This paper examines the interplay between representational and non-representational art in later prehistoric Europe, aiming to form a starting point for more detailed and expansive study on this topic. We will approach later prehistoric imagery from an angle that has not been extensively explored, focusing on the deliberate use of ambiguity and the occasional appearance of representational images – particularly human bodies and faces – against a background of predominantly abstract and geometric imagery. We will hone in particularly on the imagery of Neolithic Britain and Ireland and the imagery of the European Iron Age, drawing on examples from existing research to establish new questions, and focusing especially on the affective potentials of faces and other bodily elements in later prehistoric imagery. We argue that images should always be viewed as being ‘in the making’ and we consider how Neolithic and Iron Age images became representational, emphasizing their emergent and ambiguous characteristics.

The visual culture of later prehistoric Europe presents a conundrum. While much of the imagery archaeologists encounter is non-representational, alongside this we observe the emergence of images of bodies, particularly of faces. How are we to understand this phenomenon? And how was this early representational imagery perceived and understood? In this paper we will discuss the occurrence of faces in later prehistoric art against the wider backdrop of non-representational patterns, and consider the context and role of the face within a category of ambiguous images. We will draw on existing research on the visual culture of the British and Irish Neolithic (c. 4050–2300 BCE) and the European later Iron Age (from c. 450 BCE until c. 100 CE), using key examples to illustrate our observations and provide a starting point for more in-depth and systematic comparison.

The ideas presented in the paper broadly relate to decorative practices in later prehistoric Europe. ‘Later prehistory’ refers here to the Neolithic, Bronze Age and Iron Age, while ‘Europe’ refers to the landmass covered by the modern-day continent of Europe. The periods of time covered by later prehistory vary across Europe, particularly in Scotland and Scandinavia, which experienced ‘the long Iron Age’. During this long period of time and across a huge area, the making and use of art underwent much change. Nonetheless, we have observed potential commonalities in the way that images were conceptualized and presented, perhaps relating to broadly similar philosophies and ontologies.

The examples given in the paper relate to two more specific cases within these broad parameters: the Neolithic in Britain and Ireland (c. 4050–2300 BCE). The examples from the European later Iron Age relate to artefacts traditionally included in the category of La Tène art, which centred on Central Europe from c. 450 BCE, and spread to other areas of Europe during the later first millennium BCE, particularly westwards to Britain, Gaul and Iberia, and eastwards into Eastern Europe (although its boundaries are, of course, blurry in every sense). The term Early Celtic Art, which is often used as a synonym for La Tène art, will not be used in this paper, as there is not space to adequately summarize the issues with this term (but see for example Chittock 2014, 2021:10; Collis 2003:71–92; Gosden & Hill 2008).

The word ‘art’ is used in this paper as a form of shorthand, to refer broadly to traditions of mark-making, although both authors have written on the issues with viewing prehistoric objects, patterns or images as ‘art’ in its modern, Western sense (Chittock 2014, 2020, 2021; Jones & Cochrane 2018). More specifically, we also use the terms ‘pattern’ and ‘image’ to refer to the elements that make up the traditions we are discussing, focusing particularly on the apex between these two elements. We use ‘image’ to refer to visual representations of things and beings. ‘Pattern’ is borrowed from Cyril Fox’s seminal work, Pattern and Purpose: A survey of Early Celtic Art in Britain (1958), and in this paper it refers specifically to the non-representational. While Fox never explicitly defines the word in the context of his work, it has proven useful in describing arrangements of motifs, as it is a more specific and less loaded term than alternative possibilities, such as decoration or ornament (see Chittock 2021:20–21).

Lastly, we need to address one notable gap in our discussion: Bronze Age art. This gap occurs because substantial parts of our discussion focus on Britain where Bronze Age imagery is notably sparse unless we consider cup and ring marked imagery and the few depictions of weapons in mortuary and monumental contexts (Bradley 1997). For other regions of Europe, such as Scandinavia, Iberia and Central Europe, however, there are rich traditions of imagery during the Bronze Age (Ling et al. 2024). The regional diversity of European Bronze Age art means that incorporating all of them into our discussion would necessitate a far longer paper. The relationships between image and pattern in each tradition would require its own examination. For this paper, an analysis of the British and Irish Neolithic allows us to achieve our aim.

The authors were involved in two different research projects that analysed visual materials from the British and Irish Neolithic and the European Iron Age. The ‘Making a Mark’1 project analysed three key regions in the Neolithic of Britain and Ireland (Jones & Díaz-Guardamino 2019), while the ‘European Celtic Art in Context’2 project, analysed the archaeological and geographical contexts of ‘Early Celtic Art’ (Nimura et al. 2020). Both projects provided extensive overviews of visual materials from these two periods and locations. In both regions and periods there is a predominance of geometric and curvilinear imagery, but in both regions and periods we also observe the emergence of figurative representational imagery. As the two authors have previously analysed this imagery separately, this paper instead offers a comparison of the two regions and periods and a preliminary analysis of the processes involved in the emergent practice of representation.

The image traditions of the British and Irish Neolithic and European Iron Age are visually arresting but difficult to comprehend, being made up of three-dimensional geometric and curvilinear designs, often with no obvious representational dimensions. Recognized images of people, animals and objects make up a small proportion of Iron Age art (Chittock 2020:79), whilst in the Neolithic they are almost absent, with just a few images that might be regarded as humans and only a single set of images of deer (Valdez-Tullett et al. 2022). Despite this, inspired by the post-Renaissance traditions of representational and allegorical imagery (see for instance Eco 1959), many archaeologists have been keen to see clear imagery in designs whose representational elements are far more elusive and ambiguous. For the Neolithic period, the spectacular passage tomb art of Ireland has been regarded as representational. For example, in Co. Meath, stone 7 at the Fourknocks I passage tomb has been described as an image of a face (Hartnett 1957), and George Eogan (1986:32) described the design on the stone at the juncture between passage and chamber of the W tomb at the great Knowth passage tomb as a ‘ghostly guardian’. More recently, a design on the great Newgrange passage tomb, Co. Meath, has been regarded as resembling a whale (Hensey 2015:79–94). Meanwhile, Tim Darvill and colleagues (2005) have argued that certain incised lines on the Cronk Yn How stone, Isle of Man are images of deer.

In Iron Age art, there is a small figurative element that is seen as clearly representational, with humans and other animals depicted in bronze, gold, stone and sometimes bone; as stand-alone statues and figurines and also as elements of tools, weapons and personal ornament (see for example Aldhouse-Green 2004; Chittock 2020; Megaw & Megaw 1989). This category, however, is blurry and there is a far greater number of ambiguous images concealed within arrangements of motifs, with archaeological work dedicated to finding, identifying and discussing these hidden images (Foster 2014; Foster & Batten 2022; Megaw 1970).

In the contexts of the assemblages of later prehistoric images and patterns discussed in this paper, we must acknowledge the objects that are missing. The surviving evidence comprises a sample of the most durable finds, but the potential for images and patterns to have been created in organic materials that have since decayed is very wide-ranging. Bryony Coles’ survey of wooden figures from Britain and Ireland (1990), dating broadly from the middle of the Neolithic to the middle Iron Age, presents a small sample of anthropomorphic images made from a widely-available and easily workable material, which, in most taphonomic conditions, will not have survived. This work is simply one example that demonstrates the potential for there to have been large and varied assemblages of objects and structural remains made from organic materials and bearing images and patterns, which have not survived in the archaeological record.

It is not our aim to dispute or debate claims for representational imagery in later prehistory, especially given the recent discoveries of deer imagery at Dunchragaig cairn, Argyll, Scotland, a region with predominantly nonrepresentational rock art (see Valdez-Tullett et al. 2022). Instead, we aim to understand later prehistoric representational imagery contextually, against the background of the non-representational art it coexists with. As part of a broader program that understands images as always being ‘in the making’ (Back Danielsson & Jones 2020; also see below) our focus in this paper is on how images become representational against a background of largely non-representational mark making and imagery. We do this by examining the contexts of images, looking at images as components of objects. Rather than presenting representational imagery as inevitable, our analysis instead emphasizes representational imagery as an emergent phenomenon, where transformation occurs within a meshwork of representational ambiguity: we particularly highlight the instability of these early forms of representation. Just as the Egyptologist Rune Nyord (2021) questions what constitutes an image in Ancient Egypt, so too our analysis troubles comfortable assumptions about what constitutes a representational image in later prehistory (see also Wilkinson 2022).

Traditional archaeological studies of later prehistoric art have focused on seeking meaning within representational images, following art-historical traditions of interpretation. In an Iron Age context, there has often been a tendency to interpret anthropomorphic images as deities, largely based on written and epigraphic evidence from Roman-occupied Europe (Aldhouse-Green 2004; Green 1995:466). Hybridized and modified human images have, in the past, been referred to as named characters from Greek and Roman mythology: Janus heads and satyrs for example (Jacobsthal 1944; Megaw 1970:263;). It is sometimes argued that Iron Age deities took on the human forms of Roman Gods as Mediterranean traditions of iconography influenced Iron Age art during the later Iron Age (Green 1995:466). However, earlier Classical accounts from the later first millennium BCE are also interpreted as stating that Iron Age Europeans did not make images of their Gods (see for instance Diodorus Siculus, Library of History, XXII, 9). The biases inherent in Classical sources make this a complex issue, and the assumption that particular kinds of images shared the same meanings over large and culturally variable areas is problematic. For the British and Irish Neolithic there is less opportunity to draw on Classical literature, though in ‘The Eye Goddess’ O.G.S. Crawford (1957) attempted to utilize Near Eastern sources as a springboard to interpret the imagery of Neolithic Britain and Ireland.

During recent decades, influenced by the groundbreaking work of Alfred Gell (1998) in the anthropology of art, many prehistorians have shifted their attention towards what images and patterns do (Chittock 2014, 2021; Garrow & Gosden 2012; Gosden 2013; Jones & Cochrane 2018; Jones & Díaz-Guardamino 2019; Joy 2011, 2020). This has been a profitable avenue of enquiry which has led to the reassessment of both British and Irish Neolithic mark making practices (Jones & Díaz-Guardamino 2019) and Celtic art in Britain and Europe (Garrow & Gosden 2012; Garrow, Gosden & Hill 2008; Nimura et al. 2020). We do not wish to diverge from this approach, only reconsider how it may help us to illuminate representational imagery from this particular standpoint.

Emerging from this wider project are interests in the processes involved in making prehistoric art and, furthermore, the idea that images are constantly ‘in the making’ (Back Danielsson & Jones 2020). The argument that images are in the making stems from the recognition that traditional approaches to imagery are ‘inadequate to the task of understanding the manifold material and visual character of the images excavated and studied by archaeologists’ (Back Danielsson & Jones 2020:1). To understand images differently we foreground a series of aspects of images including their unfolding, their transformation and their multiplicity. All of these aspects of images are discussed in what follows, but here we simply wish to frame this in terms of the ambiguity of imagery.

There are many later prehistoric images that are thought of by archaeologists as being clearly representational. In the Iron Age, birds and certain mammals are depicted in bronze and sometimes in gold and non-metallic materials, adorning a wide range of weaponry, tools, vessels and personal ornament, for example (see Chittock 2020:80–81 and 83 for a summary of instances of anthropomorphic imagery on some of these object types). Anthropomorphic and, less commonly, zoomorphic images also exist as stone statues in certain areas of Europe and at certain times (Chittock 2020:81–82; Megaw & Megaw 1989:152–154, 168–171, 265–258), while bronze and chalk figurines are concentrated within parts of middle and late Iron Age Britain (c. 400 BCE–43 CE, see Chittock 2020:82). Human figures, particularly faces, constitute a fairly large proportion of these images (Chittock 2020:72). This type of uncontested imagery, however, is comparatively rare in Iron Age Europe, and exists within a much more extensive range of swirling, curvilinear patterns. Here, archaeologists are presented with a challenge caused by the boundary between image and pattern, which, in this particular art tradition, is deliberately blurred. For every ‘clearly representational’ image, there are many more arrangements of motifs that suggest figures and faces, but that cannot be confidently classified as either image or pattern, often falling into either of these two categories depending on the archaeologist viewing and writing about them. Vincent Megaw coined the term ‘pseudo-faces’ (1970) to refer to this phenomenon, describing human faces that are perhaps visible in certain lights or from certain angles, but which can quickly disappear from view, leaving only arrangements of non-representational motifs. The term refers to a category somewhere between arrangements of motifs and anthropomorphic depictions – Megaw (1970:274) also borrows the term ‘Cheshire cats’ from Jacobsthal (1941:308, 1944:19) to refer to these arrangements, making reference to the famous feline character from Lewis Carroll’s ‘Alice in Wonderland’, who could disappear at will, leaving only his smile behind. A separate, but related, phenomenon is the arrangement of motifs in ways that can be interpreted as different figures from different angles. This is illustrated by Aldhouse-Green (2004:196–197) using a fragment of gold openwork plaque from Bad Dürkheim, Germany, depicting two different faces depending on the angle from which it is viewed (see also Megaw 1970:272). The phenomenon is also well-represented in a group of Late Iron Age and Early Roman champlevé enamelled harness fittings from Britain (Figure 1; see also Foster 2014). Ginoux (2020:134) borrows Ludwig Wittgenstein’s famous ‘duckrabbit’ concept to explain the multiple ways of seeing designs. This analogy also aptly describes the shifting nature of designs between abstraction and representation.

Due to the intersection between image and pattern in Iron Age art, therefore, archaeologists are often faced with apparent representational imagery – ambiguous or hybrid creatures that clearly have limbs or eyes and mouths, but which are not readily recognizable as belonging to a specific species (Foster 2014). Moreover, an image of a creature interpreted as a particular species by one archaeologist can be seen as a different species by another – a good example of this is a pair of creatures that appear on roundels on the Witham shield from Lincolnshire, UK, which have been interpreted as horses (Megaw & Megaw 1989:198) and bulls (Hitchcock 2022:164), among other things.

As has previously been discussed (Jones & Díaz-Guardamino 2019, 2021), taking a long chronological view of the Neolithic we can observe that certain features of human anatomy have been emphasized since the beginning. For example, phalluses were carved from chalk and bone as early as the 38th to 36th centuries cal. BCE, and are found deposited in the ditches of long barrows and causewayed enclosures, such as the Trundle, Sussex and Windmill Hill, Wiltshire. This tradition continues into the later Neolithic and chalk phalli are also known from the ditches of henges, for example the large chalk phallus found at Maumbury Rings, Dorset. There seems to be an emphasis on isolated features of the human body throughout the Neolithic. Phalluses of chalk and bone are one example, another would be stylized images of faces, such as those seen on the ‘eyebrow’ motifs on the Folkton Drums, Yorkshire (Figure 2). Like the examples we discuss from the Iron Age these eyebrow motifs are decidedly ambiguous, and both a duckrabbit and Cheshire Cat effect are evident as ‘faces’ emerge from geometric designs. It is not until much later in the Neolithic that we begin to see all features of the human anatomy coalesce into figurative forms. In the case of some examples, like the ‘god-dolly’ from the Sweet Track, Somerset – a hermaphrodite with numerous gendered features (Figure 3) – we observe a degree of representational ambiguity. We also observe anthropomorphic figurines emerging in the Northern Isles of Scotland (Orkney) around the 32nd to 29th centuries cal. BCE. Recent examples at the Links of Noltland have a distinctive shouldered human form (Figure 4), and in the case of the most well-known example (the so-called ‘Orkney Venus’) we also observe the eyebrow motif in the head area of the figurine. This figure is again ambiguously gendered and could be either a male or female figure.

Difficulties in the interpretation of ambiguous representational images are, therefore, well documented in studies of later prehistoric art (see for example Chittock 2020; Foster 2014; Jacobsthal 1944:19; Jones & Díaz-Guardamino 2019:190–192; Megaw 1970). We argue here that the types of ambiguity we describe are deliberate features of later prehistoric art. They occur partially as the result of the makers of objects playing with pattern and form, and the resulting complexity has ‘the power to intrigue’ (Joy 2019), inviting the viewer to consciously try to comprehend the marks and shapes they see. At least, this is the effect they have on the archaeologists who study them. ‘Dwell-time’ is a phrase used, for example, in advertising, marketing and also in museums, referring to the amount of time a person spends looking at an exhibit or advertisement. We can perhaps apply it here in the context of later prehistoric objects – their complex and ambiguous designs increase dwell-times and perhaps contribute to the ‘technologies of enchantment’ (Garrow & Gosden 2012; Gell 1992) that make them effective and significant.

In addition to this possibility, ambiguity can also make designs open to multiple interpretations. Jones has described figurative aspects of Neolithic art that, at times, appear abstract, as ‘conditional’ and ‘situational’ (Jones 2019:191), where the appearances of objects are altered depending on their contexts of use, as well as who is using them. Chittock (2020:74) has also applied this idea to Iron Age art, with the argument that this was a deliberate aspect of Iron Age design.

Despite the difficulties faced by archaeologists when trying to impose interpretative categories onto later prehistoric art, it is widely acknowledged that faces and heads play important roles in Iron Age art traditions.

An Iron Age preoccupation with human heads has been explored by Armit (2012) in the context of headhunting, the practice of decapitating and curating human heads. Some depictions of human heads have been interpreted as representations of this practice, with a tradition of stone sculptures of this kind centring on southern France. Perhaps the most famous examples of these sculptures are from the second century BCE oppidum of Entremont (near Aix-en-Provence, France), where numerous stone sculptural fragments have been found to show representations of severed heads, some grasped by hands (Armit 2012:182). Whilst the depiction of severed heads in stone forms fit into a regional tradition of stone anthropomorphic depictions in southern France during the fifth–second centuries BCE, heads and faces appear throughout a much wider range of Iron Age designs across Europe, often on objects such as personal ornament and weaponry coalescing into other regional traditions (Chittock 2020). Anthropomorphic heads form parts of hybrid creatures that make up the forms of Maskenfibeln, a type of brooch, in Central Europe during the Early La Tène period (c. 450–380 BCE) (Bagley & Schumann 2013:131–134; Megaw & Megaw 1989:84–88). Hidden faces appear on torcs and bracelets in the Marne region of France, southwestern Germany and Switzerland during a similar period in the Middle Iron Age (Megaw & Megaw 1989:136–137). Anthropomorphic heads form the pommels of a group of swords and daggers, many of which include stylized bodies that form the handles/grips of the weapons, with these objects covering a broad geographical distribution during the Middle to Late Iron Age (Fig 5; Carlson 2011; Chittock 2020:83; Megaw & Megaw 1989:164–166; Stead 1979:61).

There is less of an emphasis on faces and heads in the British and Irish Neolithic compared to the Iron Age, though, as mentioned above, Crawford (1957) was convinced that circular motifs on passage tombs were evidence of an ‘eye-goddess’ cult that could be traced through prehistoric Europe back to the cult of Ishtar in Syria. ‘Eyebrow’ motifs do occur as carvings in chambered tomb and passage tombs, at the Holm of Papa Westray North, Orkney (Ritchie 2009) for example, though eyebrow motifs have also been discovered at the settlement site of Braes of Smerquoy, Orkney (Richards & Jones 2016:87). They need not be confined to ‘cult’ contexts, as Crawford imagined. The clearest examples of faces are the ‘eyebrow’ motifs we find on selected artefacts like the Folkton Drums (Figure 2), or the ‘Orkney venus’. Faces do also emerge in another curious context: the Knowth mace head. The mace head discovered in the eastern tomb of the great passage tomb at Knowth, Co. Meath, Ireland (Eogan 1986; Eogan & Cleary 2017) is one of the most elaborate and accomplished artefacts from the Neolithic period. The mace head has a double spiral motif on its side positioned above the central hole for the shaft. Combined with the pressure flaked areas in the upper and lower brown regions of this mottled brown and white flint artefact, when turned to face the viewer the spirals give the appearance of eyes, the hole a mouth, and the pressure flaking in the upper and lower regions the appearance of textured hair and beard. Each of these examples –Folkton, the Westray venus and the Knowth macehead – only compose themselves as faces when interacted with; the faces of these objects become most apparent when faced by a human observer.

As mentioned above, Iron Age anthropomorphic images have often been seen as depictions of deities, sometimes participating in associated mythic narratives (Megaw & Megaw 1989: 21–23, 176). Aldhouse-Green (2004) also discusses some images as depictions of mortal individuals. It has been noted, however, that the faces of Iron Age figures tend to be generic and featureless (Chittock 2020:87–88). Facial expressions are generally limited to a downturned mouth (see Figure 1 for a particularly extreme version of this, emphasized by an openwork design) and staring eyes, and identifying features limited to stylized moustaches and hair. This presents a contrast to the detailed depictions of clothing and weaponry worn on the bodies of these figures, some of which can be related directly to objects in the archaeological record (see Chittock 2020:88). It also raises a question about what these faces were designed to do, considering that they are missing the individualized features and expressions that make faces vehicles for communication and recognition. One of the aims of this paper is to examine the reasons why faces are overrepresented in later prehistoric art, in comparison to other body parts. Humans use our faces as a way of communicating, and in the recognition of other individuals. However, the generic, standardized, stylized and expressionless faces of later prehistoric art do not seem useful in either of these functions.

Human faces have, arguably, taken on new roles and been put to new uses in recent decades. Facial recognition algorithms are being implemented increasingly frequently in a wide range of domains: social media, surveillance, border control, targeted advertising and police profiling are examples given by Celis (2020), who alludes to the harmfulness and problematic nature of some of the uses of this technology through the imposition of ‘algorithmic and statistical regimes of power’ (Celis 2020:73). The COVID-19 pandemic arguably has accelerated reliance on platforms such as Zoom for communication, where visual communication occurs through the face only, removing much of the body language that contributes to in-person interactions. The face has certainly taken on heightened significance since 2020 and our analysis of faces in prehistoric contexts should be understood in this background.

The philosophers Gilles Delueze and Felix Guatarri (1987:123–164) introduce and examine the concept of facialization. Facialization describes the ‘overcoding’ of a subject with a face, a component of the ‘abstract machine of faciality’, which supplants and supersedes other semiotic systems (Deleuze & Guattari 1987:136). Their work wishes to divest the face of any auratic power (of the kind described by Walter Benjamin) conferred on it by such contemporary media as advertising, TV and film, and instead wishes us to focus on the face as an abstraction and a locus of multiple affective possibilities. We adopt these ideas with caution here. Much of Deleuze and Guattari’s discussion of the face is most applicable in a contemporary context and therefore less useful when exploring prehistory. However, we do find the idea that faces act as loci for affective possibilities useful.

There are several questions regarding faces (and other bodily components) that we could draw from Deleuze and Guattari’s discussion. We could regard abstracted faces as being diagram-like and consider what these diagrammatic faces did in prehistoric contexts. Can we understand standardized faces in this sense? We could also consider what constitutes and makes up a face in different prehistoric contexts. What are the minimal characteristics that might be regarded as a face: eyes, a mouth? How do we consider faces that emerge and dissolve, such as the ‘pseudo-faces’ mentioned by Megaw? The philosophical discussion of facialization forces us to think about the power of the face but the ambiguity of faces in prehistoric contexts also invites us to consider how the power and significance of the face emerges.

We have argued above (also see Chittock 2020) that the ambiguity in Iron Age art described above is a deliberate design feature that leaves images and patterns open to multiple interpretations. Cochrane and Jones (2018:172) have, similarly, referred to certain Neolithic objects as ‘situational’. Alongside this we also observe quite unambiguous examples of representation, such as carved phalli in bone or chalk from the British Neolithic, and birds, mammals and humans rendered in metal in the Iron Age. There seem to be two different forces at play here: one clearly focusing our attention on aspects of figuration, the other drawing our attention to the ambiguity of representation. Drawing on the work of Gilles Deleuze, particularly his work on framing and affect (particularly of faces) in cinema (Deleuze 1986), we will now consider the affects produced by these two different forces.

We recognize, in line with Darryl Wilkinson (2022), that naturalist depictions are ontologically significant and that the appearance of naturalist, or representational, imagery may signal particular ontologies of media (Jones & Díaz-Guardamino 2021 discuss this for the British Neolithic). Here we explore how these differing ontologies of media emerge over time. We find the Egyptologist Rune Nyord’s discussion of the concept of ‘presentification’ (Nyord 2021:56–57) particularly useful to consider the relationship between unambiguous images in the Neolithic and Iron Age as making an ontological connection with what they depict – there is a clear ontological connection between chalk or bone phalli and parts of the human body, while clear connections could also be established between Iron Age bronze figurines and the birds, mammals and humans they depict. These images present and draw attention to exactly what they depict. But what of the ambiguous images? It seems interesting that ambiguity occurs when human-like images are articulated together with either geometric designs or other components of the human form. It is the juxtaposition of these images that produces a sense of ambiguity. While the geometry of the face is made obvious by its positioning in the human form, the expressive qualities of the face are ambiguous, and the face may be disguised by geometric images.

It is when images are disengaged from the human form that their power and affect is most readily apparent (they are ‘presentified’ and stand in for aspects of the human), but when they are composed together in human form faces in particular become masks of ambiguity; they are rendered as locations of uncertainty, of possible differentiality, locations of becoming human? For the Neolithic we could consider the multi-gendered figurines (Figs. 3 and 4) as loci of potentiality, as forms in which what it means to be human was yet to be decided.

In the case of the Iron Age, we also observe the presentification of certain kinds of things like animals, but against this we observe several things: ambiguous faces emerge when humanlike images articulate with geometric forms. Even when we observe decidedly human forms the faces depicted are often unexpressive (Chittock 2020:87–89). Curiously, some of the most unambiguously human figures are those incorporated into ‘useful’ objects like bladed weapons (anthropomorphic sword and dagger handles for instance Carlson 2011; Fitzpatrick 1996; see also Figure 5) and bone comb handles (Fitzpatrick 2020). Similar to the Neolithic, human faces are masks of ambiguity – they are either difficult to define, or unexpressive if they are better defined. It is only when human forms are merged as components of useful objects that the potentiality of humans is brought to the foreground.

The concept of presentification allows us to underline the different ways that images work in the Neolithic and Iron Age. We can observe unambiguous presentification where images are clearly meant to evoke and stand in for aspects of humans and animals. In other cases, we can consider an ambiguous form of presentification, possibly delineating the potency and creative potential of the human form. The potency and potentiality of the human form is most apparent when we observe ambiguous human forms. In the Neolithic as multi-gendered forms, and in the Iron Age as figures inserted as components in objects like swords or bone combs. In the Neolithic the potency of the human form is enacted through the presentation of multiple gendered features, in the Iron Age this potency is enacted through the utility of objects.

In all of these cases we can observe that the human form appears to be a mutable and emergent source for understanding differing relationships with the world in Neolithic Britain and Ireland and Iron Age Europe.

In this paper, we have used examples from Neolithic Britain and Ireland and later Iron Age Europe to demonstrate potential for understanding how later prehistoric representational imagery came to be formed. It is particularly important for our account that we have problematized representation and have not taken representation and symbolism to be de facto or matter of fact. By exploring how representational imagery emerges against a background of abstract and geometric images, we are able to show how certain kinds of representational forms, particularly images of the human face and body, come to be presented. We argue that these emergent human forms acted as loci for affective potentials. The examples used, therefore, suggest ways in which we can begin to understand how representational imagery operates when we analyse images as being fundamentally mutable and in the making.