontsko-kaspijskem obmo≠ju z oblikovanjem izrazitega etno-kulturnega epicentra, proto-indo-evrop-skega kompleksa. Od mlaj∏ega neolitika dalje je indoevropeizacija Evrope in delov Azije proizvedlarazli≠ne scenarije kontakta in konflikta. Predstavljenih je sedem bistvenih dimenzij, pomembnih zapreu≠evanje kontaktov, ki so so potekali med Indoevropskimi in ne-Indoevropskimi populacijami(Ural, Kavkaz, stare populacije v Ju∫ni in Srednji Evropi). Razpravljamo o selektivnih aspektih pro-cesov kulturnega in lingvisti≠nega zlitja v ≠asu neolitika ter kasneje. Ponovno smo opredelili kontro-verzni pojem »migracija«.
KEY WORDS – formation of ethnic stocks; transition to pastoralism; early language contacts; move-
ment from the steppe zone to the west; cultural fusion and linguistic convergence
Harald Haarmann
156king environment have caused a decline in the num-
ber of speakers of languages such as Mordvin, Mari,Udmurt, Komi and others ( Abondolo 1998 ).
In the Pontic-Caspian region (that is, in the area be-
tween the Volga in the west, the Caucasus in thesouth and the Ural mountains in the north), severallanguages of Turkic affiliation (as a branch of the Al-tajic language family) are spoken, among them Tatar(Kazan Tatar), Chuvash, Bashkir, Nogay, Kumyk, Ka-rachay-Balkar and others. The presence of Turkicspeech communities in that region is due to the mi-grations of Turkic tribes during the Middle Ages(Menges 1995.19–23 ). Many of the early migrant
communities such as the Huns, Avars, Khazars, VolgaBolgars, Pechenegs, Onogurs and others that had es-tablished themselves in the steppe zone and adjacentareas have vanished from the ethnographic land-scape.
The speech communities of the minorities in the east-
ern areas of the European part of Russia, of Finno-Ugric and Turkic affiliation, have experienced a wea-kening of the social functions of their languagesand, in some regions, the younger generation has nomore command of the mother tongue, which hasbeen lost to Russian ( Haarmann and Holman 1997;
2000).
In order to understand the magnitude of the Indo-
Europeanization process in the horizon of time it issignificant to shed light on its dynamic history. Thebeginnings of that dynamic process are associatedwith the circum-Pontic region and date to the Neoli-thic. The story of human populations, their culturesand languages in the area north of the Black Sea is afascinating sequence of early sustainability, internalchange and subsequent external expansion. The ba-sic processes of human ecology can be observed inthe span of time from the immediate post-glacial pe-riod to the Late Neolithic. In the course of time, thepace of cultural development accelerates to culmi-nate in the dynamic fragmentation of the Proto-Indo-European complex. During this crucial stage, whichcovers the period between c.4500 and c.3000 BC,
the process of Indo-Europeanization is set in motion.
This process has been described as a replacement of
the ancient languages of Europe by the importedIndo-European languages ( Renfrew 2002b.6–7 ). The
idea of replacement readily associates situations ofdaily life when older equipment ( e.g., a car, a TV set
or a computer) is literally re-placed by a new ma-chine. Such a notion of replacement is far from rea-listic and even misleading in this context of culture
studies ( Haarmann 2007.ch. 5.3 ).
Ancient populations, cultures and languages do not
simply vanish. They always leave traces. This is truefor the non-Indo-European peoples and their culturesin Europe and Asia that came in contact with andunder pressure from Indo-Europeans. The Indo-Euro-pean cultures and languages that spread did not re-place the local languages of different linguistic affili-ation. They entered into a process of fusion withthem, so as to produce various locally specific pat-terns of a cultural-linguistic blend of old and newconstituents.
In this contribution, new perspectives for pinpoin-
ting the beginnings of the process of Indo-Europea-nization, the area of its irradiation and its dynamicunfolding are explored. The study of this agenda isof great complexity and requires the investigation ofaltogether seven dimensions:
●the economic dimension ( e.g., the question of the
transition from foraging to pastoralist subsistence),
●the sociopolitical dimension ( e.g., the emergence
of stratified society and statehood in southeasternEurope),
●the ethnic dimension ( e.g., configurations of geno-
mic profiles of local populations in areas that wereIndo-Europeanized),
●the cultural dimension ( e.g., fusions of divergent
cultural traditions among populations in contact,such as the Mycenaean-Minoan or Celtiberian symbi-oses),
●the linguistic dimension ( e.g., patternings of indi-
genous and borrowed elements in lexical structuresand word formation; shifts in word order),
●the visual-artistic dimension ( e.g., the spread of
diagnostic imagery related to the horse, such ashorse-headed sceptres in the steppe zone, and figu-rines depicting the horse goddess as in the Celtic tra-dition),
●the mythical dimension ( e.g., the role and func-
tions of pre-Greek goddesses such as Demeter, Hes-tia, Athena and others in Greek mythology).
In the present contribution, argumentation for a se-
lection of these dimensions is presented.
The identification of the Indo-European home-
land
The agenda of Indo-Europeanization is intrinsically
interwoven with the issue of the origins of Indo-Eu-
Indo-Europeanization – the seven dimensions in the study of a never-ending process
157ropeans, their cultures and languages. Any assess-
ment of the movements of Indo-European popula-tions depends on the geographical identification ofthe homeland. The debate about the Indo-Europeanhomeland has a history of over 150 years. Some tenhomeland candidates have been seriously discussedsince the twentieth century. Of these, two are still amatter of lively debate. The two major hypothesesstand in sharp contradiction with each other:
Alternative 1 The early Indo-Europeans were agri-
culturalists and migrated from their original home-land in western Asia (Anatolia) to the west (south-eastern Europe) and to the east (Iranian plateau, In-dia).
Alternative 2 The early Indo-Europeans were pas-
toralists and migrated from their original homelandin eastern Europe (the area north of the Black Sea)to the west (to southeastern and central Europe) andto the east (into central Asia and beyond).
The issue of the Indo-European homeland is extre-
mely complex. It is not possible to identify the home-land with any certainty while applying the methodo-logy of one single scientific discipline only. Inquiriesinto the homeland agenda have been made by histo-rical linguists, anthropologists, archaeologists, ethno-graphers, geneticists, historians of religion and cul-ture, and they all have contributed to our knowledgeof the prehistoric conditions of the spread of Indo-European populations, cultures and languages.
The Pros and Cons of the modern debate have been
mapped out in a recent study ( Haarmann 2006a.
152–170 ). The author of the present contribution
takes a stand for Alternative 2 as the original home-land of Indo-Europeans. Some of the major argu-ments in favour of the northern Pontic zone as ahomeland will be summarized in the following. Be-sides arguments for a positive identification, theanalysis will also take into consideration aspects ofa negative identification, that is, arguments of exclu-sion. One of these exclusive argumentations is theevidence that the original population in southeasternEurope was of non-Indo-European stock, thus exclu-ding the validity of this region as a possible candi-date for an extended homeland.
The genetic map for southeastern Europe and west-
ern Asia shows a pattern which has been identifiedas the ‘Mediterranean genotype’ (Map 1). This is aso-called local genetic ‘outlier’ because it differs mar-kedly from surrounding patterns.In their archaeological and cultural interpretation
of genotypes, Cavalli-Sforza and his colleagues relatethe Mediterranean genotype to the geographical dis-persal of the Greek population during the times ofcolonization in the eastern Mediterranean, that is, tothe period of the early first millennium BC ( Cavalli-
Sforza et al. 1994.290–296 ; recently repeated in Ca-
valli-Sforza 2000.119–120 ). However, upon closer
inspection of the geographical profile of this geneticoutlier, it becomes apparent that the contours of theMediterranean genotype do not coincide with thehistorical boundaries of the Greek population.
The area covered by the inner genetic gradient of
the genotype expands far beyond the gravitationallimits of Greek settlements. Greeks never settled inregions situated nowadays in Bulgaria, Romania,Serbia, Albania or Bosnia-Hercegovina. They did notsettle as far as central Anatolia either, where the in-ner gradient extends to the east. Even in the westernpart of Anatolia, on the eastern coast of the AegeanSea, the Greek population concentrated in urbancenters and only very scarcely settled in rural areas.A sizeable population, however, must have shapedthe genetic profile of the region.
When inspecting the distribution of the second gra-
dient of the genetic outlier, the assumed associationof the Mediterranean genotype with Greek settle-ment becomes even more improbable. Furthermore,the question has to be asked: why would the Greekgenotype differ so radically from the genetic profi-les of the neighbouring Indo-European populationsif not for the reason that it reflects a substantial sub-stratum in the region of divergent ethnic stock? Acomparison with the profile of other principal com-ponents shows that the gradients which cover the
Map 1. The Mediterranean genotype (after Cavalli-
Sforza 1996.63 ).
Harald Haarmann
158southern Pontic zone, respectively, form a consistent
belt stretching on either side of the Bosphorus, with-out any significant profile of the Greek stock havingbeen made.
There is a more plausible explanation for the geogra-
phical extension of this genotype, and this is that itreflects the stratum of the pre-Greek population. Inall probability, the Mediterranean genotype docu-ments the density of non-Indo-European settlementsin the circum-Pontic region. In an anthropologicalperspective it becomes apparent that the populationaround the Aegean Sea stretching on the Europeanand on the Asian side of the southern Pontic zonewas ethnically homogeneous, which does not ex-clude the possibility of cultural and/or linguistic di-versity.
There is the question of time depth. When did non-
Indo-Europeans live in the circum-Pontic region? Gi-ven the possibilities of free movement between Asiaand Europe in the pre-deluge era (that is, before c.
6700 BC), it can be conjectured that demographicdiffusion happened long before 7000 BC. The non-Indo-European population of the circum-Pontic re-gion reflects the continuous presence of foragerswho had roamed the wood- and grasslands of west-ern Asia and southeastern Europe since the Mesoli-thic Age. There is clear evidence for the continuityof populations in the region from the Upper Palaeo-lithic onwards ( Bailey 2000.16–38 ). In the light of
this assumption of a very old circum-Pontic popula-tion of non-Indo-European stock, the spread of agri-culture in southeastern Europe is understood as be-ing due primarily to idea diffusion rather than themigration of agrarian settlers from Anatolia to Eu-rope ( Haarmann 1998b; Budja 2001.29–31 ).
The data provided by human genetics do not unila-
terally favour the idea of demic diffusion from west-ern Asia into Europe. There is no genetic evidencefor one big wave of population transfer during theseventh millennium BC. On the contrary, recent re-search confirms the view that, in southeastern Eu-rope, the process of Neolithization is characterizedby several small-scale movements of populationswithin geographically limited ranges ( Semino et al.
2004; Di Giacomo et al. 2004 ).
The non-Indo-European population of the circum-
Pontic region not only left a genetic ‘footprint’, butalso linguistic traces of their presence. These tracesare best preserved by toponymy and hydronomy.The non-Indo-European elements that can be iden-tified in the names of places, rivers and phenomena
of the natural environment form part of the mostancient onomastic residue. Characteristic of the ono-mastic roots of non-Indo-European origin are certainsuffixes ( i.e., -ss-, -nd-, -nth-). The formative element
-ss- is the most frequent in this group of names: Assa
(Macedonia), Bubassos (Caria), Passa (Thrace), Sar-dessos (Troad), Termessos (Pisidia), Kabassos (Ly-cia), Larissa (Thessaly), etc. ( Otkupshchikov 1973.
7–9, 20–23 ).
Certain onomastic roots occur in names on either
side of the Aegean:
European side Asian side
Alos (Thessaly) Alinda (Caria)Bargos (Illyria) Bargasa (Caria)Kurba (Crete) Kurbasa (Caria)Leba (Macedonia) Lebinthos (Caria)Oinoe (Attica) Oinoanda (Lycia)Passa (Thrace) Passanda (Caria)Prinos (Argolid) Prinassos (Caria)Sardos (Illyria) Sardessos (Troad)Sindos (Macedonia) Sinda (Pisidia)Tegea (Arcadia) Tegessos (Cyprus)
In an onomastic survey of the circum-Pontic region,
the distribution of names containing these forma-tive elements points to a balanced dispersal in Eu-rope and Asia. In the Aegean Archipelago and in theBalkans we find altogether 181 names, in contrastto 175 names in Asia Minor. The highest concentra-tion of these pre-Greek names can be observed inthe historical areas of Caria, Crete, Thrace, Thessaly,Macedonia and Troy. The onomastic material of pre-Greek origin is most verified on the islands and inthe coastal areas of the Aegean Sea: e.g., Arakynthos
(names of mountains in Aetolia, Boeotia and Attica),Tiryns (gen. Tirynthos, town in Argolis), Titaresios(river in Thessaly), Ordymnos (mountain on Les-bos), Mykonos (island in the Cyclades), Kameiros(city on Rhodes), Skiathos (island in the Cyclades)(Kati≠i≤ 1976.42–55 ).
The assumption of a pre-agrarian population of non-
Indo-European stock in the circum-Pontic regionstands in sharp contrast to the hypothesis of a pre-sumed old Indo-European population in the area.This hypothesis which was first publicized by ColinRenfrew ( 1987) and adopted by Luca Cavalli-Sforza
and others relates the spread of agriculture to earlyIndo-European migrations from western Asia tosoutheastern Europe. According to Renfrew thesemigrations were large-scale and included a greatnumber of people. If the spread of agriculture was
Indo-Europeanization – the seven dimensions in the study of a never-ending process
159related to possible migrations, this population move-
ment must have taken place in the course of the se-venth millennium BC. However, the argumentationin favour of an old Indo-European population insoutheastern Europe (see Renfrew 1999 for a re-
make of his earlier claims) fails to give convincinganswers to crucial questions (see a-d below) concer-ning the antiquity of Anatolian languages.
The hypothesis that Indo-Europeans were the auto-
chthonous population of Anatolia brings up the que-stion of who were the bearers of the high culture atÇatalhöyük? Since this culture started to flourish asearly as c.7250 BC, it would be hazardous to asso-
ciate it with cultural activities of proto-Indo-Euro-peans. Even if the Indo-European homeland is soughtin Anatolia, there would be many difficulties to linkÇatalhöyük with Indo-Europeans. The assumption ofan Anatolian homeland places this “ within eastern
Anatolia, the southern Caucasus, and northern Me-sopotamia ” (Gamkrelidze and Ivanov 1995.791 ).
This means, if Indo-Europeans had anything to dowith Çatalhöyük, they would have had to migratefirst from their homeland to western Anatolia longbefore 7000 BC, and there is no evidence whatso-ever for such an early migration. It seems muchmore reasonable to suppose “ Çatal Hüyük was part
of a different, non Indo-European, culture ” (Du-
houx 1998.31 ).
In addition to the arguments that have been brought
forward in connection with the evaluation of the Me-diterranean genotype which contradict the assump-tion of an old Indo-European population in the cir-cum-Pontic region, there is further circumstantialevidence for the absence of Indo-Europeans and forthe presence of non-Indo-Europeans there. Severalissues will be addressed here briefly.
a.Were the Proto-Indo-Europeans sea-faring? The
answer to this question is negative. In the lexical la-yers of the protolanguage as far as it can be recon-structed there is no old vocabulary relating to sea-faring. The lexical items which are associated withwater in a natural environment refer to lakes, rive-rine landscapes, marshes and swamps, but not tothe sea. It is significant that the Greek term for sea,thalassa , is of pre-Greek (non-Indo-European) ori-
gin. Since the flood of c.6700 BC destroyed the land
bridge and separated Europe from Asia (see Haar-
mann 2006b for an outline of the consequences of
that event), sea-faring must be assumed as a precon-dition for the migrations that allegedly took place inthe post-deluge period. If there had been migrationsrequiring sea-faring at that early period, then the
people involved were definitely not Indo-Europeanspeakers.
b.Were Indo-European immigrants responsible for
the promotion of sedentary life-styles on the Euro-pean side of the circum-Pontic region? In recentyears, more and more attention has been paid to thenature of processes of acculturation that might wellhave been responsible for foragers to accustomthemselves to a sedentary life-style. According tothe acculturation hypothesis (see Whittle 1996.43–
46for this terminology), the diffusion of the idea of
food production in combination with lively tradeprovided the incentive for foragers to adopt farm-ing. In connection with the spread and regional ap-pearance of seals in the archaeological assemblages,it has been stated that
“... they may indicate more structured and inten-
sive patterns of social networks and the circula-tion of goods and people over short, medium andlong-distances in the Eastern Balkans, the Pelopon-nese and Anatolia which followed the structuraltrajectories of hunter-gatherers into farmers. ” (Bu-
dja 2005.66 ).
To explain the transition to plant cultivation among
the populations in the Balkan region, the hypothesisof an immigration of farmers is not needed. Evi-dently, there is a growing tendency among archaeo-logists to favour the acculturation hypothesis. In anumber of scholarly contributions, the assumptionof a possible interconnection between Indo-Euro-pean migrations and the spread of farming has beendiscarded (see Haarmann 1998b and Dergachev
2002 for basic arguments). The crucial question of
how long the transition from foraging to farming(that is, the acculturation process) might have lastedhas still to be investigated with more scrutiny. Itmight have lasted longer in some areas than in oth-ers. The Baltic region provides well studied settingswhere the transition to farming lasted several hun-dreds of years and was associated with lively tradecontacts and inter-ethnic social relations, includingbride purchase ( Zvelebil 1996; Haarmann 2003c ).
c.Are there traces of an old Indo-European popula-
tion in Anatolia? When referring to the period of theseventh millennium BC, the answer to this questionis no. The oldest traces of the presence of a popula-tion in Anatolia which was definitely Indo-Europeancomes from Assyrian sources c.2000 BC in which
the Hittites are mentioned for the first time. The as-
Harald Haarmann
160sumption that Anatolia was originally inhabited by
a non-Indo-European population is more consistentwith reliable data than the hypothesis of this regionbeing the homeland of proto-Indo-Europeans. Thenon-Indo-European Hatti were the previous occu-pants of the later Hittite capital of Hattusa, and theirculture is dated to c.2500–2000 BC ( Akurgal 2001.
4–18).
There is a Hattic substratum in the Hittite language,
which proves that Hattic is the older language in theregion, predating the presence of Hittite. In addition,there is a functional clue relating to the status of thetwo languages which provides evidence for the factthat the Hittites were late-comers. Hattic served as aliturgical language in the Hittite state cult, a traditio-nal function of that language which was adopted bythe Hittite priesthood. If the Hatti had come as im-migrants to an area where the majority of the inha-bitants were Hittites, the Hattic language would havenever assumed the prestigious status in Hittite soci-ety which it did enjoy.
d.Is there any evidence for an old layer of Indo-Eu-
ropean languages in Anatolia, dating to the seventhmillennium BC? In fact, there is none. The oldestIndo-European languages which can be individuali-zed in Anatolia from early inscriptions are Hittite,Luvian and Palaic (see Mallory and Adams 1997.
12–17 on Anatolian languages). In terms of their af-
filiation they form two groups: Hittite-Palaic andSouthwest-Anatolian (Luvian). There are more recentcognate languages which belong to the latter group(i.e., Lycian, Lydian, Sidetic, Pisidian, Carian). If the
Anatolian languages were the remnants of a mucholder layer of Indo-European in the region, onewould expect their structures to reflect an overallpattern of archaic features.
Indeed, there are several major features of great an-
tiquity in the Anatolian branch ( i.e., the retention of
a laryngeal phoneme, numerous heteroclita, a diver-gent verbal system), but these features do not signala time depth extending to the seventh millenniumBC. Since in Hittite, the major language of the Ana-tolian branch, cognate terms are found that date tothe fourth millennium BC (see e), the final separa-tion from the Indo-European continuum cannot havehappened earlier than about 3500 BC.
The most convincing explanation of this puzzle is the
assumption of a two-phase migration movement, notaway from Anatolia but directed toward it. The spea-kers of the ancestral language of Anatolian, the bea-rers of the Suvorovo culture ( c.4500–4100 BC) in
Moldavia and Bulgaria, came to the region with thefirst migration wave of Indo-Europeans from theeast, that is, from the northern Pontic zone. The lan-guage of the Suvorovo people “ would have been ta-
ken over and transmitted to Anatolia by the nextwave of steppe immigrants (coming with wheeledvehicles), who formed the Ezero culture ( c.3300–
2700 calBC) of Bulgaria ” (Carpelan and Parpola
2001.64 ).
e.Is there any evidence for an early separation of
the Anatolian branch of languages from the rest ofthe Indo-European stock? There is none. If Proto-Indo-Europeans had migrated from Asia to Europe,this process would be somehow reflected in the re-construction of the Indo-European protolanguage.One would expect the most archaic layer of cognateterms in the cultural vocabulary to be found in Ana-tolian. However, this is not the case. Moreover, onewould not expect lexical innovations in Anatolianwhich emerged in the fifth and fourth millennia BC.And yet, in the vocabulary of Anatolian languageswe do find cognate terms for the yoking of animals(cf. PIE * iugóm > Hit yukan ‘yoke; couple, pair’,
Greek zugon ‘yoke’, Lat iugum ‘yoke’, Lith jungas
‘yoke’, etc.), for wheel and wheeled vehicles ( cf. PIE
*Hwerg
h- > Hit hurki- ‘wheel’, Toch A wärkänt
‘wheel’; variant roots in other Indo-European langu-ages), for wool ( cf. PIE * ul-na > Hit hulana ‘wool’,
Olnd urna- ‘wool’, Lat lana ‘wool’, Goth wulla
‘wool’, etc.) and other items relating to weaving(Mallory and Adams 1997.640–641, 648–649, 655 ).
It is obvious that the Anatolian branch separated
from the rest of Indo-European rather late. In anycase, a connection between Hittite, Palaic or Luvianand the autochthonous population of Anatolia can-not be conclusively established.
The non-Indo-European languages left traces, in ma-
nifold transformations, in the lexical layers of an-cient Greek. Hundreds of terms in various domainsof the vocabulary were adopted as elements of thepre-Greek substratum in the circum-Pontic region(see Strunk 2003.86–96 for an overview of research
in this field). In the archaeological record, the nume-rous relics of Aegean cultures are the most illustra-tive reminiscence of this old terminology. The ex-pressions which are “ connected by their content
with the old Aegean culture, show clearly that thederivational types with the characteristic suffixalelements belong to a definite foreign layer in theGreek vocabulary ” (Kati≠i≤ 1976.55 ).
Indo-Europeanization – the seven dimensions in the study of a never-ending process
161Among the pertinent borrowings of pre-Greek ori-
gin, we find nouns, adjectives and verbs which rep-resent foreign derivational types ( Haarmann 1995.
44–47 ). The occurrence of verbs in the repertory of
borrowings ( e.g.ancient Greek iapto ‘to throw’, dy-
namai ‘to be capable, potent’) points to the fact that
the contacts between Greeks and the pre-Greek au-tochthonous populations were intensive. Most exten-sive is the pre-Greek layer in the domain of namesfor plants. A considerable number of borrowings arealso found in terms for natural phenomena, utensils,clothing, social relations, handicrafts, etc.
The archaeological record shows continuity of set-
tlement in the areas north of the Black Sea from theend of the Ice Age (beginning of the Holocene) intothe Neolithic period. This means that the local popu-lations were indigenous and that there was no mi-gration from outside into those regions during thatspan of time. The people that lived there left theirgenetic ‘footprints’, which testify to ethnic diversity.On the genetic maps, two distinct genomic profilesare discernible (Map 2):
●a genomic concentration in an area north of the
Azov Sea which has been identified as the putativeIndo-European homeland;
●a genomic concentration further north which has
been identified as the homeland of Uralic popula-tions.
The Neolithic cultures in the area of the Indo-Euro-
pean homeland (Seroglazovo culture) and the Uralichomeland (Agidel culture) demonstrate a continuityof lithic industries from the Mesolithic period ( Mal-
lory 1989.192–193; Parpola 1999.181–187 ). The
homeland question both for Proto-Indo-Europeansand Proto-Uralians has been much debated. As for
the Uralic homeland, a nuclear area (Volga-Kama re-gion) of more concentrated settlement and an exten-sion of a more thinly populated area stretchingfrom the Baltic to the Urals have been identified(Carpelan et al. 2001 ). There is a growing consen-
sus focusing on the Caspian depression (with an ex-tension into the region between Volga and Don) asthe area of the Indo-European homeland. This hypo-thesis is seemingly being accepted by archaeologistsand linguists alike as the most plausible of all home-land candidates (see Mallory 1997; Carpelan et al.
2001; Dergachev 2005.14–40; Haarmann 1998b;2006a.154–160 for the history of this scholarly de-
bate).
Historical linguistics has reconstructed grammatical
structures and lexical roots which are similar in bothUralic and Indo-European. These linguistic traces at-test to conditions of a long-term cultural and lingui-stic convergence when both Uralians and Indo-Euro-peans were still foragers. Since the genetic ‘foot-prints’ of the ancient populations in the northernPontic area can be made visible (see Map 2), the lin-guistic reconstructions of an early period of Uralic-Indo-European convergence gain in profile. In thelanguages of both families, there is a core vocabu-lary and a set of grammatical forms which testify toa genealogical relationship (see Haarmann 2006a.
137–146 for the reconstruction of the Nostratic su-
perphylum). These elements are not borrowed ineither language family, but belong to the core inven-tory of forms inherited from oldest times (Tab. 1).
The emergence of pastoralism in the steppe zoneThe eighth millennium BC brought about decisive
environmental changes. According to Ryan and Pit-
Map 2. The genomic profiles of ancient populations north of the Black Sea. Left: the Indo-European geno-
type (after Cavalli-Sforza 2000.117 ); Right: the Uralic genotype (after Cavalli-Sforza 2000.114 ).

Harald Haarmann
162man ( 1998.157–158, 174–178 ), the circum-Pontic
zone experienced a stage of progressive desiccationafter the mid-tenth millennium BC. The second melt-water spike (beginning about 9400 BC) never rea-ched the ancient Euxine Lake, and the aridificationof the area north of the freshwater lake proceededrapidly. The ecological preconditions for the begin-nings of pastoralism among Proto-Indo-Europeansare found in the forest-steppe zone during this pe-riod. The ongoing process of desiccation in the north-ern Pontic zone caused an extension of the steppezone in the south and a receding of the forest-steppebelt to the north.
These environmental changes had long-term reper-
cussions on human ecology. Gradually, the develop-ment in the south shifted, economically, culturallyand linguistically. This was a prolonged process thatmight have taken more than a millennium to unfold.The people in the southern steppe zone experienceda socio-economic transition from foraging to herding.Since the climatic effects of desiccation enhancedthe transition to pastoralism in the Pontic steppezone, the development there in the communities ofProto-Indo-European stock detached itself from theformer socioeconomic, cultural and linguistic basisof convergence with the Proto-Uralians, resulting inthe formation of a gravitational epicentre of Proto-Indo-European culture (as distinct from the Proto-Uralian epicentre further north).In the course of the seventh millennium BC, the dif-
ferences between foraging, as the major type of proto-Uralian economy, and pastoralism, as practiced byProto-Indo-Europeans, became more marked and thegeographical zone of each type of economy moreconcentrated. 
The process of the dissolution of the former basis of
convergence and of the formation of the Proto-Indo-European epicentre was of local coinage. This meansthat – beyond the assumed internal population mo-vement in the Pontic steppe zone after the flood –there was no population influx from either the steppezone of central Asia or from the region of agrarianpopulation of Ukraine and central Europe.
Eventually, the two epicentres with their differing
ethnic stock also became characterized by divergentproto-languages, which can be reconstructed withthe methods of historical-comparative linguistics (seeBeekes 1995.124–257 for Proto-Indo-European , Haj-
dú and Domokos 1987.179–271 for Proto-Uralian). 
At first sight, it may seem problematic to conflate a
linguistic term with the assumed speakers of a lan-guage, such as to identify the northern foragers (withUralian cultural patterns who are assumed to havespoken Proto-Uralian) as ‘the Proto-Uralians,’ and toidentify the pastoralists further south (with Indo-European cultural patterns who are assumed to have
spoken Proto-Indo-European) as ‘theProto-Indo-Europeans.’ And yet, thearchaeological record indicates thecontinuity of distinct cultural pat-terns in each area where, at a laterdate, the presence of Indo-Europeanlanguages (in the Pontic steppe re-gion) and Uralian languages (furthernorth) are documented by linguisticinterferences ( Haarmann 1996.9–
10; Koivulehto 2001 ). Since there is
no evidence of population influxfrom outside, the local Neolithic po-pulations must be ancestral to the la-ter Uralian and Indo-European spea-kers of each region.
The transition from a foraging to a
pastoralist economy was accompa-nied by changes in life-styles. As partof the process of Neolithization, thistransition has been a matter of muchdebate. There are those who explainthe shift to herding and pastoralism
Tab. 1. Linguistic convergences between Proto-Uralic and Proto-
Indo-European. a) Convergent lexical roots (after Makkay 2001.
320); b) Convergences in the pronominal system (after Hajdú and
Domokos 1987.234–235 ). 
Indo-Europeanization – the seven dimensions in the study of a never-ending process
163as resulting from the spread of technologies relat-
ing to the ‘agricultural package’ (technologies ofplant cultivation and of stock-breeding) that wereintroduced to the steppe zone from the northwest-ern Pontic area. The term ‘agricultural package’ hasbeen defined as “ the sum of traits that appear re-
peatedly in the Neolithic assemblages of SW Asia,Anatolia and SE Europe ” (Çilingiroglu 2005.3 ). Ot-
hers see a direct transition without the participationof agrarian technologies and relating forms of cat-tle-raising.
As far as the Proto-Indo-Europeans and their home-
land are concerned, two basic assumptions havebeen elaborated which stand in absolute contradic-tion to one another.
Pastoralism in the steppe zone emerged inde-
pendently and its origins are not associatedwith agriculture
Although the archaeological evidence for this early
transition is scarce, historical linguistics has recon-structed an old layer of common lexical roots for thedomain of pastoralism. This terminology forms partof the core vocabulary of Proto-Indo-European, the re-constructed common basis from which all Indo-Euro-pean languages derive. The old layer of terms for her-ding “ appear to be widespread across the entire
range of IE [Indo-European] stocks .” (Mallory and
Adams 1997.7 ) (Tab. 2).
While the Proto-Indo-Europeans ex-
perienced their shift to a pastoralisteconomy, the Proto-Uralians, in theirhomeland in the forest zone furthernorth, continued to live on foraging.Therefore, such terminology relatingto early pastoralism reconstructedfor Proto-Indo-European is absentfrom the basic vocabulary of Proto-Uralic.
An inspection of the core termino-
logy of pastoralism that can be re-constructed for Proto-Indo-Europeanreveals that the diagnostic terms re-ferring to goat and sheep – the old-est known animals that played a rolein Indo-European herding – as wellas to field and herd are either wide-spread in the branches of this langu-age family (see Tab. 2, nos. 1, 2, 5, 6and 9), or seem to be best preservedin the eastern Indo-European languages (see Tab.
2, nos. 3, 4, 7 and 12). The wide distribution is anindication of the general importance of this vocab-ulary for the early Indo-Europeans. The persistenceof the old diagnostic terms, especially in the easternIndo-European languages, points to the steppe zoneas the area of pastoralism’s origin.
Based on observations about the lack of an old la-
yer of agricultural terminology in the Indo-Iranianbranch of languages, it was assumed that the pasto-ralists who spoke such languages knew nothingabout agriculture. Given the lack of old agriculturalterminology in this major branch of Indo-European,it is tempting to deny the existence of old agricultu-ral terms for the Indo-European protolanguage.
The meaning of the lexical material referring to agri-
culture is, in many cases, diffuse and does not allowthe reconstruction of a very old layer. For example,there is no old term for ‘wheat’ and no general termfor ‘barley’. The more extensive agricultural termino-logy becomes in historical languages, the younger isthe lexical layer (often relating to stages of linguisticdevelopment of the fifth millennium BC or later).
To sum up, the linguistic and archaeological evidence
speaks in favor of pastoralism as having developedindependently of farming in the steppe zone of south-ern Russia.
Tab. 2. Diagnostic terms of pastoralist economy in the Proto-Indo-
European lexicon (after Mallory and Adams 1997 ).
Harald Haarmann
164Is pastoralism an offshoot of a farming eco-
nomy?
This view has been advocated by Renfrew ( 2002a.
4–7) and others. It is argued that hunter-gatherers
would not have experienced a transition to pastora-lism without a previous stage of animal husbandry,and this would have been intrinsically associatedwith farming practises. Renfrew categorically deniesthe possibility that hunter-gatherers might have star-ted to herd wild sheep and goats – the essential ani-mal domesticates – without the parallel stage of far-ming. It is admitted that the horse was used by hun-ter-gatherers and that the early users might havebeen horse-herders. But it is denied that these horse-herders could have been horse-breeders.
It is hazardous to discard, in a discussion of Neoli-
thic economies of the seventh and sixth millennia BCin eastern Europe, any alternative a priori ( e.g.nega-
ting a direct transition from a foraging to a pastoraleconomy). There are well known examples of a tran-sition to herding and breeding without the participa-tion of farming practises from the historical period.
The earliest traces of reindeer herding date to the
fifth millennium BC, as evidenced in rock carvingsat Alta in northern Norway (see Helskog 1988 for
the pictures of Bergbukten I). Among the Saami peo-ple of the North, reindeer herding and breeding de-veloped as an independent economic system, andthere was no influence from farming communitieswith animal husbandry which would have providedthe incentive for breeding. Similar processes of atransition from hunting and gathering to reindeerherding and breeding evolved in northern Siberiaamong the ethnic groups of Samoyedic, Altajic andPaleoasiatic stock ( Funk and Sillanpää 1999.16, 39,
62, etc. ).
In the case of the Proto-Indo-European context, a pro-
minent factor gives additional weight to this assump-tion of a direct transition, and this is the chronologi-cal continuum. Pastoralism can be readily assumedto have emerged no later than the seventh millen-nium BC. However, agriculture did not reach theeastern Pontic zone prior to 5500 BC (see the iso-chrones in the map presented by Carpelan and Par-
pola 2001.63 ). Pastoralism in the region clearly an-
tedates the practise of farming. The lexical layer ofProto-Indo-European terms for herding is older thanthe terminology of the ‘agricultural package’ whicharrived on the western fringes of the steppe zone ata later time.Diagnostic items of early Indo-European cul-
ture
The horse is of special significance for the Indo-Eu-
ropeans and their culture. There is consensus aboutthis basic fact among scholars. What is disputed,though, is the process of the domestication of thisanimal and since when it was used for riding. A le-xical root for ‘horse’ (*hekuos in simplified transcrip-tion) can be reconstructed for the Proto-Indo-Euro-pean vocabulary, and this root is common for theequivalents in all the local languages of this phylum(Tab. 3). There is a linguistic feature which makesthe issue of the horse and all that is related to it dif-ficult. In the Indo-European terminology, no diffe-rence is made between the wild and the domestica-ted horse.
According to the original version of the Kurgan hypo-
thesis, propagated by Gimbutas, it was assumed thatthe Indo-Europeans left the steppes on horseback,and that it was the military supremacy of mobilehorse-riders which gave the pastoralists the edge du-ring their expansions. However, horse-riding is notattested for the fifth millennium BC. The archaeolo-gical record of the steppe zone of southern Russiapoints to c.5000 BC as an early date for the appea-
rance of the horse motif in imagery ( Gimbutas 1991.
353). The existence of imagery relating to this pro-
minent animal as such does not entail that the horsewas already domesticated at that time. The imagerymight well relate to mythical conceptualizations ofwildlife among the early pastoralists.
In the beginnings, the wild horse might have been
hunted for its meat. Most probably, the domestica-tion of this animal to become used for riding was aprolonged process ( Levine et al. 1999 ). And yet, it
seems reasonable to assert that humans “ would ra-
pidly have recognized the greater potentiality ofthe horse as a means of transport and a powerfulcultural symbol ” (Dolukhanov 2002.18 ).
Judging from ethnographic literature and from em-
pirical observations of traditional herding in recentnomadic cultures, the role of the horse may at firsthave been marginal for the socio-economy and itssignificance might have increased gradually. Shishli-na (1997) draws attention to a certain custom among
herders of the modern steppe zone in Kalmykia whokeep horses as draught animals and, for anotherspecial purpose. In winter, when the snow cover ofthe pasture may be too hard for sheep and goats tofind fodder, the horses break the cover with their
Indo-Europeanization – the seven dimensions in the study of a never-ending process
165strong hoofs and provide access for the smaller ani-
mals to the grass below. Such a function can be con-jectured to have been the first possible use of thehorse by the Kurgan pastoralists.
At a later stage, the horse was certainly used as a
draught animal. This can be assumed for the mi-grants who, coming from the eastern steppe, arri-ved at Durankulak, and later at Varna in the north-western Pontic zone. The movement of the Kurganpeople from their homeland in the Pontic-Caspianregion to the southwest can be traced on the basisof the spread of a diagnostic cultural item, the horse-headed sceptre (Fig. 1).
Those groups of steppe people who reached the
northwestern Pontic region introduced a technolo-gical innovation: wheeled wagons. Horse-riding be-came a custom at a later date. Although this meansthat the first migrants who made their incursions inthe region where the agriculturalists settled were nothorse-riders, it does not follow that the Kurgan hypo-thesis would lose its value as an explicative model.The advance of a powerful élite imposing their or-der on the local population would be a realistic sce-nario to explain the early stage of movements of theKurgan people (Kurgan I).
Non-Indo-Europeans and Indo-Europeans: sce-
narios of contact and conflict
The drifting apart of the socio-economic systems cau-
sed a shift in culture and language, too. This meantthe gradual dissolution of the older network of No-stratic convergences, with the cultural as well as lin-guistic differences between Uralic and Indo-Europeanbecoming more marked in time. The Proto-Indo-Eu-ropeans who roamed the steppe in
search of pastures for their herdshad a life-style that differed markedlyfrom that of the Proto-Uralians whohad continued as hunters and gathe-rers in the northern forest zone.
The population of the south, the
Proto-Indo-European pastoralists,did not experience a developmentof their culture and language in iso-lation. From the earliest times of theformation of the Indo-European com-plex the pastoralists engaged in con-tacts, social interaction and trade re-lations, with their neighbors in thenorth (Proto-Uralians) and further
south (Proto-Northern Caucasians).
Conflict-free contacts between pastoralists and
hunter-gatherers in the sixth and fifth millen-nia BC
Despite the differences in life-styles, the hunter-ga-
therers from the north maintained contact with thepastoralists from the south. The social interactionof populations in the former Nostratic zone of con-vergence changed its nature and transformed intopatterns of contact between bearers of distinct cultu-res and speakers of distinct languages. The contactswhich the Proto-Indo-Europeans established withthe northern neighbours date to the sixth and fifthmillennia BC. Evidence for these early contacts ofProto-Uralians with Proto-Indo-Europeans are theloanwords which were borrowed from Indo-Euro-pean into the Uralic vocabulary (Tab. 4).
The movement of lexical borrowing was unilaterally
directed from the Indo-European south to the Uralicnorth. This observation supports the assumption thatthe language of the pastoralists and their culturewere considered to be prestigious by the Proto-Ura-lians. Judging from the distribution of the old Indo-European loanwords in the sections of the Uralic le-xicon it can be concluded that the Indo-Europeansengaged in trade (see the borrowing for the idea ‘tosell’) and had prestigious goods (such as honey) andtechnologies (such as spinning and construction) tooffer to the hunter-gatherers of the north.
There were also contacts with the southern neigh-
bours, the indigenous population of the Caucasus.Most probably, interactions between Proto-Indo-Eu-ropeans and local people in the northern Caucasus
Tab. 3. The lexical root for ‘horse’ in Proto-Indo-European and its
derivations in individual Indo-European languages (after Mallory
and Adams 1997.273–274 ).

Harald Haarmann
166began to unfold in the early fifth millennium BC.
Apparently, the same kind of prestige that the Proto-Indo-Europeans enjoyed with their Uralian neigh-bours also dominated contacts with the people inthe south. Among the Indo-European borrowings innorthern Caucasian languages, we find diagnosticterms of pastoralism such as expressions for ‘goat’and ‘cattle’, lexical evidence for trade relations ( i.e.,
‘payment’), and for the transfer of trade goods ( i.e.,
‘axe’, ‘ring’); (Tab. 5).
The scenarios of contact involving Proto-Indo-Euro-
peans, Proto-Uralians and Proto-Caucasians unfoldedunder the auspices of peaceful relations, with no re-cognizable agenda of conflict. As is known from eth-nographic literature and from comparisons of theworld’s cultures, contacts between pastoralists and
hunter-gatherers are, in principle, friendly in nature,and this is because there is no competition over re-sources. The hunter-gatherers do not need the pas-tures of the pastoralists, and there is no advantagefor the pastoralists to move with their herds into thehunting-grounds of foragers. As a rule, hunter-gathe-rers assign a higher prestige to the culture of pasto-ralists and to their trade goods.
Contacts between pastoralists and agricultura-
lists with agendas of conflict
On the western periphery of the steppe, where the
terrain that was frequented by the pastoralists andtheir herds bordered the area of arable land ( i.e., in
southern Ukraine), the boun-daries between the two eco-nomic systems of pastoralism(the eastern tradition) and ofagriculture (the western tra-dition) began to float soon af-ter c.5000 BC. The initial con-
tacts between pastoralists andagriculturalists may have beenpeaceful, but things changedwhen the socioeconomicsphere of the Cucuteni-Tripil-lye culture experienced its ex-pansion to the East and newagrarian settlements were es-tablished in areas formerlyfrequented by pastoralists.
The direct consequences of
this expansion were an in-fringement of the movementsof the pastoralists and a re-duction of their resources, thepastures that had been turnedinto fields. And yet, therewere other consequences thathad an even stronger impacton the sustainability of pas-toralism in the contact region.And this had to do with theways herding as a socio-eco-nomic system operates. A trueunderstanding of the ways ofpastoralist economy is not sel-dom hampered by stereotyp-ing views that outsiders carryin their minds.
Fig. 1. The distribution of horse-headed sceptres in the steppe zone (af-
ter Dergachev 2005.85, 88 ). The geographical expansion of di stribution
(upper) and schematic spatio-temporal trajectory (below).

Indo-Europeanization – the seven dimensions in the study of a never-ending process
167“The free-moving, chaotic ‘nomad’ is a myth. The
most complex system regulates these movementsby strict formal schedules, restrictions on numbersand types of animals, reserving or deferring pastu-res, assigning members to particular pastures andcontrolling the amount of time spent in one pas-ture. Thus, the organisational complexity variesgreatly between different pastoral groups ” (Niamir
1995.245 ).
The prehistoric expansion of Tripillye settlements
into the steppe zone was more than a territorial oc-cupation of former pastures; it caused more distur-bance than an infringement of the movements ofthe pastoralists. The consequences of this expansionculminated in a shake-up of the nomadic socio-eco-nomic system and in a threat to the accessibility ofresources. The magnitude of this threat might havebeen felt differently in the regional groups of pasto-ralists, but, in principle, the world of the agricultura-lists exposed itself to the herders as harmful to theirsocio-economic sustainability.
Given these unfavorable conditions, the expansion
of the agrarian system of subsistence to the east cau-sed increasing frictionsbetween agricultura-lists and pastoralists,stirring up ever morecompetition over theexploitation of the ter-rain. During the firsthalf of the fifth millen-nium BC, under thepressure of growing so-cio-economic stress inthe local communities,there are signs of cla-shes and even warfarebetween the westernagriculturalists and the
steppe people, as eviden-ced by layers of ashesand an increased num-ber of arrow-heads inthe archaeological re-cord of the easternmostsettlements of the Cucu-teni-Tripillye culture.
By the middle of the
fifth millennium BC,ever more settlementson the eastern periphery
of the Cucuteni-Tripolye culture (Cucuteni-Tripillye,respectively) A and B1 periods were fortified, andthe frequency of arrow-heads in the archaeologicalrecord increases. The end of the Tripillye culture ismarked by its replacement, in the northern region,by the Globular Amphora culture and, in the stepperegion, by the late Pit-Grave culture.
“In the latter case, the question inevitably con-
cerns Gimbutas’ third wave of steppe invasion.However, if the equally abrupt increase in thenumber of artificially-fortified settlements of thefinal Tripolye period is due to the threat of inva-sion by other cultures, may we infer a similarquantitative increase in fortified settlements dur-ing the Cucuteni A-Tripolye B1 period which is at-tributable to similar circumstances? Following theprinciple of analogy, there can be only one expla-nation – invasion, and therefore, this completelyconfirms Gimbutas’ idea of the first wave of steppelivestock breeders. ” (Dergachev 2002.102 )
Arguably, the migrations of the steppe people find
their ultimate motivation in elementary counterreac-tions to these scenarios of unrest.
Tab. 4. The transfer of Proto-Indo-Euro pean (PIE) terminol ogy to Uralian (U)
(after Haarmann 1996.10 ).
Tab. 5. The transfer of Proto-Indo-European (PIE) terminology to Proto-Northern
Caucasian (PNC) (after Haarmann 1996.11 ).
Harald Haarmann
168Early movements of steppe people to the west
and the nature of Indo-European migrations
It is reasonable to assert that the early pastoralists
in the Pontic-Caspian zone were challenged by theadvance of agricultural practises into the region, andtheir reactions to this ‘intruding’ factor triggered achain reaction: the great Kurgan migrations. MarijaGimbutas ( 1974; 1991; 1992 ) coined this overar-
ching term to define the movement of the steppepeople, and to identify the bearers of the earliest re-cognizable Indo-European culture, that of the peoplewho built huge burial mounds, called kurgan (a
word of Turkic origin).
Gimbutas assumed that the thrust of the migrations,
which she imagined as movements of populousgroups, was effected by groups of horse-riders. If itholds true that the pastoralists were highly mobile,then this mobility on horse-back would be a decisiveasset of any of their military operations to ensuretheir migratory advance. The assumed mobility ofhorse-riding pastoralists became the target of criti-cism intended to discredit the Kurgan hypothesis.Admittedly, there is no evidence for the use of thehorse as a riding animal in the fifth millennium BC.The validity of the Kurgan hypothesis, however, isnot at the mercy of the horse as a factor to explainthe swiftness and success of the migrations.
The most crucial factor in any approach to explain-
ing the movements of the steppe people is a refine-ment of the key concept ‘migration’. Migration is acomprehensive notion, with various conceptual fa-cets in the wide array of its overall meaning ( Bell-
Fialkoff 2000 ). Migration does not exclusively mean
‘mass movement’. The process of Indo-Europeaniza-tion of the northwestern and western Pontic regionbetween c.4400 BC (beginning of the first wave)
and c.3000 BC (end of the third wave) was not ne-
cessarily the result of massive population movements.
The cultural and linguistic changes could well have
resulted from the exertion of control of a ruling éliteover people and territory either by intermarriageinto families of local dignitaries, or by assuming po-wer through conquest. As a rule, the culture of theélite dominates, and its language is more prestigiousthan that of the local population, eventually resultingin the assimilation and language shift of the latter.In this process, elements of the local language areabsorbed as a substratum by the dominating langu-age. The same holds true for cultural patterns, likethe survival of cults of female divinities among theancient Indo-European peoples in southeastern Eu-
rope ( i.e., Thracians and Illyrians).
In the northwestern Pontic region, the incursions of
the steppe people produce permanent patterns ofchange. Judging from richly equipped graves, a newsocial élite makes its appearance at Durankulak(northeastern Bulgaria) around 4600 BC and, a hun-dred years later, the tradition of burials also chan-ges at Varna. There, insignia such as a horse-headedsceptre and other ceremonial items of political po-wer provide evidence “ of the spread of steppe tri-
bes from the east to the west and in the ‘Kurgan’model of Indo-European origins is seen to reflectthe first wave of Indo-Europeans from their home-land in the steppelands of the Ukraine and southRussia ” (Mallory and Adams 1997.557 ) (Map 3).
Durankulak, Varna and other sites provide ‘diagno-
stic’ socio-cultural profiles for the establishment ofsocial hierarchy and élite power typical of nomadicsocieties.
Fusion processes of Non-Indo-European and
Indo-European elements: patterns of domi-nance and indominance
Fusion processes of cultures and languages may be
compared to the endeavour to reconstruct a shipwhile floating in the water. You can never succeedin replacing everything old by everything new be-cause the vessel has to stay afloat. Therefore, therewill always be old elements that persist, regardlessof how many new elements are introduced. TheIndo-Europeanization of the cultures and languagesof Europe did not produce strategies of ‘replace-ment’, but patterns of fusion in which older consti-tuents (of non-Indo-European origin) and youngerelements (of Indo-European coinage) intermingled.
The results of such fusion differ greatly in the regions
where non-Indo-Europeans and Indo-Europeans in-teracted, and they are specific for any local cultureand language. According to the parameter of ‘domi-nance-indominance’ the following major configura-tions can be identified in the cultures and languagesof Europe.
Scenarios of a dominance of Indo-European
elements over non-Indo-European elements
The fusion process in which the cultures of Indo-Eu-
ropeans and of non-Indo-Europeans participated insoutheastern Europe “... not only involved those ele-
Indo-Europeanization – the seven dimensions in the study of a never-ending process
169ments that survive in the archaeo-
logical record, but affected the non-material dimensions such as sto-ries, songs, myths, rituals, and be-liefs which function within com-plex webs of meaning ” (Marler
2005.60 ). This holds true for langu-
age, also, in this case for the fusionof linguistic elements of Indo-Euro-pean and of local non-Indo-Europeanorigin. I refer here to the Indo-Euro-pean language with the longest writ-ten record, Greek. The earliest re-cords in Mycenaean Greek, writtenin Linear B, date to the seventeenthcentury BC ( Haarmann 1995.125–
126).
Greek is categorized as an Indo-European language.
However, its lexical structures and its system of wordformation differ markedly from other cognate langu-ages such as Latin, Persian, Sanskrit, etc. The reasonfor this is the pre-Greek substratum, a layer of oldlexical borrowings and formative elements from theancient non-Indo-European languages that were spo-ken in southeastern Europe before the advent of theIndo-Europeans. In ancient Greek, the old loanwordsdo not represent a lexical inventory which was isola-ted from the Greek vocabulary inherited from Indo-European. Borrowed and indigenous lexical itemsform a symbiotic network of expressions. This canbe illustrated for the various terminologies, of spin-ning and weaving for one.
Evidence for the vertical loom dating to the pre-de-
luge era cannot be given, and its appearance duringthe seventh millennium BC is scarce, although evi-dence does exist from Çatalhöyük in Anatolia, Gre-ece, and the Tisza valley in Hungary. Nevertheless,the presence of loom weights in cultural strata ofthe seventh and early sixth millennia BC demonstra-tes the existence of a similar textile producing tech-nology throughout Anatolia and southeastern Eu-rope. Later, textile production proliferated and spreadto various regions. “ In short, we might be seeing
connections southward into the Aegean, as well asnorthwestward into Hungary .” (Barber 1991.98 )
It is noteworthy that in ancient Greek weaving ter-
minology there is an abundance of borrowings thathave survived from pre-Greek times. These borro-wings of non-Indo-European origin are not isolatedin the lexicon, but have been integrated into the lan-guage, forming a broad layer of terminology that issymbiotically interconnected with expressions based
on Indo-European cognates.
In the lexical structures, two integrational patterns
can be discerned that indicate the fusion and persi-stence of pre-Greek terms within Greek terminology(Tab. 6).
One is the duality of pre-Greek (non-Indo-European)
and Greek terms (of Indo-European origin):
a.In the entire terminology relating to weaving,
there are clusters of expressions with a specializedmeaning that stem from a non-Indo-European source,and others which are inherited from the bulk ofIndo-European cognates.
b.The other integrational pattern is synonymity of
pre-Greek and Greek terms, which provides the an-cient Greek language with a great potential for stylis-tic variation.
Another domain where pre-Greek ( =non-Indo-Euro-
pean) terms have survived in ancient Greek is metal-lurgy (Fig. 2). The oldest gold treasure of the worldis known from Varna and dates to c.4500 BC, to the
times of the earliest Indo-European incursions in thenorthwestern Pontic region.
Since some basic non-Indo-European expressions are
attested for the terminology of metal-working inGreek, this is evidence that this technology was notintroduced to the region by the Indo-Europeans, buthad been in use before the Kurgan migrations. Asspecialized terms, some of these loanwords of pre-Greek origin have been mediated to our modern
Map 3. The Kurgan migrations (after Mallory and Adams 1997.339 ).
Harald Haarmann
170languages via Greek civilization, among them, metal-
lonmetal and kaminos furnace. The archaeological
term Chalcolithic Age is comprised of two elementsof the pre-Greek substratum, khalkos copper and li-
thosstone ( Hofmann 1966 ).
Another area of contacts of cultures and languages
of different stock, non-Indo-European and Indo-Euro-pean, is Tuscany in Italy. On the historical map sho-wing the spread of human genes, the genetic ‘foot-print’ of the pre-Roman population ( i.e.the Etru-
scans) is recognizable as a divergent genomic profile(Cavalli-Sforza et al. 1994.278–279 ). The most pro-
minent non-Indo-European language of ancient Italy,Etruscan, was not simply ‘replaced’ by Latin, but in-fluenced the colonial language of Roman supremacy,and later Italian, in manifold ways. In the culturalvocabulary of Latin, there is a significant number ofEtruscan expressions ( Breyer 1993 ). Among the
terms which Latin borrowed from Etruscan are at-
rium atrium house, elementum element (original
meaning: ‘letter of the alphabet’), persona person,
individual, populus people and others, and many of
the old loanwords have been transferred to the lexi-con of modern European languages.
Still today, Etruscan habits of pronouncing certain
consonants are still recognizable in the sound struc-ture of the Italian dialect in Tuscany. In the area be-tween the rivers Arno and Tiber, called ‘Gorgia to-scana’ (literally ‘Tuscan throat’), the consonants k,
p and t are regularly aspirated (to be transcribed ash, ph and th): e.g.Tuscan poho little (for standard
Italian poco), lupho wolf (for lupo), ditho finger (for
dito); (Haarmann 2003a.344–345 ). The correspon-
ding consonants in Etruscan were aspirated. Mostprobably, the habits of pronouncing among thoseEtruscans who assimilated to Latin continued amonglocal people and were transferred to Italian, thedaughter language of Latin, that originated in theearly Middle Ages.
Scenarios of a balanced distribution of Indo-
European and non-Indo-European elements
Speakers of Indo-European came in contact with Ura-
lic peoples in the southern coastal region of the Bal-tic Sea. These were long-term contacts with far-rea-ching repercussions. Gradually, the speakers of Ura-lic were driven to the Northeast or they were assi-milated. Although this meant an ethnic Indo-Euro-peanization of a region with a formerly Uralian po-pulation, in the languages that were involved in thecontact, traces of a mutual influence are clearly re-cognizable.
The stress in Germanic languages is on the first syl-
lable of a word, unless the word is a loanword or iscoined on borrowed elements from another langu-age. Deviant from the principle of the first-syllable
stress is a word such asEnglish ‘replácement’,formed on the basis ofelements of Latin ori-gin, with the stress onthe second syllable.While Proto-Indo-Euro-pean had a free stress,first syllable stress isan innovation in theGermanic languages.The change of thestress pattern is anUralic substratum, thatis, it stems from con-tacts with Uralic langu-ages, where first sylla-ble stress is the rule(Suhonen 1995 ).
The Baltic-Fennic lan-
guages that continuethe tradition of Uralicin the Baltic region
Tab. 6. The symbiosis of Greek and pre-Greek terminology in the domain of wea-
ving and textile production (after Barber 1991.278, 280 ) a. The duality of Greek
and pre-Greek terms with specific meanings; b. Synonymity of Greek and pre-Greek terminology.
Indo-Europeanization – the seven dimensions in the study of a never-ending process
171know a morphophonetic phenomenon which is cal-
led ‘gradation’ and unknown in other Uralic langu-ages. Uralic languages operate with techniques ofthe agglutinative type, which means that formativeelements are associated with the word stem in away that the structure of the stem does not change(e.g.Hungarian házhouse: házak houses: házak-
ban‘in houses’: házaimban ‘in my houses’, with the
unchanged stem form ház). In Baltic-Fennic langu-
ages, the stem of words may change like in Indo-Eu-ropean languages of the inflectional type.
Among the most prominent properties of the Finnish
sound system is regular alternation of the word stem,or to be more precise: changes within the stem whichoccur in conjunction with the addition of specific for-mative elements ( Haarmann 2003b.878–882 ). These
alternations (called in Finnish astevaihtelu ‘grada-
tion’) are governed by a multiple set of specific ruleswhich cause structural changes in the stems of words.Altogether, there are 130 stem classes. Of these, 85are declension classes (of nouns), and 45 are conju-gation classes (of verbs). Attempts to reduce thenumber of classes to a few or only one have so farbeen unsuccessful.
As for the phonetic features which underlay the ma-
nifold variations of the word stem, these can be ca-tegorized as follows: consonant gradation, total or
partial consonant assimilation, vowel mutation, andvowel loss. The operation of these realizations ofchange may occur singly (simple alternation) or in acombination of various techniques (complex alterna-tion).
The realization of systematic alternation by means
of consonant gradation is the most widely appliedtechnique. In consonant gradation, two grades aredistinguished, a strong grade and a weak grade.These correlate with specific syllable types. Thestrong grade correlates with an open syllable, theweak grade with a closed syllable. Open syllablesare those ending in a vowel, closed syllables end ina consonant. The sound changes which occur whenconsonant gradation operates may be quantitative(e.g.pp: p, piippu ‘pipe/nominative’: piipun ‘pipe/
genitive’) and qualitative ( e.g.k: Ø, joki‘river/nomi-
native’: joen‘river/genitive’).
The described alternations of the word stem are a
heritage from the times when Indo-European langu-ages exerted a strong influence on the Baltic-Fenniclanguages in their formative period.
Scenarios of a dominance of non-Indo-Euro-
pean elements over Indo-European elements
Indo-Europeanization may articulate itself in certain
ways, so that despite the massive impact of Indo-Eu-ropean culture, life-style and language, there is noshift to a predominance of Indo-European constitu-ents in a local culture and language. The scenario ofIndo-European and Uralic in contact in the Baltic re-gion illustrates such proportions of fusion.
In the course of their advance into central Europe,
the Indo-Europeans who had left their homeland aspastoralists shifted to an agrarian subsistence. Agri-culture, as practised by the ancestors of the Baltictribes, reached the southern part of the Baltic regionby about 1800 BC. The emergence of the Balto-Fen-nic branch of Fenno-Ugrian (as a major subdivisionof Uralic) falls within the span of time when the Fen-nic population in the Baltic region experienced theirtransition to sedentism and plant cultivation ( c.
1500–1000 BC).
It is significant that, in the Baltic region, an exceptio-
nally prolonged phase of transition can be observed,lasting some 700 years. Among the distinct featu-res of this phase is “ the existence of mixed hunting-
farming groups, characterized by an extended sub-
Fig. 2. Objects made of gold from the cemetery of
Varna ( c.4500 BC; after Gimbutas 1991.120 ).
Harald Haarmann
172stitution phase... ” (Zvelebil 1996.328–329 ). At an
early date, the zone of mixed hunting-farming groupswas located in an area stretching from western Prus-sia and northern Poland to eastern Prussia and south-ern Lithuania. From there it gradually shifted in anortheastern direction.
Concluding from the archaeological record and, par-
ticularly, judging from the existence of mixed hunt-ing-farming groups, contacts between the southern(Baltic) agriculturalists and the northern (Fennic)foragers were friendly. The foragers had a vast hin-terland for hunting activities where they could with-draw with the spread of sedentism and the agrarianlife-style into territories which were formerly hunt-ing-grounds. In addition to this factor of ample space,the two groups engaged in mutual trade (Map 4).
Among the commodities of the north, one was par-
ticularly preferred by men in the south, namely wo-men. These were mostly obtained via bride purchase.In a cross-cultural comparison of contacts betweenagriculturalists and foragers, it can be stated that thefarming culture is viewed by both parties as moreprestigious. The higher prestige of the farming cul-ture also created images of a more advanced societyamong foragers, in a way that the communities ofthe south became more and more attractive for wo-men of the north who had a chance to marry into theprestigious society ( Haarmann 2003c.98–100 ).
Against the background of unilaterally directed pre-
stige relations, it is not surprising to observe thatthe lively social intermingling between farmers andforagers resulted in a unilaterally directed innova-tion of the social terminology among the speakers ofFennic languages. An indicator of this is the broadlayer of loanwords of Baltic origin in two sensitivesections of the basic vocabulary of Fennic languages,in kinship terminology and, in the terminology forbody parts (Tab. 7). Since prestige values were asso-ciated with the culture of the south, conceptualiza-
tions of prestige extended to also include the langu-age of the south that was involved in the contact(i.e., Baltic).
OutlookThe stage of transition from a hunter-gatherer eco-
nomy to pastoralism in the steppe zone can as yetnot been pinpointed with any accuracy in terms ofabsolute time, except for estimates of relative time inrelation to the sequence of socio-cultural develop-ments.
At present, it is not possible to distinguish different
layers of the Proto-Indo-European vocabulary accor-ding to absolute chronology ( i.e., pastoralist termino-
logy vs. agrarian vocabulary).In this domain, only state-ments about relative chrono-logy can be made, along thelines that pastoralist termino-logy must be older than theyounger – and more scarce –agricultural vocabulary. Theexclusive application of themethods of historical linguis-tics which are available atpresent ( i.e.lexico-statistical
dating) does not produce sati-
Map 4. Trade relations and socio-economic compe-
tition in the Baltic convergence zone (after Zvele-
bil 1996.338 ).
Tab. 7. Lexical borrowings of Baltic origin in Finnish (after Haarmann
2003c.98–100 ); a) Terminology of kinship and social relations; b) Termi-
nology of body parts and bodily functions.
Indo-Europeanization – the seven dimensions in the study of a never-ending process
173sfactory results for the pinpointing of the date of the
dissolution of the Proto-Indo-European complex andits dispersal into regional cultures and languages. Es-timates range from c.4500 BC to c.3000 BC.
It is essential to correlate insights about the relative
chronology of transitions and events during the for-mative period of the Indo-European complex and ofthe fusion processes induced by contacts with non-Indo-European populations to an absolute time-frame.
This task calls for interdisciplinary cooperation, ex-ploring the chronological depth of the prehistory ofthe steppe zone in an orchestrated fashion to refinedating methods in archaeology (archaeobotany), hu-man genetics (genomic profiles of ancient popula-tions and their distribution), anthropology (humanecology), studies of cultural and linguistic fusion pro-cesses, and historical linguistics.
REFERENCES∴∴
ABONDOLO D. (ed.). 1998. The Uralic languages . Rout-
ledge. London & New York.
AKURGAL E. 2001. The Hattian and Hittite Civilizations .
Publications of the Ministry of Culture. Ankara.
BAILEY D. W. 2000. Balkan Prehistory: Exclusion, Incor-
poration and Identity . Routledge, London & New York.
BARBER E. W. 1991. Prehistoric Textiles. The Develop-
ment of Cloth in the Neolithic and Bronze Ages . Prince-
ton University Press. Princeton.
BEEKES R. S. P. 1995. Comparative Indo-European lin-
guistics. An introduction . John Benjamins. Amsterdam &
Philadelphia.
BELL-FIALKOFF A. (ed.). 2000. The role of migration in
the history of the Eurasian steppe. Sedentary civiliza-tion vs. ”barbarian” and nomad . Macmillan. London.
BREYER G. 1993. Etruskisches Sprachgut im Lateini-
schen unter Ausschluss des spezifisch onomastischenBereiches . Peeters. Leuven.
BUDJA M. 2001. The transition to farming in Southeast
Europe: perspectives from pottery. In M. Budja (ed.), 8
th
Neolithic Studies. Documenta Praehistorica 28: 27–47 .
2005. The process of Neolithisation in South-eastern
Europe: from ceramic female figurines and cerealgrains to entoptics and human nuclear DNA polymor-phic markers. In M. Budja (ed.), 12
thNeolithic Studies.
Documenta Praehistorica 32: 53–72 .
CARPELAN C., PARPOLA A. 2001. Emergence, Contacts
and Dispersal of Proto-Indo-European, Proto-Uralic andProto-Aryan in Archaeological Perspective. In C. Carpelan,A. Parpola, P. Koskikallio (eds.), Early contacts between
Uralic and Indo-European: Linguistic and archaeologi-cal considerations, Suomalais-Ugrilainen Seura . Hel-
sinki: 55–150.CAVALLI-SFORZA L. L. 1996. The spread of agriculture
and nomadic pastoralism: insights from genetics, lingui-stics and archaeology. In D. R. Harris (ed.), The Origins
and Spread of Agriculture and Pastoralism in Eurasia .
UCL Press, London: 51–69.
2000. Genes, peoples, and languages . North Point
Press. New York.
CAVALLI-SFORZA L. L., MENOZZI P., PIAZZA A. 1994. The
History and Geography of Human Genes . Princeton Uni-
versity Press, Princeton. New Jersey.
ÇILINGIROGLU Ç. 2005. The concept of “Neolithic pack-
age”: considering its meaning and applicability. In M. Bu-dja (ed.), 12
thNeolithic Studies. Documenta Praehisto-
rica 32: 1–13 .
DERGACHEV V. 2002. Two studies in defence of the mi-
gration concept. In K. Boyle, C. Renfrew, M. Levine (eds.),Ancient interactions: east and west in Eurasia . McDo-
nald Institute for Archaeological Research, Cambridge:93–112.
2005. O skipetrakh. Etjudy v zashchitu migratsionnoi
kontseptsii M. Gimbutas. Revista arheologica, serie
noua, vol. 1 . Monografie, Chisinau.
DI GIACOMO F. et al. 2004. Y chromosomal haplogroup J
as a signature of the post-Neolithic colonization of Europe.Human Genetics 115: 357–371 .
DOLUKHANOV P. M. 2002. Alternative revolutions: hun-
ter-gatherers, farmers and stock-breeders in the northwest-ern Pontic area. In K. Boyle, C. Renfrew, M. Levine (eds.),Ancient interactions: east and west in Eurasia . McDo-
nald Institute for Archaeological Research, Cambridge:13–24.
DUHOUX Y. 1998. Pre-Hellenic Language(s) of Crete. Jour-
nal of Indo-European Studies 26: 1–39 .
Harald Haarmann
174FUNK D. A., SILLANPÄÄ L. (eds.) 1999. The small indige-
nous nations of northern Russia. A guide for researchers.Social Science Research Unit, Publication No. 29 . Åbo
Akademi University, Vaasa.
GAMKRELIDZE T. V. and IVANOV V. V. 1995. Indo-Euro-
pean and the Indo-Europeans, vol. I . Mouton de Gruyter.
Berlin & New York.
GIMBUTAS M. 1974. The gods and goddesses of Old
Europe, 7000–3500 B.C. myths, legends, cult images .
Thames & Hudson. London: [2
nded. 1982 as The goddes-
ses and gods of Old Europe, 6500–3500 B.C. ].
1991. The civilization of the goddess: The world of
Old Europe . Harper, San Francisco.
1992. Die Ethnogenese der europäischen Indogerma-
nen. Institut für Sprachwissenschaft der Universität
Innsbruck. Innsbruck.
HAARMANN H. 1986. Language in ethnicity. A view of
basic ecological relations . Mouton de Gruyter. Berlin &
New York.
1995. Early civilization and literacy in Europe. An
inquiry into cultural continuity in the Mediterraneanworld . Mouton de Gruyter, Berlin & New York.
1996. Aspects of early Indo-European contacts with
neighboring cultures. Indogermanische Forschungen
101: 1–14 .
1998a. Basque ethnogenesis, acculturation, and the
role of language contacts. In Fontes Lingvae Vasconvm.Studia et documenta 77: 25–42 .
1998b. On the problem of primary and secondary dif-
fusion of Indo-Europeans and their languages. Journal
of Indo-European Studies 26: 391–419 .
2000. Russische Identität und geopolitische Realitäten
im Spiegel der postsowjetischen Sprachkontakte. In L.N. Zybatow (ed.), Sprachwandel in der Slavia. Die
slavischen Sprachen an der Schwelle zum 21. Jahr-hundert – Ein internationales Handbuch, Teil 2 . Pe-
ter Lang, Frankfurt, Berlin & New York: 725–749.
2002. Lexikon der untergegangenen Sprachen . C. H.
Beck, Munich (2
nded. 2004).
2003a. Latein. In Variationstypologie . Roelcke T. (ed.),
Walter de Gruyter, Berlin & New York: 325–358.
2003b. Finnish. In Variationstypologie , Roelcke T.
(ed.), Walter de Gruyter, Berlin & New York: 866–904.2003c. Language, economy and prestige in the context
of Baltic-Fennic contacts. Studia Indogermanica Lodzi-
ensia 5: 87–102 .
2006a. Weltgeschichte der Sprachen. Von der Früh-
zeit des Menschen bis zur Gegenwart . C. H. Beck, Mu-
nich.
2006b. On the fabric of Old World civilizations: Hu-
man response to the Black Sea flood and subsequentclimatic change. Journal of Archaeomythology 2: 27–
64.
2007. Foundations of culture. Knowledge-construc-
tion, belief systems and worldview in their dynamicinterplay . Frankfurt, Berlin & New York: Peter Lang.
HAARMANN H., HOLMAN E. 1997. Acculturation and com-
municative mobility among former Soviet nationalities.Annual Review of Applied Linguistics 17: 113–137 .
HAJDÚ P., DOMOKOS P. 1987. Die uralischen Sprachen
und Literaturen . Buske. Hamburg.
HELSKOG K. 1988. Helleristningene i Alta. Spor etter ri-
tualer og dagligliv i Finnmarks forhistorie . Alta Museum.
Alta.
HOFMANN J. B. 1966. Etymologisches Wörterbuch des
Griechischen . Wissenschaftliche Buchgesellschaft. Darm-
stadt.
KATI∞I≥ R. 1976. Ancient Languages of the Balkans, vol.
1, Mouton . The Hague & Paris.
KOIVULEHTO J. 2001. The earliest contacts between Indo-
European and Uralic speakers in the light of lexical loans.In C. Carpelan, A. Parpola, P. Koskikallio (eds.), Early con-
tacts between Uralic and Indo-European: Linguistic andarchaeological considerations . Suomalais-Ugrilainen
Seura, Helsinki: 235–263.
LEVINE M., RASSAMAKIN R., KISLENKO A., TATARINTSEVA
N. (eds.) 1999. Late prehistoric exploitation of the Eura-
sian steppe . McDonald Institute for Archaeological Re-
search. Cambridge.
MAKKAY J. 2001. The Earliest Proto-Indo-European-Proto-
Uralic Contacts: An Upper Palaeolithic model. In C. Carpe-lan, A. Parpola, P. Koskikallio (eds.), Early contacts be-
tween Uralic and Indo-European: Linguistic and archa-eological considerations . Suomalais-Ugrilainen Seura,
Helsinki: 319–343.
MALLORY J. P. 1989. In Search of the Indo-Europeans:
Language, archaeology and myth . Thames and Hudson.
London.
Indo-Europeanization – the seven dimensions in the study of a never-ending process
1751997. Indo-European homeland. In J. P. Mallory, D. Q.
Adams (eds.), Encyclopedia of Indo-European culture .
Fitzroy Dearborn Publishers, London & Chicago: 290–299.
MALLORY J. P., ADAMS D. Q. (eds.) 1997. Encyclopedia of
Indo-European culture . Fitzroy Dearborn Publishers, Lon-
don & Chicago.
MARLER J. 2005. The beginnings of patriarchy in Europe:
Reflections on the Kurgan theory of Marija Gimbutas. InC. Biaggi (ed.), The rule of Mars. Readings on the origins,
history and impact of patriarchy , Knowledge, Ideas &
Trends, Manchester: 53–75.
MENGES K. H. 1995. The Turkic languages and peoples.
An introduction to Turkic Studies . Harrassowitz. Wiesba-
den (2
nded.).
NIAMIR M. 1995. Indigenous systems of natural resource
management among pastoralists of arid and semi-aridAfrica. In D. M. Warren, L. J. Slikkerveer, D. Brokensha(eds.), The cultural dimension of development. Indige-
nous knowledge systems . Intermediate Technology Pub-
lications. London.
OTKUPSHCHIKOV Y. V. 1973. Balkano-maloaziyskie topo-
nomicheskie izoglossy. In Balkanskoe Yazykoznanie .
Moscow: 5–29.
PARPOLA A. 1999. Varhaisten indoeurooppalaiskontaktien
ajoitus ja paikannus kiellelisen ja arkeologisen aineistonperusteella. In P. Fogelberg (ed.), Pohjan poluilla. Suo-
malaisten juuret nykytutkimuksen mukaan . Suomen
Tiedeseura. Helsinki: 180–206.
RENFREW C. 1987. Archaeology & Language. The Puzzle
of Indo-European Origins . Jonathan Cape. London.
1999. Time depth, Convergence Theory, and Innova-
tion in Proto-Indo-European: ‘Old Europe’ as a PIE Lin-guistic Area. Journal of Indo-European Studies 27:
258–293 .
2002a. Pastoralism and interaction: Some introductory
questions. In K. Boyle, C. Renfrew, M. Levine (eds.),Ancient interactions: east and west in Eurasia . McDo-
nald Institute for Archaeological Research, Cambridge:1–10.2002b. “The emerging synthesis”: the archaeogenetics
of farming/language dispersals and other spread zones.In P. Bellwood, C. Renfrew (eds.), Examining the far-
ming/language dispersal hypothesis . McDonald Insti-
tute for Archaeological Research, Cambridge: 3–16.
RYAN W., PITMAN W. 1998. Noah’s Flood. The new scien-
tific discoveries about the event that changed history .
Simon and Schuster. New York.
SEMINO O. et al. 2004. Origin, diffusion, and differentia-
tion of Y-chromosome haplogroups E and J: Inferences onthe Neolithization of Europe and later migratory eventsin the Mediterranean area. American Journal of Human
Genetics 74: 1023–1034 .
SHISHLINA N. I. 1997. Zametki o kharaktere skotovodches-
kogo khozyaistva v sovremennoi republike Kalmykiia. InP. M. Kozhin (ed.), Step’ i Kavkaz (kul’turnye traditsii).
Trudy gosudarstvennogo istoricheskogo muzeia 97:106–109 .
2004. North-west Caspian sea steppe: Environment and
migration crossroads of pastoral culture populationduring the third millennium BC. In E. M. Scott et al.(eds.), Impact of the environment on human migra-
tion in Eurasia . Kluwer Academic Publishers, Amster-
dam: 91–106.
STRUNK K. 2003. Vorgriechisch/‘Pelasgisch’: Neue Erwä-
gungen zu einer älteren Substrathypothese. In A. Bam-mesberger and T. Vennemann (eds.), Languages in pre-
historic Europe . Winter, Heidelberg (2
nded.): 85–98.
SUHONEN S. (ed.). 1995. Itämerensuomalainen kulttu-
urialue/The Fenno-Baltic cultural area . Suomalais-Ugri-
lainen Seura. Helsinki.
VIDEYKO M. Yu. 2003. Tripil’s’ka tsivilizatsiia . Akadem-
periodika Kiev (2 nded.).
WHITTLE A. 1996. Europe in the Neolithic: The Creation
of New Worlds , Cambridge University Press. Cambridge.
ZVELEBIL M. 1996. The agricultural frontier and the tran-
sition to farming in the circum-Baltic region. In D. R. Har-ris (ed.), The origins and spread of agriculture and pa-
storalism in Eurasia , UCL Press, London: 323–345.