Snakes and Ladders: A technological approach to tool maintenance by-products using module flake categories

Snakes and Ladders: A technological approach to tool maintenance by-products using module flake categories | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Snakes and Ladders: A technological approach to tool maintenance by-products using module flake categories David Nora, Ariel Malinsky Buller, Boris Gasparyan, Artur Petrosyan, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6673907/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 24 Nov, 2025 Read the published version in Journal of Archaeological Method and Theory → Version 1 posted 9 You are reading this latest preprint version Abstract The study of retouching, reshaping, and rejuvenation in lithic technology has traditionally focused on finished tools, overlooking the by-products of these processes, particularly microdebitage. This has led to an incomplete understanding of the dynamic behaviours associated with tool maintenance and a lack of crucial information about prehistoric technological strategies. Here we address this knowledge gap. Specifically, we introduce a classification system for lithic by-products resulting from retouching, reshaping, and rejuvenation techniques, categorising them into five modules (M0 through M4) based on lithic technological analysis. This methodology integrates the chaîne opératoire approach to analyse flakes without size thresholds. To demonstrate our approach, we apply it, coupled with raw material sourcing, to lithic assemblages from two Middle Palaeolithic sites in Armenia, Kalavan 2 and Ararat-1 Cave. This enables a precise reconstruction of tool use-life and, in turn, the mobility strategies of Pleistocene hunter-gatherers. Our findings demonstrate that microdebitage (by-products) can contribute to a holistic view of decision-making, revealing patterns in tool maintenance and raw material provisioning. The module system provides insights into the production of ‘ghost tools,’ which are not present in the archaeological record, as well as curation behaviours and economic decisions regarding raw materials that were previously difficult to discern. By shifting the focus from finished artefacts to by-products, this framework enhances our ability to interpret lithic assemblages and understand the adaptive strategies of prehistoric hunter-gatherers. Lithic technology Tool maintenance Curation Microdebitage Module flakes Technological organisation Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 Figure 13 Figure 14 Figure 15 Figure 16 Figure 17 Figure 18 Introduction The study of retouched lithic pieces has played a key role in the systematics of Palaeolithic research since its earliest stages (Monnier, 2006; Trigger, 2006; Wargo, 2009). A well-known example is Bordes’ nomenclature for Lower and Middle Palaeolithic artefacts, which created a common language that remains foundational in the classification of lithic objects into types (Bordes, 1953, 1951, 1961). Bordes’ system focused on morphology—i.e., the form of the object—and the location of retouch as diagnostic components. Similarly, (Brézillon, 1968) defined three aspects of retouch variation: magnitude (marginal or invasive), extent (continuous or denticulated), and location (direct, inverse, alternating, and bifacial). According to Bordes (Bordes, 1951; Bordes & Sonneville‐Bordes, 1970), morphotypes were emic; i.e., configurations imbued with cultural significance. In contrast, (Debénath & Dibble, 1994; Dibble, 1995) interpreted retouched artefacts (mostly scrapers), as defined by Bordes (Bordes, 1969), as reflecting different stages along a continuum of morphological change resulting from rejuvenation or reshaping of tool edges (see also the “finished artefact fallacy” (Davidson et al., 1989)). With the adoption of technological approaches based on the chaîne opératoire in the 1990s, discussions about retouch variation and its meanings took new directions. The roots of Dibble’s perspective lie in the idea that variation in lithic assemblages resulted from adaptive strategies inherent to hunter-gatherers' modes of resource exploitation across the landscape (Binford, 1979a; Jelinek, 1976). In this view, lithic technologies were part of dynamic technological decision-making, the selection and integration of strategies for making, using, transporting, and discarding tools and the materials needed for their manufacture and maintenance (Nelson, 1991, p. 57). This aligns, to some extent, with the chaîne opératoire approach, which reconstructs sequences of actions, including knapping procedures, from raw material acquisition to tool discard. Binford had already pointed to this when he explained that such technological strategies are shaped by provisioning conditions (e.g., artefact production for future use, transport, maintenance, and recycling), all closely tied to hunter-gatherer mobility (Binford, 1973, 1979a). He coined the term ‘curation’ to describe tools that are effective across multiple tasks (Binford, 1973, 1979a; Binford & Binford, 1966, p. 66). This represents a shift in archaeological reasoning—from descriptive typology to processual analysis. At this point, a triad began to emerge: retouched tools, technological strategies, and the resulting artefacts (both products and by-products). One of the key concepts linking these domains to curation is maintenance. Maintenance is intrinsically tied to curation, as the latter refers to strategies of caring for tools and toolkits to maximise their utility over time (Andrefsky, 2009; Cornford, 1986; Frison, 1968; Shott, 1986, 1994, 1996). According to Binford, curated technologies involve tools that are maintained through repeated use (Binford, 1973). Maintenance increases efficiency by extending a tool’s use-life relative to the energy invested in its manufacture. Frison (1968) highlighted the significance of use-life and resharpening, suggesting that maintenance behaviours reflect investment in prolonging tool functionality—a concept aligned with curation. Bamforth (1986) explicitly argued that maintenance is a component of curation, proposing it as a response to raw material shortages. When raw materials are scarce, replacing worn tools is costly, making maintenance a more efficient strategy. Conversely, the frequency of tool maintenance likely decreases when raw materials are readily available. Shott (1996) defined curation as the degree of use or utility extracted from a tool throughout its life. From this perspective, maintenance contributes directly to curation by increasing the total utility obtained before discard. Repair and resharpening allow a tool to be used longer and for more tasks, thus fulfilling more of its initial utility. Maintenance includes activities such as resharpening, repairing, and recycling tools to extend their usefulness (Schiffer, 1976). These activities often involve retouching, reshaping, and rejuvenation, which result in both intentional and unintentional production of by-products—usually microartefacts (see Fig. 1). Such techniques have shaped the morphology of retouched pieces, which have transitioned from rigid typological constructs to dynamic, fluid products of economic decision-making (Geneste, 1985). The size of artefacts resulting from these maintenance actions poses a challenge. As these by-products frequently fall into the category of microdebitage, they are often under-studied or selectively sampled, focusing only on morphotypes deemed representative of specific technological signatures (Bourguignon, 1996; Cornford, 1986; Frick et al., 2024; Prévost et al., 2022). In this paper, we examine the technological interpretation and reconstruction of the operational sequence of maintenance techniques (retouching, reshaping, and rejuvenation) through two case studies from the Middle Palaeolithic of Armenia: Kalavan 2 and Ararat-1 Cave. Specifically, we propose and apply a new categorisation of microdebitage artefacts, classifying them into five modules (M0 through M4). These by-products are viewed as dynamic expressions of decision-making, complementing static tool typologies, and offer a proxy for understanding the technological organisation of Pleistocene hunter-gatherers. Microdebitage between terminology and technology The recognition and identification of microdebitage are part of the broader classification of lithic artefacts. Artefacts of this particular category typically enter the archaeological record as small, shattered objects or micro-flakes produced during operations preceding the reduction sequence (Fig. 1). These fragments relate to the process of flake formation governed by fracture mechanics, with differing morphologies resulting from specific directions of flake initiation and spalling (Cotterell & Kamminga, 1987; Li, Lin, et al., 2023; Li, Reeves, et al., 2023; Speth, 1972, 1975). From this analytical standpoint, we refer to the intentional production of microdebitage as a result of maintenance activities. At various stages of the reduction sequence, one can expect differing proportions of microdebitage—either as small, unintended by-products of fracturing during technological gestures or as intentional products created for shaping, retouching, or rejuvenating a blank or tool. For instance, higher frequencies of intentionally produced microdebitage are expected during use, retouching, and maintenance stages rather than during initial blank production (Fig. 1). Microdebitage appears throughout most stages of the reduction sequence (Fig. 1), whether produced intentionally or not (Debénath & Dibble, 1994). Torcal’s work, for example, references the production of small debris from knapping activity (labelled BP2G; Torcal et al., 1991:178). Lithic classification systems are typically based on metric criteria that distinguish macrodebitage from microdebitage by-products (Bar-Yosef, 1981; Odell, 1996, 2000, 2001), except in the case of bladelet production. To explore this further, we must first determine what threshold defines microdebitage. The size threshold has been inconsistently defined in the literature, leading to multiple terms such as microartefacts, microdebris, or chips (Clark, 1986; de la Torre et al., 2018; Dunnell & Stein, 1989; Fladmark, 1982; Frahm, 2016). These variations reflect size boundaries that range from 0.125–0.250 mm up to 20 mm or more, depending on assemblage context (e.g., chronology, site function), research feasibility, or analyst preference (see Frick et al., 2024). Fladmark (1982:205) defined microdebitage as particles smaller than 1.0 mm resulting from deliberate lithic reduction. He noted that 1.0 mm represents the smallest particle size visible to the naked eye as a conchoidal flake and used it as a practical boundary between macrodebitage (1.0–3.0 mm) and microdebitage (<1.0 mm). Many researchers subsequently revised their definitions and size thresholds accordingly. A recurring challenge is that very small flakes may resemble naturally occurring grains, making them difficult to distinguish from anthropogenic specimens (Clark, 1986; Schick, 1986; Vance, 1986). Dunnell and Stein (1989) defined microdebitage as small, shattered objects or micro-flakes produced during percussion or pressure flaking. Cornford (1986), meanwhile, emphasised the relationship between artefact size and the ratio of waste (by-products) to tools (products). He advocated for examining artefacts larger than 20 mm, whereas others argue for studying technological products irrespective of size. The technological approach rooted in the chaîne opératoire (Tixier, 1974; Tixier et al., 1980) views lithic production as guided by a cognitive plan translated into a conceptual framework and then operationalised into physical actions (Inizan et al., 1995:15). These actions leave physical traces on detached and worked pieces (e.g., dorsal scars, points of impact), which serve as markers to distinguish different technological stages. These stigmata appear in both macro- and microdebitage. As Soressi and Geneste (2011:337) emphasise, “ one advantage of the chaîne opératoire is its ability to define the "temporality" and "geography" of artefacts within the spatial and temporal context of knapping activities. Each object can be assessed in its processual context through experimentally verified attributes that clarify how prehistoric people fractured stone volumes to produce useable cutting edges—a kind of volumetric or 3D puzzle”. The arbitrariness of size thresholds is illustrated in the work of Frison and Cornford. Both scholars established metric boundaries for analysis while examining links between microdebitage and stages in maintenance procedures. Frison (1968:149–152) described five types of retouch or sharpening flakes, derived from bifaces and scrapers, into which most unworked flakes could be categorised. Cornford identified longitudinal and transverse sharpening flakes, abbreviated as LSFs and TSFs, respectively. Although both researchers omitted prior micro-/macrodebitage distinctions, they acknowledged the similar technological roles played by flakes across the reduction sequence (Cornford, 1986:341). Other scholars distinguish between "use-flakes" detached during use and "retouch flakes" removed during formal modification (Hayes et al., 2014:78). Chan et al. (2020) further divide flakes into resharpening or reworking flakes, depending on whether reshaping was intended. These classifications suggest diverse intentions, including use, retouch, reshaping, or rejuvenation—or combinations thereof. Such gestures form part of the artefact's life history, contributing to maintenance and recycling (Bamforth, 1986). Overlooking microdebitage because of its size omits a vital part of an assemblage's behavioural signal. Conversely, a technological approach to microdebitage—rather than rigid size categorisation—enables the study of stigmata that inform the reconstruction of formation processes and link these to spatial and temporal aspects of dynamic knapping activities and technological decision-making. This paper highlights edge modification practices through retouch as proxies for maintenance activities, proposing a categorisation system that transcends size-based limits. We adopt Inizan et al.’s (1995) definition: “The term 'retouch' describes removals obtained by percussion or pressure, with the intention of making, finishing, or sharpening tools.” Consequently, resharpening is a form of retouch (Iovita, 2011; Morales & Vergès, 2014) and is treated as such herein. The term (re)touch implies recurrence on the same edge. Gestures linked to (re)shaping can restore the original form (isometric) or alter it (allometric). (Re)juvenation produces a new edge. All such gestures result in by-products with distinct technological stigmata, enabling their identification. It is expected that patterned maintenance indices can be derived from formal tools in the archaeological record, whether via allometric (shape-changing) or isometric (shape-preserving) reduction (Iovita, 2014). Accordingly, by-products from these actions can complement formal tool indices (Shott, 1994). While most studies focus on final products (e.g., Dibble, 1995), fewer examine by-products (e.g., Bourguignon, 1996b; Cornford, 1986). Given the predominance of by-products in assemblages, focusing on them offers a statistically robust sample that more accurately reflects technological events. These principles can support a classification system that reconstructs tool life histories through by-product analysis, creating a heuristic, behaviourally informative methodological framework (Iovita, 2011). Study Protocol Our study protocol introduces an inclusive classification system for by-products, focusing on their technological characterisation within the context of edge modification sequences (maintenance activities). Conford (1986) has already identified and described such by-products. He showed these to Harper Kelley and Francois Bordes in Paris, who explained them to him as " trimmings of damaged larger tools." From this, it is clear that they recognized the marginal retouch to be prior, not subsequent, to the flakes' detachment.” (Cornford, 1986, p. 337) . The categorisation that we propose follows the same principles as flake identification, although we deal within the 5–20 mm size range. By identifying technical stigmata present on the lithic artefacts, caused by previous operations (i.e., the presence and location of scars on the dorsal face and the point of impact), we can place the artefacts in various steps within the space and the time of the flintknapping activity ( sensu Soressi and Geneste, 2011). We classified these items according to the full attribute list that we used for the analysis of macrodebitage. Still, items smaller than 5 mm were counted but not fully studied (Armagan, 2003, p. 103; Dunnell & Stein, 1989; Fladmark, 1982), as we were logistically unable to carry out the low magnification microscopy needed for the identification of the lithic attributes. The development of this analysis protocol was done simultaneously as part of the general lithic attribute analysis (see raw data files in https://github.com/Nora-Arch/Moduleflakescategories), and the data was gathered through an open-access entry software, E5 (https://github.com/surf3s/E5), with a personalised configuration file. All data analysis and plotting were processed with R open-source and JMP statistical software. A research compendium using the rrtools package by (Marwick et al., 2017, 2018), including detailed info on used packages, software versions, and raw and processed data, is available here: https://github.com/Nora-Arch/Moduleflakescategories. Baseline of the module identification: Flake edge modification conforms to the same principle of flake formation and fracture mechanics (Cotterell & Kamminga, 1987). The proposed identification protocol aims for a detailed reconstruction of the modes of use of the piece, maintenance, and/or recycling (Bamforth, 1986). In a previous study (Malinsky-Buller et al., 2021), we identified these flakes as "shaping flakes," an inclusive term used to define all by-products of tool modification. Their identification was based on the presence of retouch scars on their dorsal faces (that were generated by previous retouch or use), and the recognition of the point of intersection between the dorsal face and the platform, where it preserves the “parent” edge of the tool (Fig. 2). As the use/retouching/reshape/rejuvenation progresses, the technological signature of those removal leaves their mark on byproduct flakes (hereafter named “module flake”; see below). Most previously identified shaping flakes are < 20 mm in maximal dimension but are distinguished from “chips” that do not preserve visible retouch scars, or the recognition of a previous tool edge (e.g. (Malinsky-Buller et al., 2021)). In our system, a module is a set of standardised parts or independent units (by-products) that can be used to construct a more complex structure (i.e., retouched piece). The modules may consist of several distinct yet interrelated units (module flakes), which may be integrated into a holistic reconstruction of the tool's life history. Criteria Module flakes are identified and classified according to three variables: regularity, distribution, and invasiveness of the dorsal scars (maintenance techniques) in relation to the parent blank. The identification of a module flake is based on the following observations: The point of intersection between the dorsal and the platform preserves the “parent” edge of the tool- the previous working edge (Fig.2/3 red dotted line). Identification of pre-retouch scars on the dorsal face (Fig.2/3). As the use/retouching/reshaping/rejuvenation progresses, the technological signature of those removals leaves its mark on byproduct flakes, i.e., module flakes. Identification of pre-retouch scar on the parent tool. When a flake is removed during retouching, it bears on his dorsal face the remnant of the scar(s) of the “parent” blank. This is a remnant of the “original” blank, and it can be plain or have one or more negatives of previous removals. The pre-retouch scar direction in M0-M3 flakes is usually perpendicular to the striking platform; M4 may have variable pre-retouched scar directions. Invasiveness attribution. Kuhn (Clarkson, 2002; Eren et al., 2005; Eren & Sampson, 2009; Hiscock & Tabrett, 2010; Kuhn, 1990, 1992) calculate an invasiveness index for tools. Here, we adapt this approach to an invasiveness percentage in relation to the module flake (surface area). The invasiveness percentage is estimated based on the relationship between retouching scars and “parent scars”. The stratigraphy and superposition of those scars on the dorsal face of a module flake enable the distinction between the different invasiveness ratios, ranging from 1 to 100%, see Fig. 2/3b (See an example Fig. 3, c,d for the M3 and M1). According to these observations, we suggest categorizing the by-products of maintenance techniques into five flake modules of edge modification: • M0: a flake with no previous edge modification. This flake does not show previous retouch scars. The main scar present is the remnant of the previous dorsal scar of the blank from the “parent” tool (similar to a Kombewa flake). It derives from the primary edge modification. An estimated invasiveness ratio from 0-15% up to the blank ridge (Fig. 2/3). • M1: a flake stemming from the initial edge modification of M0. The initiation of retouch scars is associated with the striking platform of the module flake (i.e., the edge of the “parent” tool). Those scars stratigraphically overlay the “parent” tool plain scar(s), being a crucial criterion for their identification. An estimated invasiveness ratio from 16-25% up to the blank ridge. (Fig. 4) • M2: The scar pattern on the flake demonstrates an overlap of scars with remnants of M1, M0, and the plain scar(s) from the “parent” tool. An estimated invasiveness ratio from 26-50% up to the blank ridge (Fig.5). • M3: The scar pattern on the flake shows an overlap of scars with remnants of the M2, M1, M0, and the plain scar(s) from the “parent” tool (Fig. 6). The estimated invasiveness ratio is from 51 to 95% up to the blank ridge. • M4: The scar pattern on the flake shows an overlap of scars with remnants of the M3, M2, M1, and M0 and the plain scar(s) from the “parent” tool. The range of estimated invasiveness ratio values is from 95-100% of retouch invasiveness (Fig. 7). The presence of two ventral faces can be used to distinguish M4 flakes from the rest. By removing an M4 flake, the knapper creates a new clean edge by reshaping /rejuvenating rather than using a retouching technique (M0-M3). A M4 flake removes part of the tool's ventral face to either maintain the same shape or modify it (allometric vs. isometric morphologies (Iovita, 2014). The presence of two ventral faces can distinguish M4 flakes. Similar technological pieces were already identified throughout the Palaeolithic in different contexts and regions, either in bifacial or unifacial shaping (Bourguignon, 1996; Lamotte, 1999; Verjux & Rousseau, 1986). For example, Cornford (1986) pointed out that the removal of a Long Sharpening Flake (LSF), which we term an M4 module flake, creates a new edge with the greatest possible length and sharpness on the parent tool. The uniqueness of this type made it more distinguishable when analysing the lithic assemblages in comparison to M0-M3 (Conard & Adler, 1997; De Loecker, 2006; Fonton et al., 1991; Frick et al., 2017; Malinsky-Buller, 2014; Roebroeks et al., 1988, 1997; Zaidner & Grosman, 2015). Therefore, using module flakes terminology, we aim to deviate from a linear and staged classification sequence, where each step depends on the previous one. Rather, we want to create a dynamic classification (Fig. 8). Fig. 8 Idealised edge modification scheme illustrating the module progression of lithic tool use, retouch, and discard across edge modification modules (M0–M4). The vertical ladder represents increasing stages of utilisation of the edge (0–100%), with curved arrows denoting estimated use-life ranges for each module, and the black arrows representing the jumps either from the initial straight to the end or any other module stop. Modules correspond to characteristic retouch/reshaping/rejuvenation patterns, with M0 reflecting minimal use and M4 representing heavily modified tools nearing saturation point. The model emphasises cyclical behaviours of use, retouching/reshaping, and potential rejuvenation before final discard. When we integrate module flake categories into the lithic attribute analysis based on the technological approach, disregarding dimensions enables us to explore questions related to tool maintenance activities. This, in return, provides us with a proxy to quantify the resolution of past hunter-gatherers' mobility and their degree of tool use-life. Case studies The two case studies presented- Ararat-1 Cave and Kalavan 2 -T1/T2 - show high frequencies of artefacts that would typically fall within the category of chips/debris in a size-oriented analysis. These characteristics are known from other Middle Palaeolithic archaeological sites in the southern Caucasus (e.g., (Gasparyan et al., 2014; Glauberman et al., 2020; Malinsky-Buller et al., 2021; Yeritsyan, 1972). By utilising the novel categorisation of the module flake in both Kalavan 2 and Ararat-1 cave assemblages, we demonstrate the utility of the high-resolution analysis of microdebitage assemblages in terms of provisioning behavioural signatures. Kalavan-2 The open-air site of Kalavan 2 (UTM 40T 3821 m E, 450551 m N, 1640 m asl) is located on the northern slopes of the Areguni Mountains at an elevation of. ca. 8 km north of the shores of Lake Sevan. The chronology of the site is based on fifteen post-infrared infrared stimulated luminescence (pIRIR) samples (Malinsky-Buller et al., 2021) as well as C14 ages from micromammal remains (Rogall et al., submitted). The age range of the main archaeological layers is between 60- 45ka BP (early MIS 3). In total, 2075 lithic artefacts (49 retouched tools and 819 module flakes) from 13 sedimentary units were retrieved. Three main archaeological layers were exposed, Unit 1b in T1 and Units 4 and 7 in T2, with 990 lithic artefacts (38 tools and 275 module flakes). The recovered lithic artefacts were made from seven raw materials: obsidian, basalt, dacite, welded tuff, chert, limestone, and an unidentified metamorphic rock. The frequencies of raw material types vary from unit to unit. Obsidian is the most common material. There are no indications for the first stages of core preparation and reduction, while the debitage consists mainly of flakes < 2 cm. Retouched obsidian pieces present an extremely narrow range of tool types, the majority falling within retouched points or convergent scrapers. The non-obsidian component exhibits more substantial indications of core reduction, including flakes, Kombewa flakes, and Levallois flakes (Ghukasyan et al., 2010; Malinsky-Buller et al., 2021). Ararat-1 Cave Ararat-1 cave (39.851 N, 44.769 E, 1034 m asl) is situated 2 km east of the town of Ararat on the northeast margins of the Ararat Depression. The two archaeological horizons are radiometrically dated by 12 post-infrared infrared stimulated luminescence (pIRIR) samples and a single C14 sample to a range between 50- 35ka BP (Oikonomou et al., 2025; Sherriff et al., 2024). The lithic assemblage of Ararat-1 is mainly comprised of chert and obsidian, with mafic lava, chalcedony, and quartzite represented by a few isolated pieces. The total assemblage amounts to 1770 artefacts, of which 37 are retouched tools and 680 (38% of the assemblage) are module flakes smaller than 10 mm (108 obsidian and 26 in chert). There are few indications of obsidian blank production at the site (very few cores or primary elements); in this raw material, it is mainly the last stage of the reduction sequence that is represented in the assemblage: use, maintenance, and rejuvenation (as defined/described by (Geneste, 1985; Nelson, 1991). Typologically, the retouched pieces are dominated by several types of scrapers and retouched flakes. The techno-typological composition of the chert component attests to initial stages of the reduction sequence, including cores and cortical elements (Nora et al., submitted). Adding to the regular attribute analysis, we incorporated a proposed module for micro flake categorisation (M0–M4), we analysed key lithic attributes such as technological length, technological width, thickness, number of dorsal scars, and raw material distribution across different modules to show the difference between them and what they represent behaviourally. Results Lithic Attribute Analysis across Modules Technological Length/Width/Thickness: Kalavan 2 T1 (Fig.9) and T2 (Fig.10) display distinct patterns in flake morphology across edge modification modules, offering insights into potential differences in maintenance practices. In Kalavan 2 T1, module flakes are generally smaller and thinner, particularly in M0 through M2. For example, M0 averages 6.3 mm in length, 6.6 mm in width, and 1.1 mm in thickness, among the lowest values across all three assemblages. Similarly, the length and width values of M1 flakes in T1 average 6.7 mm and 6.5 mm, respectively. Thickness remains low and consistent in early modules. In contrast, later modules in T1, particularly M3 and M4, show a noticeable increase in size. M4, for instance, reaches an average length of 10.5 mm and a width of 12.1 mm, with thickness increasing to 2.5 mm. These values can suggest a functional shift in edge modification. Kalavan 2 T2, by comparison, consistently exhibits larger and thicker flakes across nearly all modules. M0 through M2 flakes show apparent dimensional increases relative to T1, with M2 flakes averaging 11.1 mm in length and 10.9 mm in width, accompanied by a thickness of 2.0 mm. This pattern continues through M3 and M4, the latter reaching flakes averaging 15.8 mm in width and 3.6 mm in thickness. In contrast, the Ararat-1 Cave (Fig. 11) assemblage presents a more variable profile, generally falling between Kalavan T1 and T2 in terms of size, but with less pronounced extremes. M0 through M2 exhibit smaller flakes than those in T2, with M0 averaging 8.3 mm in length and 8.0 mm in width, and M2 reaching 8.5 mm and 7.3 mm, respectively. Thickness remains consistently low in these modules, ranging from 1.0 to 1.5 mm. However, a noticeable shift occurs in M3 and M4. Module 4 averages 10.2 mm in length and 10.3 mm in width, with a thickness of 2.6 mm, closely paralleling the later module flakes observed in Kalavan T1. Overall, the Ararat-1 cave profile suggests a more moderate reduction strategy, with less emphasis on heavily retouched or robust forms. This may indicate different site functions, more expedient tool production, or variable raw material constraints compared to the Kalavan localities. Number of Dorsal Scars and the approximate Area of module flakes To explore the relationship between the estimated invasiveness ratio, we applied a multivariate analysis of variance (MANOVA) with two dependent variables: the number of dorsal scars and the approximate area (technological length x technological width). The independent variable, module, represents the categorical factor. MANOVA was chosen as a global test to assess whether flake size varied across module categories while accounting for potential correlation between the two dependent variables. This approach minimises the risk of Type I error associated with multiple comparisons and enables the detection of multivariate structures that are not visible through univariate methods alone. Following significant MANOVA results, we conducted univariate ANOVAs and Tukey HSD post-hoc tests to determine which traits contributed to group differentiation and which module contrasts were statistically significant. The combination of MANOVA and follow-up ANOVAs offers both global and trait-specific insights, allowing for a nuanced understanding of how maintenance practices could have shaped artefact form. For Kalavan 2 T1, the MANOVA revealed a significant multivariate effect of module (Wilks’ Λ = 0.56994, F (8, 1638) = 66.46, p < 0.001), indicating that the overall combination of dorsal scars and area differs across modules (for more info, please check https://github.com/Nora-Arch/Moduleflakescategories). Follow-up univariate ANOVAs confirmed that both dorsal scar count ( F (4, 820) = 150.8, p < 0.001) and approximate area ( F (4, 820) = 20.04, p < 0.001) varied significantly with module. Tukey HSD tests showed significant differences between all module levels for dorsal scars, with the most substantial increases occurring between M0 to M4 (Δ = 8.02 scars) and M2 to M4 (Δ = 3.31 scars). Even comparisons between adjacent modules (e.g., M2 vs. 3, Δ = 2.06 scars) remained statistically robust, supporting a graded intensification in edge modification. For the approximate area, significant differences emerged primarily between M0 and M3 and M4 (Δ = 45–79 mm²), indicating that size increases become apparent only with a high intensity ratio of retouch. As for Kalavan 2 T2, MANOVA also yielded a strong multivariate effect (Wilks’ Λ = 0.55014, F (8, 520) = 22.64, p < 0.001), suggesting consistent morphological structuring by module. Univariate ANOVAs showed significant differences for both dorsal scars number ( F (4, 261) = 51.91, p < 0.001) and approximate area ( F (4, 261) = 9.39, p < 0.001). Tukey comparisons again revealed a strong progressive increase in the number of dorsal scars, particularly between M0 and M4 (Δ = 7.33 scars), while M1 and M2 differed by approximately 3.3 scars. Area differences were significant between M1 and M2 through 4 (Δ ≈ 74.4 mm²), indicating that larger flakes are associated with more scars (and therefore more intense retouch), although variability in M0 limited statistical resolution in that category. MANOVA results for the Ararat-1 assemblage were also significant (Wilks’ Λ = 0.56849, F (8, 1342) = 54.74, p < 0.001), confirming that retouch intensity is reflected in morphological variation at this site as well. Univariate ANOVAs revealed significant effects for dorsal scars ( F (4, 672) = 120.4, p < 0.001) and approximate area ( F (4, 672) = 17.9, p < 0.001). Tukey tests indicated that M4 had significantly more dorsal scars than M0–2 (Δ = 5–7 scars), although the difference between M3 and M4 was not statistically significant, suggesting a plateau at high retouch levels. For the approximate area, only M3 differed significantly from M0 (Δ = 34.8 mm², p = 0.024), while other contrasts failed to reach significance. This suggests that, at Ararat-1, increases in artefact size with retouch intensity are less consistent than those seen in dorsal scars. Across all three assemblages, MANOVA results provide strong evidence that the ratio of invasiveness is associated with significant variation in flake size. Dorsal scars numbers increase progressively and significantly with module categories, showing consistent, stepwise intensification in all plots (Figs. 12–14). Approximate area shows a weaker but complementary pattern, with significant increases in higher module categories in Kalavan, and less consistent changes in Ararat-1 Cave. These data support a two-tiered pattern of morphological change: scar accumulation occurs reliably across modules, while size increases are more context-dependent. Module Flakes Frequencies and Raw Material Distance Module flakes can be further explored when considering the frequencies of module categories with the raw material units, which enables us to decipher the fragmentation of the reduction sequence, specifically when looking at the phases of tool maintenance and use (Turq et al., 2013). In the Kalavan 2 T1 assemblage (Fig. 15), the Gutansar (47 km) obsidian appears frequently in Modules 0 to 4. Hatis (50 km) has no representation in Module 0, while it is more prominent in Modules 1, 2, and 3, with few M4 pieces. The outliers in this plot are the Chikiani source (140 km), represented by a single M4 flake, Chikiani 2b on M1, and Kars Digor (158 km), which is represented by a single M2 occurrence. Tsaghkunyats sources (60 km) are present in all modules. The same trend is visible for non-obsidian raw material tuff. Module flakes made on flint and chert (green) are relatively rare, appearing in small amounts in Modules 1 and 2. At Kalavan 2 T2 (Fig.16), obsidian from Gutansar (47 km) and Hatis (50 km) are a significant raw material, and its presence is particularly prominent in Module 3. Tsaghkunyats sources (60 km) have close to zero representation on M0. In modules 1 to 3, the frequencies are up to 2%, followed by a single example in M4. Among the non-obsidian flake modules, tuff stands out. It is particularly dominant in M0, accounting for 43.8% of the material in this module, and continues to be represented across all modules, including M4. This consistent presence of tuff from M0 to M4 suggests its sustained utility throughout various stages of retouching and modification. Basalt and dacite are only minimally represented, primarily in the earlier modules. As for flint, this raw material has no representation in the first three modules, but it appears as isolated instances in the M3 and M4. Ararat-1 Cave (Fig. 17) exhibits some unique trends. Obsidian source Geghasar (35-40 km), a prominent raw material in this site, is highly concentrated in Modules 1 and 2. Gutansar (60 km) also shows a similar pattern, but its peak occurs in Module 3. Interestingly, Hatis (50 km) and its variants (Hatis alpha, beta, and gamma) appear consistently across all module categories but are less frequent in the earlier stage (Module 0). Items from the Arteni sources (100 – 130 km) appear in Modules 1 to 3, with sporadic appearances in modules 0 and 4. Non-obsidian raw materials, such as chert, basalt, and mafic lava, are rare in this assemblage and only show marginal appearances in Modules 0, 1, and 2. Discussion: Behaviour beyond module flakes The study of retouched pieces plays a key role in the systematics of the Palaeolithic record. As best exemplified by the nomenclature developed by Bordes (Bordes, 1953, 1951, 1961), pre-determined morphotypes shaped by retouching arguably conveyed cultural significance. In parallel, the perspective originally put forward by Frison (1969), later Jelinek (1976), and culminated in the works of Dibble and his students (Dibble, 1985, 1995) suggested that those same types were stages along a continuum of size change, that crossed retouching, reshaping, and rejuvenation stages. At the same time, Lewis Binford (Binford, 1973, 1979) positioned the theoretical foundation of curation, by stating that those maintenance techniques were part of a bigger concept, which emphasises maximisation of raw material utility through maintenance and recycling linked with procurement strategies. Yet despite the richness of these perspectives, their empirical application remains debated, beginning with the most fundamental issue, the basic unit of analysis. For Shott (1989, 1996) and Dibble (1995), this unit was the retouched tool. For Shott (1989:24, 1996), the definition entailed “The degree of curation, defined as the degree of use or utility extracted, expressed as a relationship between the tool's initial maximum utility and how much of that utility is realised before discard.” In contrast, (Barton & Riel-Salvatore, 2014; Clark & Barton, 2017) proposed an assemblage level approach (WABI) to study curated pieces, contrasting them with expedient technologies. Despite selecting different analytical units, those methodologies and empirical toolkits aim for similar “higher-level” questions, such as the reconstruction of the mobility patterns of the group and the extent of occupation within a locality, as well as resource exploitation intensity. Usually, analysts choose to focus exclusively on studies that link identifiable artefacts (commonly macrodebitage artefacts, in most cases attributed to tools) with non-local raw materials that could be brought from localities that exceed the daily exploitation territory (DET) to propose mobility models (Bailey & Davidson, 1983; Kuhn, 1992, 1995; Vita-Finzi et al., 1970). Nevertheless, other studies (Ekshtain et al., 2017; Féblot‐Augustins, 2009) challenge these assumptions by re-evaluating the analytical unit by looking at different proxies. Building on this, our paper shifts the unit of analysis to the by-products of the maintenance techniques, disregarding their size, as a new proxy not only to identify tools that had been removed from the site (ghost tools), but also as a new way of looking at the intensity of edge modification techniques such as retouching, reshaping and rejuvenation. Our concept of curation concurs with that of Shott (1994; 1996), in that we regard it as a behaviour designed to extend the use-life of an artefact (Shott, 1994). In this paper, our perspective enables us to fine tune the ways to document and quantify the degrees of extension of the retouched tool use-life. Our proposed nomenclature and the suggested protocol for categorising what we refer to as "module flakes" are based on fundamental principles of fracture mechanics to all artefacts regardless of their size, emphasising the identification of key diagnostic features in lithic studies, and acknowledges that tools are manufactured through processes of retouching and reshaping and rejuvenation. From a unified analytical foundation, we apply the same technological attribute analysis to microdebitage artefacts as we do to macrodebitage, serving the same analytical purposes. We establish here a module categorisation of flakes to overcome the typological diversification in the literature. “Module flakes” have been identified through several periods and contexts on lithic assemblages from the Palaeolithic to the New World in the Late Prehistoric context (Bourguignon, 1996a, 2001; Cornford, 1986a; Frison, 1968; Hays, 1992; Touzé, 2018). Synonym nomenclatures are available, either as the proposed five categories of retouch and sharpening flakes by Frison (1968) or the attribution of the LSF, TSF flakes of Conford (1986), or the several phases of decortication and post-decortication suggested by (Jérémie & Vacher, 1992). For the latest phase of rejuvenation of retouched tools (our “M4”), a wide range of synonyms were applied, all depending on the directionality of the removals, such as Tranchet Blow, or Lateral Tranchet Blow (LTB) at Nesher Ramla (Prévost et al., 2022), or the chanfren (Burney, 1967); for more morphological differences see (Frick et al., 2024). We argue that most of these different flake typologies that are technological flakes related to edge modification are targeting specific human gestures of retouch/reshape and rejuvenation applied to the tool (Tixier et al., 1980). Thus, that can be perceived as a unit instead of several typologies. Despite abundant terminologies and contextual examples, a unified technological framework for understanding these flakes has yet to be fully articulated. The progressive modification of a blank through retouching and reshaping is the main proxy used here to assess the use-life of a lithic tool. Archaeologists have made significant efforts to develop new quantitative methods based on stone tools to understand lithic reduction, more specifically, retouch intensity (Clarkson, 2002; Davis & Shea, 1998; Dibble, 1987, 1995; Eren et al., 2005; Eren & Prendergast, 2008; Eren & Sampson, 2009; Hiscock & Attenbrow, 2003; Hiscock & Clarkson, 2009; Hiscock & Tabrett, 2010; Kuhn, 1990; Marwick, 2008; Morales et al., 2015; Morales & Vergès, 2014). These quantitative methods can sometimes estimate the hypothetical initial volume of the tool blank, allowing us to predict differences in the variation of the tool along the two scaling directions (allometric versus isometric reduction trajectories (Iovita, 2009, 2010, 2011). Our approach looks at the waste material that results from this reduction, the by-product of the maintenance techniques. Our approach stems from the principle of the chaîne opératoire, that each artefact can be placed within the space and the time of the flintknapping activity, (Soressi & Geneste, 2011, p. 337). In this sense, we aim to show that the module flake categories, once incorporated into the reduction sequence, can significantly help unpack the complexities of decision-making regarding tool use and maintenance. We believe that we can now better test retouch intensity in lithic assemblages by using the quantitative approach based on tools (a static approach) and linking them to the frequencies of categories of module flakes that are representative of the dynamic behaviour of retouching, reshaping and rejuvenating. Furthermore, by binding the classification of raw material units with the module flake categories, we can attempt to quantify raw material curation in terms of the degree of extracted utility per raw material relative to the distance from their sources. The results of this study provide strong support for the interpretive validity of the module classification system as a framework for understanding edge modification in lithic assemblages. Across the datasets of all three assemblages, the combination of MANOVA and univariate analyses revealed consistent and statistically significant differentiation among module categories, indicating that each step in the classification reflects a meaningful behavioural or technological threshold in edge modification. The MANOVA results confirmed that when dorsal scar count and approximate area are considered together, retouch intensity, as categorised by module flakes, exerts a significant influence on artefact size. This multivariate structure supports the use of module flakes as a holistic categorisation of edge modification. Follow-up univariate ANOVAs revealed that the number of dorsal scars increases progressively and significantly across modules, with strong differentiation even between adjacent modules. This consistent pattern across all assemblages demonstrates that modules capture a graded process of edge modification, from primary edge modification (M0) through various degrees of retouch (M1–3) to advanced rejuvenation (M4). However, we stress that those modules still do not always reflect a linear progression, as different “stages” of retouch may occur simultaneously on different sections of a tool’s edges (Fig. 8). Approximate area followed a similar but more selective pattern. Increases in size were statistically significant primarily at higher modules, particularly M3 and M4. This suggests that while scar accumulation reflects a continuous trajectory of edge modification, increases in flake size are likely linked to intensifying maintenance and are more variable and context-dependent. The plateau in scar count observed between M3 and M4 in some cases may indicate a saturation point in edge modification, or a final phase in a tool’s use-life before discard or reshaping and transport, as evident by the “ghost tools”. When dealing with provisioning strategies, module classification can distinguish the initial approach towards the blank. M0 (primary edge modification) is conceptually different from the other modules, as shown through the Tukey test. The absence or low frequencies of M0 flakes suggest that the initial edge modification occurred off-site. This could indicate a curated provisioning strategy where pre-modified blanks or partially retouched tools were transported to the locality. In contrast, high frequencies of M0 guide us to the perspective of unmodified blanks at the locale, which can be represented by on-site blank production (when compared to raw material units, core presence, unmodified blanks, and core trimming elements) or the transportation of blanks into the locality. The presence of M1, M2, and M3 indicates that blanks are at a stage of secondary edge modification, implying the material either entered the site already retouched or was retouched on-site. For example, in Kalavan 2 T2, the high frequency of M3 made on Gutansar, Hatis, and Tsaghkunyats obsidian suggests that these pieces arrived at the site in a partially worked state; more advanced retouching (51–95% of edge modification) took place on-spot. The same pattern can be observed in Ararat-1 Cave on the module flakes from the Arteni obsidian sources. This segmented representation of the module flakes could indicate that the specific lithic raw materials became more important over time, potentially reflecting shifts in raw material availability and exploitation. The high frequency of M3 suggests raw material economisation with a significant focus on these sources during more advanced stages of edge maintenance. The interpretation of Module 4 flakes points us to a tipping point in tool maintenance when retouching leads to reshaping/rejuvenation of the edge (Fig. 8). The frequent occurrence of M4 flakes, especially in Gutansar and Hatis obsidian at Kalavan 2 T1 and T2, suggests that tools were continuously maintained on-site. (Shott, 1989, p. 24, 1996) suggested that curation is a concept directly related to the tool's initial maximum utility and how much of that utility is realised before it is discarded. Following this perspective, we can attempt to measure curation through the module flake categories, taking into consideration the raw material units (RMU) as our additional analytical unit (Odell, 2004, pp. 93–95; Roebroeks et al., 1988). The pattern we observed either from the raw material strategies and the maintenance strategies seen by the module flakes are consistent with curation strategies where the goal is to prolong the life of the tools through repeated rejuvenation, ensuring that high-quality materials like obsidian are utilised to their fullest potential (Odell 1996; Nora et al, 2025a). The single representation of M4 for the most distant raw materials, such as Chikiani obsidian in Kalavan 2 T1, suggests that some tools entered the site and were taken away after reshaping/rejuvenation. This evidence of “ghost tools” is represented only by the by-products that indicate maintenance activity at the locale. This pattern of reshaping tools to create new, sharp edges and then removing them from the site is indicative of tool provisioning for continued use elsewhere, supporting a mobile strategy where tools were maintained at key locations and then transported to other areas of activity, sensu (Bousman, 1993). The categorisation we put forward here aims to encompass the various typologies used with relation to the concept of curation and, consequently, the maintenance of lithic tools (see Fig. 18), emphasising that their behavioural signature is more significant than typological attribution (see discussion in (Bar-Yosef & Van Peer, 2009) for the use of technological traits vs. typological). The suggested categorisation reflects a behaviourally meaningful model that corresponds to measurable changes in artefact form. For example, a single M4 flake removal would create a clean edge, contrasting with a wavy edge shaped by a succession of M0/M1/M2/M3 removals. Thus, interpreting the module flakes within a technological perspective provides a nuanced reconstruction of past decision-making. The suggested categorisation can be divided into the first edge modification (M0), the retouching/reshaping (M1/M2/M3), and finally, the reshaping /rejuvenation phase (M4) (Fig. 17). Those techniques, although conceptually closer, differ in their gestures and intentionality (Forestier, 1993; Inizan et al., 1995). The different module flakes enable us to reconstruct at what stage blanks were transported to a given locality, either as unretouched blanks or as retouched items, and if the latter, to what degree. The progressive increase in dorsal scar numbers and the selective increase in artefact size confirm that the module framework captures distinct categories within the situational decision-making during the life-history of lithic tools. This reinforces the model's utility for comparative and inferential studies of lithic maintenance strategies within archaeological sites. Conclusions Applying our proposed module flake categorisation (M0–M4) to the late Middle Palaeolithic artefact assemblages of Kalavan 2 T1 and T2 and Ararat-1 Cave offers new insights into the fluidity of the technological processes underlying tool maintenance. The systematic classification of by-products, regardless of size, enabled us to reconstruct detailed edge modification sequences and tool use-life. The identification of M0 indicates the first approach to an unretouched blank, while a high frequency of M1, M2, and M3 module flakes revealed extensive initial edge modifications and ongoing retouching activities. These patterns reflect adaptive strategies for tool maintenance, which can also be used to explore retouch intensity. M4 flakes, representing advanced stages of tool use, suggest prolonged use and retouch intensity of tools that lead to a blunt or unstructured edge, as noted by Frison ( 1968 ). This dynamic module category goes beyond a linear sequence by considering edge modifications, overlapping, and an iterative nature. Rather than a linear progression of retouch, we offer a “snake and ladder” approach, emphasising the fluidity of decision-making (Fig. 8 ). The analysis underscores the critical role of large and small waste flakes in understanding past technological behaviours, stressing that by-products are integral to a holistic reconstruction of prehistoric tool use-life, rather than merely waste material. By applying this novel framework, we have expanded the interpretative potential of lithic assemblages, shedding light on the often-overlooked smaller fractions. Those pieces hold significant information about the intensity and nature of tool use, tool maintenance practices, and mobility strategies. This approach offers a more nuanced understanding of tool curation, contributing to broader discussions on Palaeolithic technological organisation and putting the decision-making processes of tool maintenance at the focal point of inquiry. Declarations Declarations Competing Interests: The authors have no relevant financial or non-financial interests to disclose. Funding Declaration: Fieldwork, laboratory analyses and syntheses were funded by the TransCause Project, Investigating Pleistocene population dynamics, in the Southern Caucasus (ERC – Starting Grant Agreement No. 948015), the Fritz Thyssen Foundation grant awarded for the project “Pleistocene Hunter-Gatherer Lifeways and Population Dynamics in the Ararat (paleo-lake) Depression, Armenia”, the Gerda Henkel Stiftung grant (n. AZ 10_V_17 and n. AZ 23/F/19) and the Leakey Foundation (awarded to AMB). Further support was provided by Gfoeller Renaissance Foundation (USA) and “Areni-1 Cave” Consortium (“Areni-1 Cave” Scientific Research Foundation (Armenia) and the Institute of Archaeology and Ethnography of the National Academy of Sciences of the Republic of Armenia. Grant support was provided to Nora by the Bina and Moshe Stekelis Fund for Prehistoric Archaeology, the Ruth Amiran Fund for archaeological research in Eretz-Israel, and the Hebrew University International PhD Talent Scholarship. The pXRF instrument used for obsidian artefact sourcing was funded thanks to generous support from Yale’s Offices of the Vice Provost for Research, Dean of Science, and Dean of Social Science and its Council on Archaeological Studies and the Department of Anthropology. Abstract: The study of retouching, reshaping, and rejuvenation in lithic technology has traditionally focused on finished tools, overlooking the by-products of these processes, particularly microdebitage. This has led to an incomplete understanding of the dynamic behaviours associated with tool maintenance and a lack of crucial information about prehistoric technological strategies. Here we address this knowledge gap. Specifically, we introduce a classification system for lithic by-products resulting from retouching, reshaping, and rejuvenation techniques, categorising them into five modules (M0 through M4) based on lithic technological analysis. This methodology integrates the chaîne opératoire approach to analyse flakes without size thresholds. To demonstrate our approach, we apply it, coupled with raw material sourcing, to lithic assemblages from two Middle Palaeolithic sites in Armenia, Kalavan 2 and Ararat-1 Cave. This enables a precise reconstruction of tool use-life and, in turn, the mobility strategies of Pleistocene hunter-gatherers. Our findings demonstrate that microdebitage (by-products) can contribute to a holistic view of decision-making, revealing patterns in tool maintenance and raw material provisioning. The module system provides insights into the production of ‘ghost tools,’ which are not present in the archaeological record, as well as curation behaviours and economic decisions regarding raw materials that were previously difficult to discern. By shifting the focus from finished artefacts to by-products, this framework enhances our ability to interpret lithic assemblages and understand the adaptive strategies of prehistoric hunter-gatherers. Author Contribution D.N. and A.M.B. are the main contributors to the conception and design of the work; the acquisition, analysis, and interpretation of data; and the drafting and critical revision of the manuscript for important intellectual content. D.N. conducted all lithic analysis, processed the data, prepared the figures, and created the open-access repository. A.M.B. reviewed and participated in the lithic analysis and contributed to writing and revising multiple drafts. B.G. and A.P. contributed to the acquisition of data through fieldwork and laboratory coordination in Armenia, ensured access to collections, and revised and approved the final manuscript. E.F. performed the pXRF analysis, contributed to data interpretation and manuscript revision, and approved the final version for publication. All authors approved the version to be published and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. Acknowledgement The authors would like to thank the Director of the Institute of Archaeology and Ethnography of the National Academy of Sciences of the Republic of Armenia, Dr. Arsen Bobokhyan, for his support regarding the permits for fieldwork investigations, sample transportation, and availability of the facilities. Nora expresses his gratitude to Prof. E. Hovers, from the Institute of Archaeology, the Hebrew University of Jerusalem, for the supervision and discussion around the paper's subject, as well as for reading too many drafts. Special thanks to N. Paz and the Human-Environment Dynamics Laboratory of the Institute of Archaeology, the Hebrew University of Jerusalem, for providing laboratory facilities and support during research. Moreover, Nora would like to thank all the volunteers who participated in the fieldwork and Doctor Marion Prevost for the brainstorming over figures design, to Sergey Alon for the drawing of tiny lithics and to A. Jallon for the good talks over the PhD. Fieldwork, laboratory analyses and syntheses were funded by the TransCause Project, Investigating Pleistocene population dynamics, in the Southern Caucasus (ERC – Starting Grant Agreement No. 948015), the Fritz Thyssen Foundation grant awarded for the project “Pleistocene Hunter-Gatherer Lifeways and Population Dynamics in the Ararat (paleo-lake) Depression, Armenia”, the Gerda Henkel Stiftung grant (n. AZ 10_V_17 and n. AZ 23/F/19) and the Leakey Foundation (awarded to AMB). Further support was provided by Gfoeller Renaissance Foundation (USA) and “Areni-1 Cave” Consortium (“Areni-1 Cave” Scientific Research Foundation (Armenia) and the Institute of Archaeology and Ethnography of the National Academy of Sciences of the Republic of Armenia. Grant support was provided to Nora by the Bina and Moshe Stekelis Fund for Prehistoric Archaeology, the Ruth Amiran Fund for archaeological research in Eretz-Israel, and the Hebrew University International PhD Talent Scholarship. The pXRF instrument used for obsidian artefact sourcing was funded thanks to generous support from Yale’s Offices of the Vice Provost for Research, Dean of Science, and Dean of Social Science and its Council on Archaeological Studies and the Department of Anthropology. Data Availability The data that supports this paper is available at https://github.com/Nora-Arch/Moduleflakescategories. References Armagan, T. (2003). Small lithic Debris Analysis, with Appendix A Alphabetic Listing of Attribute Definitions Used for the Tor Faraj Small Lithic Debris Study ; Appendix В Small Lithic Debris Analysis Results Tables. Neanderthals in the Levant. Behavioral Organization and the Beginnings of Human Modernity , 86:106; 271:275. Bailey, G. N., & Davidson, I. (1983). Site exploitation territories and topography: Two case studies from palaeolithic spain. Journal of Archaeological Science , 10 (2), 87–115. https://doi.org/10.1016/0305-4403(83)90044-4 Bamforth, D. B. (1986). Technological Efficiency and Tool Curation. American Antiquity , 51 (1), 38–50. https://doi.org/10.2307/280392 Barton, C. M., & Riel-Salvatore, J. (2014). The formation of lithic assemblages. Journal of Archaeological Science , 46 , 334–352. https://doi.org/10.1016/j.jas.2014.03.031 Bar-Yosef, O., & Van Peer, P. (2009). The Chaine Operatoire Approach in Middle Paleolithic Archaeology . http://nrs.harvard.edu/urn-3:HUL.InstRepos:2960197 Binford, L. R. (1973). Interassemblage variability-the Mousterian and the ‘functional’ argument, ed. Colin Renfrew. In C. Renfrew (Ed.), The Explanation of Culture Change (pp. 227–254). Duckworth. Binford, L. R. (1979a). Organization and Formation Processes: Looking at Curated Technologies. Journal of Anthropological Research , 35 (3), 255–273. https://doi.org/10.1086/jar.35.3.3629902 Binford, L. R. (1979b). Organization and formation processes: Looking at curated technologies. Journal of Anthropological Research , 35 (3), 255--273. Binford, L. R., & Binford, S. R. (1966). A Preliminary Analysis of Functional Variability in the Mousterian of Levallois Facies. American Anthropologist , 68 (2), 238–295. Bordes. (1953). Essai de Classification des industries ‘moustériennes’. Bulletin de La Société Préhistorique de France , 50 (7/8), 457–466. Bordes, F. (1951). Le Complexe Mousterien:Mousteriens, Levalloisien et Tayacien. L’Anthropologie , 55 , 1–23. Bordes, F. (1961). Mousterian cultures in France. Science , 134 (3482), 803–810. Bordes, F. (1969). Typologie du paléolithique ancien et moyen . CNRS. Bordes, F., & Sonneville‐Bordes, D. de. (1970). The significance of variability in Palaeolithic assemblages. World Archaeology . https://www.tandfonline.com/doi/abs/10.1080/00438243.1970.9979464 Bourguignon, L. (1996a). La conception de débitage Quina. In Reduction Processe («Chaînes opératoire») for the European Mousterian. Bourguignon, L. (1996b). Un moustérien tardif sur le site d’Ummel Tlel (Bassin d’El khowm, Syrie) ?Exemples des niveaux II base ’et III2a’. In E. Carbonell & M. Vaquero (Eds.), The last Neandertals,the first anatomically Modern Humans: A Tale About Human Diversity; Cultural Change and Human Evolution (pp. 317–336). Universitat Rovira i Virgili. Bourguignon, L. (2001). Apports de l’expérimentation et de l’analyse techno-morpho- fonctionnelle à la reconnaissance du processus d’aménagement de la retouche Quina. In Préhistoire et approche expérimentale (Vol. 5). Bousman, C. B. (1993). Hunter-Gatherer Adaptations, Economic Risk and Tool Design. Lithic Technology , 18 (1–2), 59–86. https://doi.org/10.1080/01977261.1993.11720897 Brézillon, M. (1968). La dénomination des objets de pierre taillée. Matériaux pour un vocabulaire des préhistoriens de langue française . https://www.persee.fr/doc/galip_0072-0100_1968_sup_4_1 Burney, M. (1967). McBurney: The Haua Fteah (Cyrenaica) and the stone... - Google Scholar . https://scholar.google.com/scholar_lookup?title=The%20Haua%20Fteah%20in%20Cyrenaica%20and%20the%20Stone%20Age%20of%20the%20South-East%20Mediterranean&publication_year=1967&author=C.B.M.%20McBurney Clark, G. A., & Barton, C. M. (2017). Lithics, landscapes & la Longue-durée – Curation & expediency as expressions of forager mobility. Quaternary International , 450 , 137–149. https://doi.org/10.1016/j.quaint.2016.08.002 Clarkson, C. (2002). An Index of Invasiveness for the Measurement of Unifacial and Bifacial Retouch: A Theoretical, Experimental and Archaeological Verification. Journal of Archaeological Science , 29 (1), 65–75. https://doi.org/10.1006/jasc.2001.0702 Conard, N. J., & Adler, D. S. (1997). Lithic Reduction and Hominid Behavior in the Middle Paleolithic of the Rhineland. Journal of Anthropological Research . https://doi.org/10.1086/jar.53.2.3631275 Cornford, J. M. (1986a). Specialised resharpening techniques and evidence of handedness. In J. M. Callow, P., Cornford (Ed.), La Cotte de St. Brelade, Jersey. Excavations by C.B.M. McBurney 1961–1978. (pp. 337--351.). Geo Books. Cornford, J. M. (1986b). Specialized resharpening techniques and evidence of handedness. La Cotte de St. Brelade , 1978 , 337–351. Cotterell, B., & Kamminga, J. (1987). The Formation of Flakes. American Antiquity , 52 (4), 675–708. https://doi.org/10.2307/281378 Davidson, I., Noble, W., Armstrong, D. F., Black, L. T., Calvin, W. H., Davis, W., Falk, D., Foster, M. L., Graves, P., Halverson, J., & Hewes, G. W. (1989). The Archaeology of Perception: Traces of Depiction and Language [and Comments and Reply]. Current Anthropology . https://doi.org/10.1086/203723 Davis, Z. J., & Shea, J. J. (1998). Quantifying Lithic Curation: An Experimental Test of Dibble and Pelcin’s Original Flake-Tool Mass Predictor. Journal of Archaeological Science , 25 (7), 603–610. https://doi.org/10.1006/jasc.1997.0255 De Loecker, D. (2006). Beyond the site. The Saalian archaeological record at Maastricht-Belvédère (the Netherlands). Analecta Praehistorica Leidensia 35/36 . Leiden University Press, Leiden. Debénath, A., & Dibble, H. L. (1994). Handbook of Paleolithic Typology: Lower and Middle Paleolithic of Europe . University of Pennsylvania Museum of Archaeology and Anthropology. https://doi.org/10.2307/j.ctt14btgq9 Dibble, H. L. (1985). Raw material variability in Levallois flake manufacture. Current Anthropology (Chicago) , 26 , 391–393. https://doi.org/10/bm57q4 Dibble, H. L. (1987). The Interpretation of Middle Paleolithic Scraper Morphology. American Antiquity , 52 (1), 109–117. https://doi.org/10.2307/281062 Dibble, H. L. (1995). Middle paleolithic scraper reduction: Background, clarification, and review of the evidence to date. Journal of Archaeological Method and Theory , 2 (4), 299–368. https://doi.org/10.1007/BF02229003 Dunnell, R. C., & Stein, J. K. (1989). Theoretical issues in the interpretation of microartifacts. Geoarchaeology , 4 (1), 31–41. https://doi.org/10.1002/gea.3340040103 Ekshtain, R., Ilani, S., Segal, I., & Hovers, E. (2017). Local and Nonlocal Procurement of Raw Material in Amud Cave, Israel: The Complex Mobility of Late Middle Paleolithic Groups. Geoarchaeology , 32 (2), 189–214. https://doi.org/10.1002/gea.21585 Eren, M. I., Dominguez-Rodrigo, M., Kuhn, S. L., Adler, D. S., Le, I., & Bar-Yosef, O. (2005). Defining and measuring reduction in unifacial stone tools. Journal of Archaeological Science , 32 (8), 1190–1201. https://doi.org/10.1016/j.jas.2005.03.003 Eren, M. I., & Prendergast, M. E. (2008). Comparing and synthesizing unifacial stone tool reduction indices. Lithic Technology , 49–85. Eren, M. I., & Sampson, C. G. (2009). Kuhn’s Geometric Index of Unifacial Stone Tool Reduction (GIUR): Does it measure missing flake mass? Journal of Archaeological Science , 36 (6), 1243–1247. https://doi.org/10.1016/j.jas.2009.01.011 Féblot‐Augustins, J. (2009). Revisiting European Upper Paleolithic Raw Material Transfers: The Demise of the Cultural Ecological Paradigm? In Lithic Materials and Paleolithic Societies (pp. 25–46). https://doi.org/10.1002/9781444311976.ch3 Fladmark, K. R. (1982). Microdebitage analysis: Initial considerations. Journal of Archaeological Science , 9 (2), 205–220. https://doi.org/10.1016/0305-4403(82)90050-4 Fonton, M., Lhomme, V., & Christensen, M. (1991). Un cas de « réduction » et de « transformation » d’outil au Paléolithique moyen. Un racloir déjeté de la grotte de Coustal à Noailles (Corrèze) . https://doi.org/10.3406/pal.1991.1035 Forestier, H. (1993). Le Clactonien: Mise en application d’une nouvelle méthode de débitage s’inscrivant dans la variabilité des systèmes de production lithique du Paléolithique ancien . https://doi.org/10.3406/pal.1993.1104 Frick, J. A., Herkert, K., Hoyer, C. T., & Floss, H. (2017). The performance of tranchet blows at the Late Middle Paleolithic site of Grotte de la Verpillière I (Saône-et-Loire, France). In PLoS ONE (Vol. 12, Issue 11). https://doi.org/10.1371/journal.pone.0188990 Frick, J. A., Herkert, K., Hoyer, C. T., & Floss, H. (2024). Using the Technological Feature of Tranchet Blow on Keilmesser as a Connecting Element Between Similar Middle Paleolithic Assemblages from the Côte Chalonnaise (Saône-et-Loire, France). In H. Koehler (Ed.), The Rhine During the Middle Paleolithic: Boundary or Corridor? (1st ed., pp. 287–314). Kerns Verlag. https://doi.org/10.51315/9783935751353.013 Frison, G. (1968). A Functional Analysis of Certain Chipped Stone Tools. American Antiquity , 33 (2), 149–155. https://doi.org/10.2307/278516 Gasparyan, B., Egeland, C., Adler, D., Pinhasi, R., Glauberman, P., & Haydosyan, H. (2014). The Middle Paleolithic Occupation of Armenia: Summarizing Old and New Data. In Stone Age of Armenia: A Guide-book to the Stone Age Archaeology in the Republic of Armenia (pp. 65–105). Geneste, J.-M. (1985). Analyse lithique d’industries moustériennes du Périgord: Une approche technologique du comportement des groupes humains au Paléolithique moyen . University of Bordeaux I Bordeaux. https://pascal-francis.inist.fr/vibad/index.php?action=getRecordDetail&idt=12142529 Ghukasyan, R., Colonge, D., Nahapetyan, S., Ollivier, V., Gasparyan, B., Monchot, H., & Chataigner, Ch. (2010). KALAVAN-2 (NORTH OF LAKE SEVAN, ARMENIA): A NEW LATE MIDDLE PALEOLITHIC SITE IN THE LESSER CAUCASUS. Archaeology, Ethnology and Anthropology of Eurasia , 38 (4), 39–51. https://doi.org/10.1016/j.aeae.2011.02.003 Glauberman, P., Gasparyan, B., Wilkinson, K., Frahm, E., Nahapetyan, S., Arakelyan, D., Raczynski-Henk, Y., Haydosyan, H., & Adler, D. S. (2020). Late Middle Paleolithic Technological Organization and Behavior at the Open-Air Site of Barozh 12 (Armenia). Journal of Paleolithic Archaeology , 3 (4), 1095–1148. https://doi.org/10.1007/s41982-020-00071-4 Hays, M. A. (1992). A Functional Analysis of the Lithic Material From Burrone Scierra I (Calabria, Italy) . Hiscock, P., & Attenbrow, V. (2003). Early Australian implement variation: A reduction model. Journal of Archaeological Science , 30 (2), 239–249. https://doi.org/10.1006/jasc.2002.0830 Hiscock, P., & Clarkson, C. (2009). The reality of reduction experiments and the GIUR: Reply to Eren and Sampson. Journal of Archaeological Science , 36 (7), 1576–1581. https://doi.org/10.1016/j.jas.2009.03.019 Hiscock, P., & Tabrett, A. (2010). Generalization, inference and the quantification of lithic reduction. World Archaeology , 42 (4), 545–561. https://doi.org/10.1080/00438243.2010.517669 Inizan, M., Reduron, M., Roche, H., & Tixier, J. (1995). Technologie de la pierre taillée (Vol. 4). Iovita, R. (2009). Ontogenetic scaling and lithic systematics: Method and application. Journal of Archaeological Science , 36 (7), 1447–1457. https://doi.org/10.1016/j.jas.2009.02.008 Iovita, R. (2010). Comparing Stone Tool Resharpening Trajectories with the Aid of Elliptical Fourier Analysis. In S. Lycett & P. Chauhan (Eds.), New Perspectives on Old Stones: Analytical Approaches to Paleolithic Technologies (pp. 235–253). Springer. https://doi.org/10.1007/978-1-4419-6861-6_10 Iovita, R. (2011). Shape Variation in Aterian Tanged Tools and the Origins of Projectile Technology: A Morphometric Perspective on Stone Tool Function. PLOS ONE , 6 (12), e29029. https://doi.org/10.1371/journal.pone.0029029 Iovita, R. (2014). The role of edge angle maintenance in explaining technological variation in the production of Late Middle Paleolithic bifacial and unifacial tools. Quaternary International , 350 , 105–115. https://doi.org/10.1016/j.quaint.2014.08.032 Jelinek, A. J. (1976). Form, Function and Style in Lithic Analysis. In Cultural Change and Continuity: Essays in Honor of James Bennett Griffin. New York:Academic Press. , 19–33. Jérémie, S., & Vacher, S. (1992). Le Hoabinhien en Thaïlande: Un exemple d’approche expérimentale. Bulletin de l’École Française d’Extrême-Orient , 79 (1), 173–209. Kuhn, S. L. (1990). A geometric index of reduction for unifacial stone tools. Journal of Archaeological Science , 17 (5), 583–593. https://doi.org/10.1016/0305-4403(90)90038-7 Kuhn, S. L. (1992). Blank Form and Reduction as Determinants of Mousterian Scraper Morphology. American Antiquity , 57 (1), 115–128. https://doi.org/10.2307/2694838 Kuhn, S. L. (1995). Mousterian Lithic Technology: An Ecological Perspective . Princeton University Press. https://www.jstor.org/stable/j.ctt7zv59w Lamotte, A. (1999). L’apport des remontages dans la compréhension des méthodes de débitage et de façonnage des gisements acheuléens de la Somme: Les exemples de la Ferme de l’Épinette et de l’Épinette à Cagny (Somme, France). Bulletin de La Société Préhistorique Française , 96 (2), 117–131. Malinsky-Buller, A. (2014). Contextualizing Curational Strategies at the Late Lower Paleolithic Site of Holon, Israel. PaleoAnthropology . https://doi.org/10.4207/PA.2014.ART87 Malinsky-Buller, A., Glauberman, P., Ollivier, V., Lauer, T., Timms, R., Frahm, E., Brittingham, A., Triller, B., Kindler, L., Knul, M. V., Krakovsky, M., Joannin, S., Hren, M. T., Bellier, O., Clark, A. A., Blockley, S. P. E., Arakelyan, D., Marreiros, J., Paixaco, E., … Gasparyan, B. (2021). Short-term occupations at high elevation during the Middle Paleolithic at Kalavan 2 (Republic of Armenia). PLOS ONE , 16 (2), e0245700. https://doi.org/10.1371/journal.pone.0245700 Marwick, B. (2008). What attributes are important for the measurement of assemblage reduction intensity? Results from an experimental stone artefact assemblage with relevance to the Hoabinhian of mainland Southeast Asia. Journal of Archaeological Science , 35 (5), 1189–1200. https://doi.org/10.1016/j.jas.2007.08.007 Marwick, B., Boettiger, C., & Mullen, L. (2017). Packaging data analytical work reproducibly using R (and friends) (No. e3192v1). PeerJ Inc. https://doi.org/10.7287/peerj.preprints.3192v1 Marwick, B., Boettiger, C., & Mullen, L. (2018). Packaging Data Analytical Work Reproducibly Using R (and Friends). The American Statistician , 72 (1), 80–88. https://doi.org/10.1080/00031305.2017.1375986 Monnier, by G. (2006). The Lower/Middle Paleolithic Periodization in Western Europe. Current Anthropology . https://doi.org/10.1086/506280 Morales, J. I., Lorenzo, C., & Vergès, J. M. (2015). Measuring Retouch Intensity in Lithic Tools: A New Proposal Using 3D Scan Data. Journal of Archaeological Method and Theory , 22 (2), 543–558. https://doi.org/10.1007/s10816-013-9189-0 Morales, J. I., & Vergès, J. M. (2014). Technological behaviors in Paleolithic foragers. Testing the role of resharpening in the assemblage organization. Journal of Archaeological Science , 49 , 302–316. https://doi.org/10.1016/j.jas.2014.05.025 Nelson, M. C. (1991). The Study of Technological Organization. Archaeological Method and Theory , 3 , 57–100. Odell, G. H. (2004). Lithic Analysis . Springer US. https://doi.org/10.1007/978-1-4419-9009-9 Oikonomou, I. A. K., Karambaglidis, T., Fenn, K., Gur-Arieh, S., Nora, D., Sánchez-Romero, L., Rogall, D., Vetesse, D., Gasparian, B., Petrosyan, A., & Malinsky-Buller, A. (2025). Unravelling the Formation Processes and Depositional Histories of the Middle Palaeolithic Ararat-1 Cave, Armenia: A Multiscalar and Multiproxy Geoarchaeological Approach. Prévost, M., Centi, L., & Zaidner, Y. (2022). The use of the lateral tranchet blow technique at Nesher Ramla (Israel): A new cultural marker in the Levantine Middle Paleolithic? Quaternary International , 624 , 128–147. https://doi.org/10.1016/j.quaint.2020.11.008 Roebroeks, W., Kolen, J., Van Poecke, M., & Van Gijn, A. (1997). «Site J»: An early Weichselian (Middle Palaeolithic) flint scatter at Maastricht-Belvedere, The Netherlands . https://doi.org/10.3406/pal.1997.1231 Roebroeks, W., VAN GIJN, A., VAN DE VELDE, P., & ARPS, C. E. S. (1988). From Find Scatters to Early Hominid Behaviour: A Study of Middle Palaeolithic Riverside Settlements at Maastricht-Belvédère (The Netherlands). From Find Scatters to Early Hominid Behaviour : A Study of Middle Palaeolithic Riverside Settlements at Maastricht-Belvédère (The Netherlands) , 21 . Sherriff, J. E., Petrosyan, A., Rogall, D., Nora, D., Frahm, E., Lauer, T., Karambaglidis, T., Knul, M. V., Vettese, D., Arakelyan, D., Gur-Arieh, S., Vidal-Matutano, P., Morales, J., Fewlass, H., Blockley, S. P. E., Timms, R., Adigyozalyan, A., Haydosyan, H., Glauberman, P., … Malinsky-Buller, A. (2024). Palaeoenvironmental and chronological context of hominin occupations of the Armenian Highlands during MIS 3: Evidence from Ararat-1 cave. Quaternary Science Advances , 13 , 100122. https://doi.org/10.1016/j.qsa.2023.100122 Shott, M. (1989). On Tool-Class Use Lives and the Formation of Archaeological Assemblages. American Antiquity , 54 (1), 9–30. https://doi.org/10.2307/281329 Shott, M. (1994). Size and form in the analysis of flake debris: Review and recent approaches. Journal of Archaeological Method and Theory , 1 (1), 69–110. https://doi.org/10.1007/BF02229424 Shott, M. (1996). An Exegesis of the Curation Concept. Journal of Anthropological Research , 52 (3), 259–280. https://doi.org/10.1086/jar.52.3.3630085 Soressi, M., & Geneste, J.-M. (2011). The History and Efficacy of the Chaîne Opératoire Approach to Lithic Analysis: Studying Techniques to Reveal Past Societies in an Evolutionary Perspective. PaleoAnthropology , 2011 , 334–350. Tixier, J., Inizan, M.-L., & Roche, H. (1980). Terminologie et technologie . Cercle de recherches et d’études préhistoriques. Touzé, O. (2018). Aux prémices du Gravettien dans le Nord-Ouest européen: Étude de la production des pointes lithiques à Maisières-Canal (province de Hainaut, Belgique). Bulletin de La Société Préhistorique Française , 115 (3), 455–495. https://doi.org/10.3406/bspf.2018.14920 Trigger, B. G. (2006). A history of archaeological thought (2nd ed). Cambridge university press. Turq, A., Roebroeks, W., Bourguignon, L., & Faivre, J.-P. (2013). The fragmented character of Middle Palaeolithic stone tool technology. Journal of Human Evolution , 65 (5), 641–655. https://doi.org/10.1016/j.jhevol.2013.07.014 Verjux, C., & Rousseau, D.-D. (1986). La retouche Quina: Une mise au point. Bulletin de La Société Préhistorique Française , 11–12 , 404–415. Vita-Finzi, C., Higgs, E. S., Sturdy, D., Harriss, J., Legge, A. J., & Tippett, H. (1970). Prehistoric Economy in the Mount Carmel Area of Palestine: Site Catchment Analysis. Proceedings of the Prehistoric Society , 36 , 1–37. https://doi.org/10.1017/S0079497X00013074 Wargo, M. (2009). The Bordes-Binford debate: Transatlantic interpretive traditions in Paleolithic archaeology - ProQuest . https://www.proquest.com/openview/a32bf9dadb387d8882e0a7268994373c/1?pq-origsite=gscholar&cbl=18750 Yeritsyan, B. G. (1972). ‘Некоторые особенности намеренного рассечения орудий мустьерской эпохи (по материалам Ереванской пещерной стоянки)’ [Some Features of Intentional Truncation of Mousterian Tools (Based on the Materials of Yerevan Cave Site)] . 53–60. Zaidner, Y., & Grosman, L. (2015). Middle Paleolithic sidescrapers were resharped or recycled? A view from Nesher Ramla, Israel. Quaternary International , 361 , 178–187. https://doi.org/10.1016/j.quaint.2014.11.037 Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 24 Nov, 2025 Read the published version in Journal of Archaeological Method and Theory → Version 1 posted Editorial decision: Revision requested 30 Jun, 2025 Reviews received at journal 25 Jun, 2025 Reviews received at journal 29 May, 2025 Reviewers agreed at journal 21 May, 2025 Reviewers agreed at journal 21 May, 2025 Reviewers invited by journal 21 May, 2025 Editor assigned by journal 21 May, 2025 Submission checks completed at journal 20 May, 2025 First submitted to journal 15 May, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6673907","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":459142501,"identity":"581e56e0-2a27-4a80-b0ca-64b518ffbce6","order_by":0,"name":"David Nora","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABDUlEQVRIiWNgGAWjYDCCA0DEA2YxA5kGFgwGDAwJB0B8NnaCWtgSgFokkLQw49bCANHCA1TMANYCBTi08B0/Y3jgDcMdOfP2M183/CiQYDCXSHh4gKHGjoEPhxbJMzkGB+cwPDOWOZO77WYP0GGWMxKADjuWjNNhBgfSEg7zMBxOnMGQu+0GD8gvN0Ba2A7g1nL+GVQL/5tnN//AtfzDo+VG8gGIFokctttwWxjbcGuRvPH4wME5BoeNJSSemd2WMZDgMTjzIOFAYl8yDy4tfOcTmz+8qTgsJ8Gf/Ozmmz82cgbHc5I/fPhmJyff3oBdD8R5CCYwjngSGBJgkUUkYD9AiupRMApGwSgY/gAAg3NfRX4MBjoAAAAASUVORK5CYII=","orcid":"","institution":"The Hebrew University of Jerusalem","correspondingAuthor":true,"prefix":"","firstName":"David","middleName":"","lastName":"Nora","suffix":""},{"id":459142502,"identity":"1c94ccf1-55b1-441a-9aa8-1815d613c15a","order_by":1,"name":"Ariel Malinsky Buller","email":"","orcid":"","institution":"The Hebrew University of Jerusalem","correspondingAuthor":false,"prefix":"","firstName":"Ariel","middleName":"Malinsky","lastName":"Buller","suffix":""},{"id":459142503,"identity":"45a8b21a-4411-4761-b033-eb7343b53a88","order_by":2,"name":"Boris Gasparyan","email":"","orcid":"","institution":"National Academy of Sciences of Armenia","correspondingAuthor":false,"prefix":"","firstName":"Boris","middleName":"","lastName":"Gasparyan","suffix":""},{"id":459142504,"identity":"a682342f-8942-4594-8674-7ba55d23dbcc","order_by":3,"name":"Artur Petrosyan","email":"","orcid":"","institution":"National Academy of Sciences of Armenia","correspondingAuthor":false,"prefix":"","firstName":"Artur","middleName":"","lastName":"Petrosyan","suffix":""},{"id":459142505,"identity":"11dd451f-dfa9-4b9b-84c3-b6525fc61eab","order_by":4,"name":"Ellery Frahm","email":"","orcid":"","institution":"Yale University","correspondingAuthor":false,"prefix":"","firstName":"Ellery","middleName":"","lastName":"Frahm","suffix":""}],"badges":[],"createdAt":"2025-05-15 15:08:24","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6673907/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6673907/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s10816-025-09754-0","type":"published","date":"2025-11-24T15:56:58+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":83243625,"identity":"97194954-8e10-486c-85d8-0d61d9a38130","added_by":"auto","created_at":"2025-05-21 16:30:49","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":107731,"visible":true,"origin":"","legend":"\u003cp\u003eMacrodebitage and Microdebitage within the stages of technological organisation, with the highlight metrics and definitions of microdebitage.\u003c/p\u003e","description":"","filename":"1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6673907/v1/2a90e4e48fa5619f2bfcf7e4.jpg"},{"id":83242579,"identity":"bbb265dd-f461-4e4d-adfd-587cbc0850f4","added_by":"auto","created_at":"2025-05-21 16:14:49","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":99754,"visible":true,"origin":"","legend":"\u003cp\u003eSchematic and archaeological representation of the criteria for the identification of a module flake. a. schematic representation of the parent tool. b. a cross-section of the “parent” tool with the schematic representation of the percentage ratio of increasing invasiveness. c. The main descriptive terms for module flake 3 are as follows: the colours indicate the different modules in b. d. Main descriptive terms for module flake 1. Note: Plain scar is the previous dorsal face of the initial edge (blank), so in the chronological sequence of flake removals, it is the first scar.\u003c/p\u003e","description":"","filename":"2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6673907/v1/57b2af4c211e5fe96af1bbc9.jpg"},{"id":83244354,"identity":"8ad54bbc-7093-4d0f-8b55-886b508bdec2","added_by":"auto","created_at":"2025-05-21 16:38:49","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":63257,"visible":true,"origin":"","legend":"\u003cp\u003eSchematic and archaeological representation of the criteria for the identification of a module flake. a. schematic representation of the parent tool. b. a cross section of the retouched tool with the schematic representation of the percentage ratio of invasiveness. c. Main descriptive terms for module flake 4.\u003cstrong\u003eFig. 3\u003c/strong\u003e Schematic and archaeological representation of the criteria for the identification of a module flake. a. schematic representation of the parent tool. b. a cross section of the retouched tool with the schematic representation of the percentage ratio of invasiveness. c. Main descriptive terms for module flake 4.\u003c/p\u003e","description":"","filename":"3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6673907/v1/37e83c822ae7e5181715875e.jpg"},{"id":83243627,"identity":"258cd09b-5ec5-4649-ad96-848db90e48a6","added_by":"auto","created_at":"2025-05-21 16:30:49","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":70892,"visible":true,"origin":"","legend":"\u003cp\u003eArchaeological examples of M0 and M1 flakes (from Ararat-1 Cave and Kalavan 2).\u003c/p\u003e","description":"","filename":"4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6673907/v1/7595a3315f342a40375a7fec.jpg"},{"id":83243632,"identity":"b5ddbdb7-b7ac-4938-af66-c01bb2fe8974","added_by":"auto","created_at":"2025-05-21 16:30:49","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":54872,"visible":true,"origin":"","legend":"\u003cp\u003eArchaeological examples of M2 flakes from Ararat-1 Cave and Kalavan 2(a, c-g - Obsidian; b -chert).\u003c/p\u003e","description":"","filename":"5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6673907/v1/c4eb99e462d29e54fa3667f4.jpg"},{"id":83242602,"identity":"c19f6a63-4cb2-4da5-865c-56a5097fb26a","added_by":"auto","created_at":"2025-05-21 16:14:49","extension":"jpg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":79776,"visible":true,"origin":"","legend":"\u003cp\u003eArchaeological examples M3 flakes from Ararat-1 Cave and Kalavan 2.\u003c/p\u003e","description":"","filename":"6.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6673907/v1/12dd4e6015c8e15566512a1f.jpg"},{"id":83242868,"identity":"e45a61c8-eb22-49df-bf63-bcb0073af93b","added_by":"auto","created_at":"2025-05-21 16:22:49","extension":"jpg","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":73369,"visible":true,"origin":"","legend":"\u003cp\u003eArchaeological examples of M4 flakes from Ararat-1 Cave and Kalavan 2.\u003c/p\u003e","description":"","filename":"7.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6673907/v1/3dc1c5d7f7e97068e2952e44.jpg"},{"id":83244358,"identity":"6529cc17-2894-4eb2-99da-f08c0d082ef1","added_by":"auto","created_at":"2025-05-21 16:38:49","extension":"jpg","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":58147,"visible":true,"origin":"","legend":"\u003cp\u003eIdealised edge modification scheme illustrating the module progression of lithic tool use, retouch, and discard across edge modification modules (M0–M4). The vertical ladder represents increasing stages of utilisation of the edge (0–100%), with curved arrows denoting estimated use-life ranges for each module, and the black arrows representing the jumps either from the initial straight to the end or any other module stop. Modules correspond to characteristic retouch/reshaping/rejuvenation patterns, with M0 reflecting minimal use and M4 representing heavily modified tools nearing saturation point. The model emphasises cyclical behaviours of use, retouching/reshaping, and potential rejuvenation before final discard.\u003c/p\u003e","description":"","filename":"8.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6673907/v1/53e6b279592b3f01e6cb45a7.jpg"},{"id":83242588,"identity":"32f3ac4a-097b-4012-bb2e-a90b794734da","added_by":"auto","created_at":"2025-05-21 16:14:49","extension":"jpg","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":71937,"visible":true,"origin":"","legend":"\u003cp\u003eRadar charts displaying the metric attributes (Length, Width, Thickness) of lithic flakes across edge modification modules at Kalavan 2 T1. Each chart represents a distinct module (M0 - M4), with individual observations shown as lines, polygon areas representing module averages, and shaded backgrounds indicating previously calculated averages for comparison. Numeric labels above each chart show the mean ± standard deviation for each attribute.\u003c/p\u003e","description":"","filename":"9.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6673907/v1/42f0c148fbac62e546c66b8d.jpg"},{"id":83244355,"identity":"04372e89-2770-4361-bab7-c88e8727e224","added_by":"auto","created_at":"2025-05-21 16:38:49","extension":"jpg","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":73154,"visible":true,"origin":"","legend":"\u003cp\u003eRadar charts displaying the morphological attributes (Length, Width, Thickness) of lithic flakes across edge modification modules at Kalavan 2 T2. Each chart represents a distinct module (M0–M4), with individual observations shown as lines, polygon areas representing module averages, and shaded backgrounds indicating previously calculated averages for comparison. Numeric labels above each chart show the mean ± standard deviation for each attribute.\u003c/p\u003e","description":"","filename":"10.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6673907/v1/8157dfee5576f57ea151e45a.jpg"},{"id":83242584,"identity":"25de2313-8a35-4637-98b7-e8abc2823241","added_by":"auto","created_at":"2025-05-21 16:14:49","extension":"jpg","order_by":11,"title":"Figure 11","display":"","copyAsset":false,"role":"figure","size":70756,"visible":true,"origin":"","legend":"\u003cp\u003eRadar charts displaying the morphological attributes (Length, Width, Thickness) of lithic flakes across edge modification modules at Ararat-1 Cave. Each chart represents a distinct module (M0–M4), with individual observations shown as lines, polygon areas representing module averages, and shaded backgrounds indicating previously calculated averages for comparison. Numeric labels above each chart show the mean ± standard deviation for each attribute.\u003c/p\u003e","description":"","filename":"11.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6673907/v1/d609682a410650ddbb950c13.jpg"},{"id":83242577,"identity":"4bbaa217-000e-419f-9527-52414e9751b6","added_by":"auto","created_at":"2025-05-21 16:14:49","extension":"jpg","order_by":12,"title":"Figure 12","display":"","copyAsset":false,"role":"figure","size":74156,"visible":true,"origin":"","legend":"\u003cp\u003eRelationship between the number of dorsal scars and approximate area (mm²) of module flakes across edge modification modules at Kalavan 2 T1. Each panel corresponds to a different modification module, with individual flakes plotted as coloured dots. Solid black trend lines represent linear regressions, with shaded areas indicating 95% confidence intervals.\u003c/p\u003e","description":"","filename":"12.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6673907/v1/51e64411845a7e1eee1eb0b4.jpg"},{"id":83242600,"identity":"51634af8-40ce-4831-af46-9fac5d5a8c22","added_by":"auto","created_at":"2025-05-21 16:14:49","extension":"jpg","order_by":13,"title":"Figure 13","display":"","copyAsset":false,"role":"figure","size":71257,"visible":true,"origin":"","legend":"\u003cp\u003eRelationship between the number of dorsal scars and approximate area (mm²) of module flakes across edge modification modules at Kalavan 2 T2. Each panel corresponds to a different modification module, with individual flakes plotted as coloured dots. Solid black trend lines represent linear regressions, with shaded areas indicating 95% confidence intervals.\u003c/p\u003e","description":"","filename":"13.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6673907/v1/acbd3615b0c3323bd5160ace.jpg"},{"id":83242585,"identity":"b1ac790a-8c0a-4a05-bfea-33e12d736efb","added_by":"auto","created_at":"2025-05-21 16:14:49","extension":"jpg","order_by":14,"title":"Figure 14","display":"","copyAsset":false,"role":"figure","size":75996,"visible":true,"origin":"","legend":"\u003cp\u003eRelationship between the number of dorsal scars and approximate area (mm²) of module flakes across edge modification modules at Ararat-1 Cave. Each panel corresponds to a different modification module, with individual flakes plotted as coloured dots. Solid black trend lines represent linear regressions, with shaded areas indicating 95% confidence intervals.\u003c/p\u003e","description":"","filename":"14.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6673907/v1/3e8e3eaee8c32ad70a179329.jpg"},{"id":83242595,"identity":"34a0d661-6c36-4b53-b6c2-7891d867a29b","added_by":"auto","created_at":"2025-05-21 16:14:49","extension":"jpg","order_by":15,"title":"Figure 15","display":"","copyAsset":false,"role":"figure","size":76489,"visible":true,"origin":"","legend":"\u003cp\u003eDistribution of module flakes by raw material at Kalavan 2 T1. Each bar represents the count and percentage of flakes made from a given raw material within each module (Modules 0–4). Raw materials are listed along the y-axis, ordered by approximate source distance from the site (left). The chart illustrates both the diversity of raw materials and their varying frequencies across modules.\u003c/p\u003e","description":"","filename":"15.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6673907/v1/82f5a268fb9771993a69f308.jpg"},{"id":83244356,"identity":"501b19ee-f9bb-41ad-9e62-847042837932","added_by":"auto","created_at":"2025-05-21 16:38:49","extension":"jpg","order_by":16,"title":"Figure 16","display":"","copyAsset":false,"role":"figure","size":89450,"visible":true,"origin":"","legend":"\u003cp\u003eDistribution of module flakes by raw material at Kalavan 2 T2. Each bar represents the count and percentage of flakes made from a given raw material within each module (Modules 0–4). Raw materials are listed along the y-axis, ordered by approximate source distance from the site (left). The chart illustrates both the diversity of raw materials and their varying frequencies across modules.\u003c/p\u003e","description":"","filename":"16.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6673907/v1/676c5e7c1c7bca97781c6d8d.jpg"},{"id":83242591,"identity":"4573397f-c2b6-44ce-b071-a09c941649fd","added_by":"auto","created_at":"2025-05-21 16:14:49","extension":"jpg","order_by":17,"title":"Figure 17","display":"","copyAsset":false,"role":"figure","size":110166,"visible":true,"origin":"","legend":"\u003cp\u003eDistribution of module flakes by raw material at Ararat-1 Cave. Each bar represents the count and percentage of flakes made from a given raw material within each module (Modules 0–4). Raw materials are listed along the y-axis, ordered by approximate source distance from the site (left). The chart illustrates both the diversity of raw materials and their varying frequencies across modules.\u003c/p\u003e","description":"","filename":"17.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6673907/v1/f4450250a8e4fd0a74f61577.jpg"},{"id":83242597,"identity":"342c87ed-3e88-4450-bcd0-20feb1bf690a","added_by":"auto","created_at":"2025-05-21 16:14:49","extension":"jpg","order_by":18,"title":"Figure 18","display":"","copyAsset":false,"role":"figure","size":80427,"visible":true,"origin":"","legend":"\u003cp\u003eConceptual model linking edge modification modules to tool curation and maintenance strategies. Modules 0–3 reflect retouching and reshaping behaviours, while Module 4 is associated with reshaping and rejuvenation. Definitions follow Shott (1989, 1996), with curation as utility over time and maintenance as behaviours extending tool use-life.\u003c/p\u003e","description":"","filename":"18.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6673907/v1/d9186a1138ea2cf1f2d17cf8.jpg"},{"id":97178231,"identity":"bf867253-361b-424e-b715-39bd9d343191","added_by":"auto","created_at":"2025-12-01 16:02:56","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2230001,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6673907/v1/efcb7dbf-7e60-458d-a2b2-52721c417279.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Snakes and Ladders: A technological approach to tool maintenance by-products using module flake categories","fulltext":[{"header":"Introduction","content":"\u003cp\u003eThe study of retouched lithic pieces has played a key role in the systematics of Palaeolithic research since its earliest stages\u0026nbsp;(Monnier, 2006; Trigger, 2006; Wargo, 2009). A well-known example is Bordes\u0026rsquo; nomenclature for Lower and Middle Palaeolithic\u0026nbsp;artefacts, which created a common language that remains foundational in the classification of lithic objects into types\u0026nbsp;(Bordes, 1953, 1951, 1961). Bordes\u0026rsquo; system focused on morphology\u0026mdash;i.e., the form of the object\u0026mdash;and the location of retouch as diagnostic components. Similarly,\u0026nbsp;(Br\u0026eacute;zillon, 1968)\u0026nbsp;defined three aspects of retouch variation: magnitude (marginal or invasive), extent (continuous or denticulated), and location (direct, inverse, alternating, and bifacial). According to Bordes\u0026nbsp;(Bordes, 1951; Bordes \u0026amp; Sonneville‐Bordes, 1970), morphotypes were emic;\u0026nbsp;i.e., configurations imbued with cultural significance. In contrast,\u0026nbsp;(Deb\u0026eacute;nath \u0026amp; Dibble, 1994; Dibble, 1995)\u0026nbsp;interpreted retouched artefacts (mostly scrapers), as defined by Bordes\u0026nbsp;(Bordes, 1969), as reflecting different stages along a continuum of morphological change resulting from rejuvenation or reshaping of tool edges (see also the \u0026ldquo;finished artefact fallacy\u0026rdquo;\u0026nbsp;(Davidson et al., 1989)).\u003c/p\u003e\n\u003cp\u003eWith the adoption of technological approaches based on the \u003cem\u003echa\u0026icirc;ne op\u0026eacute;ratoire\u003c/em\u003e in the 1990s, discussions about retouch variation and its meanings took new directions. The roots of Dibble\u0026rsquo;s perspective lie in the idea that variation in lithic assemblages resulted from adaptive strategies inherent to hunter-gatherers\u0026apos; modes of resource exploitation across the landscape\u0026nbsp;(Binford, 1979a; Jelinek, 1976). In this view, lithic technologies were part of dynamic technological decision-making, the selection and integration of strategies for making, using, transporting, and discarding tools and the materials needed for their manufacture and maintenance\u0026nbsp;(Nelson, 1991, p. 57). This aligns, to some extent, with the \u003cem\u003echa\u0026icirc;ne op\u0026eacute;ratoire\u003c/em\u003e approach, which reconstructs sequences of actions, including knapping procedures, from raw material acquisition to tool discard. Binford had already pointed to this when he explained that such technological strategies are shaped by provisioning conditions (e.g., artefact production for future use, transport, maintenance, and recycling), all closely tied to hunter-gatherer mobility\u0026nbsp;(Binford, 1973, 1979a). He coined the term \u0026lsquo;curation\u0026rsquo; to describe tools that are effective across multiple tasks\u0026nbsp;(Binford, 1973, 1979a; Binford \u0026amp; Binford, 1966, p. 66). This represents a shift in archaeological reasoning\u0026mdash;from descriptive typology to processual analysis. At this point, a triad began to emerge: retouched tools, technological strategies, and the resulting artefacts (both products and by-products).\u003c/p\u003e\n\u003cp\u003eOne of the key concepts linking these domains to curation is maintenance. Maintenance is intrinsically tied to curation, as the latter refers to strategies of caring for tools and toolkits to maximise their utility over time (Andrefsky, 2009; Cornford, 1986; Frison, 1968; Shott, 1986, 1994, 1996). According to Binford, curated technologies involve tools that are maintained through repeated use (Binford, 1973). Maintenance increases efficiency by extending a tool\u0026rsquo;s use-life relative to the energy invested in its manufacture. Frison (1968) highlighted the significance of use-life and resharpening, suggesting that maintenance behaviours reflect investment in prolonging tool functionality\u0026mdash;a concept aligned with curation. Bamforth (1986) explicitly argued that maintenance is a component of curation, proposing it as a response to raw material shortages. When raw materials are scarce, replacing worn tools is costly, making maintenance a more efficient strategy. Conversely, the frequency of tool maintenance likely decreases when raw materials are readily available. Shott (1996) defined curation as the degree of use or utility extracted from a tool throughout its life. From this perspective, maintenance contributes directly to curation by increasing the total utility obtained before discard. Repair and resharpening allow a tool to be used longer and for more tasks, thus fulfilling more of its initial utility. Maintenance includes activities such as resharpening, repairing, and recycling tools to extend their usefulness (Schiffer, 1976). These activities often involve retouching, reshaping, and rejuvenation, which result in both intentional and unintentional production of by-products\u0026mdash;usually microartefacts (see Fig. 1).\u003c/p\u003e\n\u003cp\u003eSuch techniques have shaped the morphology of retouched pieces, which have transitioned from rigid typological constructs to dynamic, fluid products of economic decision-making (Geneste, 1985). The size of artefacts resulting from these maintenance actions poses a challenge. As these by-products frequently fall into the category of microdebitage, they are often under-studied or selectively sampled, focusing only on morphotypes deemed representative of specific technological signatures (Bourguignon, 1996; Cornford, 1986; Frick et al., 2024; Pr\u0026eacute;vost et al., 2022). In this paper, we examine the technological interpretation and reconstruction of the operational sequence of maintenance techniques (retouching, reshaping, and rejuvenation) through two case studies from the Middle Palaeolithic of Armenia: Kalavan 2 and Ararat-1 Cave. Specifically, we propose and apply a new categorisation of microdebitage artefacts, classifying them into five modules (M0 through M4). These by-products are viewed as dynamic expressions of decision-making, complementing static tool typologies, and offer a proxy for understanding the technological organisation of Pleistocene hunter-gatherers.\u003c/p\u003e\n\u003ch2\u003eMicrodebitage between terminology and technology\u003c/h2\u003e\n\u003cp\u003eThe recognition and identification of microdebitage are part of the broader classification of lithic artefacts. Artefacts of this particular category typically enter the archaeological record as small, shattered objects or micro-flakes produced during operations preceding the reduction sequence (Fig. 1). These fragments relate to the process of flake formation governed by fracture mechanics, with differing morphologies resulting from specific directions of flake initiation and spalling (Cotterell \u0026amp; Kamminga, 1987; Li, Lin, et al., 2023; Li, Reeves, et al., 2023; Speth, 1972, 1975). From this analytical standpoint, we refer to the intentional production of microdebitage as a result of maintenance activities. At various stages of the reduction sequence, one can expect differing proportions of microdebitage\u0026mdash;either as small, unintended by-products of fracturing during technological gestures or as intentional products created for shaping, retouching, or rejuvenating a blank or tool. For instance, higher frequencies of intentionally produced microdebitage are expected during use, retouching, and maintenance stages rather than during initial blank production (Fig. 1). Microdebitage appears throughout most stages of the reduction sequence (Fig. 1), whether produced intentionally or not (Deb\u0026eacute;nath \u0026amp; Dibble, 1994). Torcal\u0026rsquo;s work, for example, references the production of small debris from knapping activity (labelled BP2G; Torcal et al., 1991:178).\u003c/p\u003e\n\u003cp\u003eLithic classification systems are typically based on metric criteria that distinguish macrodebitage from microdebitage by-products (Bar-Yosef, 1981; Odell, 1996, 2000, 2001), except in the case of bladelet production. To explore this further, we must first determine what threshold defines microdebitage. The size threshold has been inconsistently defined in the literature, leading to multiple terms such as microartefacts, microdebris, or chips (Clark, 1986; de la Torre et al., 2018; Dunnell \u0026amp; Stein, 1989; Fladmark, 1982; Frahm, 2016). These variations reflect size boundaries that range from 0.125\u0026ndash;0.250 mm up to 20 mm or more, depending on assemblage context (e.g., chronology, site function), research feasibility, or analyst preference (see Frick et al., 2024).\u003c/p\u003e\n\u003cp\u003eFladmark (1982:205) defined microdebitage as particles smaller than 1.0 mm resulting from deliberate lithic reduction. He noted that 1.0 mm represents the smallest particle size visible to the naked eye as a conchoidal flake and used it as a practical boundary between macrodebitage (1.0\u0026ndash;3.0 mm) and microdebitage (\u0026lt;1.0 mm). Many researchers subsequently revised their definitions and size thresholds accordingly. A recurring challenge is that very small flakes may resemble naturally occurring grains, making them difficult to distinguish from anthropogenic specimens (Clark, 1986; Schick, 1986; Vance, 1986).\u003c/p\u003e\n\u003cp\u003eDunnell and Stein (1989) defined microdebitage as small, shattered objects or micro-flakes produced during percussion or pressure flaking. Cornford (1986), meanwhile, emphasised the relationship between artefact size and the ratio of waste (by-products) to tools (products). He advocated for examining artefacts larger than 20 mm, whereas others argue for studying technological products irrespective of size.\u003c/p\u003e\n\u003cp\u003eThe technological approach rooted in the \u003cem\u003echa\u0026icirc;ne op\u0026eacute;ratoire\u003c/em\u003e (Tixier, 1974; Tixier et al., 1980) views lithic production as guided by a cognitive plan translated into a conceptual framework and then operationalised into physical actions (Inizan et al., 1995:15). These actions leave physical traces on detached and worked pieces (e.g., dorsal scars, points of impact), which serve as markers to distinguish different technological stages. These stigmata appear in both macro- and microdebitage. As Soressi and Geneste (2011:337) emphasise, \u0026ldquo;\u003cem\u003eone advantage of the cha\u0026icirc;ne op\u0026eacute;ratoire is its ability to define the \u0026quot;temporality\u0026quot; and \u0026quot;geography\u0026quot; of artefacts within the spatial and temporal context of knapping activities. Each object can be assessed in its processual context through experimentally verified attributes that clarify how prehistoric people fractured stone volumes to produce useable cutting edges\u0026mdash;a kind of volumetric or 3D puzzle\u0026rdquo;.\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThe arbitrariness of size thresholds is illustrated in the work of Frison and Cornford. Both scholars established metric boundaries for analysis while examining links between microdebitage and stages in maintenance procedures. Frison (1968:149\u0026ndash;152) described five types of retouch or sharpening flakes, derived from bifaces and scrapers, into which most unworked flakes could be categorised. Cornford identified longitudinal and transverse sharpening flakes, abbreviated as LSFs and TSFs, respectively. Although both researchers omitted prior micro-/macrodebitage distinctions, they acknowledged the similar technological roles played by flakes across the reduction sequence (Cornford, 1986:341).\u003c/p\u003e\n\u003cp\u003eOther scholars distinguish between \u0026quot;use-flakes\u0026quot; detached during use and \u0026quot;retouch flakes\u0026quot; removed during formal modification (Hayes et al., 2014:78). Chan et al. (2020) further divide flakes into resharpening or reworking flakes, depending on whether reshaping was intended. These classifications suggest diverse intentions, including use, retouch, reshaping, or rejuvenation\u0026mdash;or combinations thereof. Such gestures form part of the artefact\u0026apos;s life history, contributing to maintenance and recycling (Bamforth, 1986).\u003c/p\u003e\n\u003cp\u003eOverlooking microdebitage because of its size omits a vital part of an assemblage\u0026apos;s behavioural signal. Conversely, a technological approach to microdebitage\u0026mdash;rather than rigid size categorisation\u0026mdash;enables the study of stigmata that inform the reconstruction of formation processes and link these to spatial and temporal aspects of dynamic knapping activities and technological decision-making.\u003c/p\u003e\n\u003cp\u003eThis paper highlights edge modification practices through retouch as proxies for maintenance activities, proposing a categorisation system that transcends size-based limits. We adopt Inizan et al.\u0026rsquo;s (1995) definition: \u0026ldquo;The term \u0026apos;retouch\u0026apos; describes removals obtained by percussion or pressure, with the intention of making, finishing, or sharpening tools.\u0026rdquo; Consequently, resharpening is a form of retouch (Iovita, 2011; Morales \u0026amp; Verg\u0026egrave;s, 2014) and is treated as such herein. The term (re)touch implies recurrence on the same edge. Gestures linked to (re)shaping can restore the original form (isometric) or alter it (allometric). (Re)juvenation produces a new edge. All such gestures result in by-products with distinct technological stigmata, enabling their identification.\u003c/p\u003e\n\u003cp\u003eIt is expected that patterned maintenance indices can be derived from formal tools in the archaeological record, whether via allometric (shape-changing) or isometric (shape-preserving) reduction (Iovita, 2014). Accordingly, by-products from these actions can complement formal tool indices (Shott, 1994). While most studies focus on final products (e.g., Dibble, 1995), fewer examine by-products (e.g., Bourguignon, 1996b; Cornford, 1986). Given the predominance of by-products in assemblages, focusing on them offers a statistically robust sample that more accurately reflects technological events. These principles can support a classification system that reconstructs tool life histories through by-product analysis, creating a heuristic, behaviourally informative methodological framework (Iovita, 2011).\u003c/p\u003e\n\u003ch1\u003eStudy Protocol\u003c/h1\u003e\n\u003cp\u003eOur study protocol introduces an inclusive classification system for by-products, focusing on their technological characterisation within the context of edge modification sequences (maintenance activities). Conford (1986) has already identified and described such by-products. He showed these to Harper Kelley and Francois Bordes in Paris, who explained them to him as \u0026quot;\u003cem\u003etrimmings of damaged larger tools.\u0026quot; From this, it is clear that they recognized the marginal retouch to be prior, not subsequent, to the flakes\u0026apos; detachment.\u0026rdquo;\u003c/em\u003e\u003cem\u003e(Cornford, 1986, p. 337)\u003c/em\u003e\u003cem\u003e.\u003c/em\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe categorisation that we propose follows the same principles as flake identification, although we deal within the 5\u0026ndash;20 mm size range. By identifying technical stigmata present on the lithic artefacts, caused by previous operations (i.e., the presence and location of scars on the dorsal face and the point of impact), we can place the artefacts in various steps within the space and the time of the flintknapping activity (\u003cem\u003esensu\u003c/em\u003e Soressi and Geneste, 2011). We classified these items according to the full attribute list that we used for the analysis of macrodebitage. Still, items smaller than 5 mm were counted but not fully studied (Armagan, 2003, p. 103; Dunnell \u0026amp; Stein, 1989; Fladmark, 1982), as we were logistically unable to carry out the low magnification microscopy needed for the identification of the lithic attributes. The development of this analysis protocol was done simultaneously as part of the general lithic attribute analysis (see raw data files in https://github.com/Nora-Arch/Moduleflakescategories), and the data was gathered through an open-access entry software, E5 (https://github.com/surf3s/E5), with a personalised configuration file. All data analysis and plotting were processed with R open-source and JMP statistical software. A research compendium using the rrtools package by (Marwick et al., 2017, 2018), including detailed info on used packages, software versions, and raw and processed data, is available here: https://github.com/Nora-Arch/Moduleflakescategories.\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003eBaseline of the module identification:\u003c/h2\u003e\n\u003cp\u003eFlake edge modification conforms to the same principle of flake formation and fracture mechanics (Cotterell \u0026amp; Kamminga, 1987). The proposed identification protocol aims for a detailed reconstruction of the modes of use of the piece, maintenance, and/or recycling (Bamforth, 1986). In a previous study (Malinsky-Buller et al., 2021), we identified these flakes as \u0026quot;shaping flakes,\u0026quot; an inclusive term used to define all by-products of tool modification. Their identification was based on the presence of retouch scars on their dorsal faces (that were generated by previous retouch or use), and the recognition of the point of intersection between the dorsal face and the platform, where it preserves the \u0026ldquo;parent\u0026rdquo; edge of the tool (Fig. 2). As the use/retouching/reshape/rejuvenation progresses, the technological signature of those removal leaves their mark on byproduct flakes (hereafter named \u0026ldquo;module flake\u0026rdquo;; see below). Most previously identified shaping flakes are \u0026lt; 20 mm in maximal dimension but are distinguished from \u0026ldquo;chips\u0026rdquo; that do not preserve visible retouch scars, or the recognition of a previous tool edge (e.g. (Malinsky-Buller et al., 2021)).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn our system, a module is a set of standardised parts or independent units (by-products) that can be used to construct a more complex structure (i.e., retouched piece). The modules may consist of several distinct yet interrelated units (module flakes), which may be integrated into a holistic reconstruction of the tool\u0026apos;s life history.\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003eCriteria\u003c/h2\u003e\n\u003cp\u003eModule flakes are identified and classified according to three variables: regularity, distribution, and invasiveness of the dorsal scars (maintenance techniques) in relation to the parent blank. The identification of a module flake is based on the following observations: \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n\u003cul class=\"decimal_type\"\u003e\n \u003cli\u003eThe point of intersection between the dorsal and the platform preserves the \u0026ldquo;parent\u0026rdquo; edge of the tool- the previous\u0026nbsp;working edge (Fig.2/3 red dotted line).\u003c/li\u003e\n \u003cli\u003eIdentification of pre-retouch scars on the dorsal face (Fig.2/3). As the use/retouching/reshaping/rejuvenation progresses, the technological signature of those removals leaves its mark on byproduct flakes, i.e., module flakes. \u0026nbsp;\u003c/li\u003e\n \u003cli\u003eIdentification of pre-retouch scar on the parent tool. When a flake is removed during retouching, it bears on his dorsal face the remnant of the scar(s) of the \u0026ldquo;parent\u0026rdquo; blank. This is a remnant of the \u0026ldquo;original\u0026rdquo; blank, and it can be plain or have one or more negatives of previous removals.\u003c/li\u003e\n \u003cli\u003eThe pre-retouch scar direction in M0-M3 flakes is usually perpendicular to the striking platform; M4 may have variable pre-retouched scar directions. \u0026nbsp;\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eInvasiveness attribution. Kuhn (Clarkson, 2002; Eren et al., 2005; Eren \u0026amp; Sampson, 2009; Hiscock \u0026amp; Tabrett, 2010; Kuhn, 1990, 1992) calculate an invasiveness index for tools. Here, we adapt this approach to an invasiveness percentage in relation to the module flake (surface area). The invasiveness percentage is estimated based on the relationship between retouching scars and \u0026ldquo;parent scars\u0026rdquo;. The stratigraphy and superposition of those scars on the dorsal face of a module flake enable the distinction between the different invasiveness ratios, ranging from 1 to 100%, see Fig. 2/3b (See an example Fig. 3, c,d for the M3 and M1).\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eAccording to these observations, we suggest categorizing the by-products of maintenance techniques into five flake modules of edge modification:\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u0026bull;\u0026nbsp; \u0026nbsp;M0: a flake with no previous edge modification. This flake does not show previous retouch scars. The main scar present is the remnant of the previous dorsal scar of the blank from the \u0026ldquo;parent\u0026rdquo; tool (similar to a Kombewa flake). It derives from the primary edge modification. An estimated invasiveness ratio from 0-15% up to the blank ridge (Fig. 2/3).\u003c/p\u003e\n\u003cp\u003e\u0026bull; \u0026nbsp; M1: a flake stemming from the initial edge modification of M0. The initiation of retouch scars is associated with the striking platform of the module flake (i.e., the edge of the \u0026ldquo;parent\u0026rdquo; tool). Those scars stratigraphically overlay the \u0026ldquo;parent\u0026rdquo; tool plain scar(s), being a crucial criterion for their identification. An estimated invasiveness ratio from 16-25% up to the blank ridge. (Fig. 4)\u003c/p\u003e\n\u003cp\u003e\u0026bull; \u0026nbsp; M2: The scar pattern on the flake demonstrates an overlap of scars with remnants of M1, M0, and the plain scar(s) from the \u0026ldquo;parent\u0026rdquo; tool. An estimated invasiveness ratio from 26-50% up to the blank ridge (Fig.5).\u003c/p\u003e\n\u003cp\u003e\u0026bull; \u0026nbsp;M3: The scar pattern on the flake shows an overlap of scars with remnants of the M2, M1, M0, and the plain scar(s) from the \u0026ldquo;parent\u0026rdquo; tool (Fig. 6). The estimated invasiveness ratio is from 51 to 95% up to the blank ridge.\u003c/p\u003e\n\u003cp\u003e\u0026bull; \u0026nbsp; M4: The scar pattern on the flake shows an overlap of scars with remnants of the M3, M2, M1, and M0 and the plain scar(s) from the \u0026ldquo;parent\u0026rdquo; tool. The range of estimated invasiveness ratio values is from 95-100% of retouch invasiveness (Fig. 7). The presence of two ventral faces can be used to distinguish M4 flakes from the rest.\u003c/p\u003e\n\u003cp\u003eBy removing an M4 flake, the knapper creates a new clean edge by reshaping /rejuvenating rather than using a retouching technique (M0-M3). \u0026nbsp;A M4 flake removes part of the tool\u0026apos;s ventral face to either maintain the same shape or modify it (allometric vs. isometric morphologies (Iovita, 2014). The presence of two ventral faces can distinguish M4 flakes. Similar technological pieces were already identified throughout the Palaeolithic in different contexts and regions, either in bifacial or unifacial shaping (Bourguignon, 1996; Lamotte, 1999; Verjux \u0026amp; Rousseau, 1986). For example, Cornford (1986) pointed out that the removal of a Long Sharpening Flake (LSF), which we term an M4 module flake, creates a new edge with the greatest possible length and sharpness on the parent tool. The uniqueness of this type made it more distinguishable when analysing the lithic assemblages in comparison to M0-M3 (Conard \u0026amp; Adler, 1997; De Loecker, 2006; Fonton et al., 1991; Frick et al., 2017; Malinsky-Buller, 2014; Roebroeks et al., 1988, 1997; Zaidner \u0026amp; Grosman, 2015).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTherefore, using module flakes terminology, we aim to deviate from a linear and staged classification sequence, where each step depends on the previous one. Rather, we want to create a dynamic classification (Fig. 8).\u003c/p\u003e\n\u003cp\u003eFig.\u0026nbsp;8 Idealised edge modification scheme illustrating the module progression of lithic tool use, retouch, and discard across edge modification modules (M0\u0026ndash;M4). The vertical ladder represents increasing stages of utilisation of the edge (0\u0026ndash;100%), with curved arrows denoting estimated use-life ranges for each module, and the black arrows representing the jumps either from the initial straight to the end or any other module stop. Modules correspond to characteristic retouch/reshaping/rejuvenation patterns, with M0 reflecting minimal use and M4 representing heavily modified tools nearing saturation point. The model emphasises cyclical behaviours of use, retouching/reshaping, and potential rejuvenation before final discard.\u003c/p\u003e\n\u003cp\u003eWhen we integrate module flake categories into the lithic attribute analysis based on the technological approach, disregarding dimensions enables us to explore questions related to tool maintenance activities. This, in return, provides us with a proxy to quantify the resolution of past hunter-gatherers\u0026apos; mobility and their degree of tool use-life.\u0026nbsp;\u003c/p\u003e"},{"header":"Case studies","content":"\u003cp\u003eThe two case studies presented- Ararat-1 Cave and Kalavan 2 -T1/T2 - show high frequencies of artefacts that would typically fall within the category of chips/debris in a size-oriented analysis. These characteristics are known from other Middle Palaeolithic archaeological sites in the southern Caucasus (e.g., (Gasparyan et al., 2014; Glauberman et al., 2020; Malinsky-Buller et al., 2021; Yeritsyan, 1972). By utilising the novel categorisation of the module flake in both Kalavan 2 and Ararat-1 cave assemblages, we demonstrate the utility of the high-resolution analysis of microdebitage assemblages in terms of provisioning behavioural signatures.\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003eKalavan-2\u0026nbsp;\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eThe open-air site of Kalavan 2 (UTM 40T 3821 m E, 450551 m N, 1640 m asl) is located on the northern slopes of the Areguni Mountains at an elevation of. ca. 8 km north of the shores of Lake Sevan. The chronology of the site is based on fifteen post-infrared infrared stimulated luminescence (pIRIR) samples (Malinsky-Buller et al., 2021) as well as C14 ages from micromammal remains (Rogall et al., submitted). The age range of the main archaeological layers is between 60- 45ka BP (early MIS 3). In total, 2075 lithic artefacts (49 retouched tools and 819 module flakes) from 13 sedimentary units were retrieved. Three main archaeological layers were exposed, Unit 1b in T1 and Units 4 and 7 in T2, with 990 lithic artefacts (38 tools and 275 module flakes). The recovered lithic artefacts were made from seven raw materials: obsidian, basalt, dacite, welded tuff, chert, limestone, and an unidentified metamorphic rock. The frequencies of raw material types vary from unit to unit. Obsidian is the most common material. There are no indications for the first stages of core preparation and reduction, while the debitage consists mainly of flakes \u0026lt; 2 cm. Retouched obsidian pieces present an extremely narrow range of tool types, the majority falling within retouched points or convergent scrapers. The non-obsidian component exhibits more substantial indications of core reduction, including flakes, Kombewa flakes, and Levallois flakes (Ghukasyan et al., 2010; Malinsky-Buller et al., 2021).\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003eArarat-1 Cave\u003c/h2\u003e\n\u003cp\u003eArarat-1 cave (39.851 N, 44.769 E, 1034 m asl) is situated 2 km east of the town of Ararat on the northeast margins of the Ararat Depression. The two archaeological horizons are radiometrically dated by 12 post-infrared infrared stimulated luminescence (pIRIR) samples and a single C14 sample to a range between 50- 35ka BP (Oikonomou et al., 2025; Sherriff et al., 2024). The lithic assemblage of Ararat-1 is mainly comprised of chert and obsidian, with mafic lava, chalcedony, and quartzite represented by a few isolated pieces. The total assemblage amounts to 1770 artefacts, of which 37 are retouched tools and 680 (38% of the assemblage) are module flakes smaller than 10 mm (108 obsidian and 26 in chert). There are few indications of obsidian blank production at the site (very few cores or primary elements); in this raw material, it is mainly the last stage of the reduction sequence that is represented in the assemblage: use, maintenance, and rejuvenation (as defined/described by (Geneste, 1985; Nelson, 1991). Typologically, the retouched pieces are dominated by several types of scrapers and retouched flakes. The techno-typological composition of the chert component attests to initial stages of the reduction sequence, including cores and cortical elements (Nora et al., submitted).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAdding to the regular attribute analysis, we incorporated a proposed module for micro flake categorisation (M0–M4), we analysed key lithic attributes such as technological length, technological width, thickness, number of dorsal scars, and raw material distribution across different modules to show the difference between them and what they represent behaviourally.\u003c/p\u003e"},{"header":"Results","content":"\u003ch2\u003eLithic Attribute Analysis across Modules\u003c/h2\u003e\n\u003ch3\u003eTechnological Length/Width/Thickness:\u003c/h3\u003e\n\u003cp\u003eKalavan 2 T1 (Fig.9) and T2 (Fig.10) display distinct patterns in flake morphology across edge modification modules, offering insights into potential differences in maintenance practices. In Kalavan 2 T1, module flakes are generally smaller and thinner, particularly in M0 through M2. For example, M0 averages 6.3 mm in length, 6.6 mm in width, and 1.1 mm in thickness, among the lowest values across all three assemblages. Similarly, the length and width values of M1 flakes in T1 average 6.7 mm and 6.5 mm, respectively. Thickness remains low and consistent in early modules. In contrast, later modules in T1, particularly M3 and M4, show a noticeable increase in size. M4, for instance, reaches an average length of 10.5 mm and a width of 12.1 mm, with thickness increasing to 2.5 mm. These values can suggest a functional shift in edge modification.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eKalavan 2 T2, by comparison, consistently exhibits larger and thicker flakes across nearly all modules. M0 through M2 flakes show apparent dimensional increases relative to T1, with M2 flakes averaging 11.1 mm in length and 10.9 mm in width, accompanied by a thickness of 2.0 mm. This pattern continues through M3 and M4, the latter reaching flakes averaging 15.8 mm in width and 3.6 mm in thickness.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn contrast, the Ararat-1 Cave (Fig. 11) assemblage presents a more variable profile, generally falling between Kalavan T1 and T2 in terms of size, but with less pronounced extremes. M0 through M2 exhibit smaller flakes than those in T2, with M0 averaging 8.3 mm in length and 8.0 mm in width, and M2 reaching 8.5 mm and 7.3 mm, respectively. Thickness remains consistently low in these modules, ranging from 1.0 to 1.5 mm. However, a noticeable shift occurs in M3 and M4. Module 4 averages 10.2 mm in length and 10.3 mm in width, with a thickness of 2.6 mm, closely paralleling the later module flakes observed in Kalavan T1. Overall, the Ararat-1 cave profile suggests a more moderate reduction strategy, with less emphasis on heavily retouched or robust forms. This may indicate different site functions, more expedient tool production, or variable raw material constraints compared to the Kalavan localities.\u003c/p\u003e\n\u003ch3\u003eNumber of Dorsal Scars and the approximate Area of module flakes\u003c/h3\u003e\n\u003cp\u003eTo explore the relationship between the estimated invasiveness ratio, we applied a multivariate analysis of variance (MANOVA) with two dependent variables: the number of dorsal scars and the approximate area (technological length x technological width). The independent variable, module, represents the categorical factor. MANOVA was chosen as a global test to assess whether flake size varied across module categories while accounting for potential correlation between the two dependent variables. This approach minimises the risk of Type I error associated with multiple comparisons and enables the detection of multivariate structures that are not visible through univariate methods alone. Following significant MANOVA results, we conducted univariate ANOVAs and Tukey HSD post-hoc tests to determine which traits contributed to group differentiation and which module contrasts were statistically significant. The combination of MANOVA and follow-up ANOVAs offers both global and trait-specific insights, allowing for a nuanced understanding of how maintenance practices could have shaped artefact form.\u003c/p\u003e\n\u003cp\u003eFor Kalavan 2 T1, the MANOVA revealed a significant multivariate effect of module (Wilks\u0026rsquo; \u0026Lambda; = 0.56994, \u003cem\u003eF\u003c/em\u003e(8, 1638) = 66.46, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001), indicating that the overall combination of dorsal scars and area differs across modules (for more info, please check https://github.com/Nora-Arch/Moduleflakescategories). Follow-up univariate ANOVAs confirmed that both dorsal scar count (\u003cem\u003eF\u003c/em\u003e(4, 820) = 150.8, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001) and approximate area (\u003cem\u003eF\u003c/em\u003e(4, 820) = 20.04, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001) varied significantly with module. Tukey HSD tests showed significant differences between all module levels for dorsal scars, with the most substantial increases occurring between M0 to M4 (\u0026Delta; = 8.02 scars) and M2 to M4 (\u0026Delta; = 3.31 scars). Even comparisons between adjacent modules (e.g., M2 vs. 3, \u0026Delta; = 2.06 scars) remained statistically robust, supporting a graded intensification in edge modification. For the approximate area, significant differences emerged primarily between M0 and M3 and M4 (\u0026Delta; = 45\u0026ndash;79 mm\u0026sup2;), indicating that size increases become apparent only with a high intensity ratio of retouch.\u003c/p\u003e\n\u003cp\u003eAs for Kalavan 2 T2, MANOVA also yielded a strong multivariate effect (Wilks\u0026rsquo; \u0026Lambda; = 0.55014, \u003cem\u003eF\u003c/em\u003e(8, 520) = 22.64, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001), suggesting consistent morphological structuring by module. Univariate ANOVAs showed significant differences for both dorsal scars number (\u003cem\u003eF\u003c/em\u003e(4, 261) = 51.91, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001) and approximate area (\u003cem\u003eF\u003c/em\u003e(4, 261) = 9.39, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001). Tukey comparisons again revealed a strong progressive increase in the number of dorsal scars, particularly between M0 and M4 (\u0026Delta; = 7.33 scars), while M1 and M2 differed by approximately 3.3 scars. Area differences were significant between M1 and M2 through 4 (\u0026Delta; \u0026asymp; 74.4 mm\u0026sup2;), indicating that larger flakes are associated with more scars (and therefore more intense retouch), although variability in M0 limited statistical resolution in that category.\u003c/p\u003e\n\u003cp\u003eMANOVA results for the Ararat-1 assemblage were also significant (Wilks\u0026rsquo; \u0026Lambda; = 0.56849, \u003cem\u003eF\u003c/em\u003e(8, 1342) = 54.74, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001), confirming that retouch intensity is reflected in morphological variation at this site as well. Univariate ANOVAs revealed significant effects for dorsal scars (\u003cem\u003eF\u003c/em\u003e(4, 672) = 120.4, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001) and approximate area (\u003cem\u003eF\u003c/em\u003e(4, 672) = 17.9, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001). Tukey tests indicated that M4 had significantly more dorsal scars than M0\u0026ndash;2 (\u0026Delta; = 5\u0026ndash;7 scars), although the difference between M3 and M4 was not statistically significant, suggesting a plateau at high retouch levels. For the approximate area, only M3 differed significantly from M0 (\u0026Delta; = 34.8 mm\u0026sup2;, \u003cem\u003ep\u003c/em\u003e = 0.024), while other contrasts failed to reach significance. This suggests that, at Ararat-1, increases in artefact size with retouch intensity are less consistent than those seen in dorsal scars.\u003c/p\u003e\n\u003cp\u003eAcross all three assemblages, MANOVA results provide strong evidence that the ratio of invasiveness is associated with significant variation in flake size. Dorsal scars numbers increase progressively and significantly with module categories, showing consistent, stepwise intensification in all plots (Figs. 12\u0026ndash;14). Approximate area shows a weaker but complementary pattern, with significant increases in higher module categories in Kalavan, and less consistent changes in Ararat-1 Cave. These data support a two-tiered pattern of morphological change: scar accumulation occurs reliably across modules, while size increases are more context-dependent.\u003c/p\u003e\n\u003ch3\u003eModule Flakes Frequencies and Raw Material Distance\u003c/h3\u003e\n\u003cp\u003eModule flakes can be further explored when considering the frequencies of module categories with the raw material units, which enables us to decipher the fragmentation of the reduction sequence, specifically when looking at the phases of tool maintenance and use (Turq et al., 2013).\u003c/p\u003e\n\u003cp\u003eIn the Kalavan 2 T1 assemblage (Fig. 15), the Gutansar (47 km) obsidian appears frequently in Modules 0 to 4. Hatis (50 km) has no representation in Module 0, while it is more prominent in Modules 1, 2, and 3, with few M4 pieces. \u0026nbsp;The outliers in this plot are the Chikiani source (140 km), represented by a single M4 flake, Chikiani 2b on M1, and Kars Digor (158 km), which is represented by a single M2 occurrence. Tsaghkunyats sources (60 km) are present in all modules. The same trend is visible for non-obsidian raw material tuff. Module flakes made on flint and chert (green) are relatively rare, appearing in small amounts in Modules 1 and 2.\u003c/p\u003e\n\u003cp\u003eAt\u003cem\u003e\u0026nbsp;\u003c/em\u003eKalavan 2 T2 (Fig.16), obsidian from Gutansar (47 km) and Hatis (50 km) are a significant raw material, and its presence is particularly prominent in Module 3. Tsaghkunyats sources (60 km) have close to zero representation on M0. In modules 1 to 3, the frequencies are up to 2%, followed by a single example in M4. Among the non-obsidian flake modules, tuff stands out. It is particularly dominant in M0, accounting for 43.8% of the material in this module, and continues to be represented across all modules, including M4. This consistent presence of tuff from M0 to M4 suggests its sustained utility throughout various stages of retouching and modification. Basalt and dacite are only minimally represented, primarily in the earlier modules. As for flint, this raw material has no representation in the first three modules, but it appears as isolated instances in the M3 and M4.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eArarat-1 Cave (Fig. 17) exhibits some unique trends. Obsidian source Geghasar (35-40 km), a prominent raw material in this site, is highly concentrated in Modules 1 and 2. Gutansar (60 km) also shows a similar pattern, but its peak occurs in Module 3. Interestingly, Hatis (50 km) and its variants (Hatis alpha, beta, and gamma) appear consistently across all module categories but are less frequent in the earlier stage (Module 0). Items from the Arteni sources (100 \u0026ndash; 130 km) appear in Modules 1 to 3, with sporadic appearances in modules 0 and 4. Non-obsidian raw materials, such as chert, basalt, and mafic lava, are rare in this assemblage and only show\u0026nbsp;marginal appearances in Modules 0, 1, and 2.\u003c/p\u003e\n\u003ch3\u003eDiscussion: Behaviour beyond module flakes\u0026nbsp;\u003c/h3\u003e\n\u003cp\u003eThe study of retouched pieces plays a key role in the systematics of the Palaeolithic record. As best exemplified by the nomenclature developed by Bordes\u0026nbsp;(Bordes, 1953, 1951, 1961), pre-determined morphotypes shaped by retouching arguably conveyed cultural significance. In parallel, the perspective originally put forward by Frison (1969), later Jelinek (1976), and culminated in the works of Dibble and his students\u0026nbsp;(Dibble, 1985, 1995)\u0026nbsp;suggested that those same types were stages along a continuum of size change, that crossed retouching, reshaping, and rejuvenation stages. At the same time, Lewis Binford\u0026nbsp;(Binford, 1973, 1979)\u0026nbsp;positioned the theoretical foundation of curation, by stating that those maintenance techniques were part of a bigger concept, which\u0026nbsp;emphasises maximisation of raw material utility through maintenance and recycling linked with procurement strategies. Yet despite the richness of these perspectives, their empirical application remains debated, beginning with the most fundamental issue, the basic unit of analysis. For Shott (1989, 1996) and Dibble (1995), this unit was the retouched tool. For Shott (1989:24, 1996), the definition entailed \u003cem\u003e\u0026ldquo;The degree of curation, defined as the degree of use or utility extracted, expressed as a relationship between the tool\u0026apos;s initial maximum utility and how much of that utility is realised before discard.\u0026rdquo;\u0026nbsp;\u003c/em\u003eIn contrast, (Barton \u0026amp; Riel-Salvatore, 2014; Clark \u0026amp; Barton, 2017)\u0026nbsp;proposed an assemblage level approach (WABI) to study curated pieces, contrasting them with expedient technologies. Despite selecting different analytical units, those methodologies and empirical toolkits aim for similar \u0026ldquo;higher-level\u0026rdquo; questions, such as the reconstruction of the mobility patterns of the group and the extent of occupation within a locality, as well as resource exploitation intensity.\u0026nbsp;Usually, analysts choose to focus exclusively on studies that link identifiable artefacts (commonly macrodebitage artefacts, in most cases attributed to tools) with non-local raw materials that could be brought from localities that exceed the daily exploitation territory (DET) to propose mobility models\u0026nbsp;(Bailey \u0026amp; Davidson, 1983; Kuhn, 1992, 1995; Vita-Finzi et al., 1970). Nevertheless, other studies\u0026nbsp;(Ekshtain et al., 2017; F\u0026eacute;blot‐Augustins, 2009)\u0026nbsp;challenge these assumptions by re-evaluating the analytical unit by looking at different proxies.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eBuilding on this, our paper shifts the unit of analysis to the by-products of the maintenance techniques, disregarding their size, as a new proxy not only to identify tools that had been removed from the site (ghost tools), but also as a new way of looking at the intensity of edge modification techniques such as retouching, reshaping and rejuvenation. Our concept of curation concurs with that of Shott (1994; 1996), in that we regard it as a behaviour designed to extend the use-life of an artefact\u0026nbsp;(Shott, 1994). In this paper, our perspective enables us to fine tune the ways to document and quantify the degrees of extension of the retouched tool use-life.\u0026nbsp;Our proposed nomenclature and the suggested protocol for categorising what we refer to as \u0026quot;module flakes\u0026quot; are based on fundamental principles of fracture mechanics to all artefacts regardless of their size, emphasising the identification of key diagnostic features in lithic studies, and acknowledges that tools are manufactured through processes of retouching and reshaping and rejuvenation. From a unified analytical foundation, we apply the same technological attribute analysis to microdebitage artefacts as we do to macrodebitage, serving the same analytical purposes.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWe establish here a module categorisation of flakes to overcome the typological diversification in the literature. \u0026ldquo;Module flakes\u0026rdquo; have been identified through several periods and contexts on lithic assemblages from the Palaeolithic to the New World in the Late Prehistoric context (Bourguignon, 1996a, 2001; Cornford, 1986a; Frison, 1968; Hays, 1992; Touz\u0026eacute;, 2018). Synonym nomenclatures are available, either as the proposed five categories of retouch and sharpening flakes by Frison (1968) or the attribution of the LSF, TSF flakes of Conford (1986), or the several phases of decortication and post-decortication suggested by (J\u0026eacute;r\u0026eacute;mie \u0026amp; Vacher, 1992). For the latest phase of rejuvenation of retouched tools (our \u0026ldquo;M4\u0026rdquo;), a wide range of synonyms were applied, all depending on the directionality of the removals, such as Tranchet Blow, or Lateral Tranchet Blow (LTB) at Nesher Ramla (Pr\u0026eacute;vost et al., 2022), or the chanfren (Burney, 1967); for more morphological differences see (Frick et al., 2024). We argue that most of these different flake typologies that are technological flakes related to edge modification are targeting specific human gestures of retouch/reshape and rejuvenation applied to the tool (Tixier et al., 1980). Thus, that can be perceived as a unit instead of several typologies. Despite abundant terminologies and contextual examples, a unified technological framework for understanding these flakes has yet to be fully articulated.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe progressive modification of a blank through retouching and reshaping is the main proxy used here to assess the use-life of a lithic tool. Archaeologists have made significant efforts to develop new quantitative methods based on stone tools to understand lithic reduction, more specifically, retouch intensity (Clarkson, 2002; Davis \u0026amp; Shea, 1998; Dibble, 1987, 1995; Eren et al., 2005; Eren \u0026amp; Prendergast, 2008; Eren \u0026amp; Sampson, 2009; Hiscock \u0026amp; Attenbrow, 2003; Hiscock \u0026amp; Clarkson, 2009; Hiscock \u0026amp; Tabrett, 2010; Kuhn, 1990; Marwick, 2008; Morales et al., 2015; Morales \u0026amp; Verg\u0026egrave;s, 2014). These quantitative methods can sometimes estimate the hypothetical initial volume of the tool blank, allowing us to predict differences in the variation of the tool along the two scaling directions (allometric versus isometric reduction trajectories (Iovita, 2009, 2010, 2011). Our approach looks at the waste material that results from this reduction, the by-product of the maintenance techniques.\u003c/p\u003e\n\u003cp\u003eOur approach stems from the principle of the \u003cem\u003echa\u0026icirc;ne op\u0026eacute;ratoire,\u003c/em\u003e that each artefact can be placed within the space and the time of the flintknapping activity,\u003cem\u003e\u0026nbsp;\u003c/em\u003e(Soressi \u0026amp; Geneste, 2011, p. 337).\u0026nbsp;In this sense, we aim to show that the module flake categories, once incorporated into the reduction sequence, can significantly help unpack the complexities of decision-making regarding tool use and maintenance. We believe that we can now better test retouch intensity in lithic assemblages by using the quantitative approach based on tools (a static approach) and linking them to the frequencies of categories of module flakes that are representative of the dynamic behaviour of retouching, reshaping and rejuvenating. Furthermore, by binding the classification of raw material units with the module flake categories, we can attempt to quantify raw material curation in terms of the degree of extracted utility per raw material relative to the distance from their sources.\u003c/p\u003e\n\u003cp\u003eThe results of this study provide strong support for the interpretive validity of the module classification system as a framework for understanding edge modification in lithic assemblages. Across the datasets of all three assemblages, the combination of MANOVA and univariate analyses revealed consistent and statistically significant differentiation among module categories, indicating that each step in the classification reflects a meaningful behavioural or technological threshold in edge modification. The MANOVA results confirmed that when dorsal scar count and approximate area are considered together, retouch intensity, as categorised by module flakes, exerts a significant influence on artefact size. This multivariate structure supports the use of module flakes as a holistic categorisation of edge modification. Follow-up univariate ANOVAs revealed that the number of dorsal scars increases progressively and significantly across modules, with strong differentiation even between adjacent modules. This consistent pattern across all assemblages demonstrates that modules capture a graded process of edge modification, from primary edge modification (M0) through various degrees of retouch (M1\u0026ndash;3) to advanced rejuvenation (M4). However, we stress that those modules still do not always reflect a linear progression, as different \u0026ldquo;stages\u0026rdquo; of retouch may occur simultaneously on different sections of a tool\u0026rsquo;s edges (Fig. 8). Approximate area followed a similar but more selective pattern. Increases in size were statistically significant primarily at higher modules, particularly M3 and M4. This suggests that while scar accumulation reflects a continuous trajectory of edge modification, increases in flake size are likely linked to intensifying maintenance and are more variable and context-dependent. The plateau in scar count observed between M3 and M4 in some cases may indicate a saturation point in edge modification, or a final phase in a tool\u0026rsquo;s use-life before discard or reshaping and transport, as evident by the \u0026ldquo;ghost tools\u0026rdquo;.\u003c/p\u003e\n\u003cp\u003eWhen dealing with provisioning strategies, module classification can distinguish the initial approach towards the blank. M0 (primary edge modification) is conceptually different from the other modules, as shown through the Tukey test. The absence or low frequencies of M0 flakes suggest that the initial edge modification occurred off-site. This could indicate a curated provisioning strategy where pre-modified blanks or partially retouched tools were transported to the locality. In contrast, high frequencies of M0 guide us to the perspective of unmodified blanks at the locale, which can be represented by on-site blank production (when compared to raw material units, core presence, unmodified blanks, and core trimming elements) or the transportation of blanks into the locality. The presence of M1, M2, and M3 indicates that blanks are at a stage of secondary edge modification, implying the material either entered the site already retouched or was retouched on-site. For example, in Kalavan 2 T2, the high frequency of M3 made on Gutansar, Hatis, and Tsaghkunyats obsidian suggests that these pieces arrived at the site in a partially worked state; more advanced retouching (51\u0026ndash;95% of edge modification) took place on-spot. The same pattern can be observed in Ararat-1 Cave on the module flakes from the Arteni obsidian sources. This segmented representation of the module flakes could indicate that the specific lithic raw materials became more important over time, potentially reflecting shifts in raw material availability and exploitation. The high frequency of M3 suggests raw material economisation with a significant focus on these sources during more advanced stages of edge maintenance. The interpretation of Module 4 flakes points us to a tipping point in tool maintenance when retouching leads to reshaping/rejuvenation of the edge (Fig. 8). The frequent occurrence of M4 flakes, especially in Gutansar and Hatis obsidian at Kalavan 2 T1 and T2, suggests that tools were continuously maintained on-site. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003e(Shott, 1989, p. 24, 1996) suggested that curation is a concept directly related to the tool\u0026apos;s initial maximum utility and how much of that utility is realised before it is discarded. Following this perspective, we can attempt to measure curation through the module flake categories, taking into consideration the raw material units (RMU) as our additional analytical unit (Odell, 2004, pp. 93\u0026ndash;95; Roebroeks et al., 1988). The pattern we observed either from the raw material strategies and the maintenance strategies seen by the module flakes are consistent with curation strategies where the goal is to prolong the life of the tools through repeated rejuvenation, ensuring that high-quality materials like obsidian are utilised to their fullest potential (Odell 1996; Nora et al, 2025a). The single representation of M4 for the most distant raw materials, such as Chikiani obsidian in Kalavan 2 T1, suggests that some tools entered the site and were taken away after reshaping/rejuvenation. This evidence of \u0026ldquo;ghost tools\u0026rdquo; is represented only by the by-products that indicate maintenance activity at the locale. This pattern of reshaping tools to create new, sharp edges and then removing them from the site is indicative of tool provisioning for continued use elsewhere, supporting a mobile strategy where tools were maintained at key locations and then transported to other areas of activity, sensu (Bousman, 1993).\u003c/p\u003e\n\u003cp\u003eThe categorisation we put forward here aims to encompass the various typologies used with relation to the concept of curation and, consequently, the maintenance of lithic tools (see Fig. 18), emphasising that their behavioural signature is more significant than typological attribution (see discussion in (Bar-Yosef \u0026amp; Van Peer, 2009) for the use of technological traits vs. typological). The suggested categorisation reflects a behaviourally meaningful model that corresponds to measurable changes in artefact form. For example, a single M4 flake removal would create a clean edge, contrasting with a wavy edge shaped by a succession of M0/M1/M2/M3 removals. Thus, interpreting the module flakes within a technological perspective provides a nuanced reconstruction of past decision-making. The suggested categorisation can be divided into the first edge modification (M0), the retouching/reshaping (M1/M2/M3), and finally, the reshaping /rejuvenation phase (M4) (Fig. 17). \u0026nbsp;Those techniques, although conceptually closer, differ in their gestures and intentionality (Forestier, 1993; Inizan et al., 1995).\u003c/p\u003e\n\u003cp\u003eThe different module flakes enable us to reconstruct at what stage blanks were transported to a given locality, either as unretouched blanks or as retouched items, and if the latter, to what degree. The progressive increase in dorsal scar numbers and the selective increase in artefact size confirm that the module framework captures distinct categories within the situational decision-making during the life-history of lithic tools. This reinforces the model\u0026apos;s utility for comparative and inferential studies of lithic maintenance strategies within archaeological sites.\u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003eApplying our proposed module flake categorisation (M0\u0026ndash;M4) to the late Middle Palaeolithic artefact assemblages of Kalavan 2 T1 and T2 and Ararat-1 Cave offers new insights into the fluidity of the technological processes underlying tool maintenance. The systematic classification of by-products, regardless of size, enabled us to reconstruct detailed edge modification sequences and tool use-life. The identification of M0 indicates the first approach to an unretouched blank, while a high frequency of M1, M2, and M3 module flakes revealed extensive initial edge modifications and ongoing retouching activities. These patterns reflect adaptive strategies for tool maintenance, which can also be used to explore retouch intensity. M4 flakes, representing advanced stages of tool use, suggest prolonged use and retouch intensity of tools that lead to a blunt or unstructured edge, as noted by Frison (\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e1968\u003c/span\u003e). This dynamic module category goes beyond a linear sequence by considering edge modifications, overlapping, and an iterative nature. Rather than a linear progression of retouch, we offer a \u0026ldquo;snake and ladder\u0026rdquo; approach, emphasising the fluidity of decision-making (Fig.\u0026nbsp;\u003cspan refid=\"Fig15\" class=\"InternalRef\"\u003e8\u003c/span\u003e). The analysis underscores the critical role of large and small waste flakes in understanding past technological behaviours, stressing that by-products are integral to a holistic reconstruction of prehistoric tool use-life, rather than merely waste material. By applying this novel framework, we have expanded the interpretative potential of lithic assemblages, shedding light on the often-overlooked smaller fractions. Those pieces hold significant information about the intensity and nature of tool use, tool maintenance practices, and mobility strategies. This approach offers a more nuanced understanding of tool curation, contributing to broader discussions on Palaeolithic technological organisation and putting the decision-making processes of tool maintenance at the focal point of inquiry.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eDeclarations\u003c/h2\u003e \u003cp\u003e \u003cstrong\u003eCompeting Interests:\u003c/strong\u003e \u003cp\u003eThe authors have no relevant financial or non-financial interests to disclose.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eFunding\u003c/h2\u003e \u003cp\u003eDeclaration:\u003c/p\u003e \u003cp\u003eFieldwork, laboratory analyses and syntheses were funded by the TransCause Project, Investigating Pleistocene population dynamics, in the Southern Caucasus (ERC \u0026ndash; Starting Grant Agreement No. 948015), the Fritz Thyssen Foundation grant awarded for the project \u0026ldquo;Pleistocene Hunter-Gatherer Lifeways and Population Dynamics in the Ararat (paleo-lake) Depression, Armenia\u0026rdquo;, the Gerda Henkel Stiftung grant (n. AZ 10_V_17 and n. AZ 23/F/19) and the Leakey Foundation (awarded to AMB). Further support was provided by Gfoeller Renaissance Foundation (USA) and \u0026ldquo;Areni-1 Cave\u0026rdquo; Consortium (\u0026ldquo;Areni-1 Cave\u0026rdquo; Scientific Research Foundation (Armenia) and the Institute of Archaeology and Ethnography of the National Academy of Sciences of the Republic of Armenia. Grant support was provided to Nora by the Bina and Moshe Stekelis Fund for Prehistoric Archaeology, the Ruth Amiran Fund for archaeological research in Eretz-Israel, and the Hebrew University International PhD Talent Scholarship. The pXRF instrument used for obsidian artefact sourcing was funded thanks to generous support from Yale\u0026rsquo;s Offices of the Vice Provost for Research, Dean of Science, and Dean of Social Science and its Council on Archaeological Studies and the Department of Anthropology.\u003c/p\u003e \u003cp\u003eAbstract:\u003c/p\u003e \u003cp\u003eThe study of retouching, reshaping, and rejuvenation in lithic technology has traditionally focused on finished tools, overlooking the by-products of these processes, particularly microdebitage. This has led to an incomplete understanding of the dynamic behaviours associated with tool maintenance and a lack of crucial information about prehistoric technological strategies. Here we address this knowledge gap. Specifically, we introduce a classification system for lithic by-products resulting from retouching, reshaping, and rejuvenation techniques, categorising them into five modules (M0 through M4) based on lithic technological analysis. This methodology integrates the \u003cem\u003echa\u0026icirc;ne op\u0026eacute;ratoire\u003c/em\u003e approach to analyse flakes without size thresholds. To demonstrate our approach, we apply it, coupled with raw material sourcing, to lithic assemblages from two Middle Palaeolithic sites in Armenia, Kalavan 2 and Ararat-1 Cave. This enables a precise reconstruction of tool use-life and, in turn, the mobility strategies of Pleistocene hunter-gatherers. Our findings demonstrate that microdebitage (by-products) can contribute to a holistic view of decision-making, revealing patterns in tool maintenance and raw material provisioning. The module system provides insights into the production of \u0026lsquo;ghost tools,\u0026rsquo; which are not present in the archaeological record, as well as curation behaviours and economic decisions regarding raw materials that were previously difficult to discern. By shifting the focus from finished artefacts to by-products, this framework enhances our ability to interpret lithic assemblages and understand the adaptive strategies of prehistoric hunter-gatherers.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eD.N. and A.M.B. are the main contributors to the conception and design of the work; the acquisition, analysis, and interpretation of data; and the drafting and critical revision of the manuscript for important intellectual content. D.N. conducted all lithic analysis, processed the data, prepared the figures, and created the open-access repository. A.M.B. reviewed and participated in the lithic analysis and contributed to writing and revising multiple drafts. B.G. and A.P. contributed to the acquisition of data through fieldwork and laboratory coordination in Armenia, ensured access to collections, and revised and approved the final manuscript. E.F. performed the pXRF analysis, contributed to data interpretation and manuscript revision, and approved the final version for publication. All authors approved the version to be published and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003eThe authors would like to thank the Director of the Institute of Archaeology and Ethnography of the National Academy of Sciences of the Republic of Armenia, Dr. Arsen Bobokhyan, for his support regarding the permits for fieldwork investigations, sample transportation, and availability of the facilities. Nora expresses his gratitude to Prof. E. Hovers, from the Institute of Archaeology, the Hebrew University of Jerusalem, for the supervision and discussion around the paper's subject, as well as for reading too many drafts. Special thanks to N. Paz and the Human-Environment Dynamics Laboratory of the Institute of Archaeology, the Hebrew University of Jerusalem, for providing laboratory facilities and support during research. Moreover, Nora would like to thank all the volunteers who participated in the fieldwork and Doctor Marion Prevost for the brainstorming over figures design, to Sergey Alon for the drawing of tiny lithics and to A. Jallon for the good talks over the PhD. Fieldwork, laboratory analyses and syntheses were funded by the TransCause Project, Investigating Pleistocene population dynamics, in the Southern Caucasus (ERC \u0026ndash; Starting Grant Agreement No. 948015), the Fritz Thyssen Foundation grant awarded for the project \u0026ldquo;Pleistocene Hunter-Gatherer Lifeways and Population Dynamics in the Ararat (paleo-lake) Depression, Armenia\u0026rdquo;, the Gerda Henkel Stiftung grant (n. AZ 10_V_17 and n. AZ 23/F/19) and the Leakey Foundation (awarded to AMB). Further support was provided by Gfoeller Renaissance Foundation (USA) and \u0026ldquo;Areni-1 Cave\u0026rdquo; Consortium (\u0026ldquo;Areni-1 Cave\u0026rdquo; Scientific Research Foundation (Armenia) and the Institute of Archaeology and Ethnography of the National Academy of Sciences of the Republic of Armenia. Grant support was provided to Nora by the Bina and Moshe Stekelis Fund for Prehistoric Archaeology, the Ruth Amiran Fund for archaeological research in Eretz-Israel, and the Hebrew University International PhD Talent Scholarship. The pXRF instrument used for obsidian artefact sourcing was funded thanks to generous support from Yale\u0026rsquo;s Offices of the Vice Provost for Research, Dean of Science, and Dean of Social Science and its Council on Archaeological Studies and the Department of Anthropology.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eThe data that supports this paper is available at https://github.com/Nora-Arch/Moduleflakescategories.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eArmagan, T. (2003). Small lithic Debris Analysis, with Appendix A Alphabetic Listing of Attribute Definitions Used for the Tor Faraj Small Lithic Debris Study ; Appendix В Small Lithic Debris Analysis Results Tables. \u003cem\u003eNeanderthals in the Levant. Behavioral Organization and the Beginnings of Human Modernity\u003c/em\u003e, 86:106; 271:275.\u003c/li\u003e\n\u003cli\u003eBailey, G. N., \u0026amp; Davidson, I. (1983). Site exploitation territories and topography: Two case studies from palaeolithic spain. \u003cem\u003eJournal of Archaeological Science\u003c/em\u003e, \u003cem\u003e10\u003c/em\u003e(2), 87\u0026ndash;115. https://doi.org/10.1016/0305-4403(83)90044-4\u003c/li\u003e\n\u003cli\u003eBamforth, D. B. (1986). Technological Efficiency and Tool Curation. \u003cem\u003eAmerican Antiquity\u003c/em\u003e, \u003cem\u003e51\u003c/em\u003e(1), 38\u0026ndash;50. https://doi.org/10.2307/280392\u003c/li\u003e\n\u003cli\u003eBarton, C. M., \u0026amp; Riel-Salvatore, J. (2014). The formation of lithic assemblages. \u003cem\u003eJournal of Archaeological Science\u003c/em\u003e, \u003cem\u003e46\u003c/em\u003e, 334\u0026ndash;352. https://doi.org/10.1016/j.jas.2014.03.031\u003c/li\u003e\n\u003cli\u003eBar-Yosef, O., \u0026amp; Van Peer, P. (2009). \u003cem\u003eThe Chaine Operatoire Approach in Middle Paleolithic Archaeology\u003c/em\u003e. http://nrs.harvard.edu/urn-3:HUL.InstRepos:2960197\u003c/li\u003e\n\u003cli\u003eBinford, L. R. (1973). Interassemblage variability-the Mousterian and the \u0026lsquo;functional\u0026rsquo; argument, ed. Colin Renfrew. In C. Renfrew (Ed.), \u003cem\u003eThe Explanation of Culture Change\u003c/em\u003e (pp. 227\u0026ndash;254). Duckworth.\u003c/li\u003e\n\u003cli\u003eBinford, L. R. (1979a). Organization and Formation Processes: Looking at Curated Technologies. \u003cem\u003eJournal of Anthropological Research\u003c/em\u003e, \u003cem\u003e35\u003c/em\u003e(3), 255\u0026ndash;273. https://doi.org/10.1086/jar.35.3.3629902\u003c/li\u003e\n\u003cli\u003eBinford, L. R. (1979b). Organization and formation processes: Looking at curated technologies. \u003cem\u003eJournal of Anthropological Research\u003c/em\u003e, \u003cem\u003e35\u003c/em\u003e(3), 255--273.\u003c/li\u003e\n\u003cli\u003eBinford, L. R., \u0026amp; Binford, S. R. (1966). A Preliminary Analysis of Functional Variability in the Mousterian of Levallois Facies. \u003cem\u003eAmerican Anthropologist\u003c/em\u003e, \u003cem\u003e68\u003c/em\u003e(2), 238\u0026ndash;295.\u003c/li\u003e\n\u003cli\u003eBordes. (1953). Essai de Classification des industries \u0026lsquo;moust\u0026eacute;riennes\u0026rsquo;. \u003cem\u003eBulletin de La Soci\u0026eacute;t\u0026eacute; Pr\u0026eacute;historique de France\u003c/em\u003e, \u003cem\u003e50\u003c/em\u003e(7/8), 457\u0026ndash;466.\u003c/li\u003e\n\u003cli\u003eBordes, F. (1951). Le Complexe Mousterien:Mousteriens, Levalloisien et Tayacien. \u003cem\u003eL\u0026rsquo;Anthropologie\u003c/em\u003e, \u003cem\u003e55\u003c/em\u003e, 1\u0026ndash;23.\u003c/li\u003e\n\u003cli\u003eBordes, F. (1961). Mousterian cultures in France. \u003cem\u003eScience\u003c/em\u003e, \u003cem\u003e134\u003c/em\u003e(3482), 803\u0026ndash;810.\u003c/li\u003e\n\u003cli\u003eBordes, F. (1969). \u003cem\u003eTypologie du pal\u0026eacute;olithique ancien et moyen\u003c/em\u003e. CNRS.\u003c/li\u003e\n\u003cli\u003eBordes, F., \u0026amp; Sonneville‐Bordes, D. de. (1970). The significance of variability in Palaeolithic assemblages. \u003cem\u003eWorld Archaeology\u003c/em\u003e. https://www.tandfonline.com/doi/abs/10.1080/00438243.1970.9979464\u003c/li\u003e\n\u003cli\u003eBourguignon, L. (1996a). \u003cem\u003eLa conception de d\u0026eacute;bitage Quina. In Reduction Processe (\u0026laquo;Cha\u0026icirc;nes op\u0026eacute;ratoire\u0026raquo;) for the European Mousterian.\u003c/em\u003e\u003c/li\u003e\n\u003cli\u003eBourguignon, L. (1996b). Un moust\u0026eacute;rien tardif sur le site d\u0026rsquo;Ummel Tlel (Bassin d\u0026rsquo;El khowm, Syrie) ?Exemples des niveaux II base \u0026rsquo;et III2a\u0026rsquo;. In E. Carbonell \u0026amp; M. Vaquero (Eds.), \u003cem\u003eThe last Neandertals,the first anatomically Modern Humans: A Tale About Human Diversity; Cultural Change and Human Evolution\u003c/em\u003e (pp. 317\u0026ndash;336). Universitat Rovira i Virgili.\u003c/li\u003e\n\u003cli\u003eBourguignon, L. (2001). Apports de l\u0026rsquo;exp\u0026eacute;rimentation et de l\u0026rsquo;analyse techno-morpho- fonctionnelle \u0026agrave; la reconnaissance du processus d\u0026rsquo;am\u0026eacute;nagement de la retouche Quina. In \u003cem\u003ePr\u0026eacute;histoire et approche exp\u0026eacute;rimentale\u003c/em\u003e (Vol. 5).\u003c/li\u003e\n\u003cli\u003eBousman, C. B. (1993). Hunter-Gatherer Adaptations, Economic Risk and Tool Design. \u003cem\u003eLithic Technology\u003c/em\u003e, \u003cem\u003e18\u003c/em\u003e(1\u0026ndash;2), 59\u0026ndash;86. https://doi.org/10.1080/01977261.1993.11720897\u003c/li\u003e\n\u003cli\u003eBr\u0026eacute;zillon, M. (1968). \u003cem\u003eLa d\u0026eacute;nomination des objets de pierre taill\u0026eacute;e. Mat\u0026eacute;riaux pour un vocabulaire des pr\u0026eacute;historiens de langue fran\u0026ccedil;aise\u003c/em\u003e. https://www.persee.fr/doc/galip_0072-0100_1968_sup_4_1\u003c/li\u003e\n\u003cli\u003eBurney, M. (1967). \u003cem\u003eMcBurney: The Haua Fteah (Cyrenaica) and the stone... - Google Scholar\u003c/em\u003e. https://scholar.google.com/scholar_lookup?title=The%20Haua%20Fteah%20in%20Cyrenaica%20and%20the%20Stone%20Age%20of%20the%20South-East%20Mediterranean\u0026amp;publication_year=1967\u0026amp;author=C.B.M.%20McBurney\u003c/li\u003e\n\u003cli\u003eClark, G. A., \u0026amp; Barton, C. M. (2017). Lithics, landscapes \u0026amp; la Longue-dur\u0026eacute;e \u0026ndash; Curation \u0026amp; expediency as expressions of forager mobility. \u003cem\u003eQuaternary International\u003c/em\u003e, \u003cem\u003e450\u003c/em\u003e, 137\u0026ndash;149. https://doi.org/10.1016/j.quaint.2016.08.002\u003c/li\u003e\n\u003cli\u003eClarkson, C. (2002). An Index of Invasiveness for the Measurement of Unifacial and Bifacial Retouch: A Theoretical, Experimental and Archaeological Verification. \u003cem\u003eJournal of Archaeological Science\u003c/em\u003e, \u003cem\u003e29\u003c/em\u003e(1), 65\u0026ndash;75. https://doi.org/10.1006/jasc.2001.0702\u003c/li\u003e\n\u003cli\u003eConard, N. J., \u0026amp; Adler, D. S. (1997). Lithic Reduction and Hominid Behavior in the Middle Paleolithic of the Rhineland. \u003cem\u003eJournal of Anthropological Research\u003c/em\u003e. https://doi.org/10.1086/jar.53.2.3631275\u003c/li\u003e\n\u003cli\u003eCornford, J. M. (1986a). Specialised resharpening techniques and evidence of handedness. In J. M. Callow, P., Cornford (Ed.), \u003cem\u003eLa Cotte de St. Brelade, Jersey. Excavations by C.B.M. McBurney 1961\u0026ndash;1978.\u003c/em\u003e (pp. 337--351.). Geo Books.\u003c/li\u003e\n\u003cli\u003eCornford, J. M. (1986b). Specialized resharpening techniques and evidence of handedness. \u003cem\u003eLa Cotte de St. Brelade\u003c/em\u003e, \u003cem\u003e1978\u003c/em\u003e, 337\u0026ndash;351.\u003c/li\u003e\n\u003cli\u003eCotterell, B., \u0026amp; Kamminga, J. (1987). The Formation of Flakes. \u003cem\u003eAmerican Antiquity\u003c/em\u003e, \u003cem\u003e52\u003c/em\u003e(4), 675\u0026ndash;708. https://doi.org/10.2307/281378\u003c/li\u003e\n\u003cli\u003eDavidson, I., Noble, W., Armstrong, D. F., Black, L. T., Calvin, W. H., Davis, W., Falk, D., Foster, M. L., Graves, P., Halverson, J., \u0026amp; Hewes, G. W. (1989). The Archaeology of Perception: Traces of Depiction and Language [and Comments and Reply]. \u003cem\u003eCurrent Anthropology\u003c/em\u003e. https://doi.org/10.1086/203723\u003c/li\u003e\n\u003cli\u003eDavis, Z. J., \u0026amp; Shea, J. J. (1998). Quantifying Lithic Curation: An Experimental Test of Dibble and Pelcin\u0026rsquo;s Original Flake-Tool Mass Predictor. \u003cem\u003eJournal of Archaeological Science\u003c/em\u003e, \u003cem\u003e25\u003c/em\u003e(7), 603\u0026ndash;610. https://doi.org/10.1006/jasc.1997.0255\u003c/li\u003e\n\u003cli\u003eDe Loecker, D. (2006). \u003cem\u003eBeyond the site. The Saalian archaeological record at Maastricht-Belv\u0026eacute;d\u0026egrave;re (the Netherlands). Analecta Praehistorica Leidensia 35/36\u003c/em\u003e. Leiden University Press, Leiden.\u003c/li\u003e\n\u003cli\u003eDeb\u0026eacute;nath, A., \u0026amp; Dibble, H. L. (1994). \u003cem\u003eHandbook of Paleolithic Typology: Lower and Middle Paleolithic of Europe\u003c/em\u003e. University of Pennsylvania Museum of Archaeology and Anthropology. https://doi.org/10.2307/j.ctt14btgq9\u003c/li\u003e\n\u003cli\u003eDibble, H. L. (1985). Raw material variability in Levallois flake manufacture. \u003cem\u003eCurrent Anthropology (Chicago)\u003c/em\u003e, \u003cem\u003e26\u003c/em\u003e, 391\u0026ndash;393. https://doi.org/10/bm57q4\u003c/li\u003e\n\u003cli\u003eDibble, H. L. (1987). The Interpretation of Middle Paleolithic Scraper Morphology. \u003cem\u003eAmerican Antiquity\u003c/em\u003e, \u003cem\u003e52\u003c/em\u003e(1), 109\u0026ndash;117. https://doi.org/10.2307/281062\u003c/li\u003e\n\u003cli\u003eDibble, H. L. (1995). Middle paleolithic scraper reduction: Background, clarification, and review of the evidence to date. \u003cem\u003eJournal of Archaeological Method and Theory\u003c/em\u003e, \u003cem\u003e2\u003c/em\u003e(4), 299\u0026ndash;368. https://doi.org/10.1007/BF02229003\u003c/li\u003e\n\u003cli\u003eDunnell, R. C., \u0026amp; Stein, J. K. (1989). Theoretical issues in the interpretation of microartifacts. \u003cem\u003eGeoarchaeology\u003c/em\u003e, \u003cem\u003e4\u003c/em\u003e(1), 31\u0026ndash;41. https://doi.org/10.1002/gea.3340040103\u003c/li\u003e\n\u003cli\u003eEkshtain, R., Ilani, S., Segal, I., \u0026amp; Hovers, E. (2017). Local and Nonlocal Procurement of Raw Material in Amud Cave, Israel: The Complex Mobility of Late Middle Paleolithic Groups. \u003cem\u003eGeoarchaeology\u003c/em\u003e, \u003cem\u003e32\u003c/em\u003e(2), 189\u0026ndash;214. https://doi.org/10.1002/gea.21585\u003c/li\u003e\n\u003cli\u003eEren, M. I., Dominguez-Rodrigo, M., Kuhn, S. L., Adler, D. S., Le, I., \u0026amp; Bar-Yosef, O. (2005). Defining and measuring reduction in unifacial stone tools. \u003cem\u003eJournal of Archaeological Science\u003c/em\u003e, \u003cem\u003e32\u003c/em\u003e(8), 1190\u0026ndash;1201. https://doi.org/10.1016/j.jas.2005.03.003\u003c/li\u003e\n\u003cli\u003eEren, M. I., \u0026amp; Prendergast, M. E. (2008). Comparing and synthesizing unifacial stone tool reduction indices. \u003cem\u003eLithic Technology\u003c/em\u003e, 49\u0026ndash;85.\u003c/li\u003e\n\u003cli\u003eEren, M. I., \u0026amp; Sampson, C. G. (2009). Kuhn\u0026rsquo;s Geometric Index of Unifacial Stone Tool Reduction (GIUR): Does it measure missing flake mass? \u003cem\u003eJournal of Archaeological Science\u003c/em\u003e, \u003cem\u003e36\u003c/em\u003e(6), 1243\u0026ndash;1247. https://doi.org/10.1016/j.jas.2009.01.011\u003c/li\u003e\n\u003cli\u003eF\u0026eacute;blot‐Augustins, J. (2009). Revisiting European Upper Paleolithic Raw Material Transfers: The Demise of the Cultural Ecological Paradigm? In \u003cem\u003eLithic Materials and Paleolithic Societies\u003c/em\u003e (pp. 25\u0026ndash;46). https://doi.org/10.1002/9781444311976.ch3\u003c/li\u003e\n\u003cli\u003eFladmark, K. R. (1982). Microdebitage analysis: Initial considerations. \u003cem\u003eJournal of Archaeological Science\u003c/em\u003e, \u003cem\u003e9\u003c/em\u003e(2), 205\u0026ndash;220. https://doi.org/10.1016/0305-4403(82)90050-4\u003c/li\u003e\n\u003cli\u003eFonton, M., Lhomme, V., \u0026amp; Christensen, M. (1991). \u003cem\u003eUn cas de \u0026laquo; r\u0026eacute;duction \u0026raquo; et de \u0026laquo; transformation \u0026raquo; d\u0026rsquo;outil au Pal\u0026eacute;olithique moyen. Un racloir d\u0026eacute;jet\u0026eacute; de la grotte de Coustal \u0026agrave; Noailles (Corr\u0026egrave;ze)\u003c/em\u003e. https://doi.org/10.3406/pal.1991.1035\u003c/li\u003e\n\u003cli\u003eForestier, H. (1993). \u003cem\u003eLe Clactonien: Mise en application d\u0026rsquo;une nouvelle m\u0026eacute;thode de d\u0026eacute;bitage s\u0026rsquo;inscrivant dans la variabilit\u0026eacute; des syst\u0026egrave;mes de production lithique du Pal\u0026eacute;olithique ancien\u003c/em\u003e. https://doi.org/10.3406/pal.1993.1104\u003c/li\u003e\n\u003cli\u003eFrick, J. A., Herkert, K., Hoyer, C. T., \u0026amp; Floss, H. (2017). The performance of tranchet blows at the Late Middle Paleolithic site of Grotte de la Verpilli\u0026egrave;re I (Sa\u0026ocirc;ne-et-Loire, France). In \u003cem\u003ePLoS ONE\u003c/em\u003e (Vol. 12, Issue 11). https://doi.org/10.1371/journal.pone.0188990\u003c/li\u003e\n\u003cli\u003eFrick, J. A., Herkert, K., Hoyer, C. T., \u0026amp; Floss, H. (2024). Using the Technological Feature of Tranchet Blow on Keilmesser as a Connecting Element Between Similar Middle Paleolithic Assemblages from the C\u0026ocirc;te Chalonnaise (Sa\u0026ocirc;ne-et-Loire, France). In H. Koehler (Ed.), \u003cem\u003eThe Rhine During the Middle Paleolithic: Boundary or Corridor?\u003c/em\u003e (1st ed., pp. 287\u0026ndash;314). Kerns Verlag. https://doi.org/10.51315/9783935751353.013\u003c/li\u003e\n\u003cli\u003eFrison, G. (1968). A Functional Analysis of Certain Chipped Stone Tools. \u003cem\u003eAmerican Antiquity\u003c/em\u003e, \u003cem\u003e33\u003c/em\u003e(2), 149\u0026ndash;155. https://doi.org/10.2307/278516\u003c/li\u003e\n\u003cli\u003eGasparyan, B., Egeland, C., Adler, D., Pinhasi, R., Glauberman, P., \u0026amp; Haydosyan, H. (2014). The Middle Paleolithic Occupation of Armenia: Summarizing Old and New Data. In \u003cem\u003eStone Age of Armenia: A Guide-book to the Stone Age Archaeology in the Republic of Armenia\u003c/em\u003e (pp. 65\u0026ndash;105).\u003c/li\u003e\n\u003cli\u003eGeneste, J.-M. (1985). \u003cem\u003eAnalyse lithique d\u0026rsquo;industries moust\u0026eacute;riennes du P\u0026eacute;rigord: Une approche technologique du comportement des groupes humains au Pal\u0026eacute;olithique moyen\u003c/em\u003e. University of Bordeaux I Bordeaux. https://pascal-francis.inist.fr/vibad/index.php?action=getRecordDetail\u0026amp;idt=12142529\u003c/li\u003e\n\u003cli\u003eGhukasyan, R., Colonge, D., Nahapetyan, S., Ollivier, V., Gasparyan, B., Monchot, H., \u0026amp; Chataigner, Ch. (2010). KALAVAN-2 (NORTH OF LAKE SEVAN, ARMENIA): A NEW LATE MIDDLE PALEOLITHIC SITE IN THE LESSER CAUCASUS. \u003cem\u003eArchaeology, Ethnology and Anthropology of Eurasia\u003c/em\u003e, \u003cem\u003e38\u003c/em\u003e(4), 39\u0026ndash;51. https://doi.org/10.1016/j.aeae.2011.02.003\u003c/li\u003e\n\u003cli\u003eGlauberman, P., Gasparyan, B., Wilkinson, K., Frahm, E., Nahapetyan, S., Arakelyan, D., Raczynski-Henk, Y., Haydosyan, H., \u0026amp; Adler, D. S. (2020). Late Middle Paleolithic Technological Organization and Behavior at the Open-Air Site of Barozh 12 (Armenia). \u003cem\u003eJournal of Paleolithic Archaeology\u003c/em\u003e, \u003cem\u003e3\u003c/em\u003e(4), 1095\u0026ndash;1148. https://doi.org/10.1007/s41982-020-00071-4\u003c/li\u003e\n\u003cli\u003eHays, M. A. (1992). \u003cem\u003eA Functional Analysis of the Lithic Material From Burrone Scierra I (Calabria, Italy)\u003c/em\u003e.\u003c/li\u003e\n\u003cli\u003eHiscock, P., \u0026amp; Attenbrow, V. (2003). Early Australian implement variation: A reduction model. \u003cem\u003eJournal of Archaeological Science\u003c/em\u003e, \u003cem\u003e30\u003c/em\u003e(2), 239\u0026ndash;249. https://doi.org/10.1006/jasc.2002.0830\u003c/li\u003e\n\u003cli\u003eHiscock, P., \u0026amp; Clarkson, C. (2009). The reality of reduction experiments and the GIUR: Reply to Eren and Sampson. \u003cem\u003eJournal of Archaeological Science\u003c/em\u003e, \u003cem\u003e36\u003c/em\u003e(7), 1576\u0026ndash;1581. https://doi.org/10.1016/j.jas.2009.03.019\u003c/li\u003e\n\u003cli\u003eHiscock, P., \u0026amp; Tabrett, A. (2010). Generalization, inference and the quantification of lithic reduction. \u003cem\u003eWorld Archaeology\u003c/em\u003e, \u003cem\u003e42\u003c/em\u003e(4), 545\u0026ndash;561. https://doi.org/10.1080/00438243.2010.517669\u003c/li\u003e\n\u003cli\u003eInizan, M., Reduron, M., Roche, H., \u0026amp; Tixier, J. (1995). \u003cem\u003eTechnologie de la pierre taill\u0026eacute;e\u003c/em\u003e (Vol. 4).\u003c/li\u003e\n\u003cli\u003eIovita, R. (2009). Ontogenetic scaling and lithic systematics: Method and application. \u003cem\u003eJournal of Archaeological Science\u003c/em\u003e, \u003cem\u003e36\u003c/em\u003e(7), 1447\u0026ndash;1457. https://doi.org/10.1016/j.jas.2009.02.008\u003c/li\u003e\n\u003cli\u003eIovita, R. (2010). Comparing Stone Tool Resharpening Trajectories with the Aid of Elliptical Fourier Analysis. In S. Lycett \u0026amp; P. Chauhan (Eds.), \u003cem\u003eNew Perspectives on Old Stones: Analytical Approaches to Paleolithic Technologies\u003c/em\u003e (pp. 235\u0026ndash;253). Springer. https://doi.org/10.1007/978-1-4419-6861-6_10\u003c/li\u003e\n\u003cli\u003eIovita, R. (2011). Shape Variation in Aterian Tanged Tools and the Origins of Projectile Technology: A Morphometric Perspective on Stone Tool Function. \u003cem\u003ePLOS ONE\u003c/em\u003e, \u003cem\u003e6\u003c/em\u003e(12), e29029. https://doi.org/10.1371/journal.pone.0029029\u003c/li\u003e\n\u003cli\u003eIovita, R. (2014). The role of edge angle maintenance in explaining technological variation in the production of Late Middle Paleolithic bifacial and unifacial tools. \u003cem\u003eQuaternary International\u003c/em\u003e, \u003cem\u003e350\u003c/em\u003e, 105\u0026ndash;115. https://doi.org/10.1016/j.quaint.2014.08.032\u003c/li\u003e\n\u003cli\u003eJelinek, A. J. (1976). Form, Function and Style in Lithic Analysis. In Cultural Change and Continuity: Essays in Honor of James Bennett Griffin. \u003cem\u003eNew York:Academic Press.\u003c/em\u003e, 19\u0026ndash;33.\u003c/li\u003e\n\u003cli\u003eJ\u0026eacute;r\u0026eacute;mie, S., \u0026amp; Vacher, S. (1992). Le Hoabinhien en Tha\u0026iuml;lande: Un exemple d\u0026rsquo;approche exp\u0026eacute;rimentale. \u003cem\u003eBulletin de l\u0026rsquo;\u0026Eacute;cole Fran\u0026ccedil;aise d\u0026rsquo;Extr\u0026ecirc;me-Orient\u003c/em\u003e, \u003cem\u003e79\u003c/em\u003e(1), 173\u0026ndash;209.\u003c/li\u003e\n\u003cli\u003eKuhn, S. L. (1990). A geometric index of reduction for unifacial stone tools. \u003cem\u003eJournal of Archaeological Science\u003c/em\u003e, \u003cem\u003e17\u003c/em\u003e(5), 583\u0026ndash;593. https://doi.org/10.1016/0305-4403(90)90038-7\u003c/li\u003e\n\u003cli\u003eKuhn, S. L. (1992). Blank Form and Reduction as Determinants of Mousterian Scraper Morphology. \u003cem\u003eAmerican Antiquity\u003c/em\u003e, \u003cem\u003e57\u003c/em\u003e(1), 115\u0026ndash;128. https://doi.org/10.2307/2694838\u003c/li\u003e\n\u003cli\u003eKuhn, S. L. (1995). \u003cem\u003eMousterian Lithic Technology: An Ecological Perspective\u003c/em\u003e. Princeton University Press. https://www.jstor.org/stable/j.ctt7zv59w\u003c/li\u003e\n\u003cli\u003eLamotte, A. (1999). L\u0026rsquo;apport des remontages dans la compr\u0026eacute;hension des m\u0026eacute;thodes de d\u0026eacute;bitage et de fa\u0026ccedil;onnage des gisements acheul\u0026eacute;ens de la Somme: Les exemples de la Ferme de l\u0026rsquo;\u0026Eacute;pinette et de l\u0026rsquo;\u0026Eacute;pinette \u0026agrave; Cagny (Somme, France). \u003cem\u003eBulletin de La Soci\u0026eacute;t\u0026eacute; Pr\u0026eacute;historique Fran\u0026ccedil;aise\u003c/em\u003e, \u003cem\u003e96\u003c/em\u003e(2), 117\u0026ndash;131.\u003c/li\u003e\n\u003cli\u003eMalinsky-Buller, A. (2014). Contextualizing Curational Strategies at the Late Lower Paleolithic Site of Holon, Israel. \u003cem\u003ePaleoAnthropology\u003c/em\u003e. https://doi.org/10.4207/PA.2014.ART87\u003c/li\u003e\n\u003cli\u003eMalinsky-Buller, A., Glauberman, P., Ollivier, V., Lauer, T., Timms, R., Frahm, E., Brittingham, A., Triller, B., Kindler, L., Knul, M. V., Krakovsky, M., Joannin, S., Hren, M. T., Bellier, O., Clark, A. A., Blockley, S. P. E., Arakelyan, D., Marreiros, J., Paixaco, E., \u0026hellip; Gasparyan, B. (2021). Short-term occupations at high elevation during the Middle Paleolithic at Kalavan 2 (Republic of Armenia). \u003cem\u003ePLOS ONE\u003c/em\u003e, \u003cem\u003e16\u003c/em\u003e(2), e0245700. https://doi.org/10.1371/journal.pone.0245700\u003c/li\u003e\n\u003cli\u003eMarwick, B. (2008). What attributes are important for the measurement of assemblage reduction intensity? Results from an experimental stone artefact assemblage with relevance to the Hoabinhian of mainland Southeast Asia. \u003cem\u003eJournal of Archaeological Science\u003c/em\u003e, \u003cem\u003e35\u003c/em\u003e(5), 1189\u0026ndash;1200. https://doi.org/10.1016/j.jas.2007.08.007\u003c/li\u003e\n\u003cli\u003eMarwick, B., Boettiger, C., \u0026amp; Mullen, L. (2017). \u003cem\u003ePackaging data analytical work reproducibly using R (and friends)\u003c/em\u003e (No. e3192v1). PeerJ Inc. https://doi.org/10.7287/peerj.preprints.3192v1\u003c/li\u003e\n\u003cli\u003eMarwick, B., Boettiger, C., \u0026amp; Mullen, L. (2018). Packaging Data Analytical Work Reproducibly Using R (and Friends). \u003cem\u003eThe American Statistician\u003c/em\u003e, \u003cem\u003e72\u003c/em\u003e(1), 80\u0026ndash;88. https://doi.org/10.1080/00031305.2017.1375986\u003c/li\u003e\n\u003cli\u003eMonnier, by G. (2006). The Lower/Middle Paleolithic Periodization in Western Europe. \u003cem\u003eCurrent Anthropology\u003c/em\u003e. https://doi.org/10.1086/506280\u003c/li\u003e\n\u003cli\u003eMorales, J. I., Lorenzo, C., \u0026amp; Verg\u0026egrave;s, J. M. (2015). Measuring Retouch Intensity in Lithic Tools: A New Proposal Using 3D Scan Data. \u003cem\u003eJournal of Archaeological Method and Theory\u003c/em\u003e, \u003cem\u003e22\u003c/em\u003e(2), 543\u0026ndash;558. https://doi.org/10.1007/s10816-013-9189-0\u003c/li\u003e\n\u003cli\u003eMorales, J. I., \u0026amp; Verg\u0026egrave;s, J. M. (2014). Technological behaviors in Paleolithic foragers. Testing the role of resharpening in the assemblage organization. \u003cem\u003eJournal of Archaeological Science\u003c/em\u003e, \u003cem\u003e49\u003c/em\u003e, 302\u0026ndash;316. https://doi.org/10.1016/j.jas.2014.05.025\u003c/li\u003e\n\u003cli\u003eNelson, M. C. (1991). The Study of Technological Organization. \u003cem\u003eArchaeological Method and Theory\u003c/em\u003e, \u003cem\u003e3\u003c/em\u003e, 57\u0026ndash;100.\u003c/li\u003e\n\u003cli\u003eOdell, G. H. (2004). \u003cem\u003eLithic Analysis\u003c/em\u003e. Springer US. https://doi.org/10.1007/978-1-4419-9009-9\u003c/li\u003e\n\u003cli\u003eOikonomou, I. A. K., Karambaglidis, T., Fenn, K., Gur-Arieh, S., Nora, D., S\u0026aacute;nchez-Romero, L., Rogall, D., Vetesse, D., Gasparian, B., Petrosyan, A., \u0026amp; Malinsky-Buller, A. (2025). \u003cem\u003eUnravelling the Formation Processes and Depositional Histories of the Middle Palaeolithic Ararat-1 Cave, Armenia: A Multiscalar and Multiproxy Geoarchaeological Approach.\u003c/em\u003e\u003c/li\u003e\n\u003cli\u003ePr\u0026eacute;vost, M., Centi, L., \u0026amp; Zaidner, Y. (2022). The use of the lateral tranchet blow technique at Nesher Ramla (Israel): A new cultural marker in the Levantine Middle Paleolithic? \u003cem\u003eQuaternary International\u003c/em\u003e, \u003cem\u003e624\u003c/em\u003e, 128\u0026ndash;147. https://doi.org/10.1016/j.quaint.2020.11.008\u003c/li\u003e\n\u003cli\u003eRoebroeks, W., Kolen, J., Van Poecke, M., \u0026amp; Van Gijn, A. (1997). \u003cem\u003e\u0026laquo;Site J\u0026raquo;: An early Weichselian (Middle Palaeolithic) flint scatter at Maastricht-Belvedere, The Netherlands\u003c/em\u003e. https://doi.org/10.3406/pal.1997.1231\u003c/li\u003e\n\u003cli\u003eRoebroeks, W., VAN GIJN, A., VAN DE VELDE, P., \u0026amp; ARPS, C. E. S. (1988). From Find Scatters to Early Hominid Behaviour: A Study of Middle Palaeolithic Riverside Settlements at Maastricht-Belv\u0026eacute;d\u0026egrave;re (The Netherlands). \u003cem\u003eFrom Find Scatters to Early Hominid Behaviour : A Study of Middle Palaeolithic Riverside Settlements at Maastricht-Belv\u0026eacute;d\u0026egrave;re (The Netherlands)\u003c/em\u003e, \u003cem\u003e21\u003c/em\u003e.\u003c/li\u003e\n\u003cli\u003eSherriff, J. E., Petrosyan, A., Rogall, D., Nora, D., Frahm, E., Lauer, T., Karambaglidis, T., Knul, M. V., Vettese, D., Arakelyan, D., Gur-Arieh, S., Vidal-Matutano, P., Morales, J., Fewlass, H., Blockley, S. P. E., Timms, R., Adigyozalyan, A., Haydosyan, H., Glauberman, P., \u0026hellip; Malinsky-Buller, A. (2024). Palaeoenvironmental and chronological context of hominin occupations of the Armenian Highlands during MIS 3: Evidence from Ararat-1 cave. \u003cem\u003eQuaternary Science Advances\u003c/em\u003e, \u003cem\u003e13\u003c/em\u003e, 100122. https://doi.org/10.1016/j.qsa.2023.100122\u003c/li\u003e\n\u003cli\u003eShott, M. (1989). On Tool-Class Use Lives and the Formation of Archaeological Assemblages. \u003cem\u003eAmerican Antiquity\u003c/em\u003e, \u003cem\u003e54\u003c/em\u003e(1), 9\u0026ndash;30. https://doi.org/10.2307/281329\u003c/li\u003e\n\u003cli\u003eShott, M. (1994). Size and form in the analysis of flake debris: Review and recent approaches. \u003cem\u003eJournal of Archaeological Method and Theory\u003c/em\u003e, \u003cem\u003e1\u003c/em\u003e(1), 69\u0026ndash;110. https://doi.org/10.1007/BF02229424\u003c/li\u003e\n\u003cli\u003eShott, M. (1996). An Exegesis of the Curation Concept. \u003cem\u003eJournal of Anthropological Research\u003c/em\u003e, \u003cem\u003e52\u003c/em\u003e(3), 259\u0026ndash;280. https://doi.org/10.1086/jar.52.3.3630085\u003c/li\u003e\n\u003cli\u003eSoressi, M., \u0026amp; Geneste, J.-M. (2011). The History and Efficacy of the Cha\u0026icirc;ne Op\u0026eacute;ratoire Approach to Lithic Analysis: Studying Techniques to Reveal Past Societies in an Evolutionary Perspective. \u003cem\u003ePaleoAnthropology\u003c/em\u003e, \u003cem\u003e2011\u003c/em\u003e, 334\u0026ndash;350.\u003c/li\u003e\n\u003cli\u003eTixier, J., Inizan, M.-L., \u0026amp; Roche, H. (1980). \u003cem\u003eTerminologie et technologie\u003c/em\u003e. Cercle de recherches et d\u0026rsquo;\u0026eacute;tudes pr\u0026eacute;historiques.\u003c/li\u003e\n\u003cli\u003eTouz\u0026eacute;, O. (2018). Aux pr\u0026eacute;mices du Gravettien dans le Nord-Ouest europ\u0026eacute;en: \u0026Eacute;tude de la production des pointes lithiques \u0026agrave; Maisi\u0026egrave;res-Canal (province de Hainaut, Belgique). \u003cem\u003eBulletin de La Soci\u0026eacute;t\u0026eacute; Pr\u0026eacute;historique Fran\u0026ccedil;aise\u003c/em\u003e, \u003cem\u003e115\u003c/em\u003e(3), 455\u0026ndash;495. https://doi.org/10.3406/bspf.2018.14920\u003c/li\u003e\n\u003cli\u003eTrigger, B. G. (2006). \u003cem\u003eA history of archaeological thought\u003c/em\u003e (2nd ed). Cambridge university press.\u003c/li\u003e\n\u003cli\u003eTurq, A., Roebroeks, W., Bourguignon, L., \u0026amp; Faivre, J.-P. (2013). The fragmented character of Middle Palaeolithic stone tool technology. \u003cem\u003eJournal of Human Evolution\u003c/em\u003e, \u003cem\u003e65\u003c/em\u003e(5), 641\u0026ndash;655. https://doi.org/10.1016/j.jhevol.2013.07.014\u003c/li\u003e\n\u003cli\u003eVerjux, C., \u0026amp; Rousseau, D.-D. (1986). La retouche Quina: Une mise au point. \u003cem\u003eBulletin de La Soci\u0026eacute;t\u0026eacute; Pr\u0026eacute;historique Fran\u0026ccedil;aise\u003c/em\u003e, \u003cem\u003e11\u0026ndash;12\u003c/em\u003e, 404\u0026ndash;415.\u003c/li\u003e\n\u003cli\u003eVita-Finzi, C., Higgs, E. S., Sturdy, D., Harriss, J., Legge, A. J., \u0026amp; Tippett, H. (1970). Prehistoric Economy in the Mount Carmel Area of Palestine: Site Catchment Analysis. \u003cem\u003eProceedings of the Prehistoric Society\u003c/em\u003e, \u003cem\u003e36\u003c/em\u003e, 1\u0026ndash;37. https://doi.org/10.1017/S0079497X00013074\u003c/li\u003e\n\u003cli\u003eWargo, M. (2009). \u003cem\u003eThe Bordes-Binford debate: Transatlantic interpretive traditions in Paleolithic archaeology - ProQuest\u003c/em\u003e. https://www.proquest.com/openview/a32bf9dadb387d8882e0a7268994373c/1?pq-origsite=gscholar\u0026amp;cbl=18750\u003c/li\u003e\n\u003cli\u003eYeritsyan, B. G. (1972). \u003cem\u003e\u0026lsquo;Некоторые особенности намеренного рассечения орудий мустьерской эпохи (по материалам Ереванской пещерной стоянки)\u0026rsquo; [Some Features of Intentional Truncation of Mousterian Tools (Based on the Materials of Yerevan Cave Site)]\u003c/em\u003e. 53\u0026ndash;60.\u003c/li\u003e\n\u003cli\u003eZaidner, Y., \u0026amp; Grosman, L. (2015). Middle Paleolithic sidescrapers were resharped or recycled? A view from Nesher Ramla, Israel. \u003cem\u003eQuaternary International\u003c/em\u003e, \u003cem\u003e361\u003c/em\u003e, 178\u0026ndash;187. https://doi.org/10.1016/j.quaint.2014.11.037\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"journal-of-archaeological-method-and-theory","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"jarm","sideBox":"Learn more about [Journal of Archaeological Method and Theory](http://link.springer.com/journal/10816)","snPcode":"10816","submissionUrl":"https://submission.nature.com/new-submission/10816/3","title":"Journal of Archaeological Method and Theory","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Lithic technology, Tool maintenance, Curation, Microdebitage, Module flakes, Technological organisation","lastPublishedDoi":"10.21203/rs.3.rs-6673907/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6673907/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"The study of retouching, reshaping, and rejuvenation in lithic technology has traditionally focused on finished tools, overlooking the by-products of these processes, particularly microdebitage. This has led to an incomplete understanding of the dynamic behaviours associated with tool maintenance and a lack of crucial information about prehistoric technological strategies. Here we address this knowledge gap. Specifically, we introduce a classification system for lithic by-products resulting from retouching, reshaping, and rejuvenation techniques, categorising them into five modules (M0 through M4) based on lithic technological analysis. This methodology integrates the chaîne opératoire approach to analyse flakes without size thresholds. To demonstrate our approach, we apply it, coupled with raw material sourcing, to lithic assemblages from two Middle Palaeolithic sites in Armenia, Kalavan 2 and Ararat-1 Cave. This enables a precise reconstruction of tool use-life and, in turn, the mobility strategies of Pleistocene hunter-gatherers. Our findings demonstrate that microdebitage (by-products) can contribute to a holistic view of decision-making, revealing patterns in tool maintenance and raw material provisioning. The module system provides insights into the production of ‘ghost tools,’ which are not present in the archaeological record, as well as curation behaviours and economic decisions regarding raw materials that were previously difficult to discern. By shifting the focus from finished artefacts to by-products, this framework enhances our ability to interpret lithic assemblages and understand the adaptive strategies of prehistoric hunter-gatherers.","manuscriptTitle":"Snakes and Ladders: A technological approach to tool maintenance by-products using module flake categories","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-05-21 16:14:44","doi":"10.21203/rs.3.rs-6673907/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-06-30T08:11:30+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-06-26T00:37:12+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-05-29T09:35:14+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"285852868112188782946044894558794786232","date":"2025-05-22T00:21:43+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"28521465111933849436044687529425390760","date":"2025-05-21T13:38:56+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-05-21T11:53:23+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-05-21T06:59:23+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-05-20T07:15:21+00:00","index":"","fulltext":""},{"type":"submitted","content":"Journal of Archaeological Method and Theory","date":"2025-05-15T14:58:07+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"journal-of-archaeological-method-and-theory","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"jarm","sideBox":"Learn more about [Journal of Archaeological Method and Theory](http://link.springer.com/journal/10816)","snPcode":"10816","submissionUrl":"https://submission.nature.com/new-submission/10816/3","title":"Journal of Archaeological Method and Theory","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"52cb33f2-fcbe-48fa-b18e-fccd24fe2b62","owner":[],"postedDate":"May 21st, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-12-01T15:59:25+00:00","versionOfRecord":{"articleIdentity":"rs-6673907","link":"https://doi.org/10.1007/s10816-025-09754-0","journal":{"identity":"journal-of-archaeological-method-and-theory","isVorOnly":false,"title":"Journal of Archaeological Method and Theory"},"publishedOn":"2025-11-24 15:56:58","publishedOnDateReadable":"November 24th, 2025"},"versionCreatedAt":"2025-05-21 16:14:44","video":"","vorDoi":"10.1007/s10816-025-09754-0","vorDoiUrl":"https://doi.org/10.1007/s10816-025-09754-0","workflowStages":[]},"version":"v1","identity":"rs-6673907","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6673907","identity":"rs-6673907","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}