Tales of Multifunctionality: a Systematic Quantitative Literature Review of Boomerangs Used as Retouchers in Australian Aboriginal Cultures

Boomerangs are among the most recognisable elements of Australian Aboriginal technology. In the popular mindset, the prevailing image of these wooden artefacts is that of thrown implements that return to the thrower, principally used for hunting animals. However, boomerangs have a deep multipurpose role in Indigenous societies, with just a few examples of their known functions, including fighting, digging, and making music (i.e., “clap sticks”). Recently, yet another function for boomerangs has been proposed (Martellotta et al., 2021): the functional modification of lithic tools (i.e., retouching)—a form of use that is almost unknown among non-Aboriginal researchers. Here, we provide the first comprehensive review of evidence for the use of boomerangs as lithic retouching tools (percussors). A detailed lexical analysis demonstrates similarities between Palaeolithic bone retouchers used for the same purposes as the Australian hardwood boomerangs, thus confirming our traceological hypothesis and the power of using a multidisciplinary approach in investigating Australia’s deep past. This paper provides the foundation for discussions surrounding the multipurpose concept behind many Aboriginal tools by focusing on the most iconic item. We propose that, in future studies, the complex technological and functional features of boomerangs should take precedence over their famous “returning effect”.


Introduction
Boomerangs are an icon of Australian Aboriginal culture. However, from an archaeological point of view, evidence for these wooden implements is scarce (Attenbrow, 2009;Kelly, 1968;Langley et al., 2016;Luebbers, 1975;McBryde, 1977), and detailed information regarding their function has proved elusive. In popular culture, interest has long been focused on a single technological property that is widely held to be intrinsic to these artefacts-the ability of boomerangs to "come back" when thrown. Indeed, so much so that the word boomerang is commonly used as a verb in English. In Australian Aboriginal cultures, however, returning boomerangs are only a small part of the throwing stick universe. The returning boomerang, in fact, while known to have a specialised function for bird hunting in several areas, was most usually used as a plaything (i.e., child's toy) throughout most parts of mainland Australia (Brough Smyth, 1876;Davidson, 1936;Jones, 2004).
Contrarily to this popular misconception, boomerangs are (or were) among the most multipurpose tools in the Indigenous Australian toolkit. These artefacts move fluidly between various aspects of Aboriginal society, and are therefore involved in several activities, such as hunting, digging, making fire, cutting, making music, and fighting.
Regarding this multipurpose concept, a gap exists between knowledge shared by Australianists and the rest of the world. Until recent years, the typological tradition characterising western-based ethnoarchaeological research has not allowed a lot of space for the concept of multifunctionality. On the other hand, researchers closer to Indigenous traditions usually have a more inclusive approach towards the multipurpose nature of tools, a concept that is particularly present in Australian Aboriginal cultures.
Additionally, the use of boomerangs for the functional modification of stone tools seems to be a relatively new notion for both Australian and international researchers. In our previous work (Martellotta et al., 2021a), we outlined a traceological approach comparing use-wear identified on the surface of museum-curated boomerangs with those produced in the experimental use of bone blanks to retouch lithic tools (i.e., bone retouchers). This comparison with the European Palaeolithic is founded not on cultural biases but on the lack of previous traceological studies on percussion retouching in Australian contexts. While evidence for retouched lithic industries is present across the Australian continent (Holdaway & Stern, 2004;McNiven, 1993;Moore, 2004), most previous technological studies are on pressure retouching techniques (see for instance studies on Kimberley Points: Akerman et al., 2002;Hayes et al., 2014;Maloney et al., 2017;Moore, 2015), and none consider the percussion retouch. Moreover, bone and wood share similar physical and mechanical properties related to their density and elasticity ; see also Baumann et al., 2020;Bradfield, 2015;Johnson, 1985;Kamminga, 1988;Mateo-Lomba et al., 2019), and the relatively opportunistic nature of bone retouchers (i.e., "recycling" of food processing remains) well reflects the multipurpose concept behind the use of boomerangs in Australian Aboriginal cultures.
Following the traceological identification of retouch marks on boomerangs (Martellotta et al., 2021a), this study presents the first critical literature analysis regarding their use in retouching activities. The need for a systematic quantitative review lies in the fact that most previous studies on non-returning boomerangs are chaotic and rarely based on technology. Contributions undertaken in recent times, taking advantage of innovative approaches to studying the technology of ancient tools, are just a handful. Finally, this work is proposed to be a literature reference for future studies on retouching techniques by means of percussion. Such a technological behaviour left several traces in the Australian archaeological record, yet very little information is available.

Boomerangs in Australian Aboriginal Cultures: an Overview
Morphology and Classification Davidson (1936) was among the first authors to produce a systematic classification of boomerangs, and detailed geographic information regarding their distribution, with some other authors contributing until recently (Brough Smyth, 1876;Davidson, 1935Davidson, , 1936Jones, 2004).
A boomerang is composed of two arms, or wings, which intersect at the point of maximum curvature of the "stick"-defined as the elbow. The cross-section of a boomerang influences its flight patterns; most of them have a plano-convex crosssection, although some bi-convex shapes are also present. Morphological classification of boomerangs should consider the width of the angle created by the elbow and the length of each arm. When the arms have equal length, the boomerang is classified as symmetrical (Fig. 1a), whereas it is asymmetrical (Fig. 1b) when they differ in length. These two categories are used for hunting and, to a lesser extent, fighting purposes (Brough Smyth, 1876;Davidson, 1936).
The beaked or hooked shape (Fig. 1c) is a type of fighting boomerang characterised by an acute angle of the elbow and one arm much longer than the other (Davidson, 1936). This type of boomerang was manufactured in central Australia (Tennant Creek and the Tanami Desert) and traded into the Kimberley and other areas of Western Australia (Jones, 2004). Returning boomerangs (Fig. 1d) are usually symmetrical, with variable regional features. They tend to be smaller and lighter than the boomerangs used for hunting or fighting.
The peculiar cross-boomerang ( Fig. 1e) is only attested in the rainforest region around Cairns (Queensland), and it appears to have been used primarily for ritual purposes (Davidson, 1936;Jones, 2004). In general, boomerangs were not manufactured in tropical regions of Australia, probably because the dense rainforest is not a suitable environment for the use of throwing sticks. However, boomerangs were still imported into the tropics and used in rituals or percussive musical instruments (Jones, 2004).
Because of this great variety, both in shapes and uses of boomerangs, a strictly typological classification inhibits a comprehensive overview of this tool; therefore, a more technological approach is needed (Davidson, 1936).
The boomerang manufacturing process consists of procuring a hardwood preform or "blank" and adzing it until it reaches the desired morphology and dimensions. The final form of the boomerang is then attained by finely shaping with a stone or shell scraper, followed by smoothing the surface with a fine-grained grinding stone or organic incisor tools. In some areas, boomerangs were also decorated with linear or pattern designs, both engraved and painted (Akerman, 1998;Bordes, 2021;Bordes et al., 2015;Clarke, 2012;Davidson, 1936;Dawson, 1881;Hale & Tindale, 1933;Helmes, 1892;Jones, 2004;Langley et al., 2016;McCarthy, 1961;Roth, 1897;Spencer, 1922;Thomas, 1986;Wornshop, 1897).
The type of wood selected for manufacturing boomerangs varies according to the environment. Mulga, brigalow, and other species of Acacia are the most commonly used (Brough Smyth, 1876;Hawes, 1975;Jones, 2004;Kamminga, 1988;McCarthy, 1961;Paddy et al., 1993;Spencer & Gillen, 1927). Various species of Eucalyptus are also widely employed (Brough Smyth, 1876;Kamminga, 1988;Kenneally et al., 1996). Moreover, evidence shows the use of sheoak and other species of Casuarina (Brough Smyth, 1876;Jones, 2004;Kamminga, 1988;Luebbers, 1975;Spencer & Gillen, 1927;Wornshop, 1897), as well as different species of Hakea (Kamminga, 1988;Kenneally et al., 1996;Langley et al., 2016;Paddy et al., 1993). Additionally, Melaleuca and mangrove wood are known to be used (Jones, 2004;Kamminga, 1988;McCarthy, 1961;Paddy et al., 1993;Spencer & Gillen, 1927). Finally, the use of Quaercus (Hess, 1973), mistletoe tree  (Paddy et al., 1993), grey-cork wood, Ficus, and kapok tree (Kamminga, 1988) is also sporadically reported. Jones (2004) gives the most updated and reliable account of the geographic distribution of boomerangs throughout Australia, building on the work of Davidson (1935Davidson ( , 1936 and other authors over the nineteenth and twentieth centuries (Brough Smyth, 1876;McCarthy, 1961;Thomas, 1986Thomas, , 2000Wornshop, 1897). The great diversity of boomerangs makes it impossible to associate a single morphology to a specific area, and a more regional approach should be applied. We refer in this work to the division of the Australian continent based on drainage basins and languages created by David R. Horton (© AIATSIS 1996). 1 The modern-day division of Australia in states and territories might be inadequate to describe technology developed in ancient Aboriginal countries, and it is therefore not suitable for the present work. The spelling of Indigenous names reported in this article is as in the Australian Indigenous Language Database-AUSTLANG (October 2021).

Geographic Distribution
Central and northern Australia show a certain grade of uniformity in boomerangs morphologies. Only two boomerang shapes were manufactured in these areas: the hooked boomerang, used in fighting, and an asymmetrical one, longitudinally fluted ( Fig. 2a, b).
In western Australia, boomerangs are thin, mostly asymmetrical, and have rounded tips, which can sometimes be sharpened (Fig. 2c). An unusual morphology is present in the southern area, with an extra concavity on one of the arms (Fig. 2d). In the Kimberley, boomerangs have a wider elbow angle; in coastal zones of this region, the asymmetrical shape is more pronounced (Fig. 2e), whereas inland areas have more features in common with central Australia (Fig. 2f).
In the Lake Eyre and Darling River basin area, the longest and heaviest boomerangs are attested: a smaller one, thrown to kill the prey in hunting, and a longer one-up to 2 m-was used in hand-to-hand combat (Fig. 2g,h). This region is the only one where returning boomerangs seem to be absent.
Most boomerangs in south-eastern Australia are symmetrical ( Fig. 2i-j). According to some authors, a certain degree of morphological uniformity in this area could be attributed to the relative stability of environmental conditions, on the premise that a less challenging environment is less likely to result in less diversified toolkits (Jones, 2004).
The region corresponding to central and eastern Queensland has a great variety of environments and Aboriginal countries, both reflected in the variety of boomerang morphologies. Differences could be identified in the degree of curvature and the shape of the tips (Fig. 2k).

Known and Unknown Uses of Boomerangs
The variation in boomerang forms, combined with different environmental conditions, underlies the multipurpose nature of these implements, which could be related to different mobility patterns. It is a general belief that human groups living in open landscapes base their subsistence on high mobility, and they have the need to carry as few tools as possible while moving-thus, they are supposed to prefer multifunctional tools. On the contrary, mobility has a smaller effect on subsistence strategies in more temperate areas, supposedly allowing toolkits composed of more numerous and specialised tools to be an acceptable option. It is worth noting that, at the moment, no evidence certainly shows the application of these models in Australia. The relationship between human groups, tools, and the environment is still an open issue (for a discussion, see Jordan, 2015;Kelly, 1995Kelly, , 2013Shott, 1986).  (Jones, 2004, p. 82), whereas boomerangs from inland (f) have more traits in common with central Australia (Jones, 2004, p. 82). In the Eyre region, smaller boomerangs (g) were thrown, whereas longer boomerangs (h) were used in close contact fighting (Jones, 2004, p. 92). (i-j) symmetrical boomerangs from the Southeast (top) (from Jones (2004, p. 99)) and Spencer (bottom) (from Jones (2004, p. 103)) regions. (k) morphological variability of boomerang's morphologies in the region corresponding to central and eastern Queensland (Jones, 2004, pp. 96-97). Jones, P. (2004). Boomerang. Behind an Australian Icon, Wakefield Press, Kent Therefore, it is perhaps unsurprising that, besides their well-known hunting function, boomerangs were used in many other different ways. Examples include their use as a cutting tool, a hammer, a club, for digging, for making fire, as a fire-poker, for making music (Akerman, 1998;Jones, 2004;McCarthy, 1961McCarthy, , 1976Pahlow & Silady, 1986;Thomson, 1964 and references therein), and for retouching lithic tools.
Preliminary bibliographical research on boomerangs revealed a discrepancy between references on the aerodynamics of the returning boomerang and every other aspect of boomerangs themselves. Specifically, published studies on technological and social aspects of the boomerang seem to be rare, although this tool is a highly recurrent component of Australian Aboriginal toolkits. The few references on boomerangs used for retouching purposes are often presented incidentally within the description of daily activities-especially woodworking. Extracting or working timber requires specific techno-functional features of the stone implements, such as a relatively thick working edge, denticulated or serrated in its morphology, and planoconvex in its cross-section (Claud et al., 2013;González Urquijo & Ibáñez Estévez, 1994;Hardy & Garufi, 1998;Lepot, 1993). For this reason, information on methods of shaping lithic edges (i.e., retouching) is usually found in references that are primarily concerned with the general subsistence patterns of Australian Aboriginal communities rather than specific publications on boomerangs.

Understanding the Literature
We approached this review using a Systematic Quantitative Assessment technique (Healey & Healey, 2010;Petticrew & Roberts, 2006;Pickering, 2014). That is, a comprehensive method of literature search focused on highlighting the possible gaps in any given topic by quantifying the amount of existing research. Different but related topics of interest were summarised and arranged into a hierarchy. The references were then counted and categorised into one of these topics, providing a measurable degree of relevance of each reference to the main topic of the review.
Historical reports and scientific articles from the past 170 years were searched for references to and descriptions of the use of boomerangs in retouching activities and for technological comparisons with other retouching tools. References were recovered using the main web search engines for scholarly information (e.g., Google Scholar, Scopus) for the most recent contributions. For older publications, we used the Griffith University Library catalogue (Queensland, Australia), which includes the Bonus + catalogues of 12 different Australian libraries located outside of Queensland and several partner libraries from overseas. The partnership between libraries allowed us to consult old or rare publications stored in limited collections. Moreover, the digitalisation process carried out by several libraries worldwide allowed us to access old publications in PDF format. We analysed a total of 188 separate contributions published between 1842 and 2021. The references were organised in a database, each identified by the last name of the first author and the year of publication (Table 1). One to four keywords were assigned to each reference in order to sort them by main topic(s). As a result, the references were grouped into the following four categories based on the level of accordance with the main topic of this review (i.e., the use of boomerangs in retouching activities): -Class I: documents reporting evidence for boomerangs and other wooden implements being used for the functional modification of the edges of stone tools among communities throughout Australia. -Class II: information on boomerangs (e.g., manufacturing, multipurpose, woodworking, linguistic and geographical data); description of lithic tools retouched using percussion techniques.  The identification of evidence for the use of boomerangs to retouch lithic tools has been assisted by a detailed lexical analysis of the references, based on a technological parallel with Palaeolithic bone retouchers. We extrapolated descriptions of the retouching movement observed during the experimental use of bone retouchers. Then, we identified a similar lexicon among the descriptions of retouching using boomerangs and other wooden implements. The existence of a relationship between osseous and wooden raw materials is a hypothesis that deserves to be experimentally tested, but it is possible to assume that these two materials would have comparable reactions to similar mechanical stimuli. According to this hypothesis, the percussion movement associated with the retouch activity would leave similar use-wear on both osseous and wooden implements used as retouchers.

Literature Evidence of Boomerangs Used for the Functional Modification of Lithic Edges
The first approach to the bibliographic research was to look for contributions having the word "boomerang" in their title or among their keywords; as a starting point, 22 references were identified. However, only two gave information regarding the use of boomerangs in retouching activities (McCarthy, 1961;Pahlow & Silady, 1986). For the rest, few provided information on archaeological or traceological evidence for boomerangs (Bordes, 2021;Langley et al., 2016;Luebbers, 1975); and some described only boomerang typology and geographic distribution (Davidson, 1935(Davidson, , 1936Etheridge, 1897;Ferguson, 1843;Harper, 2015;Jones, 2004). The majority (11 references) were mainly about the aerodynamic properties of boomerangs, especially regarding the returning type (Anderson & Jones, 1992;Baker, 1890;Basedow, 1935;Hawes, 1975;Howitt, 1876;Nelson, 2000;Ruhe, 1982;Ruhe et al., 1986;Thomas, 1983Thomas, , 1986Thomas, , 2000. The bibliographic research then concentrated on general descriptions of the Australian Aboriginal lifestyle, such as toolkits or daily activities (e.g., hunting, woodworking). A further 166 references were then collected, for a total of 188 references. Of these, 39 were considered not relevant at any level for the use of boomerangs for retouching purposes (class IV). Ninety-two documents contained technological references to Palaeolithic bone retouchers and general discussions on organic tools in Australia and Europe (class III). Some 46 references mention tangential information on boomerangs and descriptions of lithic tools retouched through percussion in Australian Aboriginal contexts (class II). Finally, 11 documents report the use of boomerangs and some other wooden implements to modify the edges of stone tools (class I).
Literature evidence for the use of boomerangs for retouching purposes has been identified in different parts of Australia (Fig. 3a) in contexts describing woodworking activities and/or retouched lithic industries (e.g., Tula adzes).
Among the Indjilandji people at Camooweal (Desert region, between Queensland and Northern Territory), "the final step [of Tula adzes making] involved regularising the working edge by hafting the tula at a right angle to the handle and trimming the edge by soft hammer percussion using a boomerang" (Moore, 2004, p. 62). The historical photograph of an unnamed Aboriginal man retouching the stone Tula adze using a boomerang (original photo from Roth, negative V2165, Division of Anthropology, Australian Museum) is reproduced in this reference (Moore, 2004, Fig. 2) and cited elsewhere (McCarthy, 1961, Fig. 1).
Another piece of evidence comes from the Waljen people in Cundeelee (Desert region, Western Australia): "In later stages of work, lighter wooden objects were used to resharpen the adze, including boomerangs and light sticks about 14 cm long" (Hayden, 1979, p. 27).
More evidence is found among the Warlpiri in the Tanami Desert (Desert region, Northern Territory): "The sharpening is often done with taps from the flat face of a hunting boomerang" (Tindale, 1965, p. 136).
Furthermore, Roth (1904, p. 17) mentions the use of a boomerang to make the lithic cutting edge thinner after it had been already retouched with a stone hammer among the Wakaya people along the Upper Georgina River (Desert region, Northern Territory): "This is next improved upon by firmly enclosing [the stone flake's] major portion, base downwards, and at an acute angle, into a blob of gum-cement […] at the extremity of a piece of wood […] holding it in the left hand […], and then flaking off all irregularities and prominences by sharply striking them with the flat under-surface of one of the heavier kinds of boomerang".
Finally, more documents refer to boomerangs being used in retouching activities, although the authors do not specify the community to which they refer. For instance, among the descriptions of different uses of boomerangs in McCarthy  (1961, p. 347, Fig. 2) and Pahlow and Silady (1986, p. 19), their use for retouching lithic tools is cited and graphically represented (Fig. 3b). Moreover, Davis (1979, p. 55), while referring to Aboriginal people using knapped lithic tools, states that "such tools were re-sharpened by striking with a small hammerstone, the flat end of a boomerang, or forcing small flakes off with the teeth".
Literature analysis revealed that the shaping of lithic edges was sometimes performed using other types of tools unrelated to manufacturing, whose primary function, like boomerangs, was not retouching. It happened, for instance, among the Pitjantjatjara people in Aparina Springs (Desert region, South Australia): "Several times during this operation [i.e., making of a wooden spearthrower] the adze stone was retouched or given a new cutting edge, by holding the spearthrower in the right hand, with its adze stone resting against the palm, and tapping its flat face with the wooden blade of a spear until miniature flakes were broken off" (Mountford, 1941, p. 315, Fig. 2E) (Fig. 3c).
The use of wooden spears is also observed among the Ngaanyatjarra people in Gill Pinnacle (Desert region, Western Australia): "There are numerous variations in the methods of sharpening by tapping with wood. In November 1963 I saw a Ngadadjiara [i.e., Ngaanyatjarra] man at Gill Pinnacle use the flat base of the head of his composite, single-wooden-hookedbarbed-spear to tap off a flake or two from the stone blade fastened at the end of his spearthrower". (Tindale, 1965, p. 159).
The same community was also observed while using the handle of a chisel as retoucher: " […] shortly afterwards during use [of a lithic flake] in the making of the handle of a further chisel, it was sharpened. In doing this the cutting edge was tapped, using the rounded surface of one end of the second handle which Tjupurula [name of an Aboriginal man involved in the manufacturing activity] was shaping using its [the flake's] aid. The single act of resharpening resulted in the appearance on the chisel edge of a regular series of tiny semicircular stepped flake scars". (Tindale, 1965, p. 135).
Another line of evidence suggests the use of unworked pieces of wood and sticks. One comes from two communities in Papunya settlement (Desert region, Northern Territory): "Retouching of adzes done by Ngayuwa and Tapatapa using both the end of a piece of wood as described at Cundeelee […]" (Hayden, 1979, p. 56). Another example is among the Nakako people in Blackstone (Desert region, Western Australia): "When blunted by use the resharpening of these originally simple flakes is done in one of several ways: […] (b) striking with a piece of hardwood" (Tindale, 1965, p. 149).
Besides using boomerangs, the Waljen community in Cundeelee seems to use wooden sticks as well for retouching: "The most general method of resharpening adzes was to tap them with a stick while held in the hand, as noted at Cundeelee" (Hayden, 1979, p. 67; see also quote above in Hayden, 1979, p. 27). The same is observed among the Ngaanyatjarra people in Warburton (Desert region, Western Australia): "[…] the flake is first hafted and then trimmed by means of gentle blows struck with a wooden stick" (Gould et al., 1971, p. 157).
Finally, Akerman (1998, p. 3) gives a general definition of the retouch activity among Australian Aboriginal communities: "A struck flake may be used just as it comes off the core or may be retouched to improve its edge or to give it a more finished shape. Retouch may be by percussion flaking, that is by hitting it with a hammer of stone, bone or wood".

Discussion
In Australia, many early European accounts of Aboriginal life and material culture have been considered unreliable by present-day archaeologists because of the perception that cultural misunderstandings and prejudices influenced the observers. Some modern scholars have also objected to the so-called 'tyranny of the ethnographic record', arguing that a reliance on ethnographic data has dominated (and limited) our understanding of the Aboriginal past (e.g., Hiscock, 2008). Although an approach mainly based on ethnographic reports might result in cultural generalisations, it could be argued that these accounts can still be a valuable source of information if analysed with a systematic approach (see discussions in Hiscock, 2008;Wobst, 1978).
The present work aims at resolving this conflict. The proposed analysis of ethnographic evidence was carried out as a function of the technological knowledge on bone retouchers. The presence of these tools is attested in Europe, Asia, and Africa from 500 ky BP, and they started to disappear in the Upper Palaeolithic (for a synthesis, see Abrams et al., 2014;Daujeard et al., 2014;Doyon et al., 2021;Hutson et al., 2018;Martellotta et al., 2020;Martellotta et al., 2021b;Tartar, 2012;Turner et al., 2020; a complete references list on this topic is provided in Table 1 and Supplementary Information). Bone retouchers were identified among archaeological faunal assemblages between the end of the nineteenth century and the beginning of the twentieth century. In this period, lithic tools, the main focus of prehistoric technology at that time, were recognised through typological definitions. Bone retouchers, on the contrary, were only defined according to their use-wear ("retouch-induced stigmata") because no standardisation-thus, no "types"-could be recognised in their morphology (see discussion in Baumann et al., 2020;Choyke, 1997). Hence, the study of bone retouchers started from a traceological approach, whereas the technology behind their function was discovered decades later, directly through experimental approaches (Semenov, 1964). In the past 60 years, several traceological and experimental studies proved the use of bone retouchers for the functional modification of lithic edges. By means of a percussion movement in which the bone surface impacts the stone edge, and through the detachment of small, thin flakes, bone retouchers give a specific shape to the lithic edge (for an overview, see Costamagno et al., 2018;Daujeard et al., 2014;Doyon et al., 2021;Mallye et al., 2012;Mozota Holgueras, 2013, 2018Rosell et al., 2011;Tartar, 2012;Thiébaut et al., 2019;Vettese & Daujeard, 2021; a complete references list on this topic is present in Table 1 and Supplementary Information).

The Lexical Analysis
The lexical analysis carried out in this study revealed relevant similarities between the use of boomerangs for retouching purposes and the description of the experimental use of Palaeolithic bone retouchers. The same approach could be applied to less common references of other wooden implements used as retouchers-such as spears, wooden handles, or unworked pieces of wood.
The morphology of the lithic edge retouched through percussion is also described in different instances. The detachments on the retouched lithic edge are defined as "miniature flakes", which were "broken off" (Mountford, 1941, p. 315) or as "a regular series of tiny semicircular stepped flake scars" (Tindale, 1965, p. 135).
In a few instances, techno-functional information could be inferred regarding the portion of the boomerang used for retouching: "the flat face of a hunting boomerang" (Tindale, 1965, p. 136), "the flat under-surface of one of the heavier kinds of boomerang" (Roth, 1904, p. 17) and "the flat end of a boomerang" (Davis, 1979, p. 55).
The results of the lexical analysis are mirrored in several descriptions of retouching activities carried out in experimental studies on the use of Palaeolithic bone retouchers (for an overview, see Costamagno et al., 2018;Daujeard et al., 2014;Doyon et al., 2021;Mallye et al., 2012;Mozota Holgueras, 2012, 2018Rosell et al., 2011;Tartar, 2012;Thiébaut et al., 2019;Vettese & Daujeard, 2021; a complete reference list on this topic is present in Table 1 and Supplementary  Information).
The position of the hands and the lithic flake during the retouch activity is another interesting result that emerged from this study. The flake is described to be positioned "at the right angle" (Moore, 2004, p. 62) or, more precisely, "at an acute angle" (Roth, 1904, p. 17). Such information could be interpreted as a reference to the definition of retouch technique developed through experiments on Palaeolithic technologies, stating that the retoucher needs to hit the flake at an angle inferior to 90 degrees for the retouch to be successful (Inizan et al., 1995).
Finally, new information emerged regarding the technology behind the retouch activity among Australian Aboriginal communities, which has never been the object of a technological study. A few references mention the position of the retouch (i.e., the position of removals relative to the faces of the flake; Inizan et al., 1995), which, in these cases, appears to be inverse: the flake is described as positioned "base downwards" (Roth, 1904, p. 17), or the "tapping" is said the be carried out on the flake's "flat face" (Mountford, 1941, p. 315). Moreover, new information regarding the grasping techniques used during the retouching activity can be inferred through a parallel with experimental studies on bone retouchers. Retouching falls under the category of "complex bi-manual actions"; that is, actions in which both hands are employed, but with complementary roles. The preferred hand (usually related to the handedness of the knapper) performs the action-in this case, holding the retoucher-while the other hand stabilises the object-in this case, the stone flake (Mosquera et al., 2012;Uomini & Meyer, 2013, and references therein). The results of the lexical analysis revealed two different strategies for carrying out this action. In some cases, the flake is held in one hand: "[..] holding the spear-thrower in the right hand, with its adze stone resting against the palm, and tapping its flat face […]" (Mountford, 1941, p. 315); "the most general method of resharpening adzes was to tap them with a stick while held in the hand […]" (Hayden, 1979, p. 67). In other cases, the flake is first hafted and then retouched: "[…] firmly enclosing [the stone flake's] major portion, base downwards, and at an acute angle into a blob of gum-cement […] at the extremity of a piece of wood […] holding it in the left hand […], and then flaking off all irregularities and prominences […]" (Roth, 1904, p. 17); "[…] the flake is first hafted and then trimmed […]" (Gould et al., 1971, p. 157). Apart from sparking a new interest in retouching methods among Australian Aboriginal communities, these results could be considered another expression of the similarities between the analysed references and the experimental studies on bone retouchers (Fig. 3d). The resemblance in the grasping techniques can be appreciated in the few graphical representations of the use of boomerangs and wooden spearheads used as retouchers (Fig. 3b, c).

General Considerations
It is relevant to note that the retouch activity was carried out using not a specialised tool but finished tools whose primary purpose was not retouching (i.e., boomerangs, and to a lesser extent, wooden spearheads). The same can be said for the use of unworked pieces of wood, which in most cases were gathered from the ground during the woodworking activity and were both used and discarded on the spot. In some instances, moreover, both finished tools and unworked pieces of wood are used for retouching. That is the case of the Waljen people in Cundeelee, where references report the use of both sticks and boomerangs in retouching activities (Hayden, 1979, pp. 27, 67). Similar evidence comes from the Ngaanyatjarra people: Gould et al., (1971, p. 157) report the use of sticks as retouchers among these people in Warburton, whereas Tindale (1965, p. 135) refers to the use of both a spearhead and the wooden handle of a chisel to retouch lithic tools in Gill Pinnacle.
These results could suggest a relationship between multipurpose tools and high mobility patterns. Unworked pieces of wood and sticks were present in the working space where the retouching took place (e.g., woodworking site)-so the mobile group did not need to carry them. On the other hand, boomerangs and wooden spears were often carried around because wood procurement and hunting usually occurred at the same moment. However, there is also evidence for strategically placing various tools near one or more recurrent working spaces to avoid carrying them. The amount of evidence available in the literature is not substantial enough to allow for firmer conclusions about the geographic distribution of technological behaviour. However, a promising path for future research would be to focus on understanding if the use of boomerangs as retouchers is related to a high mobility subsistence strategy.
The fact that boomerangs appear to be the most common tools used for retouch activities adds more evidence to the profound multipurpose nature of these tools. It seems evident, at this point, that a typological approach is not suitable for the description of the role of boomerangs in the Australian context. Although Davidson was the first author to suggest a typology for boomerangs, he also stated-80 years ago-that any typological classification for these tools should be considered pointless due to their wide variability in morphologies and, more importantly, functions (Davidson, 1936, p. 90). Therefore, future studies on boomerangs should take into account their multipurpose nature, and the approach should be more technological rather than typological.

Conclusion
Although part of Australian Aboriginal tool-using traditions, boomerangs used as retouchers is a new concept in western-based scientific research. The present literature review and study give an empirical basis to this hypothesis, and we endorse the need for traceological approaches and experimental protocols around this topic. A more significant number of publications from this perspective would reduce the wide gap in the literature between the number of contributions on the aerodynamic properties of returning boomerangs and the deep technological and cultural value of these tools for Aboriginal communities. The present work also suggests that early European accounts of Australia are not always unreliable but can prove helpful if approached with a multidisciplinary methodology. Together, these new methods for studying boomerangs may allow us to move beyond traditional typologies and focus on understanding their technological and functional features.