166
Documenta Praehistorica XLVIII (2021)
Introduction
The site Stanovoye 4, located in the Ivanovo area of
the Upper Volga region, in western Russia, was ex-
cavated by Mikhail Zhilin, between the years 1992
and 2002. The site consists of four Mesolithic archa-
eological layers, a layer dating to the early phase of
the so-called ‘Forest Neolithic’ (characterized by the
presence of ceramic but without farming or pasto-
ralism) as well as a Middle Neolithic layer. Because
of the site’s distinct and radiocarbon-dated stratigra-
phy, as well as good organic preservation conditions,
its chronology and technological development can
be explored in detail. Furthermore, the site’s loca-
tion in the centre of the Upper Volga region makes
it relevant for studies of regional and transregional
contacts and communications from the very begin-
ning of the Mesolithic until the Middle Neolithic
(Zhilin 2002; 2009).
The Upper Volga region is likely a relevant area for
the spread of several technological concepts into
Fennoscandia and northern Continental Europe dur-
ing the Mesolithic (cf. Rankama, Kankaanpää 2007;
2008; Hartz et al. 2010; Sørensen et al. 2013; Kan-
kaanpää, Rankama 2014; Damlien 2016; Manni-
nen et al. 2021). The chronological bases for these
technological dispersions rest on more than 150
Revising the chronology of Stanovoye 4
in the Upper Volga Region (Western Russia)
Sandra Söderlind1, Mikhail Zhilin2
1 ROOTS Young Academy, Kiel University, Kiel, DE< ssoederlind@roots.uni-kiel.de
2 Institute of Archaeology of Russian Academy of Sciences, Moscow, RU< mizhilin@yandex.ru
ABSTRACT – This paper examines the Stanovoye 4 site-chronology, which was developed through
several radiocarbon dating efforts throughout the last two decades. Today, the dates indicate long-
term cultural traditions at the site, lasting over 3000 years. The goal of this paper is to understand
the site chronology holistically, which is done through a critical review of all available dates and sam-
ple characteristics. Additionally, the choice of sample materials, dating methods and preservation
practices will be discussed. Furthermore, the dating of PVA-consolidated samples is discussed from
an archaeological standpoint.
IZVLE∞EK – V ≠lanku predstavljamo kronologijo najdi∏≠a Stanovoye 4, ki je bila oblikovana v zad-
njih dveh desetletjih na podlagi ∏tevilnih radiokarbonskih datumov. Trenutni podatki ka∫ejo, da je
poselitev na najdi∏≠u del dolgotrajne, ve≠ kot 3000-letne kulturne tradicije. V ≠lanku posku∏amo ra-
zumeti kronologijo najdi∏≠a na holisti≠ni na≠in, in sicer s pomo≠jo kriti≠ne recenzije vseh dostopnih
datumov in zna≠ilnosti vzorcev. Poleg tega razpravljamo tudi o izboru materiala vzorcev, metodah
datiranja in za∏≠itnih praksah. Ob tem iz arheolo∏kega vidika obravnavamo tudi datiranje vzorcev,
ki so bili utrjeni s PVA lepili.
KEY WORDS – radiocarbon dating; Mesolithic; Neolithic; chronology; the Upper Volga region
KLJU∞NE BESEDE – radiokarbonsko datiranje; mezolitik; neolitik; kronologija; obmo≠je zgornjega
toka reke Volge
Revizija kronologije najdi[;a Stanovoye 4
iz obmo;ja zgornjega toka reke Volge (zahodna Rusija)
DOI> 10.4312\dp.48.13
Revising the chronology of Stanovoye 4 in the Upper Volga Region (Western Russia)
167
The Lahost River runs from the bog and through the
site. The stratigraphy at the site is complex with high-
ly dynamic sedimentation processes, which also dif-
fer slightly across the site’s two main trenches, i.e.
trenches 2 and 3 (Fig. 2) (Zhilin 2002; 2009; Zhilin,
Matiskainen 2003). A smaller trench (1) is situated
c. 50 meters west of trench 2 (Zhilin 2002) but will
not be dealt with here due to its small assemblage
and lack of dates. The stratigraphies from trench 2
and 3 have been subject to sediment analysis and
substantial plant microfossil analysis, which provid-
ed insights on the local sedimentation processes
(Zaretskaya et al. 2005). Pollen analysis has also
been useful for reconstructing the prehistoric envi-
ronments surrounding the site (Aleshinskaya 2001).
Environmental processes of sedimentation at
the site
The sediment layers that make up the site are num-
bered 1–9 and the archaeological layers are refer-
enced by Roman numerals I-IV. The following sedi-
mentation process at the site is based on the previ-
ous investigations by Anna S. Aleshinskaya (2001)
and Nataliya E. Zaretskaya et al. (2005).
The stratigraphy and chronology of the site starts
at the Early Mesolithic ground level, which consists
of washed moraine. The artefacts found on top of
this moraine are related to archaeological layer IV.
At the time the site was situated on, and in the near
vicinity of, the ancient shoreline. Due to rising water
levels the site was later submerged and artefacts
from archaeological layer IV became covered by se-
diment layer 8. The analysed sedimentary samples
radiocarbon dates from around 30 sites, several of
which are stratified peat bog sites (Zhilin 2009).
The Stanovoye 4 site has already been dated on se-
veral occasions: Nataliya Zaretskaya et al. (2005),
Sönke Hartz et al. (2010), Bente Philippsen (2019)
and most recently by the authors of this paper. In
this article, we calibrate and evaluate all previous-
ly dated samples, along with the newly acquired
dates. The large number of radiocarbon-dated sam-
ples, done during the last two decades, come with
various source critical aspects. Here, we review the
use of certain sample materials, varying dating me-
thods and problematic preservation practices in
order to explore and map the many source critical
factors that play into the chronology of the Stanovo-
ye 4 site, in an effort to better understand it.
This will be done by detangling various fractions of
radiocarbon samples that are characterized by dif-
ferent source critical issues and instead use these
fractions to create individual chronologies. After this
(divide-and-conquer-like) procedure, the chronolo-
gies will be compared and evaluated to reach a ge-
neral chronology for the site. This chronology, along
with the stratigraphic and archaeological evidence
from the site, will provide a more comprehensive
understanding of human activities during the early
Holocene.
The site and stratigraphic sequences
The Stanovoye 4 site is located on a sloping promon-
tory stretching into the Podozerskoye peat bog, form-
ed in the terrace of a previous glacial lake (Fig. 1).
Fig. 1. On the left, a red star marks the location of Stanovoye 4 in Russia. On the right, plan of the site
(©EuroGeographics for the administrative boundaries).
Sandra Söderlind, Mikhail Zhilin
168
from layer 8 consist of organomineral silt and gyttja,
which along with the plant compositions indicate that
sedimentation took place in an environment cha-
racterized by cooler water temperatures (in trench
2) and shallow littoral sediments in the form of gyt-
tja (in trench 3). The pollen and microfossil analy-
ses, along with the acquired dates, thus suggest that
sediment layer 8 and archaeological layer IV were
deposited at the very end of the Younger Dryas and
at the beginning of the Preboreal (Aleshinskaya
2001; Zaretskaya et al. 2005).
Gyttja (related to layer 8) continued to accumulate on
top of layer IV in the following centuries. This is due
to rising water levels of the nearby lake during the
warmer temperatures at the beginning of the Prebo-
real. This is also supported by pollen and microfossil
data (Aleshinskaya 2001; Zaretskaya et al. 2005).
Subsequent sediments, from the upper part of sedi-
ment layer 8, in trench 3, show that the water level
of the nearby lake was reduced again in the context
of a Preboreal landscape, as based on plant microfos-
sils and pollen (Zhilin, Matiskainen 2003; Zaretska-
ya et al. 2005). During this regression, archaeologi-
cal layer IIIa was accumulated. About 100 years later,
another transgression is visible in the sediments
together with changes in river activity (Zaretskaya
et al. 2005). In the same stratigraphic horizon, but
in trench 2, two thin wooden stakes were found be-
tween the two archaeological layers IV and III (Zhi-
lin 2002). This area of the stratigraphy will hence-
forth be referred to as sediment layer IV/III.
The obtained radiocarbon dates show a start of se-
dimentation of layer 8 in trench 3 in the first half of
the 10th millennium cal BC and it continues until the
first half of the 9th millennium. In trench 2, the se-
dimentation of layer 8 is dated from the second half
of the 10th millennium to the second half of the 9th
millennium (Zaretskaya et al. 2005).
Archaeological layer III is embedded within sedi-
ment layer 5 in both trenches. In trench 3, pollens
from this layer indicate that sedimentation took
place during the second half of the Preboreal. Mean-
while, pollens from the same layer in trench 2 are
characteristic of a landscape during the first half of
the Boreal (Aleshinskaya 2001). The radiocarbon
dates obtained from trench 3 show a start of sedi-
mentation of archaeological layer III around the
middle of the 9th millennium, continuing into the
first half of the 8th millennium cal BC. In trench 2,
the sedimentation of layer 5 is instead dated to the
end of the 9th millennium and continuing into the
second half of the 8th millennium cal BC (Zaretska-
ya et al. 2005).
The sediment layer 4/5 in trench 2 consists of yel-
low sand, which contains redeposited artefacts that
Fig. 2. Stratigraphies from trenches 2 and 3. Geological layers are numbered 1–9, and archaeological
layers using roman numerals I-IV. 14C ages are in uncal BP (see Table 1, available at http://dx.doi.org/
10.4312/dp.48.13).
Revising the chronology of Stanovoye 4 in the Upper Volga Region (Western Russia)
169
probably come from layer III. This layer is inter-
preted as relating to a series of flooding events, in
which finds were washed out from the slopes of the
promontory (situated north of trench 2), moved
downstream and subsequently redeposited in the
riverbanks. Pollen analysis of related sediments in
trench 2 indicates that this occurred during the sec-
ond half of the Boreal. Later on, the lake level again
rose, covering layer 4/5 (in trench 3) with gyttja.
The bottom part of this gyttja layer contains archa-
eological layer II, preserved in situ. Pollen data from
this layer indicate sedimentation during the Early
Atlantic period (Aleshinskaya 2001; Zhilin, Matis-
kainen 2003). Radiocarbon dates from the bottom
part of the gyttja layer (layer 4 in trench 2) indicate
that this sedimentation started in the first half of
the 6th millennium cal BC and continued throughout
the following centuries (cf. Zaretskaya et al. 2005).
The Middle Neolithic layer I in trench 2 is embedded
in gyttja, which was deposited during the Atlantic
period (Aleshinskaya 2001).
Cultural and technological traditions at the site
The earliest settlement phase on the Stanovoye 4 site
is represented by archaeological layer IV, in both
trenches. The archaeological materials consist of 154
lithic artefacts and c. 54 bone and antler tools. The
typology and technology of these finds indicate that
they belonged to the Early Mesolithic Butovo tech-
nocomplex (cf. Zhilin, Matiskainen 2003; Zhilin
2007; 2009).
Artefacts from this layer showcase a variety of tech-
nological concepts and techniques. The production of
larger blades was done by means of indirect (punch)
techniques while smaller blades were produced using
pressure techniques from a variety of core types
(Zhilin 2009; Hartz et al. 2010). Bone and antler ar-
tefacts were made from a variety of animal species,
including elk and beaver (Zhilin 2003; 2007; 2009;
Zhilin, Matiskainen 2003).
The subsequent archaeological layer is found solely
in trench 3 and is represented by layer IIIa. The re-
lated lithic assemblage is scarce but includes a few
chert flakes, a flint knife, an asymmetric trapeze and
an oblique one-edged arrowhead, the latter charac-
teristic of the Ienevo technocomplex, although the
round scrapers from this layer are typical for the Bu-
tovo technocomplex. The bone and antler assem-
blage includes artefacts such as a large arrowhead
with an irregular sub-biconical head, a gouge, a frag-
ment of a knife and a grooved long bone. Other or-
ganic finds include the remains of a wooden fish
trap found with a large stone sinker adjacent to it
(Zhilin, Matiskainen 2003).
The following settlement period is represented by
archaeological layer III, found in both main trench-
es. This assemblage shows typological and techno-
logical features related to the Middle Mesolithic pe-
riod, and the Butovo technocomplex, as described
above (Zhilin 2003; 2007; 2009; Zhilin, Matiskai-
nen 2003). The bone and antler assemblages from
this layer are extensive, including a variety of tools
and artefacts, along with many wooden stakes which
have been interpreted as fishing weirs (cf. Zhilin,
Matiskainen 2003).
Archaeological layer II was recovered only in trench
2. Its assemblage includes a small number of flint
and bone finds as well as thick-walled ceramic frag-
ments. This suggests the presence of an economy
based on hunting, gathering and fishing as well as
the use of pottery, which is characteristic of the ear-
liest stage of the Early Neolithic Upper Volga cultu-
ral complex (Zhilin, Matiskainen 2003).
Archaeological layer I was excavated in trench 2. It
yielded several Middle Neolithic flint artefacts and
small bone fragments (Zhilin 2002).
Materials and methods
A total of 49 radiocarbon samples are currently avail-
able from the Stanovoye 4 site (Tab. 1; available at
http://dx.doi.org/10.4312/dp.48.13). These include
four artefact samples (see Fig. 3) that have not been
published previously. In an effort to understand the
characteristics of the samples, aspects of storage and
sampling will be described. The methods involved in
the sample subgrouping will also be explained.
During and after excavation
During excavation, finds and samples were stored in
plastic bags. After the excavations the organic re-
mains were cleaned with water and subsequently
dated or stored. Some of the wooden and organic
samples were dated shortly after excavation (and
published by Zaretskaya et al. 2005), while other
samples were consolidated using one layer of a
(10%) PVA solution made up of water and carpen-
ter’s glue (bought in the late 1990s). After the conso-
lidation samples were stored in plastic bags.
Two fragments of wooden handles (KIA-35153 and
KIA-35157) that were found inside antler mattocks 
Sandra Söderlind, Mikhail Zhilin
170
(sleeves) were put in plastic bags, while still wet,
and kept there for a decade before being dated. Ac-
cording to Leopold D. Sulerzhitsky (personal com-
munication) such a situation can lead to modern car-
bon being introduced into samples from surround-
ing air as a result of fungal activity inside the wood.
There is therefore a possibility of modern contami-
nation in these samples.
Sampling and dating of artefacts
The sampling of artefacts from the site has mainly
been done by Mihail Zhilin who followed a proce-
dure that starts by mechanically removing the sur-
face of the bone artefacts in the area of sampling, in
an active effort to avoid any surface contaminators.
For some artefacts, such as larger antler sockets, a
layer of up to 3mm of the surface was removed. Sub-
sequently, a hole was drilled into the interior of the
bone, creating the bone powder used for dating. For
other artefacts, such as smaller and more delicate
beaver mandibles, the surfaces were more lightly
scraped off before drilling. Some artefacts were, how-
ever, sent to the lab for cleaning and sampling (and
subsequently published by Hartz et al. 2010).
Samples have been dated at different times and by
various labs, including the Geological Institute at the
Russian Academy of Sciences (RAS) (lab-code GIN),
Leibniz Laboratory for Radiometric Dating and Sta-
ble Isotope Research in Kiel (lab-code KIA), Aarhus
AMS centre (lab-code AAR) and the Groningen AMS
lab (lab-code GrA). The newly dated radiocarbon
samples were also dated at the lab in Kiel. The diffe-
rent labs follow similar protocols for preparing/cle-
aning the samples and collagen extraction (for in-
dividual lab procedures, see Appendix below).
The samples were calibrated using the most recent
r:5 Atmospheric IntCal20 dataset (Reimer et al.
2020) in Oxcal v 4.4.2 (Bronk Ramsey 2021).
Subgrouping the samples
In an effort to exclude certain source critical issues,
relating to various sample materials, dating meth-
ods and consolidation efforts, the samples will be
subgrouped into different datasets. Each dataset will
subsequently undergo calibration and individual
chronologies will be presented, before a more gene-
ral chronology for the site is discussed.
The attribution of samples in different datasets (Tab.
2) needs to be addressed. All samples are attributed
to one of the following datasets: 1 (sediments), 2
(artefacts) or 3 (ecofacts). Each sample can be clear-
ly attributed to one of these groups. However, for
datasets 4–7, it is possible to find individual samples
in several datasets, since one sample can be both
AMS-dated (dataset 4) and consolidated (dataset 7).
Excluded samples
Some of the previously dated sediment samples un-
derwent both cold and warm extractions, resulting in
two dates for each sample (as noted in Table 1; avail-
able at http://dx.doi.org/10.4312/dp.48.13). These
were sometimes not in agreement. In this study, we
decided to use the dates from the warm extraction
whenever the two fractions did not agree (see Ap-
pendix below).
Reviewing and calibrating the datasets
Dataset 1 – Sediments
Sediment samples and source critique
The dataset includes samples of silt and gyttja. These
samples were 2–4cm thick and were extracted from
the trench wall. In general, the use of sediments such
as gyttja, silt and peat for radiocarbon dating has
been criticized for several reasons, mainly due to
the complexity of sedimentation processes and diffe-
rent potential origins of the accumulated carbon inFig. 3. The newly dated artefacts.
Revising the chronology of Stanovoye 4 in the Upper Volga Region (Western Russia)
171
a small sample (Zaretskaya et al. 2001; Taylor, Bar-
Yosef 2014.91–92).
An issue relating to the sediment samples from Sta-
novoye 4 is the small representation of samples,
with only a few samples from each layer. Moreover,
despite the sediment layers containing archaeologi-
cal materials/layers, the sediments themselves are
not anthropogenic remains. Therefore, they don’t
directly date the occupational phases or human ac-
tivities on the site.
Calibrated sediment dates
Based on the sediment dates from Stanovoye 4 (Fig.
4), the acquired site chronology seems to reflect 4–
5 phases in a longer continuous sedimentation pro-
cess. The samples from layer IV date to the final sta-
ges of the Younger Dryas and first half of the Prebo-
real, spanning 10 500–9300 cal BC. The two oldest
dates have extensive standard deviations, amount-
ing to 500–600 years, making an exact starting point
for habitation difficult to interpret.
The samples from sediment layer IV/III date to the
Preboreal, specifically between 9300 and 8700 cal
BC. Samples from layer III seem to divide into two
clusters, one at around 8300–7900 cal BC and ano-
ther at 7800–7500 cal BC. However, there are too
few samples to confidently argue for two separate
occupational phases. Lastly, samples from layer II
date to the Atlantic period, to c. 6600–5800 cal BC.
Dataset 2 – Artefacts
Artefact samples and source critique
The artefacts from the site include bone and antler
artefacts as well as worked wooden handles and sta-
kes. In general, bone and antler samples come with
a number of source critical aspects. For instance,
aquatic animals and marine-feeding animals can
store higher amounts of aquatic carbon in their
bones, resulting in an apparently older date, known
as the marine/freshwater reservoir effect (Philippsen
2013; Taylor, Bar-Yosef 2014.71–74).
The dated bone samples from Stanovoye 4 do not
include any aquatic fauna that can be subject to the
marine/freshwater reservoir effect. Nevertheless,
many of the dated bones and antler samples come
from elk (Alces alces), which is a species that has
been argued to be at risk of the freshwater reservoir
effect due to their high consumption of aquatic
plants (Philippsen 2019). Further research on this
topic, which also included some finds from Stanovo-
ye 4, has indicated little evidence for this relation-
ship in general, although the finds from the site did
suggest some temporal differences between an elk
antler socket and its wooden handle (cf. Philippsen
2019). However, other reasons than the freshwater
reservoir effect have been proposed for this (ibid.).
These finds will also be discussed later in this arti-
cle. Other dated bones from Stanovoye 4 include
beaver (Castor fiber), which is a semi-aquatic ani-
mal with a terrestrial diet that can be excluded from
the risk of the freshwater reservoir effect.
Dataset Contents Included samples
1. Sediment samples Silt and gyttja samples. GIN-8379, GIN-10111 II, GIN-10126 II, GIN-10106 II, GIN-10112 II,
GIN-10127 I, GIN-10108 II, GIN-10122 II, GIN-10127 II, GIN-10109 II,
GIN-10125 II, GIN-10110 II, GIN-10126 I
2. Artefact\ Bone\antler artefacts AAR-22231, GIN-8854, KIA-35157, AAR-22232, GIN-8856, KIA-35158,
anthropogenic and one unworked GIN-8374, GrA-34084, KIA-39316, GIN-8375, KIA-35152, KIA-39317,
samples human bone. GIN-8376, KIA-35153, KIA-53778, GIN-8377, KIA-35154, KIA-53779,
Worked wooden finds. GIN-8378, KIA-35155, KIA-53780, GIN-8853, KIA-35156, KIA-53781
3. Ecofact samples Unworked bone\wood GIN-10113, GIN-10128, GIN-11093a
4. AMS-dated samples Artefacts dated AAR-22231, KIA-35155, KIA-53778, AAR-22232, KIA-35156, KIA-53779,
using AMS. GrA-34084, KIA-35157, KIA-53780, KIA-35152, KIA-35158, KIA-53781,
KIA-35153, KIA-39316, KIA-35154, KIA-39317
5. Conventional dated Artefacts and ecofacts GIN-8374, GIN-8378, GIN-10113, GIN-8375, GIN-8853, GIN-10128,
samples dated using conventional GIN-8376, GIN-8854, GIN-11093a, GIN-8377, GIN-8856
dating methods.
6. Non-consolidated Artefacts and ecofacts GIN-8374, GIN-8853, GIN-11093a, GIN-8375, GIN-8854, GrA-34084,
samples that were not GIN-8376, GIN-8856, KIA-35153, GIN-8377, GIN-10113, KIA-35157,
consolidated. GIN-8378, GIN-10128, KIA-39317
7. Consolidated Artefacts that were AAR-22231, KIA-35155, KIA-53778, AAR-22232, KIA-35156, KIA-53779,
samples (PVA) consolidated. KIA-35152, KIA-35158, KIA-53780, KIA-35154, KIA-39316, KIA-53781
Tab. 2. The dated samples from Stanovoye 4 are attributed to seven different datasets, characterized by
sample material, dating method and consolidation efforts. Note that each sample can be attributed to
several datasets.
Sandra Söderlind, Mikhail Zhilin
172
A human bone (GrA-34084) was also included into
the group due to its direct connection with human
activity at the site. Since Stanovoye 4 is located far
away from any marine environment, the marine re-
servoir effect can be excluded from the human bone.
However, due to the site’s location on the shore of
the nearby lake, a risk of a freshwater reservoir ef-
fect in the human bone could be considered (Philip-
psen 2013; Meadows et al. 2020).
However, we assume that the dated human bone
from Stanovoye 4 was not affected by the freshwa-
ter reservoir effect as suggested by the isotope value
(δ13C= –21.08%%) (cf. Alexandrovskiy et al. 2009).
Additionally, the large amounts of mammal bones
and smaller amounts of fish bones found in the same
layer as the human bone (layer III, trench 3) has led
to the interpretation that fish played only a minor
part in the diet of the site inhabitants at that time
(Zhilin 2004).
Generally, wooden artefacts can show an old-wood
effect. This means that an artefact can appear older
than its actual age since it is made up of several
years’ worth of tree growth and corresponding ages
(Taylor, Bar-Yosef 2014.67–68). The dated wooden
artefacts from Stanovoye 4 include several worked
stakes (often made of birch), which vary between 6
and 10cm in diameter. Additional finds include a
plank which is 1.2m long and 2cm thick and sever-
al handles (for antler sockets) which are 2–3cm in
diameter. The wooden artefacts from Stanovoye 4
were probably not affected by the old wood effect
due to their smaller sizes and fast-growing wood
species.
Wooden stakes in general come with some additio-
nal contextual ambiguity as they tend to pierce seve-
ral stratigraphic layers and archaeological contexts.
This can make their exact contextual information dif-
ficult to assume, which decreases their contextual re-
liability compared to other artefacts. The stratigra-
phic positions of dated stakes from Stanovoye 4 are
clearly defined, and their relationships with archaeo-
logical layer III (in both trenches) are reliable (Zhi-
lin 2003). The dated wooden handles are at risk of
modern contamination due to the way they were
stored prior to dating (as previously mentioned).
Calibrated artefact dates
The calibrated artefact samples from Stanovoye 4
(Fig. 5) suggest a more or less continuous chronolo-
gy at the site. Samples from layer IV span a time be-
tween 10 000 and 8700 cal BC. The oldest date dis-
plays a very large standard deviation, making the
start of this phase imprecise. Some samples from la-
yer IV match the age of two wooden stakes found in
layer IV/III (GIN-8374 and GIN-8376) as well as a
wooden stake, lying horizontally, in layer IIIa (GIN-
8377). This is likely a result of the calibration curve
Fig. 4. Calibrated sediment samples from the site. The colours indicate their find contexts: grey archaeo-
logical layer IV; light blue layer IV/III; dark blue archaeological layer IIIa; blue archaeological layer III,
and turquoise archaeological layer II.
Revising the chronology of Stanovoye 4 in the Upper Volga Region (Western Russia)
173
plateaus at this time, and would indicate that the
age of several of the artefacts from layer IV are not
significantly older than the stakes from layer IV/III.
One other sample (KIA-53780) from layer IV is much
younger than the remaining samples from that la-
yer. In fact, it instead coincides perfectly with the
samples from layer III. Perhaps the sample’s contex-
tual information was incorrectly documented, and
should instead place the find in layer III. Another
possibility is that the find represents a younger in-
trusion in older layers.
The samples from layer III suggest a long chronolo-
gy, between 8800 and 7200 cal BC. The samples clus-
ter in several groups throughout the time period,
likely due to the character of the calibration curve.
One artefact sample from layer II (GIN-8378) pro-
vides a date at around 6100–5700 cal BC, placing it
in the Neolithic period.
Dataset 3 – Ecofacts
Ecofact samples and source critique
Only three samples are included in this group, all of
which are unworked (i.e. not altered by humans).
The sample GIN-10128 come from a piece of birch
trunk (diameter 12cm), the sample GIN-10113 is a
wooden branch with a 6cm diameter and sample
GIN-11093a was taken from an elk bone. Since the
ecofacts are unworked they do not necessarily reflect
human activities on the site. Additionally, the latter
sample is the sole ecofact sample found within a clear
archaeological layer (III, in trench 3). The sample
GIN-10128 was found in a sterile interlayer between
archaeological layers IV and III in trench 2, while the
sample GIN-10113 pierced several layers, probably
as a result of a tree fall (Zaretskaya et al. 2001).
Wooden samples are, as always, at risk for old wood
effect. Though, the samples ages of the ecofact sam-
ples at Stanovoye 4 are probably not high enough to
significantly alter the chronology of the site. The same
can be said about the unworked elk bone which also
has an unknown age, but should not be older than
25 years (maximum average age of European elk).
Calibrated ecofact dates
The ecofact samples suggest a very limited chronol-
ogy (Fig. 6), based on the few available samples. The
sample GIN-10128, from layer IV/III, dates to around
9300–8800 cal BC. From the same layer, sample GIN-
10113 instead dates to c. 8200–7600 cal BC, which
more closely matches the final ecofact date from
sample GIN-11093a. This sample comes from layer
III and matches the sediment samples from the same
context, dating to 8200–7600 cal BC. Sample GIN-
10113 possibly represents a younger intrusion into
older layers.
Dataset 4 – AMS dates
AMS-dated samples and source critique
AMS-dated samples are generally more precise than
conventionally dated samples (Bronk Ramsey et
al. 2004), which makes this dataset support a more
detailed chronology in general. Additionally, the
finds from the site that were prioritized for AMS-dat-
ing are all artefacts and, thus, directly related to an-
thropogenic activities. The samples within dataset 4
are therefore characterized by both a strong contex-
tual integrity and archaeological reliability.
A critical aspect of this dataset is the lack of samples
from archaeological layers IIIa and II. Additionally,
most of the AMS-dated samples have been consoli-
dated using PVA-glue which increases the risk of
contamination (more on this in Dataset 6).
Calibrations of AMS-dated samples
The AMS-based chronology (Fig. 7) is almost iden-
tical to the previously described artefact chronolo-
gy (see Dataset 2). To summarize, archaeological la-
yer IV dates to between 10 000 cal BC and 8700 cal
BC, while layer III starts around 8800 cal BC and
ends sometime between 8300 and 7200 cal BC.
Dataset 5 – Conventional dates
Conventionally dated samples and source critique
The samples that were dated conventionally are all
ecofacts or wooden stakes. As already mentioned,
unworked ecofacts do not necessarily date anth-
ropogenic activities on the site.
Calibrations of conventionally dated samples
The chronology, based on conventionally dated sam-
ples (Fig. 8), suggests that the oldest dates from the
site come from layers IV/III and IIIa. These together
span 9300–8800 cal BC. Although these dates do
not represent the oldest settlement of the site, it ap-
pears that way since all dates from older layers were
AMS-dated.
One sample from layer IV/III groups with samples
from layer III, as was observed in dataset 3. The sam-
ples from layer III span 8600–7500 cal BC. A sam-
ple from layer II is much younger than the remain-
ing samples and dates to 6100–5700 cal BC.
Sandra Söderlind, Mikhail Zhilin
174
Dataset 6 – Non-consolidated samples
Non-consolidated samples and source critique
The non-consolidated samples can be excluded from
the risk of PVA contamination (see Dataset 7). How-
ever, other source critical aspects relate to these sam-
ples. For instance, samples consist of both ecofacts
and artefacts, which vary in the level of anthropoge-
nic relation. Additionally, the non-consolidated sam-
ples were dated using either AMS or conventional
dates, resulting in mixed levels of precision.
Calibrations of non-consolidated samples
The non-consolidated samples (Fig. 9) group in seve-
ral clusters. The first cluster of dates includes sam-
ples from layers IIIa and IV/III, which together span
9300–8800 cal BC. The two samples from layer IV,
falling within the same cluster, are discussed later
on. The samples from layer III group into several
clusters, although this is probably a result of the ca-
libration curve plateaus. It is likely that the samples
from layer III represent a continuous settlement on
the site from around 8700 until 7500 cal BC, as seen
in the other datasets. The one sample from layer II
dates the layer to c. 6100–5700 cal BC.
Dataset 7 – Consolidated samples
Consolidated samples and source critique
The consolidated samples are all AMS-dated, making
them more precise than conventional dates. The AMS
dates based on artefacts are also directly related to
anthropogenic activities on the site. Unfortunately,
all these samples come from only two of the archaeo-
logical layers, IV and III, while the remaining layers
lack representation.
Fig. 5. Calibrated artefact samples from the site. The colours indicate their find contexts: grey archaeo-
logical layer IV; light blue layer IV/III; dark blue archaeological layer IIIa; blue archaeological layer III,
and turquoise archaeological layer II.
Revising the chronology of Stanovoye 4 in the Upper Volga Region (Western Russia)
175
Several organic samples from the site were pre-
served using PVA-glue. This practice has been com-
mon since the 1950s, especially for finds from water-
logged sites. Many PVA-based consolidants exist, in-
cluding PVA, PVAc and PVAL, but the most common-
ly used variety is known as ‘Elmer’s carpenter’s glue’.
Early studies showed that the glue could be easily re-
moved in the lab prior to dating. However, in more
recent years, studies have indicated that it might be
harder to remove than initially thought. The reason
is that PVA will form cross-linking of polymers which,
over time, reduces its solubility (Johnson 1994).
An additional issue is that the details surrounding
consolidation efforts are rarely documented. Often,
it is not known which PVA-variety was used or how
the treatment was carried out. This is an issue, since
the details of these procedures play a role in the suc-
cessful removal of the consolidant. Furthermore, the
porosity of the artefact may also affect the solubility
of the PVA (cf. Brock et al. 2018).
The reliability of the samples, and the related dates,
depends on the successful removal of PVA, since re-
mains can result in dates appearing older than they
actually are. There are various strategies for removal
of PVA-contaminants in the lab, most of which in-
clude the use of repeated washes of distilled water,
acetone and methanol (as suggested by Stevens,
Hedges 2004). The samples from Stanovoye have un-
dergone similar treatments (see Appendix below).
However, several standard pre-treatments involving
acetone, chloroform and methanol have in some stu-
Fig. 6. Calibrated ecofact samples from the site. The colours indicate their find contexts: grey archaeolo-
gical layer IV; light blue layer IV/III; dark blue archaeological layer IIIa.
Fig. 7. Calibrated AMS-dated samples from the site. The colours indicate their find contexts: grey archaeo-
logical layer IV; dark blue archaeological layer IIIa.
Sandra Söderlind, Mikhail Zhilin
176
dies proven ineffective for removing common con-
solidants (Dee et al. 2011; Brock et al. 2018). There-
fore, the dating of PVA-treated finds is not recom-
mended unless alternative methods can be applied
or the consolidated areas can be avoided (ibid.). The
results by Fiona Brock et al. (2018) have been criti-
cized due to their use of modern PVA-glue (often
containing plastics and other additives), which may
not be relevant for consolidation efforts in the past
(Meadows et al. 2019). Additionally, the use of the
silica-polymer Chromosorb™ (as a base for contami-
nation experiments) might not correctly map the po-
tential chemical cross-linking between the consoli-
dant and the sample itself, due to the material’s
highly absorbent properties (Dee et al. 2011).
The use and removal of older consolidants on actual
(Mesolithic) bone artefacts has recently been investi-
gated (Meadows et al. 2019). Although the investi-
gated bones turned out to not be consolidated using
PVA, the other preservation mediums were found to
have penetrated into the bone, while the highest
concentrations were found on the surface (ibid.). If
this is true also for PVA-consolidants, it could mean
that the Stanovoye samples from further within the
bone contain lower amounts of remaining PVA-con-
tamination compared to samples from just beneath
the surface.
The issues relating to the treatment and removal of
PVA-glue from artefacts are clearly complicated. The
PVA-treated artefacts from Stanovoye must be consi-
dered somewhat less reliable, since lab pre-pre-treat-
ments may not be as reliable for PVA-removal as
previously indicated. If the consolidated samples
from the site are affected by any remaining PVA,
they could appear older than non-consolidated sam-
ples from the same contexts.
Calibrations of consolidated samples
The PVA-consolidated samples from Stanovoye 4
(Fig. 10) provide a very similar chronology to the
AMS-dated samples, mainly due to the fact that the
two chronologies largely consist of the same selec-
tion of samples. Nonetheless, the consolidated sam-
ples also follow the chronology suggested by the
non-consolidated samples. The consolidated samples
from layer IV span the time 10 000–8700 cal BC, al-
though with few overlapping dates. Samples from
layer III follow the already described pattern, which
likely represents a long-term occupation on the site
between 8800 and 7200 cal BC.
The samples AAR-22232 and KIA-53778 come from
the same artefact, an elk antler socket, which was
consolidated using PVA. The AAR-sample was the
first to be extracted and the KIA-date was taken from
the same area but from further within the artefact.
The wooden handle that belongs to the same artefact
was not consolidated, but also dated (KIA-35153).
However, this sample instead poses a risk of mod-
ern contamination due to incorrect storage after ex-
cavation. A comparison of the samples can be found
in Table 3.
Fig. 8. Calibrated conventionally dated samples from the site. The colours indicate their find contexts:
grey archaeological layer IV; light blue layer IV/III; dark blue archaeological layer IIIa; blue archaeolo-
gical layer III, and turquoise archaeological layer II.
Revising the chronology of Stanovoye 4 in the Upper Volga Region (Western Russia)
177
As is clear from the table the two samples from the
same antler artefact gave differing results, by 258 14C
years, with the AAR-sample being younger and more
closely correlating to the related wooden handle (al-
though with a difference of 92 14C years, cf. Philip-
psen 2019). Although, if the handle is contaminated
with modern carbon, perhaps the older KIA-53778
date is more reliable. The processes at play here, and
reasons for the differing dates from the one tool, are
not yet completely understood. Theoretically, rea-
sons related to PVA-contamination, slight differences
in lab procedures or other unknown factors could
play a role. Further investigations are necessary to
understand the processes behind these results.
A detailed discussion of the samples and site
chronology
Here, the multiple chronologies will be compared to
each other, and discussed in relation to the sedimen-
tation sequence and archaeological finds from the
contexts, in an effort to understand a more general
chronology of the site.
Archaeological layer IV
The first occupation phase in the site’s history relates
to layer IV in both trenches. Archaeological finds from
the context have been typologically and technolog-
ically interpreted as belonging to the early Butovo
technocomplex. At the early stage of this habitation
phase, people had settled on the shoreline of the
ancient lake.
The sediment samples from layer IV generally sug-
gest an older and shorter chronology (c. 10 500 to
9200 cal BC) compared to the artefact samples (c.
10 000–8700 cal BC). Ecofacts from layer IV have
not been dated and can thus not be compared. A clo-
ser look at the individual samples shows that the
start and end of these chronologies are rather im-
precise, mainly due to the many plateaus on the ca-
libration curve in the early Holocene.
One sediment sample (GIN-10112 II) sets the start
of the sediment chronology to 10 500–9800 cal BC.
If this lone sample is considered an outlier, the sedi-
ment chronology more closely matches the starting
point of the artefact-based chronology. However, the
oldest artefact sample (KIA-53779) also falls within
a plateau on the calibration curve, making it less
precise. Nevertheless, the early sediment samples
could be viewed as support for such an early start
to the phase. If the early artefact and sediment sam-
ples are considered reliable, the site was first set-
tled already during the end of the Younger Dryas.
Fig. 9. Calibrated non-consolidated samples from the site. The colours indicate their find contexts: grey
archaeological layer IV; light blue layer IV/III; dark blue archaeological layer IIIa; blue archaeological
layer III, and turquoise archaeological layer II.
Sandra Söderlind, Mikhail Zhilin
178
If we assume that these samples are not reliable, the
occupation should instead have started in the early
Preboreal, around 9400–9200 cal BC, as suggested by
three rather precisely dated artefact samples; KIA-
39317 (trench 2), KIA-35152 and KIA-53778 (trench
3).
The remaining artefact samples from layer IV (AAR-
22232, KIA-39316 and KIA-35153) date to 9200–
8600 cal BC, a time which is also characterized by a
calibration curve plateau, again making the end of
the occupation phase imprecise. While these samples
are all AMS-dated, the latter is at risk of modern con-
tamination. All the AMS-dated samples, except for the
one just mentioned, were also consolidated using
PVA. The dates AAR-22232, KIA-39316 and KIA-35153
also date to the same time as sediments and artefacts
from the overlaying layers IIIa and IV/III (see be-
low), which could indicate that they are unreliable.
It could also be an indicator that the temporal diffe-
rence between the two layers is small.
Due to rising water levels, people had to move away
from the shoreline and the artefacts and sediments
relating to archaeological layer IV were subsequent-
ly submerged in gyttja deposits.
Archaeological layer IIIa and sediment layer
between IV/III
During a temporary water regression during the Pre-
boreal, the site was again inhabited, this time by
people related to the Ienovo technocomplex (Zhilin,
Matiskainen 2003). It is likely that a small difference
in age is not observable in this part of the chronolo-
gy due to the long calibration curve plateau at this
time. This is assumed because of the stratigraphic se-
quence as well as a difference in typology between
the archaeological layers. Dating more (typo-chrono-
logical characteristic) samples from layer IIIa could
help determine the age of the layer in relation to ad-
jacent contexts.
The same can be argued about the wooden stakes
from trench 2, sediment layer IV/III. The samples
(GIN-8374, GIN-8376) were found in horizontal po-
sitions, stratigraphically located above the level of
layer IV and in the same level as the finds from la-
yer IIIa. Their calibrated dates fall within the same
plateau as the sample from layer IIIa. Nonetheless,
it cannot be excluded that these samples are related
to the habitation mentioned above, by people relat-
ed to the Ienovo technocomplex.
One ecofact sample from trench 3 (GIN-10113, a
thick branch) pierced several layers, perhaps as a
result of a tree fall. It is younger than the sediment
and rest of the finds, which indicates that it repre-
sents a younger intrusion into older layers.
Archaeological layer III
After another period of rising water levels, by the
nearby lake, continuous habitation in the area around
the lake and artefacts related to archaeological la-
yer III are accumulated in the water. Pollen indicates
Fig. 10. Calibrated consolidated samples from the site. The colours indicate their find contexts: grey
archaeological layer IV; light blue layer IV/III; blue archaeological layer III.
Revising the chronology of Stanovoye 4 in the Upper Volga Region (Western Russia)
179
that this occurred in the second half of the Prebo-
real (according to samples from trench 3) and in the
early Boreal (according to samples from trench 2)
(Aleshinskaya 2001).
The technology and typology of the artefacts from
layer III indicate that they are related to the Middle
Butovo technocomplex (Zhilin, Matiskainen 2003).
The artefact dates from this layer, in trench 2, sug-
gest a chronology spanning 7900–7500 cal BC. A si-
milar pattern is observable in the sediment dates
from the same context, which span the time between
8300 and 7600 cal BC.
The artefact samples from the layer in trench 3 in-
stead propose a slightly longer chronology than in
trench 2, here stretching from 8800 to 7600 cal BC.
One sample (KIA-35155) with a younger date is dis-
cussed below. Interestingly, the samples appear to
cluster into several sub-phases. However, this is pro-
bably an effect of the many calibration curve pla-
teaus during this time span, although the presence
of some temporary hiatus within this long time pe-
riod is probable. Nonetheless, it is likely that the site
was more or less continuously or repeatedly settled
from the second half of the Preboreal up to the mid-
dle of the Boreal period. This is also confirmed by
pollen data.
A closer look at some of the artefact samples indi-
cates that habitation at the site may also have taken
place during harsher climate conditions. The first five
artefact dates fall within a plateau located in the
curve around 8800–8300 cal BC. Another couple of
dates (KIA-53781 and KIA-53780) fall within the glo-
bal cold spell around the 7th Bond-event (Bond et al.
1997) around 8300–8100 cal BC. After the Bond-
event, another four dates cluster within a plateau in
the calibration curve, between 8200–7600 cal BC,
followed by another steeper part of the curve where
we also find activity on the site (samples GIN-8853)
at 7700–7500 cal BC.
The youngest sample from layer III (KIA-35155) dates
to sometime around 7500–7200 cal BC, and is also
located within a calibration curve plateau. It is signi-
ficantly younger than the other samples in trench 3,
which could indicate that it is a later intrusion. An ar-
chaeological layer dated to the second half of the Bo-
real period was excavated in trench 1 (Zhilin 2002),
and perhaps this sample relates to that occupation.
Archaeological layer II
The Neolithic layer II is dated via two sediment dates
and one artefact date, all from trench 2, to a time
span around 6600–5700 cal BC. Typologically dis-
tinct finds relating to the Upper Volga culture were
found in the same context. One of the samples (GIN-
10108 II) also falls within an extensive calibration
curve plateau, making it imprecise. More dates from
this archaeological layer would provide a better chro-
nological base for this Neolithic occupation phase.
Even though many of the dates fall within plateaus
in the calibration curve, which inherently lead to less
reliability, there are nonetheless clear signs of long-
lasting or repeated visits and activities in the area of
the site throughout the Early Holocene.
The role of AMS-dated and non-consolidated
samples from Stanovoye 4
After calibration, there is no significant difference in
the precision of the AMS-dated samples compared
to the conventionally dated samples, mainly due to
the character of the calibration curve for the early
Holocene.
However, the comparison between AMS-dated sam-
ples and conventionally dated samples showed that
certain contexts and sample materials have been pri-
oritized during dating. This creates a skewed gene-
ral chronology for the site, since the more reliable
(AMS-dated) samples all come from layers IV and III.
Meanwhile, layers IIIa and II are not represented at
all, and therefore do not contribute to the under-
standing of the site chronology as reliably as samples
from other contexts. Although it should be noted
that the conventionally dated samples, from layers
Tab. 3. Three samples from the same elk antler socket and related wooden handle.
Lab no. Part of artefact Trench Position Satrat. Uncal. BP
Cal BC Cal BC PVA-
(95.4%) Start (95.4%) End treated
KIA-35153 handle 3 square 302 Cult. lay. IV 9505±47 9126 8638 no
AAR-22232
antler
3 square 157 Cult. lay. IV 9597±43 9215 8812 yes(further outside)
KIA-53778
antler
3 square 157 Cult. lay. IV 9855±50 9447 9242 yes(further inside)
Sandra Söderlind, Mikhail Zhilin
180
IIIa, IV/III and II, tend to generally follow the chro-
nologies suggested by the AMS-dates, which indicates
that their slightly less reliable nature (due to their
dating method) plays a smaller role in this instance.
Nonetheless, a more detailed chronology of Stanovo-
ye 4 can only be possible with further AMS-dating of
reliable samples along with the further development
of the calibration curve for this time period.
Two samples (KIA-39317, GrA-34084) are AMS-dated
but not consolidated, which would make them the
most reliable samples from the site. They were also
not affected by modern or ancient processes like im-
proper storage and reservoir effects. This puts them
in a unique position for comparison with the other
samples. Nonetheless, these samples do not appear
to stand out from the chronology created by the rest
of the samples. On the contrary, they fit well into
the chronologies suggested by the artefact samples/
AMS-dated samples. This would indicate that the con-
solidated samples, which underwent cleaning pro-
cedures in the labs before dating, seem to be unaf-
fected by the PVA. However, any small differences in
age due to PVA-contamination would perhaps go un-
noticed due to the multitude of calibration curve pla-
teaus. Needless to say, further testing for PVA-con-
tamination on relevant samples would be useful in
order to resolve this issue.
Conclusion
In this contribution we discussed the many types
and common characteristics, of 14C-dated samples
that are readily available from sites excavated only
a few decades ago. It has become clear that the use
of slightly older assemblages can be a challenge, due
to the constant development of methods and tech-
niques within the field of conservation and radiocar-
bon dating. Samples taken under current best prac-
tices should, and can, still be useful for furthering
the understanding of prehistory. Therefore, it is of
little gain to simply consider them ‘bad’ by modern
standards and discard them from use. This includes
samples that were consolidated using PVA or sam-
ples of certain sample materials. It is our opinion
that it is better to work with what we have rather
than discarding information, in order to better un-
derstand the site under the given circumstances. This
is of special interest for studies of legacy materials.
In this article, we also set out to investigate the pos-
sible effect of PVA-contamination on a selection of
the samples from the site, from a strictly archaeologi-
cal perspective. Although recent studies have shown
that PVA is more difficult to remove in the lab than
was previously understood, the samples from Stano-
voye 4 do not appear to be affected by any remain-
ing PVA contamination, as based on their close cor-
relation to the non-consolidated samples. However,
extensive plateaus in the calibration curve for this
time make detailed comparisons complicated. Fur-
thermore, these are only initial indications based on
a small number of comparable samples. More studies
focused on the effects of PVA-contamination should
thus be done to confirm or refute these results. The
further dating of non-consolidated bone and antler
samples from the Stanovoye 4 site, as well as other
organic materials such as resin, would also be ben-
eficial for understanding the chronology of the site
in general.
The chronologies suggested by various fractions of
radiocarbon samples from the site Stanovoye 4 sup-
port a joint timeline which includes long standing
traditions and a continuous site history that lasts al-
most 3000 years. The oldest archaeological layer
IV on the site relates to the settlement by people re-
lated to the Early Butovo technocomplex, and took
place between c. 10 000–9500 cal BC. After some
hiatus, and shorter visits on the site by people linked
to the Ienovo technocomplex, the site is again more
continuously in use by people in relation to the Mid-
dle Butovo technocomplex. Artefacts made and im-
plemented by these people are found in layer III and
date to between 8800 and (definitely) 7600 cal BC,
possibly even until c. 7200 cal BC. Although a broad
chronology for the site has been constructed, many
details remain unclear, due to the complex calibra-
tion curve during the early Holocene.
Furthermore, it has been highlighted that certain
contexts of the site have been prioritized for AMS-
dating and consolidation, while others are hardly
represented at all. More dating efforts are needed
from layers IIIa and II in order to reach a more ho-
listic chronological understanding of the site.
Revising the chronology of Stanovoye 4 in the Upper Volga Region (Western Russia)
181
Alexandrovskiy A. L., Alexandrovskaya E. I., Zhilin M. I.,
and van der Plicht J. 2009. Mesolithic human bones from
the Upper Volga Basin: Radiocarbon and trace elements.
Radiocarbon 51(2): 637–645.
https://doi.org/10.1017/S0033822200055983
Aleshinskaya A. S. 2001. Paleographic conditions of an-
cient people settling in Mesolithic of Volga-Oka inter-
fluve (paleological data). Unpublished PhD thesis ab-
stract. Moscow. Moscow State University. (in Russian)
Brock F., Dee M., Hughes A., Snoeck C., Staff R., and Bronk
Ramsey C. 2018. Testing the effectiveness of protocols for
removal of common conservation treatments for radio-
carbon dating. Radiocarbon 60(1): 35–50.
https://doi.org/10.1017/RDC.2017.68
Bronk Ramsey C. 2009. Bayesian Analysis of Radiocarbon
Dates. Radiocarbon 51(1): 337–360.
https://doi.org/10.1017/S0033822200033865
Bronk Ramsey C., Higham T., and Leach P. 2004. Towards
High-Precision AMS: Progress and Limitations. Radiocar-
bon 46(1): 17–24.
https://doi.org/10.1017/S0033822200039308
Damlien H. 2016. Eastern pioneers in westernmost terri-
tories? Current perspectives on Mesolithic hunteregather-
er large-scale interaction and migration within Northern
Eurasia. Quaternary International 419: 5–16.
http://dx.doi.org/10.1016/j.quaint.2014.02.023
Dee M., Brock F., Bowles A., and Ramsey C. 2011. Using
a Silica Substrate to Monitor the Effectiveness of Radio-
carbon Pretreatment. Radiocarbon 53(4): 705–711.
https://doi.org/10.1017/S0033822200039151
Hartz S., Terberger T., and Zhilin M. 2010. New AMS-dates
for the Upper Volga Mesolithic and the origin of microb-
lade technology in Europe. Quartär 57: 155–169.
Johnson J. S. 1994. Consolidation of Archaeological Bone:
A Conservation Perspective. Journal of Field Archaeo-
logy 21(2): 221–233.
Kankaanpää J., Rankama T. 2014. Fast or Slow Pioneers?
A View from Northern Lapland. In Riede F., Tallaavaara
M. (eds.), Lateglacial and Postglacial Pioneers in North-
ern Europe. BAR International Series 2599. Archaeopress.
Oxford: 147–159.
Manninen M. A., Damlien H., Kleppe J. I., + 4 authors, and
Persson P. 2021. First encounters in the north: Cultural
diversity and gene flow in Early Mesolithic Scandinavia.
Antiquity 95(380): 310–328.
https://doi.org/10.15184/aqy.2020.252
Meadows J., Boudin M., Groß D., Jantzen D., Lübke H., and
Wild M. 2019. Radiocarbon dating bone and antler arte-
facts from Mesolithic Hohen Viecheln (Mecklenburg- West-
ern Pomerania, Germany). In Groß D., Lübke H., Mea-
dows J., and Jantzen D. (eds.), Working at the Sharp End:
From Bone and Antler to Early Mesolithic Life in North-
ern Europe. Untersuchungen und Materialien zur Stein-
zeit in Schleswig Holstein und im Ostseeraum Band 10.
Vachholtz. Kiel/Hamburg: 113–126.
Meadows J., Lozovskaya O., Bondetti M., Drucker D., and
Moiseyev V. 2019. Human palaeodiet at Zamostje 2, cen-
tral Russia: Results of radiocarbon and stable isotope ana-
lyses. Quaternary International 541: 89–103.
https://doi.org/10.1016/j.quaint.2019.07.017
Philippsen B. 2013. The freshwater reservoir effect in ra-
diocarbon dating. Heritage Science 1: 24.
https://doi.org/10.1186/2050-7445-1-24
2019. Approaches to determine reservoir effects in elk/
moose. Radiocarbon 61(6): 1889–1904.
https://doi.org/10.1017/RDC.2019.124
Rankama T., Kankaanpää J. 2007. The Earliest Postglacial
Inland Settlement of Lapland. In Volokitin A. V. (ed.), Ka-
mennyi vek evropeiskogo Severa [The Stone Age of
northernmost European Russia]. Russian Academy of
Sciences, Uralian Section. Komi Scientific Centre, Institute
of Languages. Literature and History. Syktyvkar.
Rankama T., Kankaanpää J. 2008. Eastern arrivals in post-
glacial Lapland: The Sujala site 10 000 cal BP. Antiquity
82(318): 884–899.
https://doi.org/10.1017/S0003598X00097659
Reimer P., Austin W., Bard E., + 38 authors, and Talamo
S. 2020. The IntCal20 Northern Hemisphere radiocar-
bon age calibration curve (0–55 cal kBP). Radiocarbon
62(4): 725–757. https://doi.org/10.1017/RDC.2020.41
Sørensen M., Rankama T., Kankaanpää J., + 4 authors, and
Glørstad H. 2013. The First Eastern Migrations of People
and Knowledge into Scandinavia: Evidence from Studies
of Mesolithic Technology, 9th–8th Millennium BC. Norwe-
gian Archaeological Review 46(1): 19–56.
https://doi.org/10.1080/00293652.2013.770416
Stevens R. E., Hedges R. E. M. 2004. Carbon and nitrogen
stable isotope analysis of northwest European horse bone
and tooth collagen, 40,000 BP–present: Palaeoclimatic in-
References
Sandra Söderlind, Mikhail Zhilin
182
terpretations. Quaternary Science Reviews 23: 977– 991.
https://doi.org/10.1016/j.quascirev.2003.06.024
Taylor R. E., Bar-Yosef O. 2014. Radiocarbon Dating. An
Archaeological Perspective. Second edition. Left Coast
Press, Inc. Walnut Creek.
Zaretskaya N. E., Ponomareva V. V., Sulerzhitsky L. D.,
and Zhilin M. G. 2001. Radiocarbon studies of peat bogs:
an investigation of south Kamchatka volcanoes and Up-
per Volga Archaeological sites. Radiocarbon 43(2B):
571–580. https://doi.org/10.1017/S0033822200041229
Zaretskaya N. E., Zhilin M., Karmanov V. N., and Uspen-
skaya O. N. 2005. Radiocarbon Dating of Wetland Meso-
Neolithic Archaeological Sites within the Upper Volga and
Middle Vychegda. Geochronometria – Journal on Me-
thods and Applications of Absolute Chronology 24: 117–
131.
Zhilin M. 2002. Stratigrafiya i planigrafiya mnogosloynogo
poseleniya Stanovoye 4 v Verkhnem Vovolzh’ye. Tverskoy
Archeologichesky Sbornik 5: 107–116. (in Russian)
2003. Early Mesolithic peat bog sites on the Upper Vol-
ga. In Bauerochse A., and Haßmann H. (eds.), Peat-
lands. Archaeological sites – archives of nature – na-
ture conservation – wise use. Proceedings of the Euro-
pean Peatland conference 2002 in Hannover, Germany.
Hannover. Verlag Marie Leidorf. Rahden: 48–61.
2004. Prirodnaya sreda i khozyaystvo mezolitiches-
kogo naseleniya tsentra i severo-zapada lesnoy zony
Vostochnoy Evropy. Moskva. (in Russian).
2007. The early Mesolithic of the Upper Volga: selected
problems. In Masoj≤ M., Płonka T., Ginter B., and Koz-
łowski S. K. (eds.), Contributions to the Central Euro-
pean Mesolithic. Institute of Archaeology. University
of Wrocław. Wroclaw: 89–104.
2009. Chronology and Evolution in the Mesolithic of
the Upper Volga-Oka interfluve. In Crombé P., van Stry-
donck M., Sergant J., Boudin M., and Bats M. (eds.),
Chronology and evolution within the Mesolithic of
North-West Europe. Proceedings of an International
Meeting, Brussels, May 30th–June 1st 2007. Cambridge
scholars publishing. Newcastle upon Tyne: 451–472.
Zhilin M., Matiskainen H. 2003. Deep in Russia, deep in
the bog. Excavations at the Mesolithic Sites Stanovoje 4
and Sakhtysh 14, Upper Volga region. 2003. In Larsson L.,
Kindgren H., Leffler D., and Akerlund A. (eds.), Mesolithic
on the Move. Papers Presented at the Sixth International
Conference on the Mesolithic in Europe, Stockholm 2000.
Oxbow monograph. Oxford: 694–702.
The organic samples from Stanovoye 4 reported in
our paper were radiocarbon dated at different
times in various labs. The pre-treatments involved
in the dating procedure can be found below along
with relevant sources.
Leibniz Laboratory for Radiometric Dating
and Stable Isotope Research in Kiel (lab-code
KIA):
Pre-treatment of the bones and collagen extraction
was done in several steps. Firstly, the pieces of
crushed, solid bone material (0,5–2mm) were de-
fatted using acetone, rinsed with demineralized
water and subsequently demineralized in HCl (c.
1%). Mobile humic acids were removed by taking
the previously demineralized bone material and
treating it with 1% NaOH (20°C, 1h) and once again
with 1% HCl (20°C, 1h). The bone collagen, to be
used for dating, was dissolved overnight as gelatine
in demineralized water at 85°C and pH = 3. Inso-
luble particles were removed using filtration through
a 0.45μm pore silver filter. The gelatine solution was
then freeze-dried, which is an adaption of the me-
thod described by Longin (1971).
Combustion to CO2 was performed in an evacuated
quartz tube with CuO and silver wool at 900°C. The
obtained CO2 was subsequently reduced at 600°C,
using H2 and iron powder as the catalyst. The re-
sulting graphite mixture was pressed into a target
holder for AMS measurement.
Geological Institute at the Russian Academy
of Sciences (RAS) (lab-code GIN):
The sediment samples dated in 2005 were subdi-
vided into two fractions, one of which underwent
a hot alkaline extraction and the other a cold alka-
line extraction. This procedure was designed to se-
parate and recover organic matter that relate to dif-
ferent stages of preservation (details in Zaretskaya
et al. 2001). This also resulted in two dates for each
sample (I and II). Nataliya E. Zaretskaya et al. (2001)
use the age difference between the two measure-
ments as a reflection of the age variation within the
Appendix – Sample treatment at different labs
∴
Revising the chronology of Stanovoye 4 in the Upper Volga Region (Western Russia)
183
1 A similar approach as described by Zaretskaya et al. (2005) has been presented as a manner of removing mobile secondary orga-
nics from the bulk sediment (Beta Analytical n.d., sections ‘Acid/alkali/acid’ and ‘Acid/alkali/acid – solubles’). These mobile orga-
nics have been shown to often reflect younger fractions of the soil (Nilsson et al. 2001, and sources therein) which, if not removed
can result in a modern contamination. Due to the higher risk of more modern carbon in the mobile organics, the cold-extraction
samples from Stanovoye were considered less reliable than warm-extraction samples, and thus excluded when the two fractions
were non-consistent. I the case that samples correspond to the same age, samples from both extraction methods remain in the study.
sediment layer. The warm-extractions were mainly
used in the current study.1
Wood and bone samples were pre-treated and dated
using routine procedures used at GIN. Wood sam-
ples underwent a standard A-A-A pre-treatment; the
samples were first cleaned by standing in hot 5%
HCl, washed with distilled water and subsequently
boiled for 20 min in a 2% NaOH solution. Alkali in-
soluble wood fragments were then digested in hot
5% HCl and finally washed with distilled water (Za-
retskaya et al. 2001).
Aarhus AMS centre (lab-code AAR):
Bone and antler samples were demineralized using
1M HCl for several days, until the end of efferves-
cence (complete dissolution of calcium carbonate)
and no observed density gradient (calcium phos-
phate completely dissolved). Humates were then re-
moved using 0.1M NaOH, with renewed NaOH until
the solution remained colourless, with each base
step lasting several hours at 4°C. After each base/
acid step, samples were rinsed three times with de-
mineralized H2O. Finally, the samples were rinsed
three times with a weak HCl-solution (pH=2), cov-
ered with that solution, and heated to 58–80°C for
48 hours, until collagen was dissolved. Samples
were centrifuged (15 minutes at 2000rpm) and fil-
tered through 0.45μm syringe filters, before being
freeze dried (Philippsen 2019).
Groningen AMS lab (lab-code GrA):
Collagen from the bones was extracted using a mo-
dernized version of the Longin method (Longin
1970). The collagen was then combusted, purified
and transferred into graphite (Aerts-Bijma et al.
2001). This graphite was pressed into target hold-
ers for the AMS ion source. The AMS measured the
isotope ratios 14C/12C and 13C/12C of the graphite,
from which the conventional 14C age is determined
(van der Plicht et al. 2000). The stable isotope ra-
tios 13C/12C and 15N/14N were measured for the
same collagen dated by AMS. These stable isotope
ratios were measured by an elemental analyser, com-
bined with a mass spectrometer (Aerts-Bijma et al.
2001; Alexandrovskiy et al. 2009).
Aerts-Bijma A. T., van der Plicht J., and Meijer H. A. J.
2001. Automatic AMS sample combustion and CO2 colle-
ction. Radiocarbon 43(2A): 293–298.
https://doi.org/10.1017/S0033822200038133
Alexandrovskiy A. L., Alexandrovskaya E. I., Zhilin M. I.,
and van der Plicht J. 2009. Mesolithic human bones from
the Upper Volga Basin: Radiocarbon and trace elements.
Radiocarbon 51(2): 637–645.
https://doi.org/10.1017/S0033822200055983
Beta Analytical (n.d.). Beta Analytic Standard Pretreat-
ment Protocols. Sections “Acid/alkali/acid” and “Acid/ al-
kali/acid – solubles”. Retrieved April 1, 2021, from https://
www.radiocarbon.com/pretreatment-carbon-dating.htm
Nilsson M., Klarqvist M., Bohlin E., and Possnert G. 2001.
Variation in 14C age of macrofossils and different fractions
of minute peat samples dated by AMS. The Holocene 11
(5): 579–586.
https://doi.org/10.1191/095968301680223521
References – Appendix
Philippsen B. 2019. Approaches to determine reservoir
effects in elk/moose. Radiocarbon 61(6): 1889–1904. 
ttps://doi.org/10.1017/RDC.2019.124
Longin R. 1970. Extraction du collagène des os fossils
pour leur datation par la méthode du carbone 14. Un-
published PhD thesis. Université de Lyon. Lyon
1971. New Method of Collagen Extraction for Radiocar-
bon Dating. Nature 230(5291): 241–242.
https://doi.org/10.1038/230241a0
van der Plicht J., Wijma S., Aerts A. T., Pertuisot M. H., and
Meijer H. A. J. 2000. The Groningen AMS facility: status
report. Nuclear Instruments and Methods in Physics Re-
search B 172(1–4): 58–65.
Zaretskaya N. E., Ponomareva V. V., Sulerzhitsky L. D.,
and Zhilin M. G. 2001. Radiocarbon studies of peat bogs:
an investigation of south Kamchatka volcanoes and Up-
per Volga Archaeological sites. Radiocarbon 43(2B):
571–580. https://doi.org/10.1017/S0033822200041229
∴