As already reported to readers of this journal, in 1992–1993 a programme to study the early Upper Palaeolithic and the late Middle Palaeolithic in Moldova, Ukraine, and Russia was carried out. Forty-four radiocarbon dates from ten sites, and the results of the programme as a whole, have been published (Hedges et al. Reference Hedges, Housley, Pettitt, Bronk Ramsey and Klinken1996; Allsworth-Jones Reference Allsworth-Jones, Orschiedt and Weniger2000). So far as Moldova is concerned, full accounts of the work done at Ciuntu and Brînzeni have been produced (Borziac et al. Reference Borziac, Allsworth-Jones, French, Medyanik, Rink and Lee1997; Allsworth-Jones et al. Reference Allsworth-Jones, Borziac, Chetraru, French and Medyanik2018a). The third site included in the programme was Buzdujeni, first investigated in the 1970s but only very briefly reported by the excavator (Chetraru Reference Chetraru1973). All these are cave sites in a quite restricted area of north-western Moldova. New small scale control excavations permitted environmental evidence to be recorded for the first time, as well as providing the opportunity for an archaeological re-assessment of the cultural remains.

Thus, the single layer of occupation at Ciuntu was originally regarded as Early Upper Palaeolithic and was assigned to the ‘Brînzeni’ archaeological culture. The three AMS radiocarbon dates obtained in the Oxford laboratory however produced a mean uncalibrated age of 20,500±180 bp, which was not consistent with the idea that the site could be taken as belonging to the beginning of the Upper Palaeolithic. A re-examination of the archaeological inventory suggested rather that it belonged to the Gravettian techno-complex, which was dominant in Europe at that time. The deposits in the cave seemingly owed their origin to a number of factors: the accumulation of fine-grained aeolian material, the exfoliation of limestone from the walls and roof, the build-up organic matter on the terrace, and at least a small anthropogenic element. Micro-morphological analysis suggested that there was some soil development which probably took place over a long period of time. Pollen analysis, on the basis of four samples, indicated that the overall environment was that of a periglacial steppe, although there were also sparse pine woods in the vicinity. In general, it was concluded that the site was occupied only seasonally by small groups of itinerant hunters and gatherers.

It is unfortunate that the publication of Brînzeni itself was so much delayed (Allsworth-Jones et al. Reference Allsworth-Jones, Borziac, Chetraru, French and Medyanik2018a). As in the preceding case, a control excavation was carried out in 1993 but by far the larger area of the site had been excavated already by Chetraru in the years between 1960 and 1975 and by Borziac in 1987. The main cultural layer was much thicker than at Ciuntu and more difficult to interpret. An analysis of the sediments in the interior of the cave by Grishchenko (Reference Grishchenko1969) already demonstrated that this layer was complex in its formation. The micro-morphological analysis carried out as part of our project reinforced that interpretation. There was no real soil formation and the loess-like sediments were subject to much freeze-thaw activity. Pollen analysis suggested that the vegetation at that time could be characterised as a periglacial wooded steppe, somewhat warmer and wetter than the colder and more arid conditions observed at Ciuntu. The AMS radiocarbon dates obtained cover a very wide spectrum but two uncalibrated results from the 1993 section of 19,780±260 and 20,140±200 bp respectively probably provide the most reliable estimate. If so, the ‘Brînzeni culture’ can hardly any longer be regarded as Early Upper Palaeolithic. A re-examination of the archaeological inventory suggests that its so-called ‘archaic’ aspect has hitherto been much exaggerated (Noiret Reference Noiret2009; Allsworth-Jones et al. Reference Allsworth-Jones, Borziac, Chetraru, French and Medyanik2018a) and this would accord with the revised chronological position suggested for it.

The third site selected for inclusion in the programme, Buzdujeni, unlike the other two, had an exclusively Middle Palaeolithic occupation. As in the first two cases, a control section was established at the site in 1993, and an investigation was carried out with the following objectives: 1) to make a new and precise study of the stratigraphy; 2) to collect samples of bone and teeth for radiocarbon and ESR dating; 3) to obtain monoliths for soil micro-morphological study and samples for palynological purposes; and 4) to provide further information with regard to the archaeological finds. A detailed report based on the work carried out at that time was published – again with a regrettable delay – in the Journal of Paleolithic Archeology (Allsworth-Jones et al. Reference Allsworth-Jones, Borziac, Chetraru, French, Mihailescu and Medyanik2018b). The study of the deposits from a sedimentological and palynological point of view by Mihailescu and Medyanik respectively, as well as French’s micro-morphological evaluation, allowed a relatively detailed picture of the Pleistocene succession to be established.

The new profile established at the back of the cave on the western side in squares 3–7 is shown at Figure 1 (see Allsworth-Jones et al. Reference Allsworth-Jones, Borziac, Chetraru, French, Mihailescu and Medyanik2018b, fig. 6). It is in excellent agreement with the profile that Chetraru established at this point, on the line з/и running south–north. The stratigraphic units are designated as cultural layers numbered according to Chetraru’s system. As in his excavations, they coincide with lithologically distinct horizons and their position agrees closely with that established for them by Chetraru at the end of his excavations. Cultural layers 1–3 could not be examined in 1993, due to the crumbly state of the section at the top, although Mihailescu was subsequently able to provide a description of them. Five intact soil blocks were removed as sediment samples and eight samples were taken for pollen analysis. In summary, it appears that there were clear environmental fluctuations over time, from the base in layer 9 upwards. After the cessation of stream activity, which probably took place during the last (Mikulino) interglacial, there was an accumulation of deposits over a time corresponding to the last (Valdai) glacial period, with predominantly mechanical rather than chemical weathering and some input of aeolian material. In general, freeze-thaw structures are much in evidence. The most favourable climatic conditions occurred in layers 6 and 5a, the latter corresponding to an early interstadial, possibly equivalent to Brørup. Layer 6 represents a phase of maximum human occupation, although here as in the other layers there were no signs of soil formation or significant anthropogenic deposits. Layer 5a is clearly of subaqueous origin, whereas layer 5b represented a return to arid and cold conditions, as witnessed among other things by the presence of many sharp-edged limestone fragments. The pale yellow loam in the upper layers is comparable in external appearance to the Valdai loesses of the region. For the most part, conditions in and around the site throughout reflected the existence of a cold periglacial steppe, with rare pine and birch woods in the vicinity.

Fig. 1. Buzdujeni 1993 drawn section

The mammalian fauna from Buzdujeni was identified by A.I. David (Reference David1980) but had hitherto been published without distinction by layer. On the basis of a manuscript supplied by him, the present author and his co-workers were able to provide a division by layer (Allsworth-Jones et al. Reference Allsworth-Jones, Borziac, Chetraru, French, Mihailescu and Medyanik2018b, table 4). The total for all the bones was given as 10,255, of which 3133 (or 30.6%) could be identified, the minimum number of individuals represented (MNI) being 138. Cave bear, by number of identified specimens (NISP), constituted 68.9% of the total but in terms of MNI they are no more than 22.5%. The second most common animal, horse (Equus latipes), accounts for only 8.8% by NISP but 10.9% by MNI, with hyaena and bison not far behind. Layer 6 was the most abundantly represented, with 27.3% of the remains by NISP. There was little marked change in species frequency in the various layers over time. So far as the archaeology is concerned, the artefacts from layers 2 and 3 were described quite comprehensively, according to the information provided by Chetraru, but, for the remainder, only the broad outlines were indicated. The entire sequence, including layers 6–8, was described by him as a Denticulate Mousterian, the sole representative of this variant in Moldova.

A new book on Buzdujeni, by Vishnyatsky and Burlacu (Reference Vishnyatsky and Burlacu2021), has now been published in St Petersburg, as indicated above. These authors and their colleagues conducted a further control excavation at the site in 2017 (Vishnyatsky et al. Reference Vishnyatsky, Burlacu, Voskresenskaya and Nekhoroshev2020) and they also re-examined the entire available material from it. In particular, they had access to N.A. Chetraru’s abundant archive at the National Museum in Chişinău, which contains much hitherto unpublished information. The new account, which is very thorough and meticulous, contains additional data not included in Allsworth-Jones et al. (Reference Allsworth-Jones, Borziac, Chetraru, French, Mihailescu and Medyanik2018b), and it also modifies those conclusions in some respects. On the other hand, the book assumes a knowledge of the work hitherto undertaken in Moldova in general and also specifically at Buzdujeni, hence the background information provided above is necessary to appreciate in what respects our understanding of the site has (or has not) been affected by the new work.

With regard to the stratigraphy of the site, there is no essential change. The basic reference remains the profile established along the western side of the cave, as recorded in 1993 (Allsworth-Jones et al. Reference Allsworth-Jones, Borziac, Chetraru, French, Mihailescu and Medyanik2018b, fig. 6). In 2017, a small extension was cut running back from square 5 for 2 m, practically to the cave wall itself. The succession revealed was identical, apart from the fact that traces were found of layers 1–3, absent in the earlier work reported by Allsworth-Jones et al. Chetraru’s archive reveals that, in the course of his excavations, seven more profiles were recorded along all the sides of his trenches and careful horizontal plans were drawn of all his layers, analogous to the one published by Allsworth-Jones et al., which corresponds to layer 3 only (Reference Allsworth-Jones, Borziac, Chetraru, French, Mihailescu and Medyanik2018b, fig. 2). (In previously publishing only a fraction of this information, it may be remarked in passing that Chetraru clearly did not do justice to himself). Vishnyatsky and Burlacu emphasise, as did Allsworth-Jones et al., that human occupation of the cave was likely to have been only sporadic. Thus, although the stratigraphic units are formally referred to as cultural layers, they should not be understood as consisting predominantly of anthropogenic material. The authors also draw attention to the effects of taphonomic change, including the fact that artefacts could and did occur in random orientations, as was clearly observed in 2017.

Two AMS radiocarbon dates were reported as a result of work carried out in 1993 (Hedges et al. Reference Hedges, Housley, Pettitt, Bronk Ramsey and Klinken1996). Vishnyatsky and Burlacu record two more, in addition to those which were reported there (Table 1).

Table 1. Radiocarbon Dates from Buzdujeni

The two new dates (Table 1), like the old ones, show an apparent stratigraphic inversion, the vertical distance between layers 2 and 6 being greater than that between 6 and 8. Both the new dates are older than the previous ones, which, as the authors say, can well be due to improved pre-treatment. Nonetheless, all the dates are fairly tightly clustered and this, in the light of the general stratigraphic situation (including the suggested age of layer 5a), seems distinctly improbable. Even if, as the authors say, the main occupation horizons can safely be equated with OIS 3, these dates can be regarded as no more than minimal, probably reflecting their closeness to the limit of the effectiveness of the radiocarbon dating method. The authors agree that further dating evidence would be desirable, using different methods, but in view of the fact that the cave sediments have now been almost entirely removed, this may not be too easy to achieve.

The details of the fauna published on the basis of a manuscript provided by A. I. David (see above) followed his indications exactly, including the division of the finds by layer (Allsworth-Jones et al. Reference Allsworth-Jones, Borziac, Chetraru, French, Mihailescu and Medyanik2018b, table 4). It appears that these details were also published in another venue, together with extra information not previously available (David et al. Reference David, Pascaru, Rusu and Găleanu2006) and this data is included in the book by Vishnyatsky and Burlacu. In respect of MNI for the large mammals, the numbers given are identical to those published by Allsworth-Jones et al., with the omission of the three small mammals included in their list (Lepus, Marmota, and Spalax) and the inclusion of wolf (Canis lupus), not hitherto mentioned. There are slight differences in terms of NISP, but nothing significant, the totals being practically the same as previously published: 10,168 (identified 3128, unidentified 7040). In addition, 310 new specimens were located during the excavation of 2017 (identified 46, unidentified 310).

What is completely new is the publication of 29 listed species of small mammals, with an NISP of 1551, including 1441 from layer 5. The most common by far is the narrow-headed vole (Microtus gregalis) (1000), followed by the steppe lemming (Lagurus lagurus) (110), the tundra vole (Microtus oeconomus) (105), and the steppe marmot (Marmota bobac) (96). The inclusion of small numbers of arctic shrew (Sorex arcticus) and collared lemming (Dicrostonyx guilielmi) in layer 5 is taken as an indication that that layer was very cold. Layer 5 was apparently not sub-divided but these conditions correspond to layer 5b as published by Allsworth-Jones et al. (Reference Allsworth-Jones, Borziac, Chetraru, French, Mihailescu and Medyanik2018b). Also identified in layer 5 (presumably 5a) were numbers of amphibians and reptiles, including 17 toads, six green lizards, and four vipers (Redkozubov Reference Redkozubov2006). It is assumed that they perished during hibernation, which would provide another indication that at that time there was little or no human occupation in the cave. All these remains were recovered by wet sieving, not a very common practice at the time, and, incidentally, a further demonstration of the care taken by Chetraru and his colleagues. Finally, it is of interest to note the special study carried out by E.N. Mashchenko and T.F. Obade concerning the mammoth remains (Vishnyatsky & Burlacu, Reference Vishnyatsky and Burlacu2021, appendix 3). They showed that the 12 mammoth teeth recovered were all deliberately broken off from an animal’s upper jaw and presumably owe their presence on the site to intentional human intervention, including transport from elsewhere, although no particular motive could be suggested for this.

The palynological results from 1993 were given in detail by Allsworth-Jones et al. (Reference Allsworth-Jones, Borziac, Chetraru, French, Mihailescu and Medyanik2018b) but it appears that another article on this subject has also been published (Medyanik & Chetraru Reference Medyanik and Chetraru2006). According to Vishnyatsky and Burlacu, this relates to the earlier results obtained on the basis of samples taken from two of Chetraru’s profiles established in the main part of the cave in the 1970s. They are severely critical of this article, on the grounds that it is incomplete and contradictory, and they do not rely on it. It is difficult to know the reasons for this but it is hard to believe that Svetlana Medyanik was in any way responsible. In any case, her results from 1993 are accepted in their totality and inform their own account of the succession.

The most extensive new information about the site relates to the artefacts. Allsworth-Jones et al. (Reference Allsworth-Jones, Borziac, Chetraru, French, Mihailescu and Medyanik2018b) state that a fairly comprehensive inventory was provided for layers 2 and 3 but the remainder was described in broad outline only. It was recognised that this record was incomplete and that a much fuller study was needed. That study has now been undertaken and the description and analysis provided by Vishnyatsky and Burlacu supersedes all preceding accounts. They have also provided a large number of illustrations. Some of these were taken from drawings in the Chetraru archive but others by Burlacu and S.N. Razumov have been added, as well as some excellent photographs by Vishnyatsky. The overall inventory is shown in Table 2.

Table 2. Lithics from Buzdujeni

The overall total includes 207 pieces from the excavation of 2017 but the great majority, 11,654 pieces, comes from Chetraru’s excavations. It can be seen that the number of finds per layer is quite variable, the most abundant by far being layer 6. Layer 4a refers to a small collection made at the mouth of the cave only. The raw material was almost entirely black and grey flint derived from outcrops along the Pruth river. According to the authors, black flint predominated in the lower layers, giving way to grey in the upper layers, perhaps somewhat easier to obtain in the increasingly harsh environment of the site. Unsurprisingly, waste flakes amount to 86% of the total. Out of 2219 striking platforms analysed, 55% were plain, 10% dihedral, 24% facetted, 8% punctiform, and 3% cortical. The sequence of core reduction, as reconstructed by the authors, proceeded from a single-platform parallel removal stage to a multi-platform discoidal radial set of removals, the shape of the core in the process often changing from 1-sided to 2-sided. Not a word is said about the presence or otherwise of the Levallois technique, which is odd, considering the prominence which it has in the accounts of the industry given by Chetraru and N.K. Anisyutkin, reflected also in the description of the situation given by Allsworth-Jones et al. Some of the drawings included in the new book look like Levallois products to this author, who described certain of the 1993 finds and others in the Chişinău museum in this way (Allsworth-Jones et al. Reference Allsworth-Jones, Borziac, Chetraru, French, Mihailescu and Medyanik2018b, figs 12 and 13). This question needs to be further addressed one way or another.

In general, the authors (like Chetraru before them) consider that the archaeological industry at Buzdujeni remained essentially the same from start to finish, notwithstanding its intermittent occurrence and no doubt lengthy time frame. In their view therefore it is appropriate to present a single unified ‘portrait’ of it, as indicated Table 3, listing all the tools recovered from the site.

Table 3. Retouched Pieces from Buzdujeni

It is immediately apparent that notches and denticulates are comparatively important, constituting 17% of the total. If they are considered together with the minimally diagnostic retouched blades, flakes, and fragments, this proportion increases to 48.5%. Hence it is easy to see why they have hitherto featured prominently in the characterisation given to the site, but the authors give good reasons why this should not be so. Many of the pieces have varying degrees of edge damage such that, as the authors say, it was even difficult for them to decide which should actually be counted as intentional tools. They attribute this damage to taphonomic factors, including trampling by large animals, such as cave bears, moving around in a comparatively small space. Various actualistic studies have demonstrated how such damage may occur and the authors quote in this context an experiment carried out at Akhshtyr cave (Kulakov & Giria Reference Kulakov and Giria2017). Newly minted flakes were scattered at this site in areas frequented by visitors and the results examined after some time clearly showed what could happen. What can happen to stone tools can also happen to bones strewn on the cave floor. It is a further tribute to the thoroughness of the work carried out previously that a record exists of the ratio of identified to unidentified fauna for each of Chetraru’s excavated squares. This information has been collated and plotted by the authors, from which it is apparent that relatively more unidentified remains are concentrated in the central areas of the cave, whereas on the peripheries and next to the walls the identified remains are much more common. This distribution cannot be accidental and it supports the idea of trampling by large animals which, for long periods, were sole masters of the cave. For the above reasons, Vishnyatsky and Burlacu reject the idea that Buzdujeni can be characterised as representative of a Denticulate Mousterian.

On the other hand, ten pieces have been identified as bifaces, described in detail. They come from layers 4, 5, and 6 only, and they are far from uniform in their appearance. Some are broken or unfinished, four are considered possible roughouts for leaf points, and one bears a resemblance to a bifacial backed knife. In addition, there are some flakes which are reasonably interpreted as biface thinning flakes. Few though they are and scattered over time, the authors accord the bifaces a special significance and, on these grounds, assign the whole site to the Micoquian. The Micoquian is certainly a well known entity, with several very well characterised assemblages in central and eastern Europe, so the idea is not implausible. In the region of south-eastern Europe the authors compare the site to, for example, Ripiceni-Izvor, on the right bank of the Pruth, and also to abri 122, in the Vâghiş river valley of eastern Transylvania (Cosac et al. Reference Cosac, Murătoreanu, Veres, Niţă, Schmidt, Hambach, Alexandru, Cuculici, Buzea, Mărgărit, Dumitraşcu, Vasile, Petculescu and Dénes2018; Veres et al. Reference Veres, Cosac, Schmidt, Murătoreanu, Hambach, Hubay, Wulf and Karátson2018). The latter site, first located in 1989 and now under re-investigation, has, incidentally, a much longer sequence of absolute dates than Buzdujeni but could very well run parallel to it. As the authors recognize, all these aspects and others do merit further research.

In the meantime, this book, taken together with our own article on the subject, surely indicates the significance of Buzdujeni in the archaeology not only of Moldova, but also the region as a whole, and it is hoped that these results will become widely known.