Britain In or Out of Europe During the Late Mesolithic? A New Perspective of the Mesolithic–Neolithic Transition

2.1 The Colonisation Approach

Colonisation, the favoured process amongst Neolithic scholars ever since the turn of the twentieth century, has incorporated an isolationist standpoint into its models. Although this school of thought took heavy criticism in the 1960s and 1980s, then again in the early twenty-first century (Clark, 1966; Dennel, 1983; Kinnes, 1988; Thomas, 1988, 1999, 2003, 2004), it dominates recent debates on the Mesolithic–Neolithic transition. In general, an abrupt shift in material culture, diet, and funerary rites (with a shift in genetic composition) at around 4000 BC has led to the suggestion that the start of the Neolithic was heavily influenced by farming migrants from Europe. This is opposed to arguments for a slower process of change which would suggest cross-channel exchange between Mesolithic and Neolithic groups – the indigenism model. Below, we will examine Alison Sheridan’s and Alasdair Whittle et al.’s colonisation models and the recent DNA evidence, before examining the indigenism approach and evaluating both in the light of new evidence from Bexhill.

Sheridan describes the late Mesolithic as a culture that chose to be extremely insular and socially introverted, with no contact to the Neolithic continent (Sheridan, 2016). Therefore, colonisation must have been the prime process that triggered the transition. Sheridan’s colonisation model focuses on four waves of migration from different parts of France (Sheridan, 2003, 2004, 2007, 2010, 2016). These waves brought about change through a rapid introduction of new material culture, the migrants acculturating isolated Mesolithic groups. These waves of migration can be described as follows. First, early colonists advanced up the Atlantic Seaboard, as a “false start” at Ferriter’s Cove, Ireland (Woodman & McCarthy, 2003, 4450–4270 cal. BC) where migration did not take hold. According to Sheridan, the reasons for this may have been a lack of critical migration mass to make the colonisation a success (Sheridan, 2016). Then, sometime between 4300 and 3900 BC, colonists built Achnacreebeag Tomb, Argyll. The motifs on the pottery in this monument resemble Late Castellic ware vessels from Brittany, the proposed origin of the migration. Though Thomas (2013) contends that this was also a failed migration because of a lack of other Neolithic material culture in the vicinity, a range of other Breton style monuments in Scotland may suggest otherwise (Sheridan & Schulting, 2020). The third wave, at around 4000 BC, (a “Carinated Bowl Neolithic” from northeastern France), spread across the south of England. Links can be made between Carinated Bowl and Chasséo-Michelsberg pottery from Les Sablins, Pas-du-Calais, suggesting a migratory route across the shortest part of the channel (Phillipe et al., 2011; Sheridan, 2016). With this pottery form came other typical items of the “Neolithic Package”: timber houses, domestic produce, new mortuary structures, and new lithic tools. These other attributes of the “Carinated Bowl Neolithic” are associated with minority elements from a wide range of continental cultures and may represent a “melting pot” of societies arriving in Britain, forming many “Neolithics.” The fourth wave, from Normandy to south-west Britain, occurred in the first quarter of the fourth millennium and is evidenced by similar tomb designs on either side of the channel. These waves were triggered by political upheaval in France – collapse of an elite, possibly patriarchal, ruling class in Brittany, and population expansion in the Paris Basin resulting in social stress (Sheridan, 2016).

Insular Mesolithic societies are believed to have been rapidly acculturated at the time of the “Carinated Bowl Neolithic” (Sheridan, 2010, 2016, 2017). This is evidenced by the disappearance of Mesolithic Microliths in the early fourth millennium. However, the mechanisms behind this are not fully understood. Why would a hunter-gatherer society suddenly and inexplicably shift its cultural position in the face of long-held tradition and a supposedly introverted, insular lifestyle? New neighbours do not necessarily mean new friends. This will be examined later by way of DIT.

Sheridan’s model deserves much credit. It brings together these unique lines of evidence to form more complex, multi-stranded Neolithics than previously thought, each with their own motivations and outcomes. However, this is juxtaposed against a more generalised concept of a single late Mesolithic described in almost purely economic terms (Sheridan, 2016). This model does not examine the heterogeneity of culture in the fifth and sixth millenniums (Conneller, 2022) and, therefore, does not allow for the possibility of these regional Neolithics to be influenced by regional Mesolithics (Clark, 1980; Conneller, 2022). Many colonisation models fall into this historiographic trap (Thomas, 1988) in which processual models of the Mesolithic are compared to more nuanced understandings of Neolithic lifeways. This creates a synthetic boundary in our perception of the past which more readily translates into binary, rather than analogue, models of transition, naturally promoting models of colonisation.

Other colonisation models of the Mesolithic–Neolithic transition have used high-quality dating evidence from Neolithic contexts to ascertain the process of transition (e.g. Whittle, Healy, & Bayliss, 2011). Through the analysis of radiocarbon dates in the Gathering Time project, Whittle et al. propose a more staggered transition, starting in the south east of England in the forty-first century BC, suggested by the early dates from Coldrum barrow and White Horse Stone. The Neolithic reached Scotland by the late thirty-nineth century BC. Interestingly, the dated introduction of megalithic architecture in Britain (c.4000–3800 BC) follows on from NW France – the first channel region to adopt this innovation (4800–4000 BC: Paulsson, 2019). The arrival of Neolithic migrants was a trigger for indigenous groups not only to accept new lifeways but to spread them throughout Britain by 3800 BC, although most of this work was likely done by colonists. However, several recent studies have pointed to an extremely low late Mesolithic population (Brace et al., 2019; Collard, Edinborough, Shennan, & Thomas, 2010; Elliott et al., 2020), which might have hindered the rapid spread of agriculture through indigenous channels. Although ostensibly beyond the remit of Whittle et al.’s Gathering Time study, little attention is given to how neolithisation spread through these Mesolithic social systems and the exact role of hunter-gatherers in this process. In reality, these hunter-gatherers would have maintained a greater deal of decision making (Schulting & Boric, 2017). This is not often acknowledged in colonisation models, leading inadvertently to a passive interpretation of the role of hunter-gatherers in the spread of Neolithic lifeways.

2.2 The Indigenism Approach

Less attention has been given to the theory of indigenism. This school of thought describes a complex, social Mesolithic with intricate networks extending to the continent (Thomas, 1988, 1999, 2003, 2004, 2008, 2013). Through these networks, hunter-gatherers sought out ideas of farming and other Neolithic lifeways. Thomas (2013), for example, argues that the “melting pot” signature identified by Sheridan could also have been generated through trade and exchange between British hunter-gatherers and the Neolithic continent. Picking and choosing the technologies that suited the indigenous best would have generated many different Neolithic signatures. This characteristic was observed in various instances throughout Europe and indigenists used it as a model for Britain. The Ertebølle is often cited as an example of hunter-gatherer/farmer interactions where items were exchanged for more typical Neolithic objects and ideas (Fischer, 1982; Hartz, Lübke, & Terberger, 2007). Through these exchanges, hunter-gatherers sought to be part of larger “Neolithic” communities (Sorensen, 2016). Similarly, in the Lower Rhine area, a tradition of technological continuity and exchange persisted between hunter-gatherer and agrarian groups (Robinson, Sergant, & Crombé, 2013). A technology exchange is also observed in western France (Marchand, 2007). Betti et al. (2020) suggest that, in times of climatic trouble, there was a more symbiotic relationship between continental hunter-gatherers and farmers. This relationship is perhaps observed in the admixture of Neolithic and Mesolithic DNA on the continent (Rivollat et al., 2020).

Identifying contact between hunter-gatherers and farmers is the cornerstone for understanding neolithisation processes on the continent. This is why characterising cross-channel connections between British Mesolithic and European Neolithic is important. These European examples demonstrate a hunter-gatherer agency not identified in British colonisation models. Within these European works, there is a clear intention to avoid placing “farming pioneers and natives in opposition” (Marchand, 2007, p. 225; Verhart, 2000). This is something perhaps lacking in British scholarship, promoted by limited evidence for similar interactions in Britain. It should be noted that the models of Julian Thomas are general at best with little exploration of regionality in the fifth millennium. Indeed, for ardent critics of indigenist approaches, this lack of evidence suggests an insular British Mesolithic community that refused to associate with Neolithic lifeways for the best part of a millennium (Sheridan, 2016). However, new evidence of cross-channel contact from Bexhill, discussed later, may help to elucidate this problem.

2.3 The DNA Evidence

DNA usage as a form of archaeological evidence has become increasingly prevalent in recent years, especially within the field of prehistory and the Mesolithic–Neolithic transition in particular (Brace et al., 2019; Cassidy et al., 2016; Fowler et al., 2021; Olalde et al., 2018). DNA analysis of British Neolithic and Mesolithic individuals clearly demonstrates continental migration of farmers from the early Neolithic onwards. The Neolithic populations studied had clear Aegean origins whilst their DNA also included a genetic signature similar to Neolithic populations from France and Spain (Brace et al., 2019; Rivollat et al., 2020), suggesting vectors of movement between western Europe and Britain. The British Neolithic genetic signatures also contained elements of hunter-gatherer DNA. However, with the possible exception of Scotland, this admixture was interpreted as belonging to hunter-gatherer/farmer interactions on the Iberian peninsula or western France prior to Neolithic groups crossing the channel. This has led to the conclusion that there was discontinuity between the genetic populations of the Mesolithic and the Neolithic of Britain (Brace et al., 2019).Curiously, this admixture signature is strongest in SE England, the area we will be exploring later in this article.

This study is revolutionary in deciphering some of the Neolithic origins in Britain. It also clearly demonstrates colonisation was prevalent. However, some thoughts will be offered up that may affect the interpretation of these findings. The first concern is the sampling and aspects of the admixture in the Brace et al. (2019) study. The DNA from only six Mesolithic individuals (all from the west of Britain) and 67 Neolithic individuals were used. Aside from an individual form Cnoc Coig, Scotland, the latest Mesolithic individual dated to the eighth millennium. The limited Mesolithic sample size was due to limited data available. However, it does lead to some unwelcome assumptions about the Mesolithic. For instance, for the conclusions in the article to be right, it assumes the genetic make-up of eighth millennium individuals would be the same as those from the fifth millennium. It also assumes that the individuals from the west of Britain would be the same as any from the east. It therefore assumes a temporal and spatial homogeneity to the Mesolithic which is often perpetuated in the transition debate (Sheridan, 2016). The Mesolithic was an extremely diverse entity that was continually changing, the cultures of the final Mesolithic being drastically different to those from the eighth millennium BC (Conneller, 2022. e.g. Bell, 2007; Bell, Brunning, Batchelor, Hill, & Wilkinson, 2016; Blinkhorn & Little, 2018; Finlay, 2004; Lewis et al., 2019). Thus, in terms of genetic mixing, this study may not be reflecting the true nature of the Mesolithic–Neolithic transition. To exaggerate the point, what would the genetics of the transition look like if Mesolithic groups of the fifth millennium BC also had connections to France or Spain?

The context from which DNA was extracted is also important to consider. For the most part, the fourth millennium DNA was extracted from individuals from “special” Neolithic contexts, i.e. burial chambers, not least because these act as preservation traps for archaeological data. However, this may also skew the narrative told to us. Only a certain type of person would be buried in these significant places. DNA studies from Hazleton North tomb suggest that this was a resting place for part of a extended family following a patriarchal line (Fowler et al., 2021). This suggests that not all members of the community were buried here. The patriarchal element at Hazleton North succinctly demonstrates that not all members of Neolithic communities were buried in these tombs. The isotope analysis further suggests that the majority of family members grew up and lived locally, suggesting a cohesive community unit. Burial chambers such as Hazleton North may therefore only reflect certain parts of “Neolithic” communities where there is less chance of burials with “Mesolithic” aspects to their DNA. One might think that the human bone fragments found under tombs (Saville, 1990) such as these might yield different histories.

The third consideration is one brought to light by Thomas (2022), who suggests that these studies are describing an event rather than a process. In this way, we are not understanding the intricacies of how the transition may have occurred. Perhaps we should consider that social intermixing of a type that would lead to an alteration of the genetic signature is only one way of changing the cultural make up of a community. If hunter-gatherer and agrarian groups of Britain were not on such intimate terms (perhaps because of taboos or other social customs), this does not mean that they could not meet, exchange ideas, and alter each other’s world view. Thus, there may still have been more processes at play than simple colonisation. These kinds of interactions are perhaps better seen in more traditional archaeological studies and material culture is still of primary importance to those studying the transition. We should, therefore, still be open minded about the dynamics at play during this period – DNA evidence is only one part of the whole story.

3.1 Problem One: The Neo-centric Transition

There is often a very “Neo-centric” approach to some aspects of the debate that almost forgets the role of hunter-gatherers in the process towards farming. This is problematic because, rather than addressing the Mesolithic–Neolithic transition, many studies instead address the arrival of the Neolithic. These are not equivalent and treating them as such may lead to an incomplete narrative. Whittle et al. (2011), for example, focus purely on Neolithic evidence, whereas Sheridan does not examine the potential of hunter-gatherer evidence in the debate. As discussed earlier, the recent DNA research is also biased towards the Neolithic, but this is due to poor survival rates in Mesolithic skeletons. Similarly, there is little acknowledgement of the role of hunter-gatherers in the north of Britain in wider transition debates (Brace et al., 2019; Finlay et al., 2019; Griffiths, 2014). Thomas (1988) suggests this perspective is exacerbated by the different approaches of Mesolithic and Neolithic scholars of the twentieth century: a dichotomy between processual models of Mesolithic adaptiveness and more nuanced research focusing on socially oriented Neolithic agency. These approaches have created a legacy that perhaps makes Mesolithic research less desirable to study for Neolithic scholars in a post-processual world. This problem is remedied somewhat by the work of Cummings and Harris (2011) who attempt to reinvigorate the Mesolithic side of the transition debate.

Garrow, Griffiths, Anderson-Whymark, and Sturt (2017) suggest Neo-centric perspectives may be more common because of a more detectable Neolithic dataset. Although this is undoubtedly often the case, we have to be careful when interpreting material culture as Neolithic. As Griffiths (2014, 2017, 2018) points out, the overly rigid use of definitions (wrongly) means that anything later than 4000 BC cannot be Mesolithic. This approach not only drastically “oversimplifies the nature of change in this period” (Anderson-Whymark & Garrow, 2015; Griffiths, 2018, p. 7) but may set a precedent for ignoring a potentially huge body of evidence (Thomas, 1988). To further the transition debate, we should avoid taking a Neo-centric perspective but rather embrace a more nuanced understanding of the transition, one that outlines the important role of both the hunter-gatherer and farmer in shaping the Neolithic.

3.2 Problem Two: An Us or Them Approach to the Transition

Another problem with current debates is that the language emphasises a dichotomous approach to the transition: colonisation or indigenism; Mesolithic or Neolithic (Sheridan, 2015; Thomas, 2008; Zvelebil & Rowley-Conwy, 1984, 1986). An emerging approach offered by Griffiths and others identify a messier picture of the Mesolithic–Neolithic transition that is both complex and local (Anderson-Whymark & Garrow, 2015; Armit & Finlayson, 1992; Cooney, 2007; Garrow & Sturt, 2011; Garrow et al., 2017). Griffith’s Bayesian approach to dating the final Mesolithic and the earliest Neolithic of Britain demonstrates that hunter-gatherers and farmers lived adjacent but perhaps separate lives in the north of England (Griffiths, 2014, 2018). Likewise, on Oronsay, pockets of hunter-gatherers likely lived alongside farming communities (Charlton et al., 2016; Finlay et al., 2019). It is possible that other hunter-gatherer enclaves survived into the fourth millennium as well. Through the examination of more temporally and geographically refined scales of evidence, these studies suggest that the Mesolithic–Neolithic transition was more piecemeal than previously thought. The full interpretive power of these studies, however, derives from the congruous and non-dualistic approach to the data. Our article will attempt to marry up two of the most prominent models of the transition (colonisation and indigenism). As a basis, it will adopt the position of European colleagues, thus not placing farmers and foragers in opposition. It is hoped this will achieve a more rounded and realistic picture of the Mesolithic–Neolithic transition.

3.3 Problem Three: Rapid Adoption and DIT

Many of the models outlined above rely on the rapid adoption of new innovations (new ideas, cosmologies, objects, etc.). This is contrary to the mechanics of innovation adoption which are rarely instantaneous (Eerkens & Lipo, 2014; Rogers, 1995). A new idea needs to percolate and permeate within the collective cultural consciousness of a new society for a long time. Only after this, can it be accepted (Rogers, 1995).

One mechanism by which a new innovation is adopted is known as diffusion of innovation theory (DIT) (Rogers, 1995). DIT, derived from sociology, is often used in modern-day settings to better understand how popular a commercial product will be. As such, it should be used with caution. At the heart of DIT, however, is the idea of cultural “norms” as barriers to change. Consequently, it has been successfully used by ethnographers to understand how different innovations have moved across wildly different cultural boundaries (e.g. Rogers & Shoemaker, 1971). Although it has its basis in culture history (e.g. Flannery, 1972), DIT echoes theories of Entanglement (Hodder, 2011, 2012), Community of Practise (Sorensen, 2016), Planned Behaviour (Ajzen, 1991; Elliott et al., 2020), and Availability Models (Zvelebil, 1986; Zvelebil & Rowley-Conwy, 1984), all used in archaeology. Rather than simply a time series, DIT fleshes out the individual processes involved in accepting innovations. As such, it goes beyond some models used in archaeology (e.g. Zvelebil, 1986). It is, therefore, a useful tool to understand the processes of the Mesolithic/Neolithic transition and will be used here to describe the spread of farming in Britain. Continental data from the sixth and fifth millenniums BC, the period of transition in France and Spain, will be compared to data from Britain to show how hunter-gatherers may have been exposed to Neolithic lifeways before these ideas took hold at around 4000 BC through colonisation.

In DIT, there are several factors that contribute to how an innovation is diffused. These include:

• Relative advantage – the degree to which an innovation is thought to be “better” than the current strategy, whether that be a new form of subsistence, such as farming, or a new cosmological system, whatever that may mean to the individual adopter. The innovation must percolate within a social system until cultural boundaries have been eroded sufficiently for the relative advantage to outweigh the perceived cost of change. This means that change is not instantaneous and that the society has to be aware of the new innovation long enough for it to percolate. In this way, DIT echoes the thoughts of Grahame Clark (1975, p. 25).

• Comparability – the degree to which a new innovation fits in with existing ideas about the world and the wants of the adopter. This reduces uncertainty in the eyes of the adopter. It is similar to “fittingness” described by Ian Hodder in which objects are more likely to become entangled when they fit in with the ever-changing historic narrative of a society or individual (Drayson & Woodward, 2015; Hodder, 2012). As a new innovation is introduced, associated new perceptions and ideas may slowly alter the world view of the society. This may erode cultural boundaries until relative advantage and comparability are at acceptable levels.

• Complexity – how easy it is to understand or learn an innovation. More complex ideas might be adopted more slowly.

• Trialability – the degree to which an invention can be used or handled before adoption. This involves physical interactions with the new technology.

• Observability – the degree to which the innovation has been observed as successful by neighbouring users. This factor can hinder uptake of an innovation if population densities are low as may have been the case in the British late Mesolithic (Brace et al., 2019; Collard et al., 2010; Elliott et al., 2020).

These are decisions that would be made on both an individual and societal level and therefore reflect agencies that are bound by and fit into the social, practical, and cosmological world of the agent.

When examining the total percentage of a population that has adopted an innovation at any one time, a time series graph will produce an S-shaped curve for a population within a single social system (Figure 1). This graph describes a snowball effect where the rate of adoption gathers momentum until a population ceiling within a social system has been reached. As momentum gathers, society and innovations become increasingly entangled and dependent on each other. This does not mean, however, that DIT is a linear process. If the total population (e.g. the Mesolithic of Britain) is made up of several social systems that are constantly in flux, as is common in human society, then the process will not be uniform. This is likely the case for the late Mesolithic which was probably extremely regionally diverse, made up of small family units (see below, Champness, Donnelly, Ford, & Haggart, 2015). Elements would be selectively adopted to meet changing local needs after a long period of adjustment (Zvelebil & Rowley-Conwy, 1986). Therefore, the time frame between the introduction and adoption of an idea is not instantaneous and varies from community to community. In sociology, the curve for one social system can be described as follows:

1. A slow uptake of the innovation (characterised by exposers and early adopters). Exposers are those who have come from a different social system and transmit aspects of an innovation. At the start of this phase, the innovation is known by the potential adopter but they have not fully accepted it. Information is exchanged and contact is maintained until a small part of the community starts to adopt aspects of it. This is similar to the availability phase in Zvelebil and Rowley-Conwy’s model (1984). It takes a while for these new ideas to permeate within a host society as the factors outlined above are consciously and subconsciously weighed. Early adopters are generally community leaders or those with a place of standing within a community, who have a greater degree of freedom of opinion. This allows them break away more easily from the norms of society. It takes just as long (if not longer) for this small population to adopt an idea as it does for the majority.

2. This slow uptake is followed by a rapid spread of the technology as the majority of the population accept the idea (early majority and late majority). These are people who frequently interact with their peers, enabling the innovation to be spread through word of mouth. The late majority are generally more sceptical of an idea.

3. Finally, after the majority of people have accepted the idea, the spread of innovation slows as latecomers to the innovation finally become aware of the idea, or give up resistance to it (the Laggards). These are the people with the most traditional views in a society.

Figure 1

A typical adoption curve following diffusion of innovation theory. The percentage of the population who have adopted an innovation is plotted against time.

In summary, DIT describes a slow process in which the new innovation must permeate into the culture of a society before it can be readily used. The adopters accept the new innovation when the relative advantage (and other factors described above) of the new technology outweighs the perceived cost of change. The relevance of DIT in an archaeological context is exemplified in the work of Eerkens and Lipo (2014). Here, the adoption of pottery in prehistoric America is successfully set within a diffusion of innovation framework where pottery is adopted successfully because of its relative advantage, comparability, complexity, and trialability.

3.4 Problem Four: Mesolithic Isolation

Perhaps the biggest problem is that late Mesolithic Britain has often been viewed as isolated (however, see Anderson-Whymark, Garrow, & Sturt, 2015; Elliott, 2015; Thomas, 1988, 2004, 2013). When considering colonisation, isolation (either geographically or socially) is commonly cited as evidence. In their comparison between Britain and southern Europe, Schulting and Boric (2017, p. 101) state that the “insular position of Britain and Ireland more readily evokes a colonisation model.” Equally, Sheridan concludes that late Mesolithic Britain was both “literally and metaphorically insular” before the arrival of Neolithic migrants from France (Sheridan, 2016, p. 228). These views are in part due to a lack of evidence that has created a survivorship bias. This is a problem with archaeology as a whole and is unavoidable, but the sparsity of fifth millennium data creates the illusion of an isolated Mesolithic and an abrupt start to the Neolithic. This has perhaps promoted a historiographic trend that describes British hunter-gatherers as passive and static (Thomas, 2013).

However, a long tradition of continental contacts stretches back to the eighth millennium BC and before. Maritime contacts between Britain and the north of France can be observed through the similarities in Honey Hill and Horsham points on either side of the channel (Cooper et al., 2017; Fagnart, Thevenin, Ducrocq, Souffi, & Coudret, 2008; Ghesquière, 2011). Contrary to the isolationist model of the Mesolithic–Neolithic transition, a growing body of evidence also suggests that some cross-channel contacts persisted into fifth millennium. Recent analysis of the seabed around the southern North Sea, for example, demonstrates that parts of Doggerland survived into the late Mesolithic (Gaffney et al., 2020; Missiaen et al., 2021; Walker et al., 2020). This suggests the east coast was more accessible than previously thought. It is exemplified by the evident use of continental T-axes in fifth-millennium Scotland (Elliott, 2015), establishing a pattern of late Mesolithic contact with Europe. These T-shaped axes are common in pottery bearing Swifterbant and Ertebølle cultures on the continent (e.g. Devriendt, 2008; Hartz et al., 2007). Further to this, a series of microliths found on the Isles of Scilly are of continental type (Anderson-Whymark et al., 2015). These microliths were predominately manufactured from flake blanks and broadly of trapeze or rhomboidal form, often asymmetric or with additional retouch on the shortest side. Many of these had an offset base (trapèze à base décalée) or retouch on the smallest edge. Comparisons could be made between these and microliths from northeast France, Belgium, and the Netherlands. Although some differences exist, it is clear that the Scilly microliths are of continental type dating from the seventh to fifth millennium (Anderson-Whymark et al., 2015). Anderson-Whymark et al. point out that the left lateralisation of the Scilly microliths is more indicative of forms from the Rhine-Meuse-Scheldt region rather than from France. Crombé (2015) suggests that several features on the minority of the assemblage may be typical of a transverse arrowhead tradition similar to Doel “Deurganckdok” dated to the second half of the fifth millennium. This may provide contemporary evidence for the Isles of Scilly. Marchand (2015, p. 976) suggests that the Isles of Scilly forms may be stylistic variants of northern European arrowheads that break the illusion of Britain’s “splendid isolation.” Used together, these strands of evidence suggest that, in the run up to the Mesolithic–Neolithic transition, Britain may not have been as isolated as first thought.

Geographically, these links between Britain and the continent suggest a movement to Britain from the east. Exciting new analysis from Bexhill demonstrates a late Mesolithic with complex, datable links to the western part of the continent. This site will be outlined below and, using DIT, forms important evidence for a more complex transition than previously thought.

4.1 Dates, Techno-Complexes, and Techno-Chronologies at Bexhill

Several of the scatters fit into the current schema for the Mesolithic of Britain found in Table 1. The radiocarbon dates for Bexhill range from Deepcar to the late Mesolithic and reaffirms the chronology put forward by Conneller, Bayliss, Milner, and Taylor (2016) and Marshall et al. (2019). For a full description of the radiocarbon dates and techno-complexes at Bexhill, please see Donnelly, Grant, Kennard, Lawrence, and Soudy (2019) and Marshall et al. (2019).

The lithic industry of late Mesolithic Britain (LM5-10: Tables 1 and 2) was, until recently, believed to be fairly homogeneous (Conneller, 2022). Except for regional variability in the north in its final stages (Griffiths, 2014), it has been difficult to create distinct technochronologies within the geometric microlith tradition of the late Mesolithic. This is due to few fifth-millennium radiocarbon dates compared with later periods (Elliott et al., 2020). Because of this, this period exudes an air of timelessness and any change after this (e.g. the transition) is thought to be abrupt. However, the plethora of late Mesolithic dates at Bexhill has allowed for substantial reinterpretation of this uniformity. By comparing the late Mesolithic microliths of Bexhill to the radiocarbon dates, a technochronology can be produced that breaks this fairly homogeneous period into eight phases (LM5-10: Donnelly et al., 2019; Marshall, 2019. Tables 1, 2 and 3). Within these phases, a new microlith typology has emerged – the Bexhill point – which may have links to continental types (see below). This preliminary re-phasing of the late Mesolithic at Bexhill, which was spearheaded by Mike Donnelly, demonstrates the cultural diversity of the late Mesolithic at Bexhill. It has potential not only for identifying regional social nuances and complexities within the late Mesolithic but also for understanding the dynamics of the Mesolithic–Neolithic transition. A larger body of dates from sites like Bexhill can be used to identify whether neolithisation was truly abrupt. Key to this is identifying the origin of the Bexhill point which dates to the final Mesolithic.

Of interest is the debitage associated with the Bexhill point industry. For the most part, it is similar to other late Mesolithic industries with a blade count of around 25–30%. However, for Site 15, the average blade count is around 15% and is associated with “globular” cores and mixed flaking patterns (Figure 4). This has so far been interpreted as a site used for the manufacture of axes, any abandoned products being used as cores (Donnelly et al., 2019).

Figure 4

Typical debitage associated with Bexhill Point industries.

This phasing emphasises that there were several technological traditions existing alongside each other at Bexhill (e.g. LM7b, LM8a, and LM8b). Groups LM8a and LM8b are geographically separate at Bexhill, exhibiting an east–west divide. This may suggest cultural diversity in the late stages of the Mesolithic and may reflect different groups with different ideologies that lived alongside each other at Bexhill (Donnelly et al. 2019). More research is being carried out to more thoroughly explore these coexisting technological traditions. The heterogeneity of culture at Bexhill is highlighted by the diversity of practice in Mesolithic Britain more generally (e.g. Bell, 2007; Bell et al., 2016; Blinkhorn & Little, 2018; Finlay, 2004; Lewis et al., 2019): the diversity in site scale (small ephemeral sites on the Somerset levels and large dense sites at Bexhill); the diverse use of ritualised watery deposits at Langley’s Lane, Blick Mead, and Bath Hot Spring; the use of pits as cultural markers or containers; the various uses for hearths; and the range of shell midden types and use in late Mesolithic Britain demonstrates this (Bell et al., 2016; Blinkhorn & Little, 2018; Blinkhorn, Lawton-Mathews, & Warren, 2016; Jacques, Philips, & Lyons, 2018; Lewis et al., 2019; Mithen, 2019). Just as the fourth millennium was made up of many Neolithics (e.g. Sheridan, 2010), we could propose, then, that the fifth millennium of Britain was made up of many Mesolithics, each with their own unique agencies and interpretations of innovations (Conneller, 2022).

4.2 The Bexhill Points

The Bexhill point, a new microlith type, is characterised by abrupt anvil (bidirectional) backing along the longest lateral side (usually the right) and an abrupt oblique notch towards one end (usually on the left and from the distal) (Figures 5 and 6). There is a small amount of inverse (and sometimes also invasive) retouch around the notch. The notch can vary in size, degree of obliqueness, and diffuseness but the majority of points have quite tight notches that angle between 45 and 90° from the main truncation. Bexhill points almost always have finer leading edge trim. They are likely a derivative of asymmetric micro-tranchet points which, although rare, occur in the final stages of the Mesolithic (Chantal Conneller pers. comm) and mostly cluster around the east of England (Conneller, 2022). Asymmetric micro-tranchet points often have quite diffuse notches compared to Bexhill points (e.g. Stainton, 1989) and do not have inverse and invasive retouch.

Figure 5

Bexhill points.

Figure 6

Bexhill points from SE England and Bexhill-esque points from N. Spain and NW France: (1–4) Bexhill; (5) North Park Farm; (6) Charlwood; (7–8) Saint Gildas 1c; (9–11) Hoëdic; (12–13) El Mazo; (14–15) Mendandia; (16) la Una; (17) Aizpea II; (18) Kobaederra; (19) Aizpea III. Redrawn from: (Alday, 2005, 2014; Alday & Cava, 2008; Campos, Fuertes, & Alonso, 2016; Cubas et al., 2016; Dupont, Marchand, Gruet, & Tessier, 2007; Ellaby, 2004; Fuertes-Prieto et al., 2021; Jones, 2013; Péquart & Péquart, 1954).

The uniqueness and importance of these points was first identified by Roger Ellaby whilst excavating at Charlwood, Surrey, but due to the nature of publication in a Neolithic monograph, little notice was taken by Mesolithic scholars (Ellaby, 2004). Aside from Charlwood and Bexhill, these points may also be found at North Park Farm (Jones, 2013) and possibly Wawcott XXIII (Froom, 2012). A further Bexhill point was recently found at Sturry (northern Kent. [Mike Donnelly pers. Comm]). These sites are also in SE England, suggesting Bexhill points are a regional phenomenon.

The chronology of Bexhill points is of interest. These microliths are present in very low quantities from 5800 BC at Bexhill. That said, the points from scatters 99 and 96 (LM6) are described as atypical and may also derive from more abundant later scatters near by. Bexhill points are most abundant from 5200 BC onwards. The microliths from Charlwood date to 4710–4330 cal. BC (Ellaby, 2004) and single sample from a hearth at Wawcott XXIII dates to 5300–4724 cal. BC (Froom, 2012). The examples from North Park Farm are not currently dated.

Bexhill points are often associated with small crescents, scalene, and backed bladelets but, in the later phases, the assemblages are more mixed and contain small isosceles and atypical trapezes. These latter types may also be continental variants (perhaps similar to Giladriere [Marchand, 1999: c.5700–5300 cal. BC]). In some instances, the Bexhill points are also associated with pottery (scatters 20B and 8E). It is possible, but as yet unproven, that the pottery is contemporary to these scatters as there is no early Neolithic activity nearby, such as Neolithic flint scatters, pits, or structures. The methodical and careful nature of the excavation means these signatures would have been picked up (assuming of course taphonomic processes were the same). Further examination and direct dating of this pottery would be needed to ascertain its true date. A very late Mesolithic Bexhill scatter, 33 (c.4400–4300 BC), that is contemporary with the Bexhill point industries, is associated with cereal grains and has the potential to be the oldest known example of cereal in Britain. However, further dating on the cereal will have to be done to determine whether or not it is intrusive (Marshall et al. 2019: 285).

4.3 Bexhill Points and Late Mesolithic Continental Connectivity

Evolutionarily speaking, the Bexhill point is not comparable to any other microlith type in Britain. Its closest typological neighbour is likely the Horsham point. However, at Bexhill, there is a gap of around 1,500 years between these two industries. This, alongside distinct morphological differences, suggests that they are not the same typology. On Horsham points, for instance, the abrupt retouch is on the shortest lateral edge, creating a concave base that has an obtuse angle to the leading truncation. The longest lateral edge is often curved with a small amount of trim. Further work will be done to see if the Bexhill point has its ancestry in the Horsham but, currently, there does not seem to be any continuity. Because Bexhill points do not seem to have evolved from any technology in Britain, We must look for similar candidates on the continent. Below, we examine several different theories and comparisons as to the origin of Bexhill points.

One theory, championed by Conneller (2022, p. 356, Conneller pers. comm.), is that the micro-tranchet tradition, to which Bexhill points belong, is a microlithised derivative of trapeze industries of the Second Mesolithic of France. The Bexhill point, with its inverse retouch, may be formed from a wide spectrum amalgamation of several characteristics from different armatures, including inversely retouched triangles and trapezoids from the final Mesolithic of northern France (e.g. Biéville-Beuville (Artur, Billard, Hervieu, & Dubrulle, 2008) or Guerny, Normandy (Prost, 2002)). Similarities to the trapezoids found at L’Essart (western France) are also striking. These have abrupt lateral and inverse and invasive basal retouch and are found in direct association with Montclus points, a proxy for possible Neolithic interactions (Marchand, 2007). Similarly, the tool composition of the latter half of the Second Mesolithic is similar to tool assemblages at Bexhill. These include oblique truncations as well as mixed assemblages of symmetrical and asymmetrical trapeze and scalene triangles (Marchand & Schulting, 2019), providing further links between the Bexhill point/micro-tranchet tradition and France.

More direct comparisons may be found in armatures with concave bases from northern Iberia and western France (Figures 6 and 7). The sites from Iberia include El Mazo, La Una, Kobaederra, Mendandia, and Aizpea. Northwest France is so far represented by just one site (Saint Gildas), though points from Hoëdic and Biochon-Est also bare comparison. The similarities between Bexhill points and armatures from these sites are striking. At El Mazo, the points are described as follows: “[a] very unique type of backed bladelet point with a notch at the base” (Fuertes-Prieto et al., 2021, p. 472). Both Bexhill points and those from northern Spain and western France have abrupt retouch along the longest lateral edge and an oblique notch (Figures 6 and 7). In some instances, as at Aizpea, there is invasive retouch around the notch, similar to examples from Charlwood. In the case of Mendandia, there is also trimming on the shorter lateral edge. This can also be observed in examples from North Park Farm and at Bexhill. At St Gildas, on the Armorican Massif (NW France), the points also have abrupt retouch along the lateral edge and a notch on the base.

Figure 7

Direct comparison between a Bexhill point from SE England (left) and a similar point from Mendandia, N. Spain (right).

There are, however, some differences between Bexhill points and continental types. The notch on the Saint Gildas points is at right angles, rather than oblique to, the leading truncation. Moreover, most of these continental variants are lateralised in the opposite direction to the majority of Bexhill points with the exception of Aizpea. This phenomenon of opposite lateralisations has been noted by Anderson Whymark et al. in their analysis of the Old Quay flints, Isles of Scilly, and may be an attempt at regional stylisation (Anderson Whymark et al., 2015). Therefore, these differences in lateralisation do not necessarily mean that those from Bexhill and the continent are separate typologies.

The chronology of these European sites is complex and covers a range of different cultural traditions. The bulk of the Spanish and French points date from c.6500 to 5300 BC overlapping with the earliest dates of Bexhill points from south-east England. El Mazo (Fuertes-Prieto et al., 2021, 6568–6435 cal. BC), an Asturian site with similarities to the Geometric Microlith tradition, is the earliest to have Bexhill-style microliths. This is followed by an uptake in these points at Mendandia, La Una, and Aizpea (5884–5573 cal. BC: Figure 8 for references) in the east. We should be cautious of the dates from Mendandia as, in some instances, bulk samples were used (Zilhao, 2011). However, the dates from the bulk samples correlate well with the stratigraphy and those from single assays (Alday, 2011). There is a geographical gap between the most easterly Spanish site (Aizpea) and Saint Gildas (5886–5730 cal. BC) in north-west France. Although the dates fit, we must be cautious of the dates from Saint Gildas because of the marine reservoir effect (Dupont & Marchand, 2008). Points from the nearby island of Hoëdic may be slightly later (c.5400–5100 BC) (Marchand & Schulting, 2019) and fit well into existing techno-chronologies of the Second Mesolithic.

Figure 8

Location and dates of the Bexhill-esque points from N. Spain and NW France. Note that the points from Hoëdic and Saint Gildas are speculative because of a lack of contextual information and the reservoir effect. Redrawn from: (Alday, 2014; Alday & Cava, 2008; Campos et al., 2016; Cubas et al., 2016; Dupont et al., 2007; Ellaby, 2004; Fuertes-Prieto et al., 2021; Jones, 2013; Péquart & Péquart, 1954).

The geographical and temporal pattern that emerges is a general trend from west to east and then northwards along the French Atlantic coast (Figure 6). On the whole, the similarities and overlapping dates suggest a fairly uniform technological tradition that may demonstrate cross-channel connectivity from the west in the sixth and fifth millennium.

4.4 Continental Points and the Neolithic

These continental points are set in a world of possible Neolithic interactions. The point from Mendandia rock shelter (c.6200–5900 cal. BC) is associated with pottery and the Kobaederra point is additionally associated with cereal. As such, it is thought that these layers represent Neolithic activity (Alday, 2014; Cubas et al., 2016). Consequently, the related armatures suggest an interplay between Mesolithic (El Mazo, La Una, and Aizpea) and Neolithic (Mendandia and Kobaederra) groups. The makers of these points were likely very aware of Neolithic practices and ideologies. On the British side, Bexhill point scatters 8E and 20B contain pottery and contemporary scatter 33 contains cereal grain. These anomalies suggest a growing Neolithic influence on Mesolithic lifestyles.

5.1 Why 4000 BC?

These possible continental connections can inform us about the processes of the British transition. This article does not suggest that there was wholesale indigenous adoption of agriculture in Britain. It hopes, rather, to explain why Neolithic colonists were successful at around 4000 BC in SE Britain, a fact succinctly demonstrated by the DNA evidence, as opposed to any other time. Related to this is the seeming rejection of microliths in the fourth millennium BC. Below we hope to explain that this abandonment of Mesolithic things and practices (which is as much a part of the transition process as the introduction of Neolithic ones) was tied to a slow and gradual acknowledgement of the Neolithic world. In earlier sections, we have demonstrated that Bexhill style points on the continent were made by hunter-gatherers aware on Neolithic cultures. We propose here that as Bexhill points reached Britain in the sixth- and early fifth-millennium BC, so did this awareness of agrarian communities abroad. In accordance with DIT, Neolithic ideas from Spain and France were able to permeate Mesolithic culture through a continuous negotiation between new lifeways and existing belief systems, changing views on relative advantage, comparability, complexity, and possibly observability of Neolithic innovations before they reached Britain wholesale after 4000 BC.

For those initial agricultural colonists to have been successful, they had to understand the lay of the land in all its intricacies (Case, 1969). Hunter-gatherers had a lot to offer in this regard: they knew the land, its resources, and its idiosyncrasies. They knew its weather patterns. They likely also had extensive contacts between neighbouring groups (Champness et al., 2015). They could be a safety net for new migrants who did not know the land or how their crops and cattle would be affected by it. We can note that the earliest Neolithic sites from the United Kingdom (those associated with megalithic burial mounds and “houses,” e.g. White Horse Stone and Coldrum in SE Britain) fit in seamlessly with our Bexhill point dates (Figure 10). The geographical proximity of these early Neolithic sites to the Bexhill point flint scatters may reflect a desire to arrive close to communities who were already aware of the ideologies brought by these new colonists. However, cooperation is often reliant on cultural synergy (Gatcher, Herrmann, & Thoni, 2010). By this it is meant, if two cultures understand each other in some way, they are more likely to cooperate. As suggested by Betti et al. (2020), this awareness between cultures may have fostered symbiotic relationships between hunter-gatherers and the new migrants during the transition. This meant that, by 4000 BC, hunter-gatherers were sufficiently open to new changes brought by colonists from the continent for cooperation to happen. In turn, any surviving Mesolithic groups in the fourth millennium could more easily integrate into new societies if they wished.

This theory perhaps also gives an opposing explanation to the possible earlier failed colonisation of Ferriter’s Cove, Ireland, at c.4300 BC described by Sheridan (2016). This resulted in a “false start” (Sheridan, 2010) rather than a mass spread of agriculture more than 300 years earlier than thought. Neolithic ideas had not sufficiently permeated into Mesolithic culture by this point for permanent change to take place. Conversely, lithic technology associated with Mesolithic continental connections (e.g. Horsham points, Honey Hill points, and Bexhill points) are all a southern British phenomenon, the first successfully colonised area by Neolithic migrants. The initial “failed” colonisations fall outside of this semi-acculturated zone suggesting the importance of this gradual exposure to Neolithic life through cross-channel contact.

With the nuances that emerge from this diffusion of innovation model and the new genetic and typological evidence, one can therefore propose that the two most prolific models of the Mesolithic–Neolithic transition (colonisation and indigenism) are not conflicting but rather complementary interpretations for neolithisation: Interactions between hunter-gatherers and farmers that brought about an awareness and gradual acceptance of Neolithic things and practices, and colonisation of Britain by Neolithic migrants, their practices more palatable to indigenous people who had previously been exposed to Neolithic lifeways. Without earlier contact, there would not have been a successful colonisation; without migration, there would not have been a wider introduction of Neolithic things and practices.

We do not suggest that this was necessarily a linear process and applicable to the whole of Britain. Indeed, DIT argues for a non-linear messy process which occurs unevenly. The preliminary phasing at Bexhill suggests late Mesolithic Britain may have been more regionally and culturally diverse than previously thought. Different Mesolithic communities with differing cosmological beliefs and lifeways would have reacted to new Neolithic ideas differently. The relative advantage, comparability, complexity, trialability, and observability of these innovations would have been different for different communities. As such, the Neolithic would have been adopted at different times, if at all. This is exemplified by the messy picture of the transition proposed by Griffiths (2014) and the time-transgressive adoption of Neolithic “things and practices” proposed by Whittle et al. (2011).