Middle Palaeolithic lithic tools: Techno-functional and use-wear analysis of target objects from SU 13 at the Oscurusciuto rock shelter, Southern Italy

: The Oscurusciuto rock shelter (Ginosa, Puglia, southern Italy) is a Middle Palaeolithic site characterized by a significant stratigraphy made up by several anthropic levels. The stratigraphic unit 13, consisting of a sandy compact deposit mixed with pyroclastic sediment, is a short palimpsest situated on a layer of tephra, identified as Mt. Epomeo green tuff (dated Ar/Ar ~ 55 ka BP). From a technological point of view, the aims of the production were backed flakes, convergent flakes, and other flakes obtained by means of a Levallois debitage, plus (less represented) bladelets produced by an additional volumetric reduction system. Our aim in this research was to examine a selection of the above-mentioned target objects produced by debitage in order to understand the manufacture and life-cycle of each single tool from a dynamic perspective. We integrated techno-functional and use-wear analyses: the first was implemented to globally comprehend each tool, identifying each single techno-functional unity (prehensile and transformative portions), whereas the second revealed the way in which these tools had been used, proceeding to identify the activity involved ( e.g. , piercing, cutting and/or scraping), and the type of material (vegetable/animal, soft/hard) on which these activities had been carried out. The combined use of these two approaches allows us to ascertain the intention of the prehistoric craftsmen, the gestures and procedures involved in making the tools, and the way they had been used. From one single object we are thus able to reconstruct a series of complex behaviours, encompassing the creation, the life-cycle and finally the ‘death’ or repurposing of the tool in question.


Introduction
The Oscurusciuto site is a Middle Palaeolithic rock shelter located in Southern Italy with a very rich record, essential for the definition of Neanderthals technical behaviour, as related both to the management of the raw material on the territory and the crafting and use of lithic tools. The lithic collections of Oscurusciuto show substantial uniformity even though each level displays its own peculiarity. The acquisition of raw material is local; Neanderthals used pebbles of jasper, chert, cherty limestone and quartz sandstone available in the terraces and areas near the site. The recurrent Levallois method is the most commonly noted concept of debitage especially in its unipolar modality aiming at producing elongated supports, both convergent or not. Furthermore, there is an additional volumetric exploitation aiming at producing bladelets (Boscato et al., 2011;Marciani, 2013;Marciani et al., 2016;Ranaldo, 2005;Ronchitelli et al., 2011;Spagnolo, 2013;Spagnolo et al., 2016;Villa et al., 2009).
As it is widely known, the Levallois reduction sequence is an integrated concept that allows for the production of a great quantity of predetermined products with specific characteristics and dimensions (Boëda, 1991;1994;1995;Schlanger, 1996;Van Peer, 1992). The degree of predetermination of this concept regards both obtaining of a certain quantity of target products (resulting from the lineal\preferential or recurrent Levallois debitage) and their quality (Boëda, 1991;1994;1995;. It also testifies to a specific economic strategy with regard to the maximization of cutting edge productivity obtained from cores (Brantingham & Kuhn, 2001;Lycett & Eren, 2013).
The Levallois production at the Oscurusciuto stratigraphic unit 13 was aimed at producing a clear set of debitage goals: flakes, convergent flakes and backed flakes. In previous studies we have noted the peculiar and recurrent features of these objects and their particular role in the economy of the level (Marciani, 2013;Marciani et al., 2016). We assumed that these target objects had been imported as finished objects into the site, or had resulted from an in situ debitage process (Marciani, 2013;Marciani et al., 2016;. As the target objects are the answer to specific needs and necessities which motivated the flaking activity itself (Boëda, 2013), we examined the role played by Levallois target objects in the society where they were produced. In particular, our main objective for this work was to focus on the Levallois target flakes produced at Oscurusciuto SU 13. We wanted to test (1) if the target objects of a reduction sequence (the production-aim), were actually used by Neanderthals to perform their activities (the functional-aim); (2) if each production-aim corresponded to a single functional-aim i.e. each tool was used solely for one activity, or, on the contrary, if it was used for a multitude of purposes; and finally (3) if tools used for a specific activity also had a specific structure.
To accomplish these goals we integrated techno-functional and use-wear analyses: the first was implemented to globally comprehend each tool, identifying each single technofunctional unity (prehensile and transformative portions), whereas the second revealed the way in which these tools had been used, proceeding to identify the activity involved (e.g., piercing, cutting and/or scraping), and the type of material (vegetable/animal, soft/hard) on which these activities were carried out.
From a methodological point of view, we note that both techno-functional and use-wear analyses have been applied on several lithic assemblages coming from diverse archaeological contexts and periods (e.g., Boëda, 1997;2001;Soriano, 2000;Da Costa, 2017;Lourdeau, 2010), however few works have attempted to combine them Boëda et al., 2015;Bonilauri, 2010;Pedergnana, 2017). In the case of the site of Ficoncella (Rome -Italy; dated back to 500,000 years BP), the combined use of techno-functional and use-wear approaches was essential to determine the technical structure and the peculiarities of some active unities (trihedral, mini-rostrum and brute cutting edge) barely described before in lower Palaeolithic literature . Another outstanding example is the case of Umm el Tlel site (central Syria -dated back to 40,000 years BP). Here the authors were able to evidence the technical role of the bladelets. These very well-known tools actually resulted from different production systems, performed several tasks, and were hafted in different ways (Boëda et al., 2015).
The challenging point about applying these two approaches together is that there is an added value by their combined use. In fact, the technological analysis permits us to understand the production procedure of lithic tools, and the techno-functional analysis allows us to understand the structure of the tools, as well as the functional potential of the parts constituting the tools. Finally, the use-wear analysis verifies or rejects these hypotheses. Tracceologists can also use the techno-functional study as a proxy for the selection of their samples, i.e. defining significant criteria to select the pieces to analyse, and examining traces guided by technical parameters. It is worth remembering that these approaches work in a continuous dialogue, comparing and integrating the observations from each approach.
As a matter of fact our contribution with this paper is to demonstrate the complementarity and added value of combination of these approaches, which produce new and stimulating insights into "the ways of existing" of prehistoric technical objects (Du mode d'existence des objets techniques Simondon, 1958)

The site
The Oscurusciuto rock shelter is situated in the region of Puglia, Southern Italy, namely in the ravine of Ginosa (Taranto) (Figure 1). The site, opened into the Pleistocene calcarenite (Calcareniti di Gravina), stands at an elevation of 235 m above sea level, at about 15 meters from the current bottom of the ravine, and at about 20 km far from the actual Ionic coastline (Figure 2).
The archaeological deposit measures 60 m 2 at its base, and has a thickness of more than 6 meters. This sequence extension downwards gradually increases because the hill erosion damaged the deposit in the shelter, especially on the upper layers ( Figure 3). Since 1998 until the present day, the first 3 meters of the sequence corresponding to nine main Middle Palaeolithic occupation phases have been investigated by the U.R. "Preistoria e Antropologia" under the "Dipartimento di Scienze Fisiche, della Terra e dell'Ambiente -University of Siena, Italy".
The chronological limits of the Neanderthal occupation of the rock shelter are obtained by two dates. The first, obtained by the C14 method (on collagen), is referred to the bottom of  (Ramsey & Lee, 2013;Reimer et al., 2013). The former derives from the identification of the tephra layer (SU 14) as Mount Epomeo green tuff Spagnolo et al., 2016) datable to about 55,000 BP (Allen et al., 2000). The faunal associations found at Oscurusciuto are characterized by the main presence of Bos primigenius, Equus ferus, Cervus elaphus and Dama dama in different variations and associations, which show both little paleo-climatic fluctuations and the coexistence of different biomes, possibly related to a micro-scale landscape variability linked to the ravine environment. Essentially, the studied samples (from SU 15 to SU 1) show that the Neanderthal hunters exploited two different environments: a forest-steppe area, probably present on the hilly relieves, and a moister territory with wooden coverage, inside the ravine (Boscato & Crezzini, 2012;Boscato, 2017).
SU 13 (Figure 4) consists of the sedimentary interface between the SU 11, sandy layer above, and the SU 14, the proper tephra deposit. It represents the first stable re-colonization of the rock shelter which occurred during the final phase of volcanic ashes deposition (Marciani Journal Figure 3). It is a short palimpsest excavated for the extension of 11 square meters. The level is characterized by the presence of quite abundant faunal remains; however, only six elements were identified, due to high fracture grade (1 = Equus ferus and 5 = Bos primigenius). Above SU 13 a series of structured hearts made in prepared dimple was exposed which seems to be arranged in a line that divides the space of the rock shelter into two portions: inside and outside the line of fires .

Lithic production
In SU 13, 7504 well preserved lithic artefacts were found. As in all the upper stratigraphic units, the dominant raw materials are jasper and cherty limestone in their fine granulometry, found in the form of pebbles, which can still be found on the sea-terraces and river deposits around the site (nowadays almost between few tens of meters and over than 20 km far from the site) (Marciani, 2013;Marciani et al., 2016).
The occupation of SU 13 is a short palimpsest that was possible to disentangle in at least two (if not most) settlement events Spagnolo, 2017). The lithic collection corroborates this idea because there are several fragmented reduction sequences. Namely, the lithic material was entered into the site at different stages of debitage, which means in the form of rough objects (pebbles), or as semi-finished items (decortication happened outside the rock shelter), and as finished tools (target objects or retouched tools). However, there is evidence of pieces exported from the site, especially the target objects. Only 30 pieces have been retouched .
Only two concepts of debitage were utilized at the level 13: the Levallois concept and the additional volumetric reduction sequence. Namely, the latter refers to an additional production aimed at producing bladelets where the striking platform is prepared but the convexities are not prepared and only the natural convexities of the raw block are used, i.e. C2 type of cores according to Boëda classification (Boëda, 2013). (2018)  In this paper, we will focus only on the Levallois products: flakes, convergent flakes and backed flakes. From a totality of 385 target pieces (Marciani, 2013;Marciani et al., 2016) we selected 312 target objects with clearly recognizable technical features (fragmented or postdepositional altered pieces were excluded from the study). In the text we refer to the objective of debitage\end-products with the expression "target objects" (target flakes, target blade, etc.) because we think that even if not very common in bibliography it expresses very well the strong degree of intentionality used in order to produce them.

Techno-functional analysis
The techno-functional analysis is the study of the structure and potentiality of tools which are defined as objects consisting of three main parts: transformative part, transmitting part and prehensile part. These parts are defined as techno-functional unities: the UTFttransformative techno-functional unit corresponds to the active portion of the tool plus the edge, which is the part which actually enters into contact and modifies the material; the UTFp -prehensile techno-functional unit is the holding portion. Between those there is the transmitting techno-functional units (UTFtr) which is conceived as an intermediate factor transmitting the force from the handle to the transformative portion (Boëda, 1997;2001;Bonilauri, 2010;Da Costa, 2017;Lepot, 1993;Lourdeau, 2010;2015;Soriano, 2000).
The techno-functional unities are indispensable for the tool operation, as is the synergy between them that makes the tool capable of realizing an action. Identifying these parts and understanding their synergy enables us to understand the potential capacity of each single tool. Furthermore, in addition to prehensile, active and transmitting parts, we should not forget the role of the gesture. Namely, the specific gesture involved in moving a tool is indispensable for the proper functioning of the tool, whereby we mean the kind of action, as well as the actual movement, both of which made the tool technically worthwhile (Leroi Gourhan, 1973). The identification, location and characterization of UTFt and UTFp lay in empirical observation of the objects and collection of objective data. Whereas the identification of the transmitting part is much more difficult to perceive, for this reason in this work we consider together the prehensile and transmitting parts (Boëda, 2001;Bonilauri, 2010;Da Costa, 2017;Lourdeau, 2010;Soriano, 2000).
After analysing each single piece, comprising the identification, characterization and localization of each UTF, we have identified groups of pieces that share the same structural composition, i.e. techno-type. Going more in detail, a techno-type is defined by pieces with the same number and characteristic UTFt; and the same position and number of UTFp. The sub-types are inner variations of techno-types in our case expressed by the delineation and position of cutting edges. After defining the techno-types and sub-types, i.e. the combination of UTFt and UTFp, we have identified how these combinations were arranged on each blank.

Defining the blank
The blank is actually the support on which the UTFt and UTFp are installed. At Oscurusciuto SU 13 we defined blanks on the basis of four parameters: shape, section, elongation index and size of the pieces.
The shape is rectangular, when there are 2 parallel\sub parallel edges plus a transversal edge (these edges could be sharp of backed; in this group trapezoidal and oval shapes are also considered). Alternatively, it is described as triangular when there are 2 edges converging into a point ( Figure 5). The section of the piece takes into consideration the section-shape of the items but also the number of cutting edges. This means that triangular and trapezoidal pieces have 2 cutting edges, whereas rectangular-triangle and rectangular-trapeze have only one cutting edge opposed to a backed side ( Figure 5). The elongation index is given by the ratio between height and width. If the result is less than 1.4 it is a flake, if the result is comprised between 1.5 and 2 it is a long flake, if the result is major then 2.1 it is a blade ( Figure 5). The size of the pieces is based on the graphic of dispersion of the height of these pieces of Oscurusciuto SU 13. We define a piece as "small" if its maximum height is until 20 mm, "medium" if it is comprised between 21 and 34 mm and "big" if its height is higher than 35 mm ( Figure 5).
The trihedron is defined as a geometric figure composed of three planes meeting at a single vertex. As UTFt the trihedron is a punctual UTF, which means that it does not have an extension. In order to define the trihedron we consider its location, surface relations, angles, that is the angle made by the surfaces (section plan between ventral and dorsal surface of the item) and the openness angles which is the angle made by the two lateral sides (left and right side of the items) Rocca, 2013) (Figure 6).  If there is more than one UTFt on the same blank we register which kind of relation exists between them: contiguous if the two (or more UTF) are in an adjacent position and in contact between them, and opposed if the two (or more UTF) are in front of each other. All the parameters are considered with the flake oriented in the technological axis. (2018)

Defining prehensile parts (UTFp)
In the absence of evident grip-traces or glue residues, so far, we consider as potential UTFp all the backed sides, the cortical edges and more in general a discontinuous and irregular edges (Soriano, 2000). The butt is also considered as potential grip, in fact especially for Levallois chapeau de gendarme, the ergonomic features of its shape are already attested in experimental studies (Baena Preysler et al., 2016). We consider the thickness, location, angles and extension of UTFp. A notable parameter when defining the structure of a tool is the relation between backed sides i.e. a single backed side, two continuous backed sides, two opposed backed sides, or three framed backed sides (Da Costa, 2017).
LPA is focused on analysis and interpretation of macro use-wear (edge-removals, edgerounding) while HPA is based on the observation of micro use-wear (micro edge-rounding, polishes, striations).
Macro use-wear was observed at low magnification (20x -80x) by means of a Hirox KH 7700 3D digital microscope, using a MX-G 5040Z body equipped with an AD-5040Lows and an AD-5040HS lens. Micro-wear analysis was performed using the mentioned Hirox microscope fitted out with a MXG-10C body and an OL-140II lens (140x-480x). The microscope enables us to obtain in-focused pictures through the overlapping of planes taken at different focus levels (Arrighi et al., 2016;Moretti et al., 2015).
The traces on the archaeological lithic tools were interpreted by means of the comparison with the experimental reference collection of the U.R. Preistoria e Antropologia -University of Siena, Italy.

Blank
The majority of blanks (almost 50% of the complex) are of medium size (Table 1). Based on the elongation index, the main objects pursued were flakes ( Table 2) and most of them could be considered as rectangular (with at least two cutting edges) (Table 3); 34 pieces have a backed side (Table 4).   Combining these technical traits, we obtained the blanks on which the active and prehensile parts of the instrument are imposed (Figure 7). The majority of Oscurusciuto SU 13 target objectives of Levallois debitage are rectangular flakes of medium and small size, followed by rectangular big and medium long flakes and big blades. We note that elongated products are mostly big and medium sized, whereas flakes are mostly small. Convergent flakes are scarcely represented, no matter what the size. Except for convergent medium flakes (19 pieces), other sizes do not account for more than ten unities (Figure 7).
According to these characteristics, we can sub-divide the macro-class of blanks into: type-blank A, B or C. Type-blank A is made by rectangular blanks with one cutting edge opposed to a backed side. This group contains small, medium and big rectangular flakes, long flakes and blades with rectangular-trapeze and triangular-trapeze sections, which means one cutting edge opposed to a back (34 pieces). Type-blank B is made by the rectangular blanks with at least two cutting edges. This group comprehends small, medium and big rectangular flakes, long flakes and blades with trapezoidal and triangular sections, which means pieces with at least 2 cutting edges (233 pieces). Type-blank C is made by convergent blanks with at least two cutting edges. This group comprehends small, medium and big convergent flakes, long flakes and blades with trapezoidal and triangular sections, which means pieces with at least 2 cutting edges (42 pieces).
Having defined these three classes of blanks, in this paper, we decided to focus only of the type-blank A in order to give extensive attention to each category (further work will focus on the other two blank type, B and C) (Figure 7). (2018)

Type-blank A
Type-blank A encompassed 34 pieces, most of them are elongated supports (blades or long flakes) (Table 5). On these supports several UTFt are imposed: 21 UTFt trihedrons (from now UTFt T or just T) and 39 UTFt cutting edges (from now UTFt C or just C) ( Table 6). The most represented UTFt is the cutting edge. It could be found alone, in association with a trihedron, or with other cutting edges ( Table 7). The majority of cutting edges show a rectilinear delineation, less represented are denticulate, convex and concave forms (Table 6). Usually cutting edges are in a lateral position (32) and occupy 3\4 or the totality of the edge of the tool.
The 21 UTFt trihedron are made up by 3 planes encompassing the dorsal and ventral faces of the flakes plus the butt, or plus a third face constituted by a rib made up of two negatives on the dorsal surface or by a broken portion of the flakes, which lets us suppose an intentional use of the fracture. Consequently, these UTFt are slightly difficult to identify, as their features, (when the rib or butt is the third face), mostly occur during the production of the pieces and not through intentional action (like in the case of intentional fracture). This issue could have caused an overestimation of this UTFt.
The UTFp are found in several different combinations: the most relevant one is the presence of 2 backed sides (most of them proximal plus lateral) and the 3 backed sides that actually forms a frame around the piece (from now on we refer to backed side with D) ( Table  6, Table 7).
Going into more detail, we focused on the combination among the active parts (UTFt C and UTFt T), plus their association with the prehensile parts (UTFp). In this way we were able to individuate classes of tools that have the same structure (techno-type). For the nomenclature, we describe each tool based on the combination of UTFt and UTFp (UTFt cutting edge = C, UTFt Trihedron = T; UTFp backed side = D), the repetition of each letter represents the frequency of the unity. The techno-type is indicated with a letter and the subtype is indicated by a number. It this way each piece is described by a code i.e. CDD-A1 means a piece with one UTFt cutting edge: C + two backed sides: DD, belonging to technotype A, subtype 1. Table 7. Techno-type and subtype defined by the combination of UTFt and UTFp. The column "plus UTF trihedron" indicates how many trihedrons are present on the pieces (please note that this column refers the number of UTFt Trihedrons and not to the number of pieces). We recognized 5 techno-types, presenting a variety of sub-types ( Figure 8, Table 7). To sum up the most common techno-type was A: 1 cutting edge + 2 backed sides (CDD), present in sub-types 1, 2, 3, 4 and 5. Sub-types 1, 2, 4, and 5 have the cutting edge opposed to the lateral backed side, whereas sub-type 3 is the only one that has just one cutting edge in a transversal position (Figure 8, Table 7).

Techno -Type and Sub-type N. of Pieces Plus UTFt trihedron
Highly represented is also techno-type D: 1 cutting edge + 3 backed sides (CDDD), whose peculiarity the 3 backed sides constituting a frame for the cutting edge. Not surprisingly this is where we find the major number of trihedrons as the backed side is actually the third surface that enhanced the creation of the trihedron itself. The pieces of techno-type B: 1 cutting edge + 1 backed side (CD), are peculiar because in both cases the UTFp is made through retouch. Finally, we have only one case of techno-type E: 1 trihedron + 3 backed sides (TDDD). This piece is very interesting as it is the smallest tool in the collection; here the intention of creating 3 backed sides as prehensile portion leaving free only the distal trihedron become clear. This is also the only case were the trihedron is found alone and not connected with other UTFt (Figure 8, Table 7).
These above-mentioned techno-types and sub-types could be found imposed on various blanks. Starting from the blades, these pieces are found in big and medium sizes; generally, we noted a recurrence of 2 baked sides and an opposite cutting edge. Incidentally, each tool has its own peculiarities; they are supports for different sub-types (Table 8, Figures 8, 9, and 10). The case of the pieces ID 919 and ID 989 (techno-type B) is interesting because the retouch is implemented to improve the adherence potential of the grip, and it is adjacent to the cutting edge opposed to the backed side. This configuration of the tool introduces the hypothesis of a hafting (Figure 9).  Table 8. Synthetic table of techno-types and sub-types (C: cutting edge, T: trihedron: D: backed side; n. refers to the total number of pieces of each type), how they are disposed on the blank, plus information regarding the use: action, worked material and angles of the active edge (for the blank: BB: big blade, MB: medium blade, BL: big long flake, ML: medium long flake, SL: small long flake, MF: medium flake, SF: small flake).

Techno-type
Subtype N.  Considering the medium blades, the pieces ID 852 and ID 853 refits, in this case it is interesting that the same configuration of the tool is created in the same core. Namely, pieces that technologically play the role of predetermining-predetermined pieces from a technofunctional point of view present identical the same prehensile and active portions (sub-type C2) ( Table 8, Figures 8 and 10).

Blank
As for the long flakes, we encountered specimens of every sizes: big, medium, and small in varying numbers (Figures 11 and 12). This is the blank with the most variations in technotypes, in fact of 9 pieces we noted 8 different sub-types (Table 8, Figures 8, 11, and 12). Flakes are represented only in small and medium sizes (Figures 13 and 14). The main defining feature in these pieces is the presence of 3 framed backed sides (techno-type D) ( Table 8, Figures 8 and 13). Probably because of their small dimensions, these small pieces, show a particular prehensile need that could be met by the 3 backed sides, artifice that permit the creation of trihedrons.

Use-wear analysis
Of the examined sample (34), 21 tools show evidence of use, 9 pieces display unclear or uncertain traces because of post-depositional modifications, 4 artefacts do not reveal use-wear at all. We noted that in the majority of the cases use-wear traces were found on the portion of the pieces recognized as active by the techno-functional analysis. Only in one case traces were found on another edge which was not recognized by techno-functional analysis (ID 991, Figure 13). In other cases, there was not enough microscopic evidence to prove the technofunctional reading.
Going more into detail, 22 UTFt were used, in particular we noted 20 cutting edges (UTFt C) and 2 trihedrons (UTFt T). In addition, traces were visible on 2 prehensile portions (UTFp). Each artefact showed evident traces on only one UTFt, with the exception of the piece ID 1384 where its 3 UTFt showed traces ( Figure 15).
Due to the fact that micro-wear traces (polishes) are by nature not very evident, for several pieces it was inferred only the action carried out and general information about the hardness of the worked material. Nevertheless, as a general functional scene, we may deduce that techno-type analysed was regularly used for various tasks (Tables 8 and 9). The tools were used for processing hard (9), semi-hard (8) and soft materials (5). Such a variety is confirmed also when worked materials have been detected, as both vegetal (4) and animal tissues (4) were processed (Table 9, Figure 16).   (2018)  UTFt C are used mainly for longitudinal actions and to a lesser extent in transversal movements. The transversal actions are carried out by cutting edges with wider angles, in particular when hard materials were processed. Comparing worked materials and motions, we may conclude that the transversal actions were mostly performed on hard material, while the longitudinal ones on all types of material (Tables 8 and 9, Figures 9 to 14 and 16).

Journal of Lithic Studies
Few items are employed in mixed actions. A single UTFt C, of the flake ID 1012, was used with both longitudinal and transversal movements for processing vegetal material ( Figure 13, Tables 8 and 9). Whereas two different UTFt C of the big blade ID 1384 were both used for longitudinal actions and UTFt T was employed in a rotational motion. In both cases animal tissue was processed ( Figure 15, Tables 8 and 9).
By means of a techno-functional analysis, 21 trihedrons were identified, but traces were found on only two UTFt T of which one was involved in a rotational motion (pieces ID 1384 - Figure 15), and the other in a longitudinal action (pieces ID 882 - Figure 17) (Tables 8 and 9). These tools were used for processing hard and semi-hard material.
Evidence of probable hafting traces on UTFp is detected on two pieces (ID 1020 and ID 919), confirming the hypothesis of the techno-functional analysis. In both cases, the traces are located on the backed sides of the tools (Figures 9 and 10). In both cases, the traces are few bright spots located on the backed sides of the tools.

Combined use of techno-functional and use-wear analysis
The result obtained by the combined use of the techno-functional and use-wear analysis gives a consistent result. In fact, in the sampled pieces, we note that in the majority of the pieces the use-wear analysis confirmed the interpretation given in the techno-functional study.
We note that sometimes the same techno-type are installed on different blanks, as in the case of sub-type A2 and D2, a recurrent combination in almost all blanks. Then again, other techno-types seem to be specific to some blanks, such as techno-type B. Moreover, some blanks are the support for several techno-types, e.g., the big blades present quite a number of different combinations, and they seem to be made to purpose unique tools. In contrast, other blanks support few techno-types, as is the case with medium flakes where most of them support the techno-type D (Table 8). A very remarkable fact is that the techno-types are consistent with their actual use, in fact we noted that some techno-types were used for specific activities, as is the case with technotype A, which is mostly used for cutting (longitudinal movement). On the other hand, there are techno-types, such as C and D, which are employed in a variety of activities, such as cutting and scraping (respectively longitudinal and transversal movements). In addition, some techno-types met a particular need, as is the case with techno-type E, which can be considered unicum in the collection both from a structural and a functional point of view, in fact this piece (a UTFt T framed by 3 backed sides) was used for carving ( Figure 17). Finally, in techno-type B a very specific construction of the piece (also produced by retouch) does not correspond the same activity, one is used for scraping and the other for cutting, however both pieces are related to butchering activities (Table 8).

Conclusions
A first methodological observation was that the integrated use of techno-functional analysis and use-wear analysis yields consistently matching results to the benefit of both technologists and tracceologists. Thanks to this combined method we get much closer to the human realities where there were several needs to solve, and several tools to be manufactured.
This work answered the three main queries posed at the beginning of this research. 1. We proved that the target objects (production-aims) were also actually used as functional objects (functional-aims), i.e. a large number of items shows use-wear. 2. We noted that each production-aim actually comprised several techno-types employed for different purposes. Namely, through the technological approach we learned that the production-aims was flakes, convergent flakes and backed flakes. Thanks to this study, we perceive great diversity in both structure and functionality of these tools, noting that to each class of target objects corresponds a set of different functional objects, as each category actually comprises several blanks, and each blank could be the support for several techno-types. 3. Finally, we also proved that tools with a specific structure aimed to solve a specific task. (2018)  The encouraging results of this study motivated us to continue in this direction, analysing the other two categories (type-blanks B and C) in order to obtain a larger set of statistically significant data.

Journal of Lithic Studies
Lastly, one question remains: the trihedron was identified by techno-functional analysis but presented traces in only 2 cases. For this reason, we want to know why there is such little evidence of usage on these pieces. Is it because the action did not leave strong visible traces, or because this part was not used? We could also hypothesise an artificial increase due to possibly misleading criteria of identification. To be able to answer these questions we are planning to set up an experimental protocol to verify the functional potential or the trihedron, possibly applying more restrictive criteria for its identification, based on the technical criteria present on the two items which showed clear use-wear traces.