Aurelian Rusu
Citește mai multe articole de la acest autor.
Read more articles by this author.

Descarcă acest document în format PDF.

Download a PDF file of this document.

Keywords: dry-stone walls, ‘trapezoidal house’, ‘Lepenski Vir – Schela Cladovei culture’, Mesolithic, Neolithic


This paper discusses the recorded and published finds from Stubica, a site of the ‘Lepenski Vir – Schela Cladovei culture’, emphasizing the presence of trapezoidal stone structures which are characteristic of this culture. The importance of the site and its architectural features within its broader archaeological context of the Iron Gates Mesolithic and Neolithic is also addressed.


The site of Stubica has been used by its discoverer BorislavJovanović[1] as a tool in enforcing his interpretation of Mesolithic and Early Neolithic of the Iron Gates area. However, the site’s features were somewhat overlooked by the researcher himself in his arguments, and thus got lost in the ‘mists’ of the Danube’s Iron Gates archaeological record. The aforementioned author has attributed the site to the ‘Neolithic Lepenski Vir culture’ (Jovanović 1971, 36), while other researchers[2] who followed his work also pursued his interpretation of the Stubica site. They accepted its attribution to the Neolithic. That is until now. While I cannot produce any unpublished record for this site[3], by analysing the provided one in contrast with the rest of the published archaeological record for the aforementioned archaeological periods and region, I can bring a different perspective on Stubica and its relationships to the ‘Lepenski Vir – Schela Cladovei (LV – SC) culture’[4]. This ‘culture’, which spanned for several millennia c.9500–6000 cal BC roots deeply into the European Mesolithic. In its final stage it was contemporaneous with the South East European Early Neolithic (Rusu 2011).

History of research

Research on the Mesolithic and Early Neolithic settlement of the Iron Gates region, rarely mention Stubica[5]. The site is located in the Upper Gorge of that area, on the right (Serbian) bank of Danube River, across the river from Buffalo’s Rock[6], between Padina and Lepenski Vir sites. The excavations were carried out by B. Jovanović in the autumn of 1970. They were rescue excavations conducted as a part of the project for construction of the Iron Gates Dam I. At present the parts of the site researched then are submerged.

The site[7] was mentioned by the excavator in his 1971 paper: “Stubica, located between the dissected ridges of Bivolja Stena, breaking through the flow of the Danube at this place, represents a settlement of larger dimensions. It was situated within a deep valley surrounded by steep mountains, with the base of massive layers of loess. Sounding works of that locality, of very limited size, have discovered along the riverbank area (110 m of length) a cultural layer of 1,50 to 1,80 might, with at least two dwelling horizons. Ceramics found here in a great number, corresponds according to its basical, typological quantities to Padina B–2 phase, although some earlier elements are also possible. The position of Stubica in relation to the Upper Gorge is also not without any significance – it is situated nearly at the half-way distance between Lepenski Vir and Padina (Pl.I.1)” (Jovanović 1971, 35). The author also noted the similarities between these horizons and those from Padina B–1 and B–2, and Lepenski Vir I–II, which he interpreted as a “phase of settlement’s development for the fishing-agricultural – cattle-breeding period of the Iron Gates group of the Early Neolithic” (Jovanović 1971, 36). The same relative periodization between the sites, including Stubica, was reported again in more recent papers (Jovanović 1973, 293; 1974, 21).

The first and last paper on Stubica written by its discoverer was published in 1984. In this paper the author reported that the size of Stubica was similar to that of Padina B III[8]. There are descriptions of remains of habitation structures similar to those from Padina. They are partly interred, with trapezoidal foundations. Their front, back and sides are delimited by stone walls, with thresholds and hearths (Jovanović 1984, 177–178). No plans, drawings, or photographs of the excavations, or material cultural remains of the site have ever been published. The researcher focused the short publication on the Starčevo type pottery and the similar finds from Padina and Hajdučka Vodenica, excavated by the same author (Jovanović 1966; 1968; 1971; 1974), and also on the ones from Lepenski Vir. It was mentioned that the large quantity of pottery could not be studied in its totality, because of technical problems[9]. Another topic of the paper was on the fact that climatic and anthropological factors at the time did not allow a more detailed research. The anthropological factor was the construction of the first hydroelectric dam of the joint Yugoslav – Romanian project that was opened in 1971.

The best report on the site was later published in 1996 by Ivana Radovanović. This author attributed the site to the Early Neolithic according to the ideas of its discoverer. A distinction was made for the stone constructions that are identical to those from Padina B3, and the pottery was comparable to that from Padina B2, Lepenski Vir III and Donja Branjevina, that is Starčevo type pottery (Radovanović 1996, 345). Since then the researchers working on the Mesolithic and Neolithic of the Iron Gates, indiscriminately refer to an Early Neolithic site when referring to Stubica. It is worth noting that the reference is just in the maps of some papers (Bonsall 2008; Bonsall et al. 1996; 1997; Borić 2001; Borić et al. 2008) as well as in a catalogue of the discoveries made in the Iron Gates (Radojičić and Vasić 2003, 40–41). Apart from the publication of Adina Boroneanț that follows Jovanović’s interpretation (Boroneanț 2012, 218–219), no other researcher[10] gave it a second thought

Analysing the site’s discoveries

Two elements are attributed to this site. They are first, ‘Starčevo culture’ pottery, second, habitation ‘stone structures’. Out of the two, the one that can and will be worked through is the latter, because the other is self explanatory (the pottery being of Starčevo type, as it is reported to be)[11]. If one follows the arguments of B. Jovanović, we can see that the reason why he attributed the site exclusively to the Early Neolithic was based on the presence of the characteristic pottery of ‘Starčevo-Criș culture’[12] type. According to the excavator the pottery was associated with the ‘trapezoidal structures’, as he had already noted at Padina[13]. Thus for him, the proximity with Padina, and the similarities with the finds was the key element to attribute Stubica to the Early Neolithic (Jovanović 1971, 35; Jovanović 1984, 177–178).

‘Trapezoidal structures’ and their constitutive elements

Now let us focus on the ‘stone structures’ from Stubica. These habitations have a trapezoidal foundation. Their front, sides and back are bordered by dry-stone (technique type) walls. Inside, they have a sunken rectangular hearth with a typical threshold. ‘Trapezoidal structures’ were recorded at Alibeg, Padina, Stubica, Lepenski Vir, Vlasac, Icoana, Ostrovul Mare km 875, Kula (Radovanović 1996, 316–351) and Schela Cladovei (Bonsall 2008, 256). In the next paragraphs I will discuss their characteristics, the shape, floors, hearths and drystone walls.

Trapezoidal foundations

Rectangular, irregular rectangular and oval foundations were also recorded. Considering the complex stratigraphical sequence recorded by all the researchers who excavated sites in the Iron Gates (Srejović 1972, Jovanović 2008, Sladić 1986, Păunescu 1996, Boroneanț 2000, Borić et al. 2008, Boroneanț and Bonsall 2013), and the fact that the foundations consist of different types of floors, those which are directly on the ground surface, and those which are represented by a line of stones that rendered the trapezoidal shape of the foundation, it seems likely that some of the ‘rectangular, irregular rectangular and oval shaped foundations’ could in fact have been ‘trapezoidal shaped foundations’. Evidently this is mere speculation and the ones that we are strictly interested in are the ones declared as having ‘trapezoidal shaped foundations’. Nevertheless the ‘rectangular irregular rectangular and oval shaped foundations’ were recorded at Alibeg, Padina, Lepenski Vir, Vlasac, Hajdučka Vodenica, Răzvrata, Icoana, Ostrovul Banului, Schela Cladovei, Ostrovul Mare and Kula (Radovanović 1996, 124–125). The sites with ‘trapezoidal shaped foundations’ were already named on the previous paragraph.


Regarding floor surfaces which were ‘reported’ to be ‘directly on the ground[14], we have to be somewhat cautious. In my opinion there were cases when what our fellow researchers considered to be a sterile deposition was in fact a floor. As an example I shall discuss a dwelling recorded at Vlasac by Dušan Borić and his collaborators in 2006: “The dwelling had a levelled floor area with several large blocks of construction stones at the floor level. The thin flooring was red-burned throughout and it possibly contained some limestone inclusions, which the micromorphological analyses should determine in more detail. There were no artefacts on the floor level and it seems that, upon abandonment, the floor was intentionally covered with a 10cm layer of hard-packed and largely sterile soil. The layer above this deposit contained a concentration of burden wooden poles and charcoal, possibly from the upper construction of the dwelling. A broken projectile point found in this association is dated in the range 6654 to 6484 cal BC 95 per cent probability (OxA–16540) and likely represents the terminus ante quem for the occupation of the dwelling” (Borić 2006, 12). As one can notice within the quoted paragraph, it is not that a floor was not recognized. From the fact that ‘there were no artefacts on’ what the researchers thought to be ‘the thin red-burned floor’, and that one artefact was found over the ‘largely sterile soil’ that covered the floor area uniformly (10cm layer), along with the ‘concentration of burden wooden poles and charcoal’ that was resting on the ‘largely sterile layer’; I argue[15] that the actual floor surface was in fact the ‘10cm layer of hard-packed and largely sterile soil’ that was applied over the ‘thin red-burned[16] ground.

We should also consider that perhaps such was the formation and the utilization of the habitation that a distinct floor could not have been archaeologically recorded. Nevertheless, it must have been there considering the fact that in some situations it was marked by a row of stones, or that hearths or traces of fire within a habitation were recorded. Even without evidence of special preparation from an archaeological perspective, we have to consider that people, when constructing a habitation, did in fact prepare the ground beneath their feet, the ground on which they sat and slept most of the time. In its most elaborate form (Lepenski Vir I, Vlasac), the floor[17] of the structure was made out of a concrete-like substance (limestone as base material) which incorporated a central rectangular hearth (Srejović 1972; Srejović and Letica 1978). An alternative form was a floor made of clay [Padina (Jovanović 2008), Vlasac (Srejović and Letica 1978), and Kula (Sladić 1984; 1986)], which also incorporated a rectangular hearth. The incorporated rectangular hearths were made of stone blocks (rectangular prism-like) or stone plaques. Both types of floor had thresholds made of stone slabs that linked the entrance with the hearths. A floor of burned clay was recorded at Schela Cladovei (Păunescu 2000, 443)[18] with no hearth in relation to it; while one in relation to a rectangular hearth was mentioned for Hajdučka Vodenica (Radovanović 1996, 322). As for Icoana, a floor of pounded clay with large stone slabs was mentioned. The stones were considered not to belong to the platform; close to, but outside of the structure was a rectangular hearth (Boroneanț 2012, 124, 126).


Stone bordered hearths (rectangular, oval, round and trapeze) linked by a concrete like substance or clay were reported at Alibeg, Icoana, Ostrovul Banului, Ostrovul Mare km 875, Kula (Radovanović 1996, 137–138) and Schela Cladovei (Boroneanț 2012, 173). They were associated with oval or trapeze shaped structures. At Alibeg the two rectangular hearths were inside the trapezoidal shaped structures (Boroneanț 2000, 108–110). If we consider their binding material, these hearths were either part of floors made from the same material (floors that were destroyed prior to the archaeological discovery), or models of a specific kind of hearths, where the constructors were focused only on the durability of the hearths and gave little attention to the floors. Either way, the two binding materials are similar to the ones used for the two types of floors mentioned above. Rectangular hearths made of stone slabs, some of them with a specific threshold, others without, but all associated with the trapezoidal dwellings were recorded at Alibeg (Boroneanț 2012), Padina (Jovanović 2008), Lepenski Vir (Srejović 1972), Vlasac (Srejović and Letica 1978), Hajdučka Vodenica (Jovanović 2008), Icoana (Boroneanț 2012), Ostrovul Mare (Boroneanț 2012) and Kula (Sladić 1984; 1986). Rectangular shallow ditches intended for the stone slabs were recorded at Hajdučka Vodenica (Jovanović 2008, p. 321, fig. 42–43).

Dry-stone walls

Remains of dry-stone walls were recorded at the back, front and sides of the trapezoidal structures at Padina (Jovanović 1987, 1–16; 2004, 55–61; 2008, 289–324), Stubica (Jovanović 1984, 177–178), Lepenski Vir II (Srejović 1966, 13–17; 1969, 13–21; 1972) and Vlasac (Srejović and Letica 1978, p. 25, fig. 12; Borić 2006, 12). The documentation of the stone walls, and their relationship with the ‘trapezoidal structures’, is rather insufficient, even where ample documentation was possible, as the following paragraphs will show.

At Padina, the dry-stone walls are recorded as follows: “…House 13, for instance, was the only building at Padina with its rear zone limited by blocks of rock placed vertically, similar to a massive stone wall. Bringing such large rocks (1,50–0,80 m),…” (Jovanović 2008, 301). However, the lack of documented photographs and drawings is compensated by artistic reconstructions (Jovanović 2004, fig. 1, 2).

At Lepenski Vir, there is only one documented representation of the stone walls, a photograph (Srejović 1972, 87, fig.  28), although they were described in detail as: “…arched supporting walls of stone blocks and slabs were set up to a height of about one metre. These constructions, executed in a drystone technique, are solidly built and in some layers have been preserved complete” (Srejović 1972, 74).

At Vlasac, one could argue that the ‘stone structures’ recorded there, made of stone slabs, could have been parts of fallen walls (see Srejović and Letica 1978, plates XIV, XXV, XXXI, XXXII), especially where they were in the proximity of the hearths (see Srejović and Letica 1978, plates XVI, XX, XXI, XXV). The information given for the remains of the dry-stone walls at Stubica has already been presented above.

There are some observations to be made concerning the types of floors (limestone or clay) related to the dry-stone walls. It is to be noted that these walls were only recorded at the second settlement LV II at Lepenski Vir. However, the researcher failed to mention what specific material the floors were made of, aside from that ‘they were no longer floored with limestone mortar’ (Srejović 1972, 75). This omission seems a bit reckless, since the author gives a thorough description and reasoning for the appearance of the walls at LV II; it was a consequence of the previous LV I settlement that modified the slope of the site, creating anthropic terraces that needed reinforcement with walls to protect the backs and the fronts of the new structures, in order to prevent earth slides (Srejović 1972, 73–77). The impression that one could get is that the floors were made of naturally levelled soil[19] that covered the previous structures of LV I, and apart from levelling, nothing else was needed to be done by the inhabitants of LV II. Perhaps the quality of the soil was sufficient for the inhabitants of LV II, who according to D. Srejović[20] lacked the technical knowledge of their predecessors to produce limestone floors.

At Padina the floors of the structures with drystone walls were made of pounded clay, as was reported by B. Jovanović (1971; 1972). This site lacks structures with a floor made of limestone plaster.

At Stubica there is no mention of a floor in association with the remains of the stone walls. Most likely that is an archaeological omission[21]. From the data presented above, it is understandable that the elements of the trapezoidal structures were documented throughout the sites of ‘LV – SC culture’. Where only partial data has been recorded there are several factors to explain it. Namely, the condition of preservation, or more accurately put – the state of degradation of the uncovered structures. Secondly, the archaeological method that was used, some of which was erroneous in execution and thus in recording the data. However, it is clear that the stone structures discovered at Stubica belong to a ‘LV – SC culture’ type site.

Discussion on the cultural nexus of the ‘trapezoidal structures’

Some of the ‘trapezoidal structures’ at Padina and Lepenski Vir were reported as being archaeological contexts for ‘Early Neolithic’ type pottery[22]. As such they were the key note for the ‘debate phase’ of research on the archaeological findings within the Iron Gates area (Rusu, 2016). The debate focussed on whether these structures were Mesolithic or Neolithic[23], and as such on the ‘LV – SC culture’ itself.

One researcher, D.  Srejović, who excavated Lepenski Vir (and Vlasac) considered the presence of Neolithic pottery to be intrusive[24] from the upper levels which were recorded as belonging to Neolithic settlements (Srejović 1966, 15; 1968, 85–87; 1972, 134–135). The other, B. Jovanović, who excavated Padina (Hajdučka Vodenica and Stubica) considered that the Neolithic pottery found at this site was in direct context with ‘trapezoidal structures’ (Jovanović 1971, 1972, 1973). The fact that the Early Neolithic[25] settlements recorded in South-Central Europe did not display this type of architecture for habitations[26], did little for the debated upon the cultural nexus of the sites from the Iron Gates region. Those were seen as a local type of Early Neolithic manifestation by B. Jovanović and D. Borić[27]. These researchers were in fact the most vocal of such interpretation. However, when the evidence from their research is considered in the context of other available data on this subject, the interpretation annuls itself[28]. The fact that the settlements in question did not display pottery[29] of Early Neolithic type throughout their habitations – 15 out of c.85 at Lepenski Vir I (Srejović 1972, 49, 134); 8 out of 15 at Padina B III (Radovanović 1996, 280–281) were reported with pottery – did not produce at least some restrain in considering them as Early Neolithic. To prove they belong to Early Neolithic they should expose those defining traits[30] throughout all or most of the habitations and their respective sites.

B.  Jovanović asserted in his last paper on the subject that »The stratigraphic position of concentrations of pottery within the cultural layer is uncertain, due to the formation of large middens upslope« (Jovanović 2008, 303) and also the assemblage found in house 18 at Padina III[31] »…it is more similar to an unexpected (sic!) phenomenon of the “ProtoVinča pottery”… « (Jovanović 2008, 309). These facts do not support the ‘in situ debate side’ i.e. the ones that claimed that ‘LV-SC’ is a ‘Neolithic culture’. The »Settlements of the Lepenski Vir culture in Upper gorge of the Iron Gates (Padina, Stubica, Lepenski Vir, Vlasac) are situated in the closed microgeographical area which complete mileage is only 15km. It means that all these settlements were at the walking distance from each other« (Jovanović 2004, 55). Out of the four only Vlasac reported no pottery within its ‘trapezoidal structures’ (Srejović and Letica 1978; Borić et al 2014, 26–27) even though Vlasac is the farthermost downstream, a fact that should had been relevant to B. Jovanović considering that he saw Padina as » the settlement at the western most periphery of Lepenski Vir culture to which interactive contacts with Starčevo culture settlements downstream from the Iron Gates (Ključ region) arrived most lately « (Jovanović 2004, 58), and that the » only accessible communication was running along the narrow littoral zone or by the Danube whose rapids and whirlpools were significant obstacle for navigation « (Jovanović 2004, 55). However, not only did this not create a problem in the reasoning of the respective researcher, but also he had not acknowledged the fact that Vlasac had trapezoidal structures, or the fact that Stubica[32] also had these types of structures, as he stated that » Padina and Lepenski Vir are (the) only settlements from both Danube banks in the Iron Gates gorge that have architecture declared justly as the symbol of this culture « (Jovanović 2004, 55).

Furthermore, one should notice that at Vlasac and Lepenski Vir, settlements of Early Neolithic type[33] were recognized, overlaying on those of LV – SC culture’s settlements (Srejović 1968, 1972; Srejović and Letica 1978). On the majority of the sites attributed to this ‘culture’ this situation was recognized (Radovanović 1996, 316–351), with the exception of the sites excavated by B. Jovanović, namely Hajdučka Vodenica, Padina and Stubica, a situation that is rather questionable (to say the least), considering that those were the basis[34] on which he constructed his interpretations of the ‘Iron Gate group of the Early Neolithic’ as a whole. It becomes strikingly clear that B.  Jovanović simply did not recognize[35] whilst excavating, the presence of any Early Neolithic structure that could have over imposed upon the trapezoidal structures. Thereby, he considered the ‘trapezoidal’ ones, based on the presence of portable Early Neolithic material, to be an Early Neolithic architecture.

The fact that most of the archaeological material[36] and the burials associated with these structures were of Mesolithic timeframe and tradition[37] (Srejović 1966, 1972; Jovanović 1971; 2004; 2008; Kozłowski and Kozłowski 1982; 1984; Radovanović 1996; 2000; 2006; 2006a; Roksandić 1999; 2000; 2006; 2008; Borić 1999; 2011; Bonsall et al. 1997; 2000; 2002; 2004; 2008; Borić and Miracle 2004; Borić and Dimitrijević 2009; Borić et al. 2004; 2008; 2009; 2014; Boroneanț 2000; Bonsall 2008; Lazarovici 1979; 1979a; 1983; 2006; Mihailović 2004) was somehow overlooked by B. Jovanović and D. Borić. They took that to be evidence of a local type of Early Neolithic – one that used Starčevo culture’s pottery and some types of stone tools[38] and other portable artefacts specific to the Early Neolithic, together with the use of predominant Mesolithic artefacts and Mesolithic burial manifestation. In this sense, the first researcher considered it a ‘kind’ of Neolithic culture (Jovanović 2008), while the last researcher considered it to be evidence for a ‘transformational phase’ (Borić and Dimitrijević 2009) of the Iron Gates Mesolithic to Neolithic. As I have already stated (2011) a ‘culture’ or a ‘transformational phase’ will expose its archaeological features on a site throughout the majority of its habitations, not only on (roughly) half of them as it was reported at Padina for example. Furthermore, more than one material needs to be considered (pottery in this case) when identifying cultural trait. Moreover, if we are to consider a cultural manifestation, we are to see its traits, if not in all archaeological contexts[39], then at least in the majority of them. Therefore, the archaeological traits for what they understand to be a ‘Lepenski Vir culture’ of Mesolithic tradition that incorporated some Neolithic traits should be altogether and undoubtedly in close relation to one another. That is not the case for ‘LV – SC culture’. It is (only) of Mesolithic tradition.

For the trapezoidal structures the chronological dates prove that this type of architecture was in use a ‘millennia’ prior to the Early Neolithic phenomenon reaching Iron Gates[40]. Not only were they in use then, as it was previously stated, but also the materials and burials associated with them were of Mesolithic tradition, thus demonstrating that this architecture was a local innovation, a reaction to a specific environment[41]. Even though the plaster technology has a wider context and was related to the Neolithic in Near East region, for the same time frame as the one used for LV-SC structures, it is a fact that the European Neolithic on the other hand, lacks this technology (Nandris 1988). These facts should annul previous[42] claims that these types of architecture were of Early Neolithic tradition.

If one follows the discourse of D.  Borić over time, one can notice that if one supposition fails to sustain itself on the data, then another is being proposed for the same idea i.e. that the ‘trapezoidal structures’ from the Iron Gates were influenced by those from ‘Neolithic Anatolia’. The argument has not been modified despite D.  Borić’s[43] over 15 years of research on the subject, coupled with more than 50 years since the discovery and the subsequent research of these structures in the Iron Gates area; the consequential data produced no evidence that the floor technology was imported. Nor do the arguments proposed by this researcher sustain this hypothesis. It was suggested first that the floor technology reached the region as part of the ‘Neolithic package’ (Borić 1999; 2002); then, it was correlated with burials underneath the floors at Lepenski Vir (Borić 2004; 2008); now a third correlation is being proposed, with Cyclopeneritea (Borić 2008; 2011). The problem with all these scenarios is that they do not correlate with each other, nor do they sustain the hypothesis on their own. The first scenario fails as the ‘Neolithic package’ was later in the Iron Gates region than the technology for the floors. The correlation with burials underneath the floors at Lepenski Vir also fails as the practice was documented at other sites – Vlasac, Schela Cladovei – as being earlier than the ‘arrival’ of the Neolithic in the region. And, if one follows the researchers own arguments (2011, 170), the scenario with the Cyclopeneritea just does not stand, because » The habitat of these marine gastropods is deltas of big rivers to the sea, and those found in the Danube Gorges might have come from the Black Sea and the Danube delta. «; meaning they are not part of a network related to ‘Neolithic Anatolia’ but rather of a network constructed on the Danube river. A network that ‘LV-SC’ was part of.

It is worth noting that Hajdučka Vodenica was considered by B.  Jovanović an Iron Age site (Jovanović 1966; 1966a; 1968) and that D. Srejović (excavator of Lepenski Vir and Vlasac sites) was the one who observed that at least for the first level of habitation there were elements that belonged to the Late Epipaleolithic manifestations (Srejović 1969, 16–17). A fact that B. Jovanović was reluctant to accept (Jovanović 1971, p. 37 note 28) until 2004 when he conceded that it belonged to the ‘Lepenski Vir culture’, with the distinction that he considered it » a kind of ‘fishing Neolithic’ « (Jovanović 2004). The elements in question were rectangular hearths, two stone decorated boulders, at least one floor made of pounded clay, tools made of bone, antler and boar tusks, stone tools (anvils, hammers, axes and fishing weights) and burials. However, what is more intriguing about Hajdučka Vodenica is that even though Early Neolithic pottery was associated with some of the elements in question, following the finding of Iron Age material, the author of the excavations considered the structures an Iron Age phenomenon. Perhaps because that was the first site out of all three (Hajdučka Vodenica, Padina, Stubica) that B. Jovanović excavated or because his specialisation was Bronze/Iron Age. Whatever the case may be, one cannot help noticing that this researcher initially considered the first site (that he excavated for the Iron Gates dam project) Hajdučka Vodenica as belonging to the Iron Age period on the grounds of finding material from that period on the lower most level of the excavation; and that the same researcher considered his second site as belonging to Early Neolithic on the ground of finding pottery from that period in what he believed to be closed archaeological complexes – which now, according to his own last publication on the subject, are far from it (see Jovanović 2008, 303); and that the same researcher, on basis of the findings from his second site (Padina) and noticing the similarities, attributed his third site – Stubica – to what he believed was a local type of Early Neolithic. It is worth noting that Iron Age archaeological remains were recorded at the majority of the sites that also exhibited ‘LV – SC culture’ manifestation. As, previously stated, they also had Early Neolithic remains. More precisely the sites presented occupational layers and features dated to the Epipaleolithic/Mesolithic (i.e. ‘LV-SC culture’), Neolithic, Iron Age, Roman and Medieval times.

The novelties of ‘LV-SC’ archaeological manifestations proved challenging to most of the archaeologists involved in the excavations at the time. After some consideration the majority of them acknowledged that it was a local Mesolithic phenomenon[44]. B. Jovanović was the only researcher involved in the excavations of the remains of ‘LV – SC culture’ that considered it from the start (1968) as being a Neolithic phenomenon, and who maintained his interpretation all the way to his last publication on the subject (2008). However, the fact that the majority of the researchers involved with the excavations agreed on what ‘LV-SC culture’ was, in itself is not conclusive, as it happens more does not mean better. Nevertheless, in this case, B. Jovanović got his interpretations wrong[45]. And, as much as his follower D. Borić tried in the recent years to build the evidence to support B. Jovanović interpretation for this ‘archaeological culture’, the data[46] just does not support it. Both of them, in trying to understand the phenomenon, focused mainly on the sites from the gorges of the Iron Gates region, and only on the ones from the right bank of the Danube, namely Padina, Lepenski Vir, Vlasac and Hajdučka Vodenica. It is clear that the sites chosen for debate were the ones excavated by B. Jovanović (Padina, Hajdučka Vodenica), versus the ones excavated by D. Srejović (Lepenski Vir, Vlasac). However, the phenomenon was present on at least 17 sites (Rusu 2010, 12) so far, a fact recognized by many other researchers, including the two in question, though in a sense it eluded them. For if they would have always considered the archaeological phenomenon in its wider context, perhaps their interpretations could have been more accurate. Or perhaps not, considering that, unfortunately, perception[47] is what governs us in understanding the past human experience, more than any other cognitive function.

This ‘LV-SC’ is a complex phenomenon in all its forms of subsistence strategies, architecture, burial ritual, art and tools. And as such it is more diverse in its architecture than just this form of trapezoidal ground plan structures. The one in question is a distinct trait, which was discussed here because its remains were discovered at Stubica. Through all its diversity of manifestation this ‘culture’ blends its forms in a way that makes it a whole. It seems it is more correct to understand this particular phenomenon, or more precisely to work with it from the perspective of its cultural manifestations as a whole, rather than fragmenting it in to local units – its sites. Sites that were discussed on the principle of their degree of resemblance (Padina vs. Lepenski Vir), rather than their traits that can be found in a lesser quantity in others, and that make them part of the same phenomenon. Therefore, it is appropriate to approach those traits from a wider archaeological construct[48].


From the data that has been collected and published so far and discussed in this paper, one can understand that the ‘trapezoidal structures’ within the Iron Gates region was a ‘LV – SC’ cultural phenomenon seen as a local Mesolithic manifestation and therefore the remains that were recorded at Stubica belong to that phenomenon. The Stubica site preserved two cultural levels of habitation. One was that of ‘LV – SC culture’, which consisted of parts of the ‘trapezoidal structures’ such as its stone walls, rectangular hearths and stone slab thresholds. The other was that of ‘Starčevo culture’, which was represented by the ceramic pottery.

The fact remains that B. Jovanović considered Stubica as part of the ‘LV – SC culture’ manifestation. That has never been an issue, but rather what type of culture he considered it to be. Thus the object of this paper is to highlight that at Stubica two cultural levels of different distinctive type were recorded. Perhaps this paper will encourage the publication of the presumed field documentation of the site and of the entire material[49] retrieved from it.

If all we have has already been published, then this paper will stand alone as a demonstration that the remains of the stone structure which were discovered at Stubica belong to a ‘Mesolithic community’.


I am grateful for their contributions on this paper to Lorena Vog (providing a translation of B.  Jovanović’s 1984 paper), Cristina Kiru (reader), Raluca Pușcău (text correction), Ivana Radovanović (critical comments and suggestions), Lorraine Houseago (text correction), Paolo Biagi (active reviewer), Dragoș Diaconescu and Florin Drașovean (editorial imput) and Adina Boroneanț (editorial reviewer).

Footnotes and citations

  1. Borislav Jovanović, the man, has recently deceased; Borislav Jovanović, the archaeologist, will continue to influence future generations of researchers.
  2. They will be mentioned throughout the present paper.
  3. It is this author’s intention to prompt fellow researchers from former Yugoslavia that can reach/find the documentation and material excavated at Stubica, to do so and publish it.
  4. For the use of the terminology ‘Lepenski Vir – Schela Cladovei (LV – SC) culture’ and how the present author understands it see Rusu 2011, 2016.
  5. See all the papers and authors mentioned in Rusu 2010 and complete that with Borić 2011 and Boroneanț 2012.
  6. Stânca Bivolului (Buffalo’s Rock) is a rock (class T – Hypsographic) in Județul Mehedinți (county), Romania (Europe) with the region font code of Eastern Europe. It is located at an elevation of 166 meters above sea level. Stânca Bivolului is also known in Serbian as Bivolja Stena. Its coordinates are 44°34’60” N and 22°1’60” E in DMS (Degrees Minutes Seconds) or 44.5833 and 22.0333 (in decimal degrees). Its UTM position is EQ83 and its Joint Operation Graphics reference is NL34–12 (http://www.getamap.net/maps/romania/mehedinti/_bivol_stinca/); at 8:37h on 09.06.2016.
  7. In respect to B. Jovanović’s own approach of publication of the site, I will not include a map within this paper. However, maps with the location of the site have been published in Jovanović 1971, Radovanović 1996, Bonsall et al. 1997, Borić 1999 and in most papers on the Iron Gates archaeology that followed those already mentioned (for an extensive bibliography see Rusu 2010).
  8. Padina – the site; B – part of the two periodization phases of the site; where B is divided in B–1, B–2, B–3; III – one of the four sectors of the site as they were proposed and annotated by B. Jovanović.
  9. The material was retrieved more than ten years prior, and the implication stands as another example of how hard was/is the process of dealing with the discoveries made within the whole Iron Gates Dams I and II Project.
  10. Except for the present author, for whom this site proved challenging. At first it seemed, following the description of the habitation structures, to represent a type site for ‘LV – SC culture’. Later on I retracted (see the note 5 from p. 7 of my 2011 publication), although the excavator himself attributed it to the ‘Lepenski Vir culture’. That happened because of the terminology employed by B. Jovanović who understood it as a local type of Early Neolithic culture, while through my understanding it is a culture of Mesolithic tradition. After I had done a more detailed analysis upon the Mesolithic and Neolithic of the Iron Gates region, I found it necessary to address the information given for the Stubica site
  11. The pottery might be of different periods and that is not the concern of the present author. As it clearly indicates a specific behavior the author is interested only in its connection with the ‘trapezoidal structures’
  12. Aside for the reasonings expressed already, no other explanation for using this terminology was given by B. Jovanović in any of his works known to the present author. Most likely it was used as such as a communication device with the Romanian colleagues, as Criș was the translated form of the Hungarian Körös which stands for the same archaeological culture (Lazarovici 1979, 15) i.e. the Serbian Starčevo.
  13. Padina was first excavated in 1968, then in 1969, and in two more campaigns in 1970, (Jovanović 1968, 1969, 1970); “out of which the ones from 1970 overpassed the other two in extent and results, which were dedicated mainly to preparatory works” (Jovanović 1971, 23).
  14. Actually they were not even mentioned as such, they were not mentioned at all. This implies they ‘were set directly on the ground’.
  15. The arguments are based on the knowledge that similar ‘hard-packed’ floors were recorded at Padina (Jovanović 2008), Vlasac (Srejović and Letica 1978), and Kula (Sladić 1984; 1986).
  16. From my perspective it was made so for technological and thermic reasons, thus creating a stable layer on which the floor was to be made, and also creating the isolation it needed.
  17. Approx. 20  cm thick nearby the hearth, decreasing towards the edges of the floor to a thickness of c.5 cm (Srejović 1972, 54).
  18. Even though Al. Păunescu claimed he was quoting V.  Boroneanț – the principal investigator of Schela Cladovei, the information offered could not be verified with the one published by V. Boroneanț himself. Thereby, for the time being, with caution Al. Păunescu is used as the only source of reference
  19. By colluvium, alluvium and aeolian processes.
  20. See the same quoted author’s work.
  21. I see this as a possibility considering the relationship between the stone walls, hearths and thresholds, and the floors on the other sites mentioned in this paper.
  22. See Garašanin and Radovanović 2001 and Jovanović 2008 on the specifics of the pottery
  23. For the present author ‘Mesolithic’ and ‘Neolithic’ are seen as archaeological constructs that incorporate each on their own a certain type of human behaviour. While others will focus on the time frame of each of these archaeological devices, I focus on their material evidence recovered through archaeological excavations, and all that follows after such an endeavour i.e. archaeological interpretation, debate and reinterpretation of the findings. And that is why I opted for ‘LV-SC culture’ (where I keep ‘culture’ just for its archaeological historicity) as this ‘construct’ is the best device when dealing with ‘Mesolithic’ and ‘Neolithic’ of the ‘Iron Gates’ region.
  24. Even though that was the intention, the paper of M. Garašanin and I. Radovanović (2001) does not make a strong case in proving that the pottery found in Lepenski Vir I structures was not intrusive from the overlying Neolithic levels (Lepenski Vir III), which according to the aforementioned authors did in fact contain the same type of pottery
  25. And all Neolithic for that matter
  26. See E. Banffy’s analysis on house structures of the Mesolithic and Early Neolithic in South-East and Central Europe, with references made also to the Near East (Banffy 2004, 49–71).
  27. One of the most ardent followers of B. Jovanović’s line of thoughts regarding the Iron Gates archaeological manifestation of Mesolithic and Early Neolithic periods (see both authors’ works on this subject).
  28. Although if only their own research is to be considered, it also gives way to a series of loose ends that cannot be tied up to sustain their interpretation
  29. The most potent item used by some researchers that want to understand some cultural manifestations as Neolithic.
  30. Those defining traits are: habitations’ architecture, burial rituals, subsistence strategies – tools and related food resources and artistic manifestations.
  31. That assemblage was so often and ‘potently’ used to prove the presence of Early Neolithic i.e. ‘Starčevo culture’ type pottery in context with the ‘trapezoidal structures’ (Jovanović 1971, Borić 1999).
  32. This is the omission made by B. Jovanović towards Stubica that I mentioned in the introduction, and this is somehow baffling considering that he was the one who considered Stubica I–II, Padina B I–II and Lepenski Vir I–II as part of the same cultural phase on the ground of the structures that he discovered at Stubica
  33. That had specific habitations, hearths and ovens.
  34. The basis was the presence of Early Neolithic pottery in association with trapezoidal structures found at Padina and Lepenski Vir
  35. One should carefully read (see Jovanović 2008, 289–290) the presentation of the geographical/stratigraphical formation of Padina site, to understand the difficulty to which the researcher was exposed in excavating and interpreting the discoveries. A similar situation existed for most of the sites from that region (for an example see Borić et al. 2014 7–9 discussion of Vlasac).
  36. Lithic: quartz – which is predominant with a stable percentage throughout the sites of ‘LV-SC’; flint – present (also as source) more on the Upper Gorges; and silicate rocks present (also as source) more on the Lower Gorges; Bone – used for spears, arrowheads, hooks and also probably as flutes; Antler – a dominant item for ‘LV-SC’; Boar tusks – apparently specific to ‘LV-SC’ (Srejović 1969; 1972; Boroneanț 1980; Srejović and Babović 1981, 1983; Boroneanț 2000).
  37. The use of bipolar technique for chipped stone tools; Microlithic industries; Body position of the dead for burials, where the dominant one is extended supine and with a specific one for ‘LV-SC’ which is the ‘sitting position’ – the dead were placed as if they were sitting in à la turque position; Food resources and dietary consumption, with a strong reliance on the aquatic resources of the Danube river and also on terrestrial ones, specifically red deer (in correlation with antler) and with a special one represented by dog consumption (Srejović and Babović 1981, 1983; Radovanović 1996, 1999; Boroneanț 2000; Bonsall 2008; Radovanović 2006).
  38. Present in even lesser quantity than the pottery (see Antonović 2006, 128–129; Mihailović 2004, 62–67).
  39. Archaeologically, we have been dealing with damaged past human records
  40. See D. Borić discussion of the trapezoidal structures at Vlasac dated cca.7000 cal BC, versus the earliest structures at Lepenski Vir that were dated cca.6300/6200 cal BC (Borić 2011, 170), and consider that only around 6000 cal BC complete Early Neolithic life style traits were recorded for Iron Gates region (Bonsall 2008, 267-and next pages).
  41. Bonsall et al. 2002, suggested that the appearance of limestone plastered floors was a result of the ‘8200 BP event’ (c.  200 cal BC). However, considering that the first plastered floors appeared millennia prior, then we have to understand that this type of floor was created as a reaction to the specific local conditions of the Iron Gates region.
  42. Relying on the first chronological data related to these structures, the respective authors made a case in trying to prove that since Early Neolithic was close to Iron Gates by 6200 cal BC, it meant that the technique was brought from the Near East. Afterwards the discourse focused on the possibility that perhaps the Vlasac structures, dated c.7000 cal. BC. might had been influenced from the same region even earlier (Borić 2002; 2005; 2008; 2011; Borić and Dimitrijević 2007).
  43. His research contributed greatly in providing necessary data for understanding the archaeological phenomena of Iron Gates area. See all his works on this subject
  44. Discussions on the debates over these intriguing discoveries can be read in Radovanović 1996, Rusu 2010, Boroneanț 2012
  45. I am aware of the colloquial term, but since it is a human error, using academic terminology would not make up for it, and as such it is the most accurate word to describe it.
  46. To build up an understanding start with the synthesis works of Radovanović 1996, Boroneanț 2000, Bonsall 2008 and Borić 2011, and afterwards follow each way they may lead you.
  47. Perception that is, for the most part, a mixture of ‘the first impression’ over the discoveries with past experience of one researcher. Or some sort of personal impression that one researcher made to another that could be called ‘teaching’ and ‘learning’, regarding the same discoveries
  48. A ‘culture’ in this case part of a wider framework – Mesolithic – that finds its last stages of evolution contemporaneous with the Neolithic beginnings in Europe.
  49. It can be expected that not only Neolithic pottery was discovered, and that there could very well be tools from stone, antler, bones and tusks. Considering the discussions of this paper, one should be sceptical in considering that only Neolithic pottery was recovered or could have been recovered as portable material.