Archaeological evidence for thinking about possibilities in hominin evolution

Abstract

The emergence of the ability to think about future possibilities must have played an influential role in human evolution, driving a range of foresightful behaviours, including preparation, communication and technological innovation. Here we review the archeological evidence for such behavioural indicators of foresight. We find the earliest signs of hominins retaining tools and transporting materials for repeated future use emerging from around 1.8 Ma. From about 0.5 Ma onwards, there are indications of technical and social changes reflecting advances in foresight. And in a third period, starting from around 140 000 years ago, hominins appear to have increasingly relied on material culture to shape the future and to exchange their ideas about possibilities. Visible signs of storytelling, even about entirely fictional scenarios, appear over the last 50 000 years. Although the current evidence suggests that there were distinct transitions in the evolution of our capacity to think about the future, we warn that issues of taphonomy and archaeological sampling are likely to skew our picture of human cognitive evolution.

This article is part of the theme issue ‘Thinking about possibilities: mechanisms, ontogeny, functions and phylogeny’.

1. Introduction

The human ability to think about possible futures, or even entirely fictional scenarios, has been argued to have played a critical role in human evolution [1,2]; but when, where, and how did these abilities evolve? The archaeological record provides the means by which to determine when these capacities first appeared. While there has been extensive debate surrounding how to interpret individual artefacts, assemblages and landscape-wide patterns, it nevertheless remains clear that archaeological remnants are the most direct way to establish facts about the evolution of the human mind [3–7]. Here we examine the current archaeological evidence for the emergence of these capacities in pre- and early Homo sapiens.

2. Thinking about possibilities: in the deep past

The future matters to all animals and many have evolved behavioural mechanisms that increase the likelihood of future survival and reproduction (such as instincts for caching food, building nests, preparing for regular changes such as the day–night cycle, etc.). Furthermore, through conditioning, animals can learn about the regularities of their local environments and so form predictions about what is likely to happen next (e.g. [8]). While immensely powerful, such learning is about the immediate future only, rather than about distant events. A hiatus of minutes, let alone days or weeks, tends to make associating events impossible. However, at least some species appear to be able to do more than just learn by association. For instance, human's closest animal relatives, the great apes, can pick an appropriate tool for a problem that is out of sight (e.g. [9]), can reason about the trajectory of an object even when a hidden event could not be observed (e.g. [10]), and can communicate with each other to coordinate behaviour in collaborative problem-solving tasks [11]. Such data suggest that great apes have some rudimentary capacity to think about events other than what is currently perceived.

There have even been rare experimental claims about animals thinking hours ahead (e.g. [12,13]), but these studies have attracted criticism for not having ruled out simpler explanations for the observed behaviour (e.g. [14–16]). Naturalistic observations have also only rarely led researchers to propose that animals are thinking about distant futures. Perhaps the most extraordinary claim has been that male orangutans may make loud calls in the direction they are headed to communicate travel plans to others [17]. But this case is not entirely compelling given that the orangutans tend to face both in the direction they are travelling as well as the direction they are calling—whether or not they mean the calls to inform others. While there continues to be considerable debate and future research may well establish more sophisticated cognitive capacities in other animals (e.g. [18–22]), current evidence suggests that there is something unique about the human ability to think about the future. In particular, humans appear to be distinct in their capacity to mentally travel to distant future events and to consider multiple, even mutually exclusive possibilities [23–26].

Thought about the distant future and about multiple future possibilities manifest in many distinct human behaviours. For instance, imagining what might be needed in future allows people to make effective plans (e.g. gather dry firewood and tinder before starting a fire) and to prepare for threats and opportunities before they are upon us (e.g. making a weapon before it is needed; carrying a container to gather resources). Merely imagining possibilities, or hearing others relay them, can lead individuals to experience corresponding emotions (e.g. anxiety; elation) which may drive them to act to prevent disaster or to increase the likelihood of success in their own situation [27,28]. We can mimic events or paint scenes—but it is spoken language that we most widely use to communicate our memories of past events, imaginings of future possibilities and thoughts about options [29]. People relentlessly ask questions and give advice reflecting our distinct urge to exchange mental scenarios [22]. Such exchanges have led us to construct shared imaginary realities, where ornaments are symbols with meaning and dead people keep living in some other world, where calendars structure our lives and fictional stories can be more important than fact.

We socially inherit cultural tools that are essential to our everyday lives and ultimate survival (e.g. [30–32]). Though cultural evolution may at times proceed without much higher cognition [33], in our ancestors foresight played a critical role in at least two ways: teaching and innovation [26]. We do not just socially learn through copying, but also by deliberately instructing others in what we think they may need in future. We can use words, maps, and calendars to communicate about what to expect where and when; in a sense such tools can ‘extend our mind' [34]. Also, we can deliberately set out to find new solutions to recurring problems and we can recognize solutions worth retaining for the future. Consider the technologies upon which we so depend today. Many animals use tools, and some even make them (e.g. [35]). However, a key to human innovation is that we can recognize not just the present but also the future use of a tool [36]. This recognition is why we do not throw away tools once we are done with the immediate tasks at hand. Instead, we are driven to retain items for future deployment, carrying them with us—often within mobile containers—to wherever we might need them. Recognition of future use may drive us to refine tools further, say, making them sharper, more efficient, or more reliable. It can also motivate us to intentionally share them with others now or in future (e.g. as heirlooms). It drives us to teach others about the idea. The capacity to think about possibilities has played a critical role in driving humans to develop ever more powerful tools and customs that permeate much of our lives, and along the way, these processes have left identifiable clues in the archaeological record that can tell us about the origins of this critical attribute of the human mind. In the following sections, we consider the currently available archaeological evidence in terms of: (i) technology and behaviour that may reflect foresight by a hominin; and (ii) artefacts and features that reflect the communication of thoughts about the future. We then consider: (iii) the material evidence for the creation and sharing of narratives.

3. Technology and behaviour reflecting foresight

One of the earliest archaeological indicators of increased foresight is control of fire. Being able to produce and maintain fire at will was undoubtedly one of the most significant technological advancements in the human story. With fire, people could keep warm, keep away predators, smoke out prey, cook, and eventually manipulate raw material properties for tool creation (e.g. [37,38]). Making fire requires considerable planning and persistence. Whether using firesticks or striking stones, resulting embers or sparks need to be carefully caught and fanned into a flame, requiring fire-feeding materials to be at the ready. Maintenance of fire also requires ongoing attention, particularly if certain temperatures are required for cooking or tool-making. While there are glimmers of evidence for controlled use of fire reaching back some 1.6 Ma, habitual use of fire (systematically repeated use of fire in specific sites and/or regions [39]) is only evident from about 400 000 to 300 000 years ago [39–41] (figure 1). A particularly useful archaeological record of change in fire-use is found at the Israeli site of Qesem Cave [42]. Here, a well-preserved deposit recording the period between about 420 000 down to 200 000 years ago shows a distinct pick up in the frequency of fire use from 300 000 years ago, implying a change in human behaviour at this time. Shahack-Gross et al. [42] suggest that this change may be in the size of the human group gathering to work next to the fire, as well as how groups were organizing their space.

Figure 1.

Temporal appearance of archaeological evidence for foresight and fiction. Lighter shading indicates sporatic appearance of evidence while darker shading indicates stronger evidence. The last 100 000 years is stretched out to better represent the period of sporatic appearance of evidence in this most recent period, while the last 1.5 Myr is compressed.

Stone tools have long been considered hard evidence for tracking hominin planning capacities, particularly owing to their ubiquity and survivability across enormous spaces and temporal periods. Engagement with such tools may have played an important role in the evolution of mind (e.g. [43]). While Oldowan technologies hint at developing cognitive advances from about 2.5 Ma (e.g. [44]), it is not until the emergence of Acheulean technology some 1.8 Ma that the earliest sign of ‘distinctly human' foresight begins to become evident (e.g. [45]). Analyses of lithic data indicate that making the characteristic symmetrical handaxes and cleavers required considerable skill and dedication, and that they were retained for multiple future uses—suggesting recognition of future use (e.g. [46]). A recent study found that even after 90 h of instructed learning, modern people were still not able to produce a good quality teardrop-shaped handaxe such as those found archaeologically [47]. Therefore, it appears that extensive social learning and practice was required in transferring this flintknapping knowledge from one generation to the next. Whether teaching was involved or not, deliberate practice by an individual to improve future skills is an important expression of foresight [8]. The regularity by which these tools were made suggests that early Homo were acting with a standardized plan in mind, though questions have been raised about what cognitive capacities are actually involved here (e.g. [48–52]). The persistence of the technology for many hundreds of thousands of years implies that this plan did not change much, indicating that the foresight of these tool-makers (at least as reflected in stone tool innovation) was not yet like that of people today.

Changes in stone technology, involving a reduction in the focus on making large cutting tools (that is, handaxes and cleavers) and instead the production of smaller and more diverse types of tools, occur from about 300 000 years ago. In Africa, Acheulian technologies transition into Middle Stone Age (ca 500–280 000 to 50–25 000 years ago) technologies, while in Europe the Acheulian (Lower Palaeolithic) transitioned into the Middle Palaeolithic (ca 300 000 to 40 000 years ago). On both continents, this transition involved the introduction of Levallois approaches to reducing stone nodules into tools [53]. This method requires complex preparatory steps shaping the stone core so as to produce flakes of specific, pre-determined forms. Often cores were reduced to the point where these desired flakes could be immediately struck off, the core then carried or stored around the landscape in readiness for tool replacement. Investigating exactly how individuals or groups organized their tool manufacture and use routines (or lack of routines) has therefore been another pathway to assessing planning abilities [54–57]. In this Middle period, we see not only a development in the complexity of stone tool production, methods which required more steps to create a pre-determined tool type, but also a development in the management of raw material supplies.

Through geological mapping and diagnostic chemical signatures in lithic raw materials, analysts have been able to track the movement of materials across landscapes and debate how such evidence reflects the cognition and/or behaviour of the hominin/s involved (e.g. [58–64]). Rare instances of long-distance transport of obsidian have been documented from the Oldowan (HWKE, Olduvai Gorge at 1.75 Ma, greater than 270 km transport) and Acheulean in eastern Africa [5,65]. For example, Hallos ([46], p. 169) neatly provides instances of knapping for both immediate needs, where ‘almost complete reconstruction of nodules and the production and discard of flake tools in the same location attest to knapping activities carried out in a reactive context' as well as examples where the toolmaker transported not only finished Acheulean bifaces, but bifaces in various stages of production, suggesting that ‘the production of bifaces was not in response to some immediate ‘cue’, such as the presence of a carcass' [46]. Nonetheless, long-distance transport of materials does not become a regular feature of the archaeological record before the second half of the Middle Stone Age (after about 140 000 years ago) [5,65]. Furthermore, once the long-distance transport of raw materials is regularly appearing in the H. sapiens archeological record, we have indications from other datasets that this movement of materials is reflecting large-scale social networks (where items are transported from one person or community to the next before reaching their final destination) rather than the movement of individuals [65]. Such large-scale social networks require extensive thinking about the future, and as such, these data again suggests an underlying change in foresight capacities. By contrast, research into Neanderthal mobility patterns in their European territories found that these populations, while practising a variety of strategies depending on their environmental and cultural context, tended to not travel extensive distances (over 60 km) for the whole of the Middle Palaeolithic (e.g. [66–69]). Part of this difference has been argued to reflect smaller social networks maintained by Neanderthals as opposed to contemporary and later H. sapiens [70].

How people moved these raw materials around the landscape is also of interest here. An individual can only carry so much in their hands and be actively using them at any one time. As such, containers may have been used to move at least some of the materials transported long distances we see recorded archaeologically. Today, containers are often central to long-term storage of both raw materials, food stuffs, and material culture items—such storage, like transport, reflects thinking of the future. We recently reviewed the physically surviving evidence for containers [71] and found that the oldest examples date back to about 100 000 years ago (Blombos, shell ochre containers [72])—where a ‘roof' in preservation appears to be reached. While many speculate that containers would have been useful at much earlier periods, concrete evidence for them is thus far lacking.

Hunting itself may reflect foresight and planning—especially where dangerous or otherwise difficult-to-catch prey was involved [5,73]. For example, when hunting large and/or dangerous animals, modern hunter–gatherers typically first observe the animal to learn its behavioural patterns (such as what path it takes to waterholes or rivers), before they plan the attack and prepare accordingly (which might include prepping weapons, recruiting other hunters, digging a pit-fall or building a corral). Analysis of faunal remains suggest that by 780 000 years ago, hominins were regularly killing large game [74], but direct indications of the technology involved do not appear until some 300 000 years later in the archeological record. Physical evidence for hunting spears is first seen in the form of a horse scapula exhibiting a semicircular perforation from a 500 000-year-old context at Boxgrove, England [75], while similarly-aged stone artefacts from the South African site of Kathu Pan 1 suggests that Homo heidelbergensis—probable common ancestor of Neanderthals and H. sapiens [76]—were constructing stone-tipped spears from about 500 000 years ago ([77] but see [78]). Another example of early hunting weaponry is provided by extraordinary preservation conditions at the German site of Schöningen where several wooden spears used to hunt horses around 300 000 years ago were recovered [79]. Their manufacture and evidence for transport, use, and repair have been interpreted as reflecting considerable planning depth [80].

The Kathu Pan 1 stone artefacts are particularly important here as they exhibit evidence for having been hafted to a wooden shaft, that is, they represent a composite technology. Evidence for hafting (among other composite technologies) has traditionally been considered a marker of increasing foresight (e.g. [81–85]). Demonstrating that a stone tool was hafted can be challenging where clear damage from this process is not abundantly evident, so while the Kathu Pan 1 examples indicate that this practice stretches back to 500 000 years ago, it is not until about 300 000 years ago that spear tips with clear traces of having been hafted are commonly present in African Middle Stone Age and Eurasian Middle Palaeolithic records (e.g. [5,86–89]).

Indeed, hafting represents a radical departure from all previous tool manufacture in that composite tools involve adding, rather than just subtracting, something. Hafting typically required the use of ligatures and/or adhesives in order to connect the spear tip to the shaft, these parts needing to be manufactured individually before being brought together into a new whole. Adhesives often required a number of steps to prepare for use (e.g. [90–92]). For example, adhesives could be made from heating particular types of tree sap or bark, or stem from animal-based materials. They frequently require heating to certain temperatures for specific times and can have ochre or other materials mixed in to help develop the adhesive qualities of the natural raw materials. Material evidence for hafting in the form of surviving pieces of adhesives only date back to about 200 000 years ago and have been found in a number of Middle and Upper Palaeolithic contexts in Eurasia and Africa, indicating that H. heidelbergensis, Neanderthals, and H. sapiens had mastered this technology (e.g. [89,93–96]).

Another hunting technology that adhesives can be deployed for is the construction of traps, such as the use of sinews in creating snares. Setting a trap clearly implies thoughts about the possibility of a prey animal being caught in the future. As snares tend to be made up of organic components (e.g. ligatures made from materials such as plant fibres) which typically do not survive long periods in the archaeological record, discovering when such technology was first used is exceedingly difficult. Having said this, evidence has been recovered from Sibudu in South Africa suggesting that tiny blue duikers were trapped with snares some 70 000 years ago [86], the snares possibly making use of the adhesive mixture of acacia gum, beeswax, and powdered ochre found used for hafting stone points at this same site [84].

Finally, a technological advance that opened up oceans of possibility were watercraft. The earliest signs of sea crossings come from the island of Flores [97], however, this migration might have been the result of an involuntary journey, as individuals may simply have been hanging on for dear life onto something that floats. It has been argued that the colonization of the ancient continent of Sahul (Australia and New Guinea) required watercraft technology in order to travel between the islands from Sunda to the shores of Sahul [98], with the first humans arriving prior to 65 000 years ago [99]. In this case, we still do not know exactly what route was taken by the first explorers to reach the continent, though geographical research has determined that inter-island visibility was better, following the ‘northern route' from Borneo, through Sulawesi, and into the bird's head of New Guinea [100]. Compelling evidence for watercraft comes from Japan where some 38 000 years ago, people transported obsidian from the island of Kozushima to the mainland [101]. This raw material acquisition required a 20 mile journey across the sea and a return back with heavy load—impossible to do without a watercraft and considerable planning. Many of these technological advances are undoubtedly intertwined not only with foresight, but also with an increased capacity to communicate and coordinate thoughts about the future.

4. Artefacts and features which communicate thoughts about the future

Language enables humans to broadcast their expectations, plans, and reflections, and much of human conversation is about events displaced in time. Mental time travel and language are closely related [102], and both are open-ended because they draw on the capacity to recursively embed elements into larger structures [103,104]. So, evidence for the origin of language may conceivably inform us also about the evolution of foresight. However alas, in spite of considerable scholarly attention, the facts of language evolution remain to be established (e.g. [105–108]). For the present purposes, we therefore focus on aspects of communication about the future that have left traces in the archaeological record.

Archaeologically, the classic evidence for thinking about possible futures has been burial [109]. While it is possible that communities buried their dead simply to deter scavengers or effectively dispose of a decaying corpse, the careful placement of a body and especially the inclusion of grave goods is generally considered evidence that people were burying the body out of shared concerns for the deceased [110]. Potentially the careful burial of infants, including fetuses, with grave goods being most suggestive of particular concern for possible futures as such young individuals cannot have yet achieved earned status (see summary in [111]). While such cases may indicate inherited status, they may also represent care for the child into an afterlife. The oldest examples of burials widely accepted date to about 130 000 to 100 000 years ago at Mt Carmel (Israel) and include both H. sapiens and Neanderthals [112,113], though there has been some suggestions that H. heidelbergensis may have deliberately dropped deceased individuals into the Sima de los Huesos cave system in Spain [114] with similar suggestions for Homo naledi at Rising Star Cave in South Africa ([115], but see [116]). Burials of the very young are known in both H. sapiens (Krems–Wachtberg [117]) and Neanderthal contexts (Amud, La Ferrassie [111,118]). Indeed, the oldest burial in Africa is that of a 3-year-old child dating to 78 000 years ago from the Middle Stone Age layers of Panga ya Saidi (Kenya) [119]. A slightly younger burial (dating to about 74 000 years ago) of an infant (known simply as BC3) has also been found, this time from Border Cave in South Africa. This latter child was found associated with a single perforated and ochred Conus shell [120].

We might also consider the development of semi-permanent forms of ornaments—beads and pendants made from hard animal materials and stone—as possible evidence for forethought as such items play a central role in message transmission. In recent times, ornaments are regularly used to exosomatically carry information about the wearer. So ingrained is this behaviour that archaeologists have been able to use the appearance and distribution of beads and pendants to study past cultural development and interactions. When this use of personal adornment for communication began is unknown, though it is widely believed that easy to manipulate materials (such as plant-parts and natural earth colourants such as ochre) were first used before people moved to investing time and effort into the more durable forms [121]. While the impetus for a move to using more durable materials to create ornamentation (marine shell beads and later animal teeth, bones, antler, ivory, stones, etc.) is also unknown (but may be to do with increasingly large and internally differentiated societies), the advantages of this behavioural change meant that messages were able to be transmitted beyond a single interaction. They could now last for significant periods of time, including across lifetimes (being transferred from one individual to another [122]), and could be altered or rearranged to transmit new messages [123]. Durable ornaments were also able to be accummulated in greater and greater numbers, allowing increasing visual impact and an amplification of the messages they convey across generations. Because beads and pendants can transmit information about the labour costs involved in their creation, these items provided an excellent format for social competition, conspicuous displays of individual and group wealth, social networks, and community cohesiveness. As such, durable ornaments could become more and more complex in their use and be designed to benefit wearers not only now, but in the future. The earliest examples of such ornaments appear in the archaeological record from around 142 000 years ago, where a collection of 33 Tritia shell beads from Bizmoune Cave, Morocco demonstrates a long-standing tradition of shell bead use had already developed by that time in northern Africa [124]. This use of durable body adornment was not restricted to H. sapiens, with a number of finds in Neanderthal sites suggesting that this species was also practising this behaviour at least as far back as 115 000 years ago [125].

Eventually, symbols were deployed to communicate what can be found where and so equip people for future journeys. A small number of complex Gravettian (ca 30 000 to 20 000 years ago) engravings from Moravia (specifically, the sites of Pavlov and Predmostì) may be maps or hunting plans. Possibly the most convincing example is one from Pavlov which shows the site itself (represented by a double circle) as it is found in front of a mountain. This image also shows the nearby river and surrounding valley and slopes—information that might be useful for creating a hunting strategy [126]. An even clearer example was found in younger Late Magdalenian contexts at Abauntz Cave (Navarra, Spain). Here, the engraving dating to 13 660 cal. BP shows the landscape surrounding the site—still recognizable today—along with routes to different parts of that landscape [127]. As the evidence currently stands then, maps appear at least from the very end of the Pleistocene epoch, though it is worth remembering that map-making can take many forms, some may not leave any traces (such as songlines) or not be easily recognizable to archaeologists. For example, the Ammassalik carved wooden maps which are tactile and represent the coast lines of Greenland may not have been interpreted correctly were it not for them having been collected from the makers themselves [128].

Finally, people also constructed maps of time. The earliest potential example was recovered from Aurignacian levels of Abri Blanchard (France) and has been argued to track the lunar cycle [129]. Dating to about 32 000 years ago, this artefact is an ovaloid plaque of bone about 10 cm long. One face of this piece features a series of small incisions of comma-like and circular shapes along with equally spaced notches along the top edge. Careful study of these intentional marks led Marshack to argue that the incised design represents a lunar calendar, a suggestion upheld by ensuing analysis [130] (see summary of Palaeolithic evidence for astronomy in [131]). Later examples include a 10 000-year-old arrangement of 12 pits in Scotland (Warren Field) which is claimed to be the remnants of a lunar calendar [132], as well as an apparent solar calendar dated to 7000 years ago at Goseck in modern Germany. This last example comprises of two arcs which appear to track the changing location of sunrise and sunset over the year [133]. Using this calendar, the length of future days and the change of the seasons can be predicted. Structures like these can be found around the world (e.g. Wurdi Youang stone arrangement [134]; or Nabta Playa [135,136]), with the most famous probably being Stonehenge [137]. With space and time mapped out, people could ever more effectively teach each other, coordinate plans, and shape the future to their own design.

5. Evidence for storytelling and thinking about fictional happenings

Material evidence for early humans having shared narratives is exceedingly hard to identify archaeologically owing to the ephemeral nature of storytelling. Having said this, collaborations between geological researchers and Indigenous communities are finding ways to date stories. For example, Nunn and Reid collected Aboriginal stories of coastline inundation from 21 locations around the Australian coast and tied them to post-glacial sea-level rise events which occurred more than 7000 years ago [138]. Similarly, Matchan et al. were able to provide a date for the Gunditjmara story of Budj Bim [139]. In this oral tradition, four giant beings arrived in southeastern Australia. Three of these beings strode out to other parts, but one crouched in place, transforming into a volcano—Budj Bim—and spat out his teeth (lava). This lava flowed to the sea and created the coastal wetlands that has sustained the Gunditjmara. Geological study and dating of the twin eruptions of the Budj Bim and Tower Hill volcanoes using 40Ar/39Ar found that this event occurred some 37 000 years ago and did flow to the sea—mirroring the story as told. This eruption created a significant impression on the community, leading to a story powerful enough to continue into the present day. These examples suggest that, in some contexts at least, scientists may be able to connect Indigenous stories with geological events dating back thousands of years.

A line of evidence which is on less steady ground is the habitual use of hearths. Polly Weissner found in her study of conversations among modern Ju/‘hoansi Bushmen that 80% of conversation at night (around hearths) were stories—as opposed to day time conversations which typically revolved around daily tasks [140]. Interestingly, the night-time stories often involved religious experiences, myths, and folktales. We saw above that habitual use of hearths appear from around 300 000 years ago, and so one might speculate that the telling of stories date to sometime after this period. Alas, the evolution of language, let alone when it was used to convey stories about real or fictional possibilities, has remained unresolved and this is not the space to review that vexing issue (e.g. [105–108]).

Another speculative indicator may be evidence for music, as music is often intertwined with storytelling and religious activity [141]. Music is defined as ‘sound that conveys emotion', and consequently, can be a powerful tool for conveying messages and stories [142], potentially aiding group cooperation, social cohesion and alliance signalling in early communities [143–145]. It may also have been used as a mnemonic device for transmitting larger amounts of information. Musical instruments become visible in the archaeological record in the form of flutes made from bird bone and mammoth ivory recovered from Geißenklösterle Cave (Swabian Jura, Germany) around 43 000 and 42 000 years ago [146]. However, given how well-designed and crafted these tools were, they appear to have been the product of a long tradition of flute-making. Therefore, musical instruments must have first emerged at some earlier point in time, and of course, singing or whistling allow us to make music without leaving any traces. As it stands, we can only be confident about fictional stories once people began to create visual representations of fictional creatures.

Images of therianthropes—creatures that combine human and animal features—are the best evidence for fictional narratives in the deep past [147–149]. They imply some creative thought as such creatures cannot be simply reproduced from real observations, and so it would seem reasonable to assume that they represent characters that featured in fictional stories. Up until recently, the famous ‘lion man' sculpture in mammoth ivory recovered from Hohlenstein-Stadel, southwest Germany has been the oldest evidence for such an image [148]. Found in an Aurignacian context, it has been dated to about 40 000–39 000 years ago, though a recent reinterpretation of this enigmatic artefact suggests that it is not a lion-headed therianthrope, but rather of a realistic standing bear [150]. The only similarly-aged example of therianthropes from this Aurignacian period is the possible ‘part woman-part bison' painting at Chauvet (Rhône-Alpes, France) [151], with later Magdalenian (21 000 to 14 000 years ago) examples including the ‘bison-man' at Le Gabillou (Dordogne, France) [152], a ‘bison-man' and a ‘deer-man' at Les Trois Frères (Ariège, France) [153], and a possible ‘bird-man' at Lascaux [154]. On the other side of the world, ongoing research in Sulawesi, Indonesia has discovered a painted panel older even than the infamous lion man. At Leang Bulu’ Sipong 4, a panel dated to at least 43 000 years ago shows several therianthropes hunting wild pigs and anoa [147]. Similar to the ‘bird-man' at Lascaux, each of the humanoid figures appear to have bird-like heads. By this point in time, then, it seems likely that fictional narratives were part of human culture.

6. Discussion and conclusion

The capacities to think about the remote past, future, and even about fictional situations is widely assumed to be unique to humanity. Even if comparative psychological research was narrowing this gap, humans clearly demonstrate considerably more competence than even our closest animal relatives. The ability to think ahead underlies many critical human behaviours, such as teaching and innovation, which together drive human cultural evolution [26].

In reviewing the temporal appearance of the archaeological evidence summarized herein, it is immediately apparent that issues of taphonomy need to be considered. The earliest evidence for forward planning is dominated by the most durable of artefact forms: stone tools. Study of these almost indestructible artefacts allows us to see a change in behaviour at about 1.8 Ma (with the appearance of symmetrical bifacial handaxes), and then another change somewhere around 500 000 years ago involving smaller stone tools and the appearance of hafting (figure 1). Most other lines of evidence rely more heavily on organic artefacts (those made from shell, bone, or any other hard animal material) being archaeologically visible. While exceptional preservation conditions at some locations has allowed for the rare survival of organic-based material culture to remain for extended periods (such as the Schöningen spears), most date to well after 140 000 years ago. As such, it is possible that the technologies themselves and the cognitive capabilities that underlie them were present much earlier in our past, but material traces of them are unlikely to survive the ravages of time.

These issues just focus on the ability of artefacts to preserve based on their raw material alone, the resolution of the archaeological record is made significantly more complicated by differential preservation between site types and regions. For example, many activities which would allow us insight into hominin cognition (like hunting or large gatherings) did not occur in the cave environments which better preserve archaeological materials and which archaeologists preferentially target for excavation. Further, some whole regions are not particularly conducive to preserving discrete archaeological layers which allow us to compare one time period to another in the first place. For example, the European continent is characterized by episodic sedimentation and erosion events which often mixed and destroyed various contexts, and as such archaeologists are frequently presented with a series of snapshots which can be missing long sections in periods of interest. Africa, on the other hand, includes landscapes which were largely stable over very extended periods, but this stability resulted in another problem where enormous numbers of artefacts were able to be laid down over many, many thousands of years, accumulating as an apparently single archaeological layer which can be difficult to separate for meaningful temporal analysis [155,156].

Another factor we need to keep in mind is sample size. Some of the archaeological evidence which provides the clearest picture of foresight and storytelling, are items probably made in far fewer numbers than everyday technologies. Specifically, ornaments, musical instruments, calendars, maps, and art pieces could be expected to have been made only occasionally, and therefore the likelihood of them being recovered by archaeologists is significantly lower [157]. Furthermore, it should be remembered that once a form of evidence becomes visible in the archaeological record, the behaviour that underpins those particular artefacts or features was probably practised long before the date of the record. In other words, the archaeological signature tends to record when certain behaviours are commonly occurring—not their first appearance in the population. Such first appearances are extremely unlikely to enter the archaeological record and be found by archaeologists, and so the dates of emergence we discuss are likely to underestimate the initial appearance of a behaviour.

Having outlined these caveats, the currently available archaeological evidence suggests that there were three broad transitions in the evolution of human foresight (figure 1):

• (i) hominins could conceive of future needs: ca 1.8 million + individuals retained tools for future use, multi-step technologies developing;
• (ii) hominins acted cooperatively with others to shape the future: ca 500 000 + flexible planning and cooperation facilitating growing technological and social possibilities; and
• (iii) hominins were capable of open-ended foresight and communication about possibilities: ca 140 000 + increasing use of material culture to record and exchange thoughts about the future.

These three phases are not to suggest that the evolution of human foresight occurred in a step-like manner. If anything, the archaeological record appears to indicate that the opposite is true: that these cognitive abilities developed gradually over time and in a number of hominin species. We suggest these three transitions to highlight the nature of the apparent changes that occurred over long periods of time. From what we can see, the first signs of hominins considering a more remote future are linked to preparation for future needs, as indicated by the appearance of symmetrical bifaces requiring a complex series of steps to execute, as well as the glimmers of control over fire and the long-distance movement of materials for future use. By around 500 000 years ago, there are the first signs hominins moved beyond this, developing technologies which require additional planning and considerable cooperation. The period between 500 000 and 140 000 years ago is characterized by everyday use of controlled fire, big game hunting, advancing methods of core preparation, diversification of lithic technology, and the advent of hafting methods. Such evidence indicates that individuals were cooperatively working to shape the future in their favour. Finally, from 140 000 years ago, ‘occasional' evidence appears in the archeological record (those artefacts that were made in far fewer numbers than everyday tools) indicating that multiple forms of material culture had been co-opted into helping manipulate future possibilities through human interactions. Signs of mobile containers, traps, and other novel tools begin to appear, indicating the presence of a feedback loop where foresight leads to innovations, that in turn increase the capacity to plan and prepare for possibilities [26]. Ornaments and burials, and eventually cave art and other symbols indicate increased communication about events other than the here and now. Indeed, in the latter part of this phase the first signs of entirely fictional scenarios in the form of theriothropes appear in the record.

The human capacity to construct mental scenarios and embed them into narratives is not a discreet, encapsulated cognitive module, but seems to depend on a host of cognitive faculties that need to work in concert, as illustrated by Suddendorf and Corballis' theatre production metaphor [45]. To entertain events that may occur tomorrow and beyond, one needs a mental stage to imagine the scenario offline, and recursive capacities to combine basic elements, representations of objects and actors, into narratives. Furthermore, to take advantage of any such mental plays one needs something akin to a director to evaluate, and something like an executive producer to put plans into action. Finally, we can learn from other's mental scenarios and reflections, and coordinate our actions accordingly, because we can broadcast our mental plays to each other. The theatre metaphor has been used to argue that use the theatre metaphor to argue that human mental scenario building rests on many complex cognitive processes and mechanisms (e.g. working memory, recursive embedding, theory of mind, causal reasoning, metarepresentation, executive functions and language) and that shortcomings in any one of these components may severely restrict one's ability to imagine, evaluate, and share mental events. For instance, without considerable working memory capacity—long highlighted as critical in human cognitive evolution (e.g. [158,159])—our forebears could not have mentally staged and compared multiple versions of future events. Here, we restricted our focus on archaeological evidence for advanced foresight itself, but future studies may additionally want to assess traces of changes to these purported component capacities (cf. [160] for a similar strategy examining the development of foresight in children).

Despite the vagaries of the archaeological record, it appears that the origins for the human ability for thinking about future possibilities stretches back beyond 1.8 Ma. From this early period of humanity, up through an increasingly complex family tree [161], the ability to conceive of the future and act to mitigate its risks or maximize a potential outcome developed gradually. These abilities cumulated in both individuals and groups cooperating on an ever-increasing scale and co-opting in increasing types of material culture to manipulate and track the future. People eventually exchanged ideas about potential situations, driving them on to avoid some possibilities while striving to turn others into reality.

Data accessibility

This article has no additional data.

Authors' contributions

M.C.L.: conceptualization, data curation, formal analysis, investigation, methodology, writing—original draft, writing—review and editing; T.S.: conceptualization, methodology, writing—original draft, writing—review and editing.

Both authors gave final approval for publication and agreed to be held accountable for the work performed therein.

Conflict of interest declaration

We declare no competing interests.

Funding

This study was supported by ARC DP210101572 to T.S.