The West Tofts handaxe: A remarkably average, structurally flawed, utilitarian biface

The West Tofts handaxe is a small British Acheulean biface well known for its cortical preservation of a fossilised bivalve shell. The shell ’ s retention, its prominent central placement, and perceptions of the tool ’ s broader aesthetic-value have resulted in it being described as an example of early hominin aesthetic intent. When combined with its mid-to-late Pleistocene age, the handaxe plausibly has implications for our understanding of hominin cognitive evolution and the origins of Palaeolithic art. Crucial to the assignment of aesthetic intent are a series of assumptions concerning the exceptionality of the tool ’ s design, production and use. Here, we test those assumptions. The West Tofts handaxe is revealed to be technologically and morphologically unremarkable for the British late Acheulean, was produced on a tabular flint nodule that did not require invasive (central) flake re-movals, and displays remarkably average flaking investment. High-resolution micro-CT scans of its internal structure reveal numerous fractures and air pockets, and at least one – potentially two – additional fossilised shells. These inclusions and flaws feasibly impacted the tool ’ s production, potentially resulting in the shell ’ s retention. Limited microwear traces were identified on its distal flaked edges, potentially hinting at the tool ’ s expedient use. These and other data underline the importance of restraint when attributing aesthetic intent or ‘over-investment ’ to the West Tofts handaxe. Today we may consider it a remarkable artefact worthy of elevation


The West Tofts handaxe and questions of aesthetic intent
The West Tofts handaxe (WTH) is a pointed bifacially flaked large cutting tool produced on a tabular nodule of flint (Fig. 1).Named after the Norfolk (UK) village where it was discovered, it potentially dates to marine isotope stage (MIS) 11 (374-424 thousand years ago [ka]) or 9 (300-337 ka) and was likely produced by an early neanderthal (Homo neanderthalensis) population or an alternative Middle Pleistocene hominin species (H.heidelbergensis s.l.).Currently on display in the University of Cambridge's Museum of Archaeology and Anthropology (MAA), it is one of the most recognisable Palaeolithic artefacts in the world.
Such visibility and recognition stems from a fossilised bivalve shell centrally preserved in the handaxe's cortex (Fig. 1).Early records attest to the shell's importance, with the Cambridge Antiquarian Society (CAS) (1915: 91) describing the handaxe as having been "carefully chipped so as to expose in the centre of its convex face a fossil pecten shell".This came four years after the handaxe was discovered in 1911 (Oakley, 1973), and one year after its purchase by the MAA in 1914.CAS set a precedent for more than a century of Palaeolithic discussion; numerous texts describe the tool as having been 'carefully' or 'deliberately' knapped to retain, centrally frame, and emphasise the fossil (e.g., Oakley, 1973Oakley, , 1981;;Hayden, 1993;Jelínek, 2001;Hodgson, 2006;Wynn and Berlant, 2019;Assaf and Romagnoli, 2021).Overwhelmingly, causation for the fossil's retention and placement is ascribed to the aesthetic sense and visual values of the knapper.In turn, the WTH has captured the imagination of wider non-academic audiences, with it regularly being exhibited for its aesthetic and 'artistic' value (Museum of Archaeology and Anthropology, 2023).
The WTH is not unique in retaining a fossil shell within its cortex (Nobel and Davidson, 1996).Flint frequently displays internal and external inclusions, and these are regularly retained or exposed during the production of Acheulean handaxes.What aids the West Tofts' distinctiveness is the large size of the shell, the shell's long axis being near-parallel to the flint nodule's surface, and the lack of damage to the fossil (Fig. 1).This combination of naturally occurring characteristics is uncommon relative to most flint nodules, but not wholly unexpected.
What elevates the West Toft handaxe to being near-unique among artefacts is how these natural elements are perceived when combined with its hominin-imposed formal properties.
The morphological and technological properties of hominin material culture (i.e., artefacts) provide archaeologists with insight into past behaviour, behavioural variation, and how these changed through time and space (Lycett, 2009;Okumura and Araujo, 2019).Within a Middle Pleistocene context, where hominin behaviour is tightly linked to cognitive capabilities and evolutionary trajectories (Gowlett et al., 2012;Stout et al., 2014;Nowell, 2017;Lombard and Hogberg, 2021;Coolidge et al., 2023), stone tools provide the only abundant source of evidence for understanding how the hominin brain evolved.Part of this discussion Fig. 1.Superior (A), inferior (C) and side profile views of the West Tofts handaxe (B,D).Note the large fracture at the distal tip on Fig. 1D, as evidenced by the exposed non-patinated flint.This suggests the distal symmetry of the tool (left side of the superior view) to have originally been different relative to the (likely taphonomically damaged) artefact that we see today.Images courtesy of the University of Cambridge Museum of Archaeology and Anthropology.
E. Flanders and A. Key concerns when hominins, most often late Homo, started to display evidence of an aesthetic sense, and in turn, artistic material culture (see Martín-Loeches [2023] for a recent review).Art, and the aesthetic values underpinning it, play an important role in our understanding of hominin cognitive evolution.Indeed, the Middle Stone Age of Africa, Middle/Upper Palaeolithic of Asia, and Upper Palaeolithic of Europe witness a strong association between our species and the presence of cave, rock, and figurative art (d 'Errico and Nowell, 2000;Barham and Everett, 2021;Coolidge et al., 2023;Martín-Loeches, 2023;Bebber, in press).This does not mean we are the only hominin species to have created art; potentially, far from it (e.g., Joordens et al., 2015;Hoffman et al., 2018;Bello, 2021;Marquet et al., 2023).However, because there is little doubt that artistic expression and aesthetic intent are present from an early point in our own species, it raises the potential for art to have an earlier point of origin, beyond our c.300,000 years.
Herein lies the value of the WTH to debates surrounding hominin cognitive evolution and the origins of art.In arguing that an aesthetic motive underpins the handaxe's formal properties, a great deal is implicated regarding the cognitive capabilities, aesthetic sense, and artefactual motivation of hominins during MIS 9 to 11 (Mithen, 2003;Nowell, 2017;Straffon, 2019;Wynn and Berlant, 2019;Gowlett, 2021;Martín-Loeches, 2023;Bebber, in press).It suggests that a Middle Pleistocene hominin species, potentially on a different evolutionary lineage to our own, was not only capable of the complex sensorimotor, emotion-valuation and meaning-knowledge neurological interactions underpinning human aesthetic values and behaviour (Wynn and Berlant, 2019), but occasionally overrode otherprincipally utilitarianpressures determining the form of Acheulean bifaces (Lycett et al., 2016).Simply, when aesthetic intent has previously been assumed for the WTH, a range of cognitive capabilities and aesthetically driven behavioural choices are then linked to the hominin knapper, and by proxy, wider Middle Pleistocene populations.
What has always been clear for the WTH is that aesthetic intent is assumed by the present-day beholder, and not supported by objective evidence.This is despite some stating the fossil's "aesthetic sensibility" and preferential placement to be "fact" (Hodgson, 2005: 33;Hodgson, 2006: 63). 1 Even if the central location of the fossil can be demonstrated, it does not aid our understanding of the knapper's motivation.Occasionally, as justification, it is stated that the biface's characteristics "exceeded functional requirements" (Hayden, 1993: 123;Wynn and Berlant, 2019 [also see : Currie, 2011]), but objective evidence to support this claimbeyond the fossil's presence or neurological theoryis not presented.
In suggesting that the handaxe was carefully flaked to preserve the shell and associated cortex, and shaped to centrally locate the fossil, several key assumptions are made.First, it is assumed the final aesthetic form of the tool was not only in the mind of the knapper during its production, but the ability to intentionally create such a specific form was within their technical repertoire.This speaks to wider discussions concerning the cognitive complexity of visualising and producing Acheulean LCTs (Gowlett, 2006(Gowlett, , 2021;;Herzlinger et al., 2017;Lycett and Eren, 2019;Pargeter et al., 2019;Stout et al., 2021;Hodgson, 2023), but takes on additional technical and visuospatial demands due to the minimal room for error and/or mid-production adjustment.Indeed, to remove large, thin flakes yet intentionally retain the shell, the Acheulean individual would have had to exhibit remarkable control over the complex interacting variables known to impact flaking outcomes (Dibble and Pelcin, 1995;Nonaka et al., 2010;Dogandžić et al., 2020;Li et al., 2022).
Therein lies the second assumption; the knapper consciously (actively) chose to keep the shell, and thus crafted the handaxe in a way to avoid its removal (Oakley, 1973(Oakley, , 1981;;Jelínek, 2001;Assaf, 2018;Assaf and Romagnoli, 2021;Wadley, 2021).Further, there is the assumption the central location of the fossil was intentional to frame, and therefore emphasise, the fossil (Hayden, 1993;Hodgson, 2006;Bredekamp, 2019).That is, "by knapping around the shell, they were able to accentuate it and draw attention to it" (Wynn and Berlant, 2019: 290).These latter two assumptions can be extended further by explicitly linking these choices to the knapper's aesthetic sense and visual reward systems (Hayden, 1993;Hodgson, 2000Hodgson, , 2005;;Bredekamp, 2017;White and Foulds, 2018;Deane-Drummond and Fuentes, 2020).It has even been stated that "the fossil played a role in the selection of this particular" flint nodule; the assumption being that without the shell, the nodule may not have been chosen for the creation of a handaxe (Oakely, 1981;Jelínek, 2001;Assaf, 2018;Assaf and Romagnoli, 2021: 286).Finally, if one or more of the above are put forward as an explanation for the WTH's formal properties, then one must assume that in each instance, a more plausible alternative explanation does not exist.
In the absence of any alternative evidence, and potentially in the absence of wider support from the archaeological community (Wymer, 1985;Chase and Dibble, 1987;Nobel and Davidson, 1996;Gowlett, 2021), these intuitive assumption-based interpretations of the handaxe are pervasive in the literature.Importantly, it is present-day H. sapiens intuition creating these assumptions and it is our own aesthetic appreciation, training and education being projected onto past hominins (Bebber, in press).To highlight one of many examples, Bredekamp (2019: 37) states "by framing this object [the fossil], and as such transforming it into a picture … he [ 2 ] brought the genre of the 'picture within a picture' into the world.The fossil can be regarded as the impetus to develop, through the experience of iconic difference, a pictorially based consciousness of distance.Already graspable here may be … the "conceptual space of rationality", the conditio sine qua non [necessary condition] for the development of a sphere of reflection".
In a world where the necessity of testing archaeological assumptions via hypothesis-driven investigations has been explicitly (Eren et al., 2019;Magnani et al., 2019) and repeatedly (Chase andDibble, 1987, 1992;Outram, 2008;Clarkson and Shipton, 2015;Eren et al., 2016;Lin et al., 2018;Eren and Bebber, 2019) made, it is high time that, insofar as is possible, the assumptions concerning the aesthetic intentionality of the WTH be tested.In other words, given the inculcated importance of the WTH, its value to wider discussions of hominin cognitive evolution, and the leaps in logic required to infer aesthetic intent, there is a clear need to more fully investigate how and why the artefact was created.

How can instances of 'careful knapping' and aesthetic intent be identified?
Central to claims that the WTH was 'carefully' flaked to preserve and 'frame' the fossil are the inferences that the original flint nodule was chosen because of the shell, and that its cortex patterning and retention, and associated flaking investment and strategy, deviates from Acheulean norms (Oakley, 1973;Hayden, 1993;Jel Ínek, 2001;Hodgson, 2005;Bredekamp, 2017Bredekamp, , 2019;;Assaf and Romagnoli, 2021;Wadley, 2021).Including the idea that the tool is over-engineered, "exaggerated" (White and Foulds, 2018: 316) and "exceeded functional requirements" (Hayden, 1993: 123) relative to knapping sequences observed during the production of utilitarian tools (be the function solely mechanical [i.e., cutting] or inclusive of social roles [Key and Lycett, 2017]).This does not mean the handaxe was never used, but instead, utility was not the leading consideration during its production, and functional selective pressures (c.f., Lycett et al., 2016) were not guiding its design.Given the Acheulean spans c. 1.8 million years, three continents, and diverse 1 "The proposition that tools served as conveniently available objects for the realisation of an 'aesthetic' sensibility, is supported by the fact that fossils were preferentially placed within the overall form of handaxes, as in the West Tofts example."As quoted from Hodgson (2005:33) and Hodgson (2006: 63).
hominin species (Gallotti and Mussi, 2018;Wynn and Gowlett, 2018;Key, 2022), the exceptionality of the West Tofts should be defined relative to its deviation from centrality when compared against other northern European MIS 9 to 11 flint handaxes.
Claims of 'over-engineering' based on proximally and centrally located cortex retention can lie in juxtaposition to other interpretations of Acheulean handaxes.Crucially, the shell's intentional retention should not necessarily be synonymous with aesthetic or symbolic reasoning.Some suggest high-cortex retention or minimal flake removals to reflect an expedient knapping strategy, where the goal is to establish a length of (often tip-focused) working edge but minimise flaking investment (White, 1998;Li et al., 2015;Lee, 2017;Muller et al., 2022;García-Medrano et al., 2022).Others link central and basal cortex retention to the handaxe's early stage of reduction as part of an extended production or use-life sequence (Newcomer, 1971;McPherron, 2003;Emery, 2010;García-Medrano et al., 2022).Further, cortical butts increase the cutting forces and efficiency of handaxes, making the proximal retention of cortex a demonstrably utilitarian trait (Gowlett, 2006;Key et al., 2016;Wynn and Gowlett, 2018;Key and Lycett, 2023).Pointed handaxessuch as the WTHare also often more cortical than ovate alternatives (Wymer, 1968;White, 1998;White and Foulds, 2018;Herzlinger et al., 2021).Zaidner et al. (2006: 253) generalised pointed handaxes as displaying "cortex on both faces … intensive tip preparation, high amount of cortex on the proximal end, and … a cortical butt".
There are several additional alternative explanations for the WTH's central cortex.It could stem from the considerable skill and manual demands required to 'thin' bifaces by driving flakes over the tool's midpoint (Stout et al., 2014;Key and Dunmore, 2018;Liu et al., 2023).It may also be derived from a decision to aid manual friction and/or adjust the tool's centre of gravity for ergonomic and mechanical (torque) reasons (Shimelmitz et al., 2017;Wynn and Gowlett, 2018;Wynn, 2020;Key and Lycett, 2023).The knapper may have also recognised there was no a priori reason to remove long 'thinning' flakes.Indeed, the WTH is made from a thin tabular nodule of flint, an ideal blank for bifaces because there is no requirement to thin the tool and remove flakes across its central surface (Newcomer, 1971;Whittaker, 1994).As with many lithic artefacts, the handaxe's original form may also have been remodified (McPherron, 1999;Dibble et al., 2017;Kuhn, 2020), meaning the shell may not initially have been centrally framed.Finally, there may be additional non-aesthetic reasons for retaining the WTH's central cortex, and in turn the fossil, that are not immediately observable or clear to us as H. sapiens with experimental knapping as our primary point of reference (Eren et al., 2016;Dibble et al., 2017).
To test whether the formal properties of the WTH can be used to support claims of late Acheulean aesthetic intent, we investigate: 1) The exceptionality of the WTH's 2D outline shape via geometric morphometric techniques.Outline shape is important for understanding diverse Acheulean phenomena (Lycett, 2009;Lycett et al., 2016), including biface flaking strategies and the extent to which tool-design decisions were driven by utilitarian selective pressures (McPherron, 1999(McPherron, , 2000;;Costa, 2010;Key and Lycett, 2017;Herzlinger and Goren-Inbar, 2020;Shipton and White, 2020;Caruana et al., 2022;García-Medrano et al., 2022;Biermann Gürbüz and Lycett, 2023;Clark, preprint).If the WTH is comfortably located within the shape space of other handaxes from the same East Anglian context then the inference that it was not designed or flaked (reduced) in an exceptional manner, and was instead likely guided by the functional considerations that drove or limited the majority of handaxes (Crompton and Gowlett, 1993;Lycett et al., 2016;Key and Lycett, 2017;Finkel and Barkai, 2018;Wynn and Gowlett, 2018), can be supported.2) Cortex placement and whether the WTH's retention of a cortical central regionwhich includes the shellcan be considered exceptional or unusual.If a large number of alternative East Anglian handaxes also display centrally located cortex then the retention of the shell would conceivably have been part of a regularly observed reduction strategy.3) Flaking intensity on the superior (shell-side) surface of the WTH and whether there is evidence of over-investment by the knapper.If the WTH displays a high number of flake removals relative to other similarly-sourced handaxes (when controlling for size), then superfluous time, energy and care may have been spent producing the tool (Shipton and Clarkson, 2015).Arguably, this could have been to retain and 'frame' the shell.If the WTH displays a similar number of flake removals to other handaxes, then it is difficult to support suggestions of knapper over-investment.Instead, there would be evidence of investment enough to produce a serviceable utilitarian tool, in-line with a majority of other handaxes.4) Whether the thickness and width to thickness ratio ('refinement' [i.e., relative thinness]) of the unmodified tabular flint nodule on which the WTH was made is within the range observed for noncortical (fully flaked) British MIS 9 to 11 Acheulean flint handaxes.
If either is, the knapper may have been under little-to-no pressure to thin the WTH by driving long flakes over the central portion of the tool.Indeed, the most straightforward way to produce a thin late Acheulean handaxe is to select a tabular nodule which is already of suitable thickness and width (Whittaker, 1994).Gowlett's (2006Gowlett's ( , 2021) ) imperatives could then have been achieved via minimally invasive, edge-focused, flaking and shaping.5) The handaxe's internal microstructure using high-resolution microcomputed tomography (micro-CT; μCT).Micro-CT allows the internal structure of the flint to be analysed, revealing features that may have been evident to the Acheulean knapper and impacted their flaking decisions, but are obscured to us due to the curatorial protections given to Palaeolithic artefacts (Abel et al., 2011).Most notably, we are interested in identifying inclusions and fractures within the flint that could feasibly have been visually identified at the point of production (i.e., prior to patination and increased cortex development), or audibly identified during hammerstone strikes (Patten, 1999;DeForest, 2006;Smith et al., 2021).We were also keen to identify how deeply embedded the shell is within the flint, as this could impact the integrity of the flint nodule.6) Whether the handaxe was used as a utilitarian tool prior to its discard.As plausible social functions cannot be substantiated (e.g., Pope et al., 2006 [see also : Key and Lycett, 2017]), we focus on mechanical utility and investigate whether microwear traces created via cutting processes are present on the WTH's flaked edges (See: Ollé and Vergés, 2014;Stemp et al., 2016).

Shape analysis 2.1.1. Comparative sample
To understand whether the WTH's 2D plan view (outline) shape is typical relative to other northern European late Acheulean assemblages, we compared its form to 99 Acheulean bifaces with a similarly vague East Anglian (Suffolk and Norfolk, UK) collection history.Sites include Mildenhall, Lakenheath, Elveden, Weeting, Broomhill, Feltwell, and Shrub Hill.All are curated at the MAA.We investigated handaxes with a similar collection history to the WTH to ensure the sample was not biased toward specific cultural groupings, reduction strategies, or ecological contexts (Lycett et al., 2016;Shipton and White, 2020;Ashton and Davis, 2021).All are known to be from MIS 9, 11 or 13, or would be classified as such based on their form (as with the WTH).When identifying handaxes we gave no consideration to their shapes, but it is important to acknowledge that we were aware of the hypothesis being tested when the sample was selected and all tools were required to be complete (unbroken).We aimed to select a sample where ~50% retained at least one cortex patch measuring >15% of the total surface area of the tool.We did this to facilitate the WTH's comparison to 1) a E. Flanders and A. Key generalised East Anglian Acheulean assemblage (n = 99), 2) more intensively flaked handaxes with minimal cortex (n = 54) and 3) handaxes retaining similar levels of cortex to the WTH (n = 45).The handaxes with minimal and >15% cortex are hereafter known as the non-cortical and cortical assemblages, respectively.Data for the comparative sample and WTH are in Supplementary Data 1 or as described below.

Data collection procedure
The outline shape of the superior surface of the WTH was recorded from the image in Fig. 1A.The superior surface of the comparative handaxe sample (n = 99) was defined as the most convex side of the tool, as this is the surface retaining the fossil in the WTH.High resolution digital photos of each handaxe were recorded using a lightbox and black velvet background (available on reasonable request via the MAA).In each image a scale was placed level with the handaxe's bifacial edge.Shape data were collected via elliptical fourier methods.We refer the reader to Clark (preprint) for detailed descriptions of the shape analysis procedure.
In brief, each handaxe was orientated via its line of maximum symmetry (Costa, 2010;Lycett and von Cramon Taubadel, 2008;Schillinger et al., 2014).Subsequently, each handaxe image was converted into black (handaxe) and white (background) silhouettes and collated into a single.tpsfile from which the outline of each artefact was extracted using tpsDig2 (version 2.32).150 equidistant landmarks were plotted around each outline and each landmark was assigned XY coordinates (starting from the tip).We used a greater number of landmarks relative to previous studies (e.g., Iovita, 2010;Caruana et al., 2022;Courtenay, 2023;Clark, preprint) as in this instance it was found to better convey shape differences within the assemblage.Size variation in coordinate data was removed via 2D generalised procrustes analysis performed in PAST v.3 (Hammer et al., 2001).Subsequently, EFA was conducted using 30 harmonics which were subsequently processed via a principal component analysis (PCA) in PAST.

Cortex location
The 45 East Anglian handaxes forming the cortical assemblage were investigated.In addition to the silhouettes created during the shape analyses these handaxes (along with the WTH) also had their superior surface cortical patches isolated by hand, and a silhouette generated (again using GIMP 2.10.32) (Fig. 2C and D).The two silhouettes could then be stacked on top of each other to create a grayscale image highlighting cortical coverage over the surface of each handaxe (Fig. 2E and  F).With the cortical silhouette images created, the frequency of central cortex retention could be investigated.To do this we used the 'centroid measurement' function within ImageJ (version 1.53) which automatically locates the central coordinate of a selected object; in this case, the complete superior surface of the handaxe (Fig. 2D).ImageJ defines the centroid location as the average of the x and y coordinates of all pixels in the image selected.Once the central coordinate was identified, it was possible to visually identify which handaxes retained cortex at this location (Fig. 2I and J).

Flaking investment
Flaking intensity is a relative measure, and when applied to handaxes, it typically refers to the number of flakes removed relative to a core's volume or surface area (Kuhn, 1991;Shipton and Clarkson, 2015;Muller et al., 2022).In this way, it is a record of the investment (energy, time, risk of breakage [Torrence, 1989]) put into the design and production of a tool relative to the gross size of the object being shaped.For a given core size, the more flakes removed then the greater the flaking intensity and the greater the time and energy investment expended by the tool-producer.Here, flaking investment is calculated as 1) the surface area (mm 2 ) of the non-cortical (flaked) portion of a handaxe's superior surface divided by the number of flake scars observed on that surface and 2) the surface area (mm 2 ) of a handaxe's entire superior surface divided by the number of flake scars observed on that surface.The greater the resulting number, the lower the flaking intensity and the lower the inferred knapping investment.
To do this we used the cortex and plan-view silhouettes created from the cortical assemblage (n = 45) in Section 2.2.Using these images, the area of the cortex and the overall surface of each handaxe was calculated in pixels using the colour threshold and 'measure' functions in ImageJ, with thresholding values being adjusted to select either the cortex, flaked surface, or both (i.e., the whole superior surface area) (Fig. 2G  and H).From these data it was possible to calculate the percentage of cortical and flaked surfaces on each handaxe.The scale bars in the original photos were then imported into the silhouette images and used to inform ImageJ of each handaxe's size, subsequently allowing the previously calculated pixel areas to be converted to mm 2 .We then recorded the number of flakes removed on each handaxe's superior surface.Flake removal counts were collected directly from the artefacts and not from digit photos.Following Lycett et al. (2006) we only recorded flake scars ≥10 mm in length as smaller removals are less easily attributed to intentional flaking decisions.

Nodule thickness and width to thickness ratio ('refinement')
If the flint nodule on which the WTH was produced is of a similar thickness to handaxes in the non-cortical comparative sample (n = 54)where their thickness can be considered hominin-derivedthen there would have been little utilitarian motivation for the WTH knapper to remove the central cortex and shell.The original, un-modified thickness of the tabular flint is preserved across large portions of the tool where the cortex is symmetrically retained on both the inferior and superior surface (compare Fig. 1A and C).We recorded the maximum thickness observed between two cortical portions of these surfaces, with thickness defined as the maximum distance between the superior and inferior surfaces of the tool perpendicular to it's width and length.While we cannot rule out the possibility that thicker portions may have existed at the tool's tip or lateral edges prior to flakes being removed, it is clear that the majority of the WTH is produced from a similarly thin piece of tabular flint.
In a similar respect, if the width to thickness ratio ('refinement') of the original unmodified WTH flint nodule is similar to fully-flaked handaxes, then the knapper would again have been under little-to-no pressure to thin or alter the width of the original flint nodule.This may have resulted in the few shallow (edge focused) flake removals evident on the tool.The width of the original unmodified flint nodule may still be proximally preserved on a portion of the tool (most clearly evidenced on its inferior surface [Fig.1C] and one lateral view [Fig.1B]), but it is likely the nodule was slightly wider towards its midpoint.As a result, we are limited to calculating the width to thickness ratio of the final tool form and not the original nodule.We use the ratio of width divided by thickness to define the relative thinness of the WTH and the comparative handaxe sample (cortical and non-cortical [n = 99]).Width was recorded as the maximum distance on the superior surface of the tool perpendicular to the line of maximum symmetry (Fig. 3).The maximum thickness of the tool was observed between two cortical surface areas and is therefore the same as the original nodule thickness measure recorded above.The ratio therefore represents a maximum value, which would decrease should the original flint nodule be wider than the data used here.

Internal structure via Micro-CT
Micro-CT is a non-destructive technique for analysing the internal structure of an object at the microscale.Widely used in palaeoanthropological and Palaeolithic faunal research (e.g., Skinner et al., 2008;Bello et al., 2013Bello et al., , 2016;;de Jager et al., 2022), its application within lithic-based analyses is substantially more limited (Abel et al., 2011;Falcucci and Peresani, 2022;Macdonald et al., 2022;Stemp, 2023;Key et al., in press).Micro-CT scanners capture a series of 2D planar x-ray images and reconstruct the data into 2D cross-sectional slices, which can later be processed into a 3D model.
Our intention was partly explorative, to see if there were any features present in the flint that may have impacted the production of the tool but are not externally visible.We were also keen to assess how deeply the shell was embedded into the flint, and by extension, whether the shell could have been detached via flake removals without weakening the core's integrity or increasing the likelihood of diving fractures (see: Whittaker, 1994).
Here, we used a Bruker Skyscan 1273 benchtop micro-CT scanner with a 0.3 mm copper filter.The WTH was secured in plastic and scanned with a rotational step of 0.150 and voxel size of 50 μm.Due to the relatively large size of the WTH two partial scans were required before their overlapping segments were digitally stitched together.In total, 2400 slices were collected and reconstructed into cross-sectional images of the handaxe.We used Avizo 3D (v.2022.2) to process the resulting 3D model, allowing the completed 3D model and individual cross-sectional slices to be examined as required.Within the grayscale slices inclusions and fractures could be observed due to changes in said grayscale values, with denser material appearing brighter and less dense material appearing darker.Any inclusions were isolated using the Avizo 3D segmentation editor.Groups of voxels differing in grey value from the principal internal structureand thus representing inclusions of differing density to the surrounding flintwere selected using the 'magic wand' tool and extracted as a new material within the model.Lighter and darker material were segmented separately (Supplementary Information 1) (the external micro-CT model is available on request from the MAA).

Microwear analysis
Microwear analysis seeks to identify whether stone tool artefacts were used to process (usually cut) materials in the past, and ideally, to pinpoint the type of material worked (Ollé and Vergés, 2014;Stemp et al., 2016).Taphonomic processes and the development of patination do, however, have potential to mask or remove microwear traces (Keeley, 1980).Given the WTH was potentially recovered from a fluvial context (Oakley, 1971), and patination has started to develop on the flaked flint surfaces, wear traces may be difficult to identify or accurately record.When combined with the subjectivity of light-microscopy microwear techniques, and the lack of relevant micro-3D comparative databases, we limited our investigations to identifying whether the WTH was used to process materials, and not the type of material that may have been worked.
A Keyence VH-6000 microscope was used to record the 3D

Shape analysis
Principal component 1 (PC1), PC2 and PC3 explained 57.96%, 18.43% and 7.09% of the shape variation observed in the combined handaxe assemblage (n = 100), respectively.As highlighted in Fig. 4, the Fig. 3.The distal tip of the WTH displays evidence of a fracture incurred after the tool was produced (purple shading).A not insignificant amount of material appears to have been removed during this breakage.While it is difficult to predict how much material was lost (dotted white line) and how this impacted the overall shape of the tool, it is clear that the WTH's original form would have differed to the artefact we observe today.The location of the WTH's 'maximum width' is also highlighted, along with the potentially preserved width of a portion of the original unmodified nodule ('nodule width').(For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)E. Flanders and A. Key WTH is comfortably located within the shape space of both the noncortical and cortical handaxe assemblages, indicating its 2D plan-view shape to be typical of East Anglian late Acheulean handaxes.Arguably, the WTH is more centrally located within the cortical assemblage, potentially hinting at a greater degree of shape similarity with other cortical bifaces.PC1 represents variance between more pointed (− 0.2) and ovate (0.2) handaxe forms, revealing the WTH to (unsurprisingly) represent one of the more pointed handaxes in the analyses.PC2 principally describes plan-view skew, with the analyses revealing most handaxes (including the WTH) to display low levels of skewness.
During our morphometric analyses it became clear that the WTH's outline shape (superior surface) varies relative to the form originally imposed by the Acheulean knapper.A ~20 mm fracture with no clear patination is present at the tip of the tool (right side, superior view), and is large enough to indicate that an important piece of flint has been removed (Fig. 3).It is difficult to estimate the tool's original outline shape, but given the high symmetry currently seen on the tool, it is our reasoning that the original outline profile may have been slightly less symmetrical (Fig. 3).Given the patination observed elsewhere, it is reasonable to assume that this fracture occurred after the handaxe was made.

Cortex location
Within the cortical handaxe sample 64% (n = 29) retained cortex at the centre of their superior surface.Thus, it is common for handaxes in East Anglia to retain centrally located cortex.Our analyses support previous texts highlighting the shell to be centrally placed within the handaxe.We do, however, go one step further; we independently calculated the central location of the shell (yellow cross, Fig. 2H) and it lies in very close proximity to the geometric centre of the handaxe (red cross, Fig. 2H).The two points are separated by approximately 1 mm.In terms of cortex patterning, it is clear that the retention of central and proximal cortical regions is not unusual.Eight handaxes from the sample display a similar cortical surface structure to the WTH, with non-flaked basal edges and a central line of cortex running proximally (Fig. 2G).

Flaking investment
On average the cortical handaxes retained lower cortex percentages relative to the WTH (Table 1; Fig. 5).Only nine handaxes (20%) displayed greater cortex retention.Thus, while the WTH retains a large amount of cortex on its superior surface compared to most comparable artefacts, it is in no way an outlier in this regard.This relationship can also be seen in Fig. 6A, where relative to its size the WTH displays a small flaked area; but again, this disparity is observed in a number of other handaxes.
The flaking intensity observed on the WTH is demonstrated to be close to the mean values returned from the n = 45 cortical assemblage (Table 1; Fig. 6B).This result is similar for both the 'flaked surface' and 'whole superior surface' data.Therefore, the flaking intensity observed on the WTHand in turn the investment put into its flaking and shaping processappears to be distinctly average for East Anglian late Acheulean handaxes.

Nodule thickness and width to thickness ratio ('refinement')
The maximum thickness of the WTH is 36 mm.The original nodule thickness record is the same.When compared to the cortical and noncortical handaxes, the WTH is just above the 75% interquartile range, indicating it to be slightly thicker than most, but well within expectations for an East Anglian late Acheulean handaxe (Fig. 7A).The relative thinness of the WTH is comfortably within the 25-75 interquartile range of the comparative handaxe sample (Fig. 7B).

Table 1
Descriptive data concerning the flaking intensity observed in the cortical handaxe assemblage along with the same traits observed on the West Tofts handaxe.

Internal structure via Micro-CT
A high-resolution 3D model of the WTH's external and internal structure was created (Fig. 8).2D slices through the handaxe highlight the heterogeneity of the internal raw material, which was strikingly diverse.Contrary to our initial prediction, there was no distinction in density between the flint and fossil shell, making it difficult to determine the depth at which the latter is embedded.An air pocket 2 mm beneath the surface of the shell and one third of the way from its most distal tip was, however, evident (Fig. 8A).Notably, the air pocket is level with the lowermost edges of the shell as they make contact with the flint's cortex.The presence of the air pocket inline with the edge of the shell (white dotted line, Fig. 8A) suggests the void may have been created when the shell was fused to the nodule during the flint's formation.In turn, it may depict the original contact surface between the flint and fossilised shell and indicate the fossil to be superficially located on the flint nodule.
Somewhat unexpectedly, the micro-CT scan revealed other features concerning the quality of nodule from which the WTH is produced.Fig. 9D and H highlight the frequency with which additional low density inclusions (voids) are present within the handaxe.The majority are small and spherical in nature, which is consistent with the presence of small volumes of air that are created naturally during the formation of some flint bands.
Importantly, a relatively substantial natural fracture is also present within the flint.At a depth of 6.5 mm and located just above the shell, a thin rhombus shaped hairline air volume is evident (Figs.8B and 9D).Examination of successive 2D slices and the 3D segmentation (Supplementary Material 1; Fig. 9D and H) reveal this to most likely be a fracture that runs beneath but parallel to the superior surface.It is difficult to identify how large the fracture is as the micro-CT scan will only identify air voids and the presence of the fracture between two touching surfaces will not be visible.The fracture can be visibly confirmed as being at least 13.3 × 6.5 mm in size, but we anticipate it being bigger.An additional but slightly smaller fracture is also present in the proximal base of the tool and is most clear in the bottom right hand side of Fig. 9D.It is again characterised by a hairline air volume that is roughly square in shape.
High density objects were also clearly visible within the flint (Fig. 9C  and G).As with the air bubbles there are a high number of small inclusions that may feasibly be small quartz crystals or pebbles of other rock types that became trapped in the flint as it formed.These are scattered broadly evenly through the flint.Of most interest, is the presence of at least one other fossil shell just beneath the larger surface shell.As highlighted in Figs.8C, 9C and 9G, but most clearly seen in Supplementary Material 1, the shape of this high-density inclusion matches a small bivalve with a relatively steeply curved shell.The shell is too small to match the superior surface shell.Proximal to the surface shell another possible fossil shell inclusion is present (Fig. 8C).Again, this is a high-density inclusion, but presents as a flat roughly crescentshaped object (Supplementary Material 1).Potentially, it is a fragment of a larger shell that became trapped in the flint as it formed.Alternatively, it could be a thin fragment of stone.

Microwear
Microwear traces in the form of striations and abrasive wear (polish) were present on the edges of the WTH.Striations were the most obvious feature and in a few instances were evident in spite of the presence of patination (Fig. 10C and D).This indicates that hard materials scraped across the edge of the WTH.Potentially, this was during the tool's use, where small chips of flint or other grit could have been pushed into the tool's surface during cutting actions.Alternatively, stones and other sediment may have been pushed against the WTH after it was discarded, likely during fluvial movement.One point in favour of the former explanation for some traces is that multiple parallel striations were at times present, indicating repetitive linear motions.
The development of patination hindered the ease with which abrasive (smoothing) wear traces could be identified.Nonetheless, some limited evidence was present (Fig. 10A) with the few 'clear' instances being located in the distal half of the tool.These were few and far between, but are suggestive of the presence of use-derived polish.Abrasive wear was observed on the raised ridges between flake scars, which could easily be derived from post-depositional movement.Occasionally, there were indications that wear resulted in the removal of patination (Fig. 10B), which would strongly suggest post-depositional alteration.
Microfractures (microflakes) and edge rounding were the clearest and most frequent micromorphological features (Fig. 10E to H).Both could be observed around the circumference of the tool.Edge rounding was at times visible on surprisingly thin sections of edge (Fig. 10E); had this rounding been created during fluvial processes one would have expected the edges to snap.Some edge portions displayed particularly high micro-fracturing densities.Moreover, in some of these high-density areas the damage resulted in concave and/or blunted edges (Fig. 10H).

Discussion
The West Tofts handaxe is unusual among curated Acheulean bifaces.Its large well-preserved fossilised shell near-perfectly centred and flush with a handaxe's superior surface may even be unique.Singularity and distinctiveness in an artefact's formal properties is not, however, tantamount to artistic expression or aesthetic intent (Straffon, 2019;Bebber, in press).Aesthetic value and artistic appreciation on the part of present-day archaeologists, artists, art historians or other specialists is similarly uninformative concerning the motivation of an alternative, potentially non-ancestral hominin species c. 300,000 to 400,000 years ago.To be clear, there is no doubt the WTH raises aesthetic sentiments for some present-day individuals (Oakley, 1973;Hayden, 1993;Hodgson, 2005;Wynn and Berlant, 2019), or that aesthetic intent and appreciation has a potentially deep prehistoric origin (Gowlett, 2011;Straffon, 2019;Gowlett, 2021;Martín-Loeches, 2023).Nor do we question the tool's outline symmetry or the fossil's central location as sought after artistic criteria (Bebber, in press).
Our focus here has been to assess whether the Acheulean individual responsible for producing the WTH was motivated by aesthetic and artistic design principles, or alternatively, whether other explanations for its formal properties exist.Our data lead to the conclusion that the WTH's design was most likely not guided by aesthetic intent.Instead, on the balance of evidence, it was most likely expediently produced in response to the utilitarian pressures informing the overwhelming majority of handaxe production events during the British late Acheulean (Roe, 1981;White, 1998;Lycett et al., 2016;Wynn and Gowlett, 2018;Kuhn, 2020).We therefore follow Chase and Dibble (1987), Nobel and Davidson (1996), and Gowlett (2021) and call for restraint when using the WTH to infer the presence of aesthetic intent in the Lower Palaeolithic.
The data outlined above demonstrates that the WTH: 1) Is unexceptional in terms of its plan-view outline shape relative to a large sample of alternative East Anglian late Acheulean handaxes.
Although more pointed than some forms, it is comfortably located within the shape space of both cortical and non-cortical samples.While few would claim the WTH to be unusually shaped, this finding supports the tool's form being imposed according to the functional pressures guiding and limiting the overwhelming majority of Lower Palaeolithic handaxes.If retaining and centrally 'framing' the shell was the overriding concern during the tool's production, then its shape is more likely to have been exceptional in some way; be it as a clear outlier or just further removed from centrality.In this scenario, cultural mechanisms could still have guided towards a 'typical' handaxe form (Lycett et al., 2016), but due to the technological demands of retaining and centrally locating the shell, the outline shape could easily have deviated (consciously, unconsciously, or by mistake) without there being a cost to the main shell-focused aesthetic goal.Quite unexpectedly, it became clear that the tip portion of the WTH is damaged and a portion of flint has been removed.It is difficult to say how this impacted the tool's plan-view shape and symmetry, but it is clear the tool's original form would have been slightly different to what we see today.Future work should take this into consideration.Further, it reinforces the need to consider whether the handaxe went through several stages of shaping (Dibble et al., 2017).It is entirely possible that an earlier form of the WTH did not place the shell at its centre.The tool was reported to be 5″4 inches (137 mm) long when purchased by the MAA in 1914 (Cambridge Antiquarian Society, 1915), which is very close to its present length (132 mm), meaning that any damage at its tip likely occurred before its curation.2) Does retain the fossilised shell at its geometric centre.For the first time, we also provide evidence that the geometric centre of the shell is in the same location.Intentional or not, it is remarkable that the two points align within a few millimetres.This does not, however, add much to our understanding concerning aesthetic intent as these findings are in-line with previous discussion points (Hayden, 1993;Hodgson, 2006;Bredekamp, 2019;Wynn and Berlant, 2019).If, in the future, others demonstrate the frequency with which fossils, holes, pebbles and other inclusions are centrally framed in handaxes under neutral selective expectations (Lycett, 2008), then it would be possible provide additional commentary on this feature.More interesting, perhaps, is the fact that 64% of the cortical comparative handaxe sample also retain cortex at their geometric centre.The WTH is, therefore, far from exceptional in this regard.In roughly half of these instances, the central cortex remains despite flake removals having being struck from two or three directions (as seen in the WTH).If retaining the central cortex (and by extension the shell) was a conscious decision by the flintknapper, then it was feasibly part of a widely observed reduction strategy focused on producing a length working edge and minimising additional superfluous flake removals, as often observed on pointed handaxes (Wymer, 1968;White, 1998;Zaidner et al., 2006;Herzlinger et al., 2021).3) Is distinctly average in terms of flaking intensity and the knapper's inferred investment.It does display a relatively high level of cortex retention (38%), but 20% of the comparative sample was higher still, meaning this feature cannot be used to support suggestions of aesthetic intentionality without other, non-shell retaining bifaces, being treated similarly.In total, 19 flake scars above 10 mm in length were observed on the WTH's superior surface.This could be considered low compared to some British Acheulean assemblages (Shipton and Clarkson, 2015;García-Medrano et al., 2019), and when considered in relation to its flaked and superior surface areas (i.e., flaking intensity), the WTH is close to the cortical comparative sample's median values.Thus, the WTH does not appear to be over-engineered.Insofar as we can tell from flake scaring patterns, no greater time or energy was invested in this tool relative to the late Acheulean norm.Of course, it is difficult to assess the care dedicated to each flake's removal, but the highly sinuous edge, the presence of multiple step and hinge terminations, and the asymmetric, seemingly-hard hammer-shaped (i.e., generally thick and short), flake removals could hint at a lower level of consideration.The WTH could even be considered expediently produced and used (Shipton and Clarkson, 2015), potentially by an individual of average skill (contra Bredekamp [2017;2019]) (Stout et al., 2014;Pargeter et al., 2019); at least compared to many other British Acheulean handaxes (Roe, 1981;White, 1998;Shipton and Clarkson, 2015;García-Medrano et al., 2019;Ashton and Davis, 2021).Although, importantly, a lack of knapping skill is not conclusively tantamount to a lack of aesthetic intentionality.Simply, there is little evidence to suggest greater time, energy or care was taken when shaping the WTH relative to any other flint biface.4) Was produced on a flint nodule only marginally thicker than the average East Anglian Acheulean handaxe.Likewise, the final tool's relative thinness ('refinement') is comfortably within expectations for the same comparative dataset.It is, therefore, clear that the tabular nodule on which the WTH was produced did not require the removal of centrally invasive 'thinning' flakes.Simply, there was no need to remove the central cortex, and in turn, the fossil.Given the cortex retained on the final artefact (see 'original nodule' width), the original nodule may not even have been much wider than the finished tool form.Meaning there would also have been little need for the knapper to alter the width of the tool.The nodule was already thin enough, and potentially wide enough, to meet Gowlett's (2006) imperatives and create an effective, ergonomic, utilitarian tool.In fact, it is difficult to think of a more perfectly shaped nodule from which to easily and expediently create a serviceable handaxe.Only a few shallow, edge-focused flakes needed to be removed, and shallow, edge-focused flakes is what we see on the tool (bar one proximal flake removal on the inferior side).It is possible, therefore, that the tabular nature of the flint was the overriding consideration for the hominin when initially selecting the flint nodule, not the presence of a shell.5) Is produced from a surprisingly heterogenous flint.The WTH contains numerous small low-density (likely air) volumes and highdensity inclusions, several internal fractures, and at least onepotentially twoadditional fossil shells.Potentially, the air pockets derive from organic and soft inorganic structures embedding within the flint as it formed, before subsequently decaying/disintegrating.The high density features may be fossilised shell fragments, crystals, or other small stones.
East Anglian flint varies highly in terms of inclusions and past exposure to fracture initiation (thermal shock, fluvial activity, etc.), so it is not unexpected to see these features.What is intriguing is their potential role in the tool's production.Stone nodules containing fractures and large inclusions can be audibly distinct from more homogenous alternatives.When struck with a hammerstone or softhammer billet, more homogenous flint should produce a higher pitched sound for a longer duration (Patten, 1999;DeForest, 2006;Smith et al., 2021;DeForest and Lyman, 2022).Without striking the WTH it is difficult to test, but the knapper possibly became audibly aware of the presence of fractures and inclusions in the flint.This feasibly explains why the tool retains cortex (including the shell) and displays a sinuous edge with only six large flake removals on its superior surface.Had the tool been flaked more heavily, the risk of an internal fracture being reached, and the core breaking, would have increased.Thus, the knapper created a serviceable handaxe, but may have decided to keep flake removals to a minimum (see also below).This does raise the question of why a lower quality nodule was initially selected, if not for the presence of the shell.Nevertheless, the thin tabular nature of the flint would have been highly desirable and the knapper evidently had the skill to make a pointed, edge worked tool irrespective of any structural flaws.
We also reveal the fossil shell to be superficially embedded into the flint's surface, meaning it could have been removed via the detachment of reasonably thick flakes.However, had this occurred, it is very likely that the largest internal fracture, located just above the distal tip of the shell and at a depth of 6.5 mm, would have been reached.Patination makes it difficult to see whether this feature was externally visible at the point of production, but if the knapper was aware of the fracture's presence, it could easily have deterred any central flake removals.
Further, to set-up a thick flake removal capable of 'getting under' the fossil, a steep platform at the tool's edge would have been necessary, which would have reduced the tool's breadth and, in turn, relative thinness (Fig. 11).Had the removal of such a thick flake been successful, the likelihood of the tool breaking during transport and handling would also have increased (Fig. 11).It would also have made the tool considerably smaller, which is important given the WTH can already be considered a small Acheulean handaxe.
Additionally, the micro-CT scan reveals the shell to be situated in a concavity, meaning the centrally-directed flakes that were removed had no option but to terminate where they did.In turn, there is every possibility the knapper did attempt to remove the fossil shell, but the concavity prevented flakes from reaching it (Fig. 11).If this were the case, the retention of the shell and its cortical surrounding would reflect a lack of skill and understanding concerning the establishment of platforms appropriate to remove the shell.The presence of the shell in a depression could equally have caused the knapper to purposefully avoid detaching long, centrally invasive flakes.The knapper had no way of knowing how deeply embedded the shell was, and given the flint was so thin, had a flake detachment reached the shell, it could have broken the tool in two (Whittaker, 1994).
In sum, the shell can be considered a structural flaw in the raw material.Intentional avoidance of the shell, combined with the availability of all other areas for flaking, could have led to the fossil's near-perfect central location without there needing to be an aesthetic intent.Effectively, the knapper would have been free to flake around the shell, including the asymmetric removal of flakes/material from each surface (as seen on the WTH), but would have been under pressure to make sure flakes never reached it.In turn, the retention of the central cortex and/or the shell was possibly a strategy to avoid breaking the biface.If this were the case, it would be a fascinating example of a hominin being aware of structural limitations in a core and still producing a serviceable tool.Alternatively, the knapper may have actively tried to remove the fossil, but each flake naturally terminated at the central concavity's perimeter (Fig. 11).6) Retains microwear traces feasibly linked to its use as a cutting tool.
Importantly, no one feature by itself is conclusive.Abraded surfaces, linear striations, rounded edges and localised microfracturing can easily be produced through natural processes.The abrasion is, however, very localised and limited, multiple parallel striations are often present, and the microfractures are more heavily concentrated on some areas relative to others.The irregular, localised nature of the abrasion and microfractures could lean away from a natural cause, while the multiple parallel striations could lean towards their production through repetitive cutting motions.The presence of rounding on thin, easily broken edges could suggest its creation through cutting activities.But again, it is impossible to rule out the impact of natural taphonomic processes.Whatever the case, these features can generally be considered 'light' (with the exclusion of some microfracture concentrations).
Overall, microwear evidence from the WTH is not conclusive and the features discussed above may plausibly all be naturally derived.
Equally, there is reason to think that some traces could stem from the WTH's use as a cutting tool.If the latter were the case, the limited traces would suggest the tool was used for a relatively brief period before being discarded.If the tool did have a short use-life trajectory, it questions why a hominin would invest in the tool's aesthetics.
These data highlight multiple alternative, non-aesthetically derived routes for the fossil shell to have been retained, and even centrally located.Some support the intentionality of the fossil's retention, but build on well established theory concerning the utilitarian and mechanical (conchoidal fracturing) mechanisms guiding stone tool design and production during the Acheulean.Others present the shell as a redundant feature, with the tool's design being influenced by other factors and the fossil's presence and centrality being incidental.These lines of evidence are arguably more persuasive than the shell being retained and centrally 'framed' because of its aesthetic value.
Particularly when little indisputable evidence supports the presence of aesthetic intent in the Lower Palaeolithic (Straffon, 2019;Gowlett, 2021;Martín-Loeches, 2023), and any objective evidence in support of aesthetic design intent or choice (see: Kuhn, 2020) is lacking.Hayden's (1993: 123) claim that the WTH's form, proportions and symmetry "exceeded functional requirements" is vaguely described (see also : Currie, 2011;Bredekamp, 2017Bredekamp, , 2019)), but nonetheless, we want to highlight how the above data demonstrates the opposite.It is remarkable how the WTH conforms with a large number of alternative handaxes in multiple functionally-linked formal properties (see: Key and Lycett, 2017).To the extent that we cannot pick out a single featurebesides the shellwhich deviates from a utilitarian explanation or human behavioural ecology expectation.Be it tool shape, flaking intensity, nodule selection, edge sinuosity, cortex symmetry and relative thinness.Even the tool's outline symmetry, which is reasonably high, will match a large number of other British late Acheulean handaxes (Cole, 2015;McNabb and Cole, 2015); most of which will have been produced to complete cutting tasks.Moreover, the shell may have even had a social utility removed from any aesthetic consideration (e.g., Pope et al., 2006;Straffon, 2019).If this study has demonstrated anything, it is that the WTH is remarkably average in terms of its functionally-linked morphometric properties.Fig. 11.A micro-CT slice demonstrating hypothetical flake removals that could potentially have removed the fossil shell, but would have resulted in a substantially smaller tool (top left, centre).Depicted in the bottom left and on the right are the realised flake removals surrounding the shell.In each instance, the flake terminates as it reaches the central concavity which holds the shell.The knapper may have been trying to remove the shell, but because it was located in a concavity, each flake had no choice but to terminate just before reaching it.
A number of additional lithic-relevant principles of art and design have not been investigated here (including composition/balance, rhythm/movement, proportion/ratio, colour [Bebber, in press]).However, many of these features may be widely observed in flint Acheulean handaxes, as they are in other biface tools (Bebber, in press).Discussion of the WTH's exceptionality in these regards could, therefore, be limited.It is well known, for example, that a large number of handaxes have length/width ratios approximately adhering to the 'golden ratio' of 1.618, a form considered by some to have aesthetic qualities (Pope et al., 2006;Gowlett, 2011).
In light of the 'added value' messaging potentially observed in some extreme handaxe forms (Mithen, 2003;Cole, 2017;Wynn and Gowlett, 2018;Wynn and Berlant, 2019;Gowlett, 2021), including during the British late Acheulean (White and Foulds, 2018;Gowlett, 2021), it is important to ask why the WTHif it is an aesthetic object meant to capture attention and provoke visual feedbackdoes not display similar attributes.It would not have been difficult to remove additional flakes and/or cortex, make the cortex symmetrical (both the inferior and superior cortex are decidedly asymmetrical), more heavily 'refine' the tool's edges, or create an outline form that stands out relative to most handaxes (for example, an extreme ficron).Simply, why display the shell on an "otherwise … not particularly remarkable" (Wynn, 2018: 20:23) handaxe when more visually distinctive and/or appealing forms are well known from the British late Acheulean?Gowlett (2005) highlights one potential answer by demonstrating morphologically extreme handaxes to only be extreme in one or two-related attributes.But given the shell is a naturally occurring feature and not a hominin-imposed morphometric attribute, the applicability of these findings to the WTH is unclear.
Finally, the WTH is not unique in preserving a shell, hole, fossil, crystal, or inclusion negative on its surface.Occasionally, these features are centrally located (Oakley, 1973(Oakley, , 1981;;Hayden, 1993;Wynn and Berlant, 2019;Gowlett, 2021), even if the WTH can be considered unique in how all of its natural and hominin-imposed formal properties combine.Wynn and Berlant (2019) and Jelínek (2001) interpret the repeated but very low-frequency of biface central inclusions as hominins accentuating and drawing attention to these features.While we do not discount the potential for Middle Pleistocene hominins to have such capabilities, we interpret the temporal and spatial porosity of these artefacts differently.In light of the hundreds of thousands of handaxes that have now likely been discovered, and the fact inclusions are not rare in many conchoidally fracturing rock types, it is overwhelmingly likely that a small subset of this enormous sample will display centrally framed inclusions (Nobel and Davidson, 1996;Gowlett, 2021).It is for this reason, we believe, that despite a multiple-continental tour of museum collections, relatively few centrally-framed inclusions appear to have been discovered by Wynn and Berlant (2019;Wynn, 2018).To our knowledge, no individual sites have multiple artefacts displaying highly central, potentially intentionally 'framed' inclusions.As a result, claims in favour of inclusion-based aesthetic intent draw on similarities between a few poorly contextualised handaxes potentially produced >100, 000 years apart and often discovered in different countries (if not continents) (Oakely, 1981;Hayden, 1993;Wynn, 2018;Wynn and Berlant, 2019).
Given how easily skilled knappers can determine the final form of handaxes (Eren et al., 2014), if centrally framed inclusions were aesthetically important to Acheulean hominins, and neurologically and sensorily derived feedback was strong enough to influence tool-production decisions (White and Foulds, 2018;Hodgson, 2023), then we should expect to see a higher frequency of these artefacts in the archaeological record, and at least some sites with multiple clear examples.Should a Palaeolithic site display multiple instances of this phenomena at rates above that naturally observed in relevant raw material sources, then a stronger case for aesthetic intent can be made.This does not mean the hominin who made the WTH could not appreciate visually appealing tool forms or gain some form of visual pleasure or stimulus (Hodgson, 2006;White and Foulds, 2018;Gowlett, 2021;Bebber, in press) as a side product of the tool's production, but it does mean the central retention of the fossil can be viewed as a statistical inevitability.

Conclusion
The West Tofts handaxe is a remarkable artefact.It is just not reliable evidence for the presence of aesthetic intent or artistic capabilities in the Lower Palaeolithic.The above data and discussion highlight how there are numerous, diverse additional explanations for why the handaxe retains a centrally located fossil shell.Many of these explanations align with well established theory and evidence concerning the design and production principles adhered to by late Acheulean hominins.If aesthetic intent and the intentional 'framing' and placement of the fossil shell is to ever again be presented as fact, then the evidence underpinning these alternative explanations needs to be objectively and robustly challenged.To be clear, none of the evidence we present can falsify the aesthetic intent hypothesis, but there is clear evidence in support of other explanations.Of course, aesthetic, functional and other pressures are not mutually exclusive (Machin, 2009;Lycett and von Cramon Taubadel, 2015;Wynn and Gowlett, 2018), and similarly to Chase and Dibble (1987) and Gowlett (2021) we are not ruling out the possibility of mid-to-late Acheulean hominins displaying aesthetic intent.Parsimony suggests this may even be likely.Nevertheless, on the balance of evidence the West Tofts handaxe should not be used to support such a capability.At the very least, debate must move away from a purely aesthetic-focused perspective.In sum, we present evidence demonstrating the WTH to be morphologically and technologically unremarkable, incomplete, and produced on a structurally flawed but conveniently thin flint nodule.Potentially, it was used as a cutting tool and the knapper may even have attempted to remove the fossil.Together, these data reinforce the likelihood of the shell's retention and central location being an incidental byproduct of an Acheulean hominin's intention to produce a utilitarian tool.

Declaration of competing interest
None.

Fig. 2 .
Fig. 2. The step-by-step process involved in the identification of each cortical handaxe's centre point (red crosses), along with the extent of cortex coverage on the superior surface (A to G; see Section 2.2).Steps K and L highlight (in red) the calculation of the cortical surface are relative to the overall superior surface area (Section 2.3).Note the centre of the shell in yellow (H).Image G illustrates the cortex placement observed on the 45 East Anglian handaxes forming the cortical comparative sample.(For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

E.
Flanders and A. Key   microstructure of the handaxes' edge.All bifacially flaked portions of edge on the superior and inferior sides of the tool were initially investigated at 50x magnification.At this stage, our intention was to identify whether any wear traces were present and where they were located.Attention focused on the 20 mm of surface closest to the edge apex.If any microwear traces were identified, their location was noted.Portions of surface with visible wear traces, be it linear striations or patches of abrasion (smoothing), were subsequently scanned at a higher magnification.In some instances, a 3D model of the surface's structure was recorded, along with high resolution images being captured.Depending on the size and extent of the wear, we used magnifications of 100× to 1000×.If these wear traces deviated from the natural (unworn) surface morphology observed on the flint then the presence of utilitarian use-related wear traces on the WTH can be supported.It was not our intention to record every striation or patch of abrasive wear in detail, but instead to determine whether use-related microwear traces are present on the WTH, and in turn, whether it may have been used as a utilitarian object during the Acheulean.

Fig. 4 .
Fig. 4. PCA results from the shape comparison between the WTH (red diamond) and the cortical (square, dark teal) and non-cortical (circle, pale teal) East Anglian handaxes.(For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 5 .Fig. 6 .
Fig. 5. Plot depicting the number of flake scars observed on the superior surface of each cortical handaxe relative to the tool's flaked surface area (%).The WTH is indicated in red.(For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 7 .
Fig. 7. Jitter plots displaying the thickness (A) and relative thinness ('refinement') (B) of the WTH relative to the cortical (squares) and non-cortical (circles) comparative handaxe samples combined.For each, the 25-75 interquartile range defines the box and the line represents the median value.A scatter plot of width against thickness is also displayed (C).In all instances the WTH (red cross) displays values towards the sample's central tendencies.(For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 8 .
Fig. 8. Micro-CT scan of the WTH. Figure A is a 'slice' taken through the shell.Visible 2 mm beneath the shell's surface is a small air pocket.Figures B, C and D highlight the presence of an internal fracture (B), a second shell fossil (C), and a potential third shell fossil (D), respectively.The dotted white line in image A highlights the predicted original point of contact between the fossil shell and flint surface.

Fig. 9 .
Fig. 9. Visualisation of segmented volumes from within the WTH.Images A and E demonstrate the position of the handaxe in the subsequent images.Images B and F highlight all volumes segmented; this includes high and low density materials as described in Section 2.5.Segmentation of high-density material is highlighted in C and G, while D and H depict low density materials (likely air pockets).Note the two larger pink curved flat surfaces in image C as these highlight the two potential internal fossil shells.The large flat square shape in the top of image D highlights the largest internal fracture.(For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 10 .
Fig. 10.Composite image illustrating micromorphological features present on the edge of the WTH.Image A demonstrates abrasive wear indicative of use (cutting) or taphonomic processes.Image B illustrates the role of taphonomic processes in the creation of some features via evidence of patination having been removed.Striations in the surface of the flint can be seen in images C and D, while edge rounding is evident in E and F. Images G and H demonstrate examples of microchipping present on the edge of the tool, including those with evidence of patination.