TrendsFrom the Bible to Chomsky, languagehas been commonly viewed as repre-senting a profound change in thought,appearing suddenly and uniquely inhumans, with communicative lan-guage as a byproduct.

Modern research increasingly sug-gests mental continuity, more consis-tent with evolutionary theory, with

OpinionLanguage Evolution: AChanging PerspectiveMichael C. Corballis1,*

From ancient times, religion and philosophy have regarded language as afaculty bestowed uniquely and suddenly on our own species, primarily as amode of thought with communication as a byproduct. This view persists amongsome scientists and linguists and is counter to the theory of evolution, whichimplies that the evolution of complex structures is incremental. I argue here thatlanguage derives frommental processes with gradual evolutionary trajectories,including the generative capacities to travel mentally in time and space and intotheminds of others. What may be distinctive in humans is the means to commu-nicate these mental experiences along with knowledge gained from them.

nonhuman animals showing evidencefor mental functions such as mentaltime travel and theory of mind thatare critical to communicativelanguage.

Language can then be viewed as adevice for sharing thoughts andexperiences rather than as a vehiclefor those mental contents themselves.

Nonhuman animals such as great apesand dogs are more capable of under-standing human-like language than ofproducing it, so communicative lan-guage in humans depended on adapt-ing output systems to be flexibleenough to convey the generativity ofthought.

Such a system may have originated inbodily gesture before incorporatingvocal signals that eventually providedus with speech.

1School of Psychology, University ofAuckland, Private Bag 92019,Auckland 1142, New Zealand

*Correspondence:m.corballis@auckland.ac.nz(M.C. Corballis).

Is Language Uniquely Human?According to the Old Testament, language was a gift to Adam and was for a time the commonlanguage of all humans.When people built the Tower of Babel so that theymight reach closer toHeaven, the Lord thought them disrespectful and dispersed them, and confounded theirlanguages so they could no longer communicate. Echoing this story Noam Chomsky hasargued that language was bestowed on some individual, whom he whimsically names Pro-metheus, within the past 100 000 years, well afterHomo sapiens is thought to have emerged asa distinct species [1]. Languages subsequently proliferated into the 6000 or so languages of theworld, the vast majority mutually incomprehensible.

The notion that language emerged in humanswell after our species itself evolved receives somesupport from archaeology, based on an apparent explosion of artistic and technologicalinnovation within the past hundred millennia followed by dispersal from Africa in which H.sapiens eventually replaced other large-brained hominins such as the Neanderthals andDenisovans. Human artifacts from this period included cave art, musical instruments, sculpture,bodily embellishments, and notations for record keeping and ended an apparent stasis in toolmanufacture toward more innovative and varied forms. This surge of activity has been variouslylabeled the ‘human revolution’ [2], the ‘great leap forward’ [3], and the last of seven ‘majortransitions in evolution’ [4] that began with the emergence of replicating molecules. Thepaleoanthropologist Ian Tattersall writes ‘The entirely novel competitive entity representedby behaviorally modern [258_TD$DIFF]Homo sapiens appeared on the planet far too rapidly to be accountedfor by the slow workings of natural selection at the individual level’ [5].

The Evolutionary ChallengeSuch views are antithetical to the theory of evolution. Charles Darwin himself wrote ‘If it could bedemonstrated that any complex organ existed, which could not possibly have been formed bynumerous, successive, slight modifications, my theory would absolutely break down. But I canfind no such case’ [6].

Could language be the case that Darwin feared? Chomsky has often characterized language as‘an organ of the body’ and ‘on a par with the digestive and immune systems’ [7]. Language is

Trends in Cognitive Sciences, April 2017, Vol. 21, No. 4 http://dx.doi.org/10.1016/j.tics.2017.01.013 229© 2017 Elsevier Ltd. All rights reserved.

GlossaryEvo–devo: derived from evolutionarydevelopmental biology andincorporates the idea that regulatorygenes, highly conserved acrossspecies, can influence embryonicdevelopment through varying thetiming and expression of othergenes. For example, althoughdifferent species have dissimilarlimbs, such as legs, flippers, orwings, their development iscontrolled by very similar genes, withthe differences due to the way theirexpression is regulated.Extended Evolutionary Synthesis:extensions of the Modern Synthesis,dating from the mid-20 century,incorporating further developments ingenetics, biology, and populationthinking; includes aspects such asevo–devo, punctuated evolution,spandrels, and niche construction.Modern Synthesis: early 20-centuryintegration of Darwinian evolutionarytheory with Mendelian inheritanceand population-level ideas.Niche construction: process bywhich organisms construct their ownenvironment, which can then alterthe selection of traits that areadaptive. For example, traits adaptedfor survival in a hunter–gatherersociety may differ from thoseadaptive in city life.Punctuated evolution: the idea thatnew species evolve relatively rapidlyfollowed by a period of stasis withlittle genetic change.Spandrel: an architectural termreferring to the triangular spacebetween an arch, a pillar, and aceiling that is sometimesappropriated by artists for addeddecoration. The term is borrowed inbiology to refer to a byproduct of afeature that evolved for differentreasons. A simple example is the useof the nose to hold spectacles. Thebrain has been claimed to provide arich source of spandrels, withactivities such as music or chessperhaps being byproducts of acomputational faculty selected forother adaptive functions.

also undeniably complex, given that its rules of operation are still not well understood after some60 years of Chomskian linguistics. The emergence of an organ as complex as language in asingle step seems clearly counter to Darwinian theory.

Evolutionary theory, however, has itself undergone changes, with theModern Synthesis (seeGlossary) and the later Extended Evolutionary Synthesis incorporating ideas from genetics,embryology, and population thinking, raising the possibility of more rapid change than thatimplied by the selection of randommutations [8]. However, it remains problematic whether thiscan drastically alter the Darwinian scenario, let alone explain the emergence of language in asingle step. Pinker and Bloom discuss a number of neo-Darwinian possibilities including theview that language is a ‘spandrel’ – a byproduct of the brain’s evolved capacity for computa-tion [9] – but conclude that ‘there is every reason to believe that language has been shaped bynatural selection as it is understoodwithin the orthodox “synthetic” or “neo-Darwinian” theory ofevolution’ [10]. Gould and Eldredge wrote of ‘punctuated evolution’, implying that evolutioncan proceed in spurts, but they were clear that it is ‘not about ecological catastrophe or suddengenetic change’ [11]. Chomsky himself [1] appealed to evo–devo, whereby complex changescan occur through mutations in regulatory genes that orchestrate the timing of developmentrather than through the emergence of new genes. Such mutations, however, do not addcomplexity, suggesting they cannot explain the sudden emergence of a system as complex asgrammar [12].

Another prediction of the synthesis is that evolutionary change can be hastened through ‘nicheconstruction’, whereby organisms modify their own environments, biasing the process ofselection [8]. This might have influenced and perhaps accelerated evolutionary change duringthe Pleistocene, which saw the emergence of what has been termed the ‘cognitive niche’ [13],but this would imply a gestation period of some 2 million years rather than a sudden changewithin the past 100 000 years. Some have also contested the archeological evidence for the‘human revolution’, suggesting either that the changes were continuous and date back toprecursors ofH. sapiens [14] or that the impression of change was more a question of sporadicregional variation than of discontinuity through time [15].

Language and ThoughtA common theme among those who argue for the sudden emergence of language is thatit signaled a change in the manner of thought itself. Thus, Tattersall writes that from about100 000 years ago ‘we start finding plausible indications that members of the new species werestarting to think symbolically’ [5]. Chomsky writes of ‘internal language’ (I-language) as thefundamental basis of human symbolic thought with communication merely a byproduct [1,7]. Itis, then, through a secondary process of ‘externalization’ that the diverse languages, spokenand signed, are formed. I-language is considered common to all humans, underlying whatChomsky calls ‘universal grammar’ – a term that goes back to 17th-century scholars whosought to identify aspects of language common to all languages [16]. There have been manyattempts to specify such a grammar, once satirized by James D. McCawley in his book ThirtyMillion Theories of Grammar [17], but Chomsky’s most recent and most economical account isthe Minimalist Program [18]. The main ingredient is ‘unbounded Merge’, whereby elements aremerged recursively to generate a potential infinity of structures.

The burden of explaining language evolution is lessened, however, if language is regarded ascommunication, not thought. In this case thought can be considered to have precursors innonhuman animals (e.g., [16,19]) rather than to have appeared in a single step in H. sapiens.Language can then be considered primarily as a device for sharing our thoughts and expe-riences – more a tool than a fundamental cognitive shift [20,21]. In moving the emphasis tolanguage as communication rather than thought, this more commonsense view, in the words of

230 Trends in Cognitive Sciences, April 2017, Vol. 21, No. 4

Dor, ‘turns the Chomsky proposal on its head’ [21]. The ‘great leap forward’, if indeed a reality,can then be attributed to an enhanced capacity to share knowledge and experience rather thanthe thoughts themselves.

This is not to say that thought is independent of language. Indeed, many if not most of ourthoughts and knowledge are derived from what other people tell us. However, thoughtsthemselves, whether derived from personal experience or from communication with others,are not the same as the language in which they are expressed. We often cannot find words forparticular thoughts we have or for individuals we can envisage although their names escape us.Thought cannot be tied to any specific language, and when we translate from one language toanother the thought remains – at least ideally – despite being expressed in different words [22].Even in retelling stories within the same language, we typically do not use the same words orexpressions. As a source of knowledge, however, language creates a ratchet effect so thatknowledge can build over time and spread among individuals, allowing specializations todevelop [4]. In these ways language may have been influential in a human revolution, butthrough its power of communication rather than from a change in the innate structure ofthought itself.

If expressive language is designed for the communication of thought, its properties must be atleast partly shaped by theway in which thought itself is structured. In this respect the account tobe given here is in accord with Chomsky’s. The difference lies in the suggestion here that thenature and structure of thought have a long and gradual evolutionary history.

Several authors have speculated about possible precursors of language, including the manu-facture and use of tools [23,24], navigation [25,26], reciprocal altruism [27], and socialunderstanding [28]. Here I focus on two mental capacities that seem especially critical tothe recursive, generative nature of language itself and that may ultimately obviate the need forany special mechanism unique to language, or indeed to our own species.

Mental Time Travel‘Displacement’ is a design feature of language enabling reference to events at other places andother times, past or future [29], and is increasingly recognized as critical to its evolution (e.g.,[30,31]). This in turn must depend on the ability to imagine such events, such as what you didyesterday or plan to do tomorrow. Although some, including myself, have claimed that mentaltime travel itself is uniquely human [32,33], the evidence increasingly suggests that it may have along evolutionary pedigree, with behavioral evidence from species as disparate as scrub jays[34], rats [35], and chimpanzees [36,37]. Behavioral evidence, however, can be ambiguous orflawed [38], and my own shift to an acceptance of evolutionary continuity was prompted mainlyby evidence from neuroscience [39].

It is well established that the hippocampus in humans is activated when people consciouslyremember past episodes or imagine future ones [40,41] or even imagine purely fictitious ones[42]. In the rat, ‘place cells’ in the hippocampus record where the animal is located in a spatialenvironment [43] and sometimes trace out trajectories even when the animal has beenremoved. The trajectories are sometimes ‘replays’ of trajectories previously taken, sometimesthe reverse of those trajectories [44], and sometimes trajectories the animal did not take[45], some of which may be anticipations of future trajectories [46]. Reviewing this evidence,Moser, Rowland, and Moser write that ‘the replay phenomenon may support “mental timetravel” . . . through the spatial map, both forward and backward in time’ [47]. Mental travelsthrough known spatial locations may be a common capacity of animals that move aboutphysically in space and need to know whether they are, where they have been, and where theymight go next.

Trends in Cognitive Sciences, April 2017, Vol. 21, No. 4 231

In humans the hippocampus seems to be involved in language itself [48,49]. Even in the rat,however, hippocampal activity has some language-like properties. It is influenced by activity inthe neighboring entorhinal cortex in a modular fashion, adjusting for parameters like spatialscale, direction of the head, and proximity to borders. Different combinations of modules canresult in thousands of combinations, comparable to ‘that of an alphabet in which all words of alanguage can be generated by combining only 30 letters or less’ [47]. In short, spatialimagination, like language, is generative. I can imagine myself in different locations, such asmy office, in many different ways and at different spatial scales. I can zoom out from the office toits location in the house, the location of the house in the city, the city in the country, and so on.Even in the rat, the spatial scale zooms out as recording shifts from the rearward to the forwardend of the hippocampus [50] and this arrangement is mirrored in human hippocampal activityas people process narratives linked to videos, with increasingly forward activation as the focusshifts from detail to more global understanding [51].

Generative grammar itself, then, may depend on the generative nature of spatiotemporalimagination rather than on any property unique to language itself. The zooming propertyimplies recursion, in which spatial maps are nested in maps at larger scales. It has even beenproposed that Chomsky’s Minimalist Program and the concept of Merge can be applied tosimple sensorimotor actions such as grasping and manipulating an object [52]. We areprofoundly spatial creatures and even our non-spatial thoughts, such as reasoning and logic,may be grounded in spatial metaphor rather than abstract symbols [53]. The concept ofuniversal grammar has been doubted [54,55], but if it can be said to exist its universality mayreside in the common experience of the spatiotemporal world rather than in the innate structureof language.

Theory of MindWe travel mentally not only in time and space but also into the minds of others. The capacity tounderstand what others feel, think, or believe is known as theory of mind and underlies thehuman obsession with storytelling, whether in the form of novels, plays, movies, TV soaps,gossip, or, in earlier times – and among present-day African tribes – stories told around thecampfire at night [56]. It has been suggested that storytelling drove the evolution of languageitself [57].

Theory of mind is required for language in a capacity other than storytelling. Language is notmerely a matter of words; it requires that the speaker and listener (or signer) know what is ineach other’s mind and that each knows that the other knows this – requiring metacognition[58,59]. In this sense language is underdetermined [60]. Even individual words may carrymultiple meanings depending on context. An extreme example is the English word ‘set’, which,according to the Chambers Dictionary, has 105 different meanings.

Whether nonhuman species possess theory of mind has been much disputed sincePremack and Woodruff in 1978 asked ‘Does the chimpanzee have a theory of mind?’[61] Thirty years later there was still disagreement; in one view, the answer was negative andthe very idea that chimpanzees might have theory of mind was deemed ‘Darwin’s mistake’[62], while others found evidence that chimpanzees could be shown to understanding thegoals, intentions, perceptions, and knowledge of others but not their beliefs or desires [63].A more recent study, however, showed that great apes look in anticipation of where ahuman agent will falsely believe an object has been hidden [64] – they seem to pass thefalse-belief test, often regarded as the gold-standard test of theory of mind [65]. This studyjoins a chorus of studies gradually showing greater mental continuity between humans andother species than is commonly assumed. Even rhesus monkeys may be capable ofspontaneous metacognition [66].

232 Trends in Cognitive Sciences, April 2017, Vol. 21, No. 4

A neural network for mental orientation in ‘space, time, and person’ is now fairly well establishedin humans and includes widespread regions in the frontal, parietal, and temporal cortices [67].This network overlaps extensively with what has been termed the ‘default mode network’,which is active in ‘mind wandering’ – spontaneous internal activity unrelated to the concerns ofthe moment [68]. Homologous networks have been identified in monkeys [69] and rats [70],again suggesting evolutionary continuity. Of course, internal thinking may be less complex inrats andmonkeys than in humans, but we should bear in mind Darwin’s edict: ‘The difference inmind between man and the higher animals, great as it is, certainly is one of degree and not ofkind’ [71].

In summary, comparative research is increasingly drawn to the conclusion that some nonhumananimals are capable ofmore ‘advanced’ cognition than hitherto realized, suggesting evolutionarycontinuity rather than a sudden shift in cognition that somehow made humans special. Mentalexcursions in time and space and into the minds of others may be more comprehensiveand sophisticated in humans than in other species in part because of communicative languageitself, which feeds our mental travels through storytelling and our minds through the exchangeof knowledge. Their origins, however, may go well back in our evolutionary history.

The Communicative AspectAccording to a shifting tide of opinion, it is in our ability to communicate our thoughts, ratherthan our thoughts themselves, that we differ most profoundly from other species and it is our

Box 1. The Gestural Theory of Language Origins

Among primates only humans appear capable of vocal learning [77] and therefore of speech. Language, however, is notrestricted to speech; sign languages have all the hallmarks of true language [78]. This raises the possibility that languageoriginated, not from vocal calls, but frommanual gestures. This idea goes back to the 18th century [79–81] and has beenincreasingly endorsed over the past few decades (e.g., [82–90]).

Captive great apes have never developed anything resembling speech but have learned to use manual gestures [72,91]in ways that have language-like features; great apes in the wild also communicate bodily in ways that are morelanguage-like than their vocalizations [92]. People, too, gesture manually while speaking [93,94], and like speech,gesture involves the hippocampus [95]. Some have supposed that visible gesture and vocal language have always beenequal partners [96,97] (but see [98] for a contrary view).

Gestural communication may derive from the primate ‘mirror system’, a brain circuit active during both the productionand the perception of intentional movements, suggesting a natural basis for communication [85]. In macaques thesystem responds to actions perceived acoustically as well as visually but not to animal calls, implying that theincorporation of vocal action evolved relatively late [99]. Brain imaging of a homologous system in humans suggeststhat it expanded to encompass the language circuit [100].

Bodily actions provide a natural basis for relaying events in our predominantly spatial world, especially through hand andarmmovements as exemplified in sign languages. During the Pleistocene, with the emergence of the genusHomo, brainsize increased and the hands were increasingly freed by the shift to obligate bipedalism, allowing communicationthrough gesture and mime [101]. In the interests of efficiency, mimed gestures were probably simplified and con-ventionalized over time, losing their iconic quality [102]. Tomasello writes that human languagemight have ‘evolved quitea long way in the service of gestural communication alone, and the vocal capacity is actually a very recent overlay’ [89].

To some, a transition from the visible to the acoustic mode seems too extreme to be credible (e.g., [103,104]). Speech,however, is itself a system of gestures of the tongue, lips, and larynx [105]. Movements of the hand andmouth are linkedneurally, phylogenetically, developmentally, and behaviorally [106–109], suggesting that mouth gesturesmight plausiblyhave blended into manual ones [110], and the visible movements of speech remain verbally informative [111], althoughadding vocalization makes the gestural content more accessible as audible speech. The acoustic code, however, ismore arbitrary and is sustained largely through culture, although it does carry non-arbitrary associations [112,113].

The transition might be an early example of miniaturization � tucking the burden of communication into the mouth andfreeing the hands and arms for activities such as carrying or making and using tools. It was perhaps this switch, and notlanguage itself, that produced the ‘human revolution’ [114], but it did not stop there. Subsequent shifts, from writing tothe Internet, have profoundly shaped and widened human culture.

Trends in Cognitive Sciences, April 2017, Vol. 21, No. 4 233

Outstanding QuestionsCanwe identify aspects of thought thatare present only in humans and thathave no counterparts in nonhumananimals?

If so, can we find evidence that suchaspects evolved suddenly within thetime span of Homo sapiens or is itmore plausible to suppose that theyevolvedmore gradually, such as duringthe Pleistocene?

Can we find anatomical, genetic, orfossil evidence that bears on the ori-gins of speech as distinct from lan-guage itself?

Or does the evidence further supportthe origins of expressive language inmanual gestures rather than vocalcalls?

capacity to communicate that gives an inflated view of human cognition. The difference hasmore to do with production than with understanding. Great apes [72] and even dogs [73,74]can understand and respond to spoken requests involving sequential processing and[259_TD$DIFF]substantial word vocabularies, which implies a degree of symbolic understanding but a lackof themeans to produce anything approaching humanproductive language.Great apes, at least,may come closer through bodily movement and gesture than through vocalization (Box 1).

The pressure to develop output systems flexible enough to communicate our thoughts andexperiences probably grew during the Pleistocene when our hominin forebears wereincreasingly forced from a forested environment to more open territory with dangerouspredators, and survival depended on cooperation and social interaction – the aforemen-tioned ‘cognitive niche’ [13]. The initial impetus may have come from the sharing ofepisodic events, whether remembered, planned, or fictitious but perhaps also increasinglyincluding information about territory, danger, food sources, tool making, and the habits andabilities of individuals. Its course of evolution is probably best indexed by the threefoldincrease in brain size that began some 2 million years ago and is probably attributable tothe emergence of grammatical language [75] and the vast increase in knowledge that itafforded. The increase in brain size was incremental through this period, again suggestinggradual evolution rather than the sudden appearance of a prodigious Prometheus withinthe past 100 000 years.

Concluding Remarks: Is Rapprochement Possible?There are some signs that the longstanding divide over language evolution may be softening.Through progressive modifications, Chomsky’s theory of syntax has become simpler. This issuggested in the title of the current Minimalist Program, whose primary ingredient is Merge,recently described as the ‘simplest case’, requiring a ‘slight mutation’ producing a ‘slightrewiring of the brain’ [7]. These descriptions now seem at odds with earlier notions of languageas a bodily organ producing a major transition in evolution and are perhaps an attempt to bringthe evolution of language within the bounds of evolutionary credibility.

At the same time, Minimalist theory seems to place a greater burden on externalization toexplain how a concept as apparently simple as Merge can account for the complex and variedgrammars among the 6000 or so languages, spoken or signed, in the world. A recent articlepoints out that the parsing of a sentence cannot be based simply on the string of words or signsbut must depend on an internal device to impose hierarchical structure; in its simplest form, thatdevice is Merge, applied recursively. The question then is whether this internal structuring is‘uniquely human (species-specific) and uniquely linguistic (domain-specific)’ [76] or whether itdepends on the way that experience and knowledge have been incrementally structured overmillions of years of evolution (see Outstanding Questions).

AcknowledgmentsThree unnamed reviewers provided extremely helpful input.

References

1. Chomsky,N. (2010)Somesimpleevodevo theses: howtruemight

they be for language? In The Evolution of Human Language(Larson, R.K., ed.), pp. 45–62, Cambridge University Press

2. Mellars, P. and Stringer, C., eds (1989) The Human Revolution:Behavioral and Biological Perspectives on the Origins of ModernHumans, Princeton University Press

3. Szathmáry, E. (2015) Toward major evolutionary transitionstheory 2.0. Proc. Natl Acad. Sci. U. S. A. 102, 10104–10111

4. Diamond, J. (1992) The Third Chimpanzee: The Evolution andFuture of the Human Animal, HarperCollins

234 Trends in Cognitive Sciences, April 2017, Vol. 21, No. 4

5. Tattersall, I. (2016) A tentative framework for the acquisition oflanguage and modern human cognition. J. Anthropol. Sci. 94,157–166

6. Darwin, C. (1859) On the Origin of Species by Means of NaturalSelection, John Murray

7. Chomsky, N. (2015) Some core contested concepts. J.Psycholing. Res. 44, 99–104

8. Laland, K.N. et al. (2015) The extended evolutionary synthesis:its structure, assumptions and predictions. Proc. Biol. Sci. 282,20151019

9. Gould, S.J. and Lewontin, R.C. (1979) The spandrels of SanMarco and the Panglossian paradigm: a critique of the adapta-tionist programme. Proc. Biol. Sci. 205, 281–288

10. Pinker, S. and Bloom, P. (1990) Natural language and naturalselection. Behav. Brain Sci. 13, 707–784

11. Gould, S.J. and Eldredge, N. (1977) Punctuated equilibria: thetempo and mode of evolution reconsidered. Paleobiology 3,115–151

12. Christiansen, M.H. and Chater, N. (2008) Language as shapedby the brain. Behav. Brain Sci. 31, 489–558

13. Pinker, S. (2010) The cognitive niche: coevolution of intelligence,sociality, and language. Proc. Natl Acad. Sci. U. S. A. 107,8993–8999

14. McBrearty, S. and Brooks, A.S. (2000) The revolution thatwasn’t: a new interpretation of the origin of modern humanbehavior. J. Hum. Evol. 39, 453–563

15. Shea, J.J. (2011) Homo sapiens is as Homo sapiens was. Curr.Anthropol. 52, 1–35

16. Fitch, W.T. (2010) The Evolution of Language, Cambridge Uni-versity Press

17. McCawley, J.D. (1982) Thirty Million Theories of Grammar,University of Chicago Press

18. Chomsky, N. (1995) The Minimalist Program, MIT Press

19. Hurford, J.R. (2007) The Origins of Meaning, Oxford UniversityPress

20. Everett, D.L. (2012) Language: The Cultural Tool, Pantheon

21. Dor, D. (2015) The Structure of Imagination: Language as aSocial Communication Technology, Oxford University Press

22. Jackendoff, R. (2011) What is the human language faculty? Twoviews. Language 87, 586–624

23. Steedman, M. (2014) On dendrophilia: Comment on “Toward acomputational framework for cognitive biology: unifyingapproaches from cognitive neuroscience and comparative cog-nition” by W. Tecumseh Fitch. Phys. Life Rev. 11, 382–388

24. Stout, D. and Chaminade, T. (2012) Stone tools, language andthe brain in human evolution. Philos. Trans. R. Soc. Lond. B Biol.Sci. 367, 75–87

25. Hauser, M.D. et al. (2010) The faculty of language: what is it, whohas it, and how did it evolve? Science 298, 1569–1579

26. Berwick, R.C. and Chomsky, N. (2016)Why Only Us: Languageand Evolution, MIT Press

27. Bickerton, D. (1995) Language and Human Behavior, Universityof Washington Press

28. Seyfarth, R.M. and Cheney, D.L. (2014) The evolution of lan-guage from social cognition. Curr. Opinion Neurobiol. 28, 5–9

29. Hockett, C.F. (1960) The origins of speech. Sci. Am. 203, 88–96

30. Bickerton, D. (2014) More than Nature Needs: Language, Mindand Evolution, Harvard University Press

31. Gärdenfors, P. and Osvath, M. et al. (2010) Prospection as acognitive precursor to symbolic communication. In The Evolu-tion of Human Language (Larson, R.K., ed.), pp. 103–114,Cambridge University Press

32. Tulving, E.R. (2002) Episodic memory: frommind to brain. Annu.Rev. Psychol. 53, 1–25

33. Suddendorf, T. and Corballis, M.C. (2007) The evolution offoresight: what is mental time travel, and is it unique to humans?Behav. Brain Sci. 30, 299–351

34. Clayton, N.S. et al. (2003) Can animals recall the past and planfor the future? Trends Cogn. Sci. 4, 685–691

35. Wilson, A.G. et al. (2013) Event-based prospective memory inthe rat. Curr. Biol. 23, 1089–1093

36. Janmaat, K.R.L. et al. (2014) Wild chimpanzees plan their break-fast time, type, and location. Proc. Natl Acad. Sci. U. S. A. 111,16343–16348

37. Osvath, M. and Karvonen, E. (2012) Spontaneous innovationfor future deception in a male chimpanzee. PLoS One 7,e36782

38. Suddendorf, T. and Corballis, M.C. (2010) Behavioural evidencefor mental time travel in nonhuman animals. Behav. Brain Res.215, 292–298

39. Corballis, M.C. (2013) Mental time travel: a case for evolutionarycontinuity. Trends Cogn. Sci. 17, 5–6

40. Addis, D.R. et al. (2007) Remembering the past and imaginingthe future: common and distinct neural substrates duringevent construction and elaboration. Neuropsychologia 45,1363–1377

41. Buckner, R.L. (2010) The role of the hippocampus in predictionand imagination. Annu. Rev. Psychol. 61, 27–48

42. Hassabis, D. et al. (2007) Using imagination to understand theneural basis of episodic memory. J. Neurosci. 27, 14365–14374

43. O’Keefe, J. and Nadel, N. (1978) The Hippocampus as aCognitive Map, Clarendon Press

44. Foster, D.J. and Wilson, M.A. (2006) Reverse replay of behav-ioural sequences in hippocampal place cells during the awakestate. Nature 440, 680–683

45. Gupta, A.S. et al. (2010) Hippocampal replay is not a simplefunction of experience. Neuron 65, 695–705

46. Pfeiffer, B.E. and Foster, D.J. (2013) Hippocampal place-cellsequences depict future paths to remembered goals. Nature497, 74–79

47. Moser, M.B. et al. (2015) Place cells, grid cells, and memory.Cold Spring Harb. Perspect. Biol. 7, a021808

48. Covington, N.V. and Duff, M.C. (2016) Expanding the languagenetwork: contributions from the hippocampus. Trends Cogn.Sci. 20, 869–870

49. Duff, M.C. and Brown-Schmidt, S. (2012) The hippocampusand the flexible use and processing of language. Front. Hum.Neurosci. 6, 69

50. Strange, B.A. et al. (2014) Functional organization of the hippo-campal longitudinal axis. Nat. Rev. Neurosci. 15, 655–669

51. Collin, S.H.P. et al. (2015) Memory hierarchies map ontothe hippocampal long axis in humans. Nat. Neurosci. 18,1562–1564

52. Knott, A. (2012) Sensorimotor Cognition and Natural languageSyntax, MIT Press

53. Lakoff, G. (2014) Mapping the brain’s metaphor circuitry: met-aphorical thought in everyday reason. Front. Hum. Neurosci. 9,958

54. Evans, N. and Levinson, S.C. (2009) The myth of languageuniversals: language diversity and its importance for cognitivescience. Behav. Brain Sci. 32, 429–492

55. Tomasello, M. (2009) Universal grammar is dead. Behav. BrainSci. 32, 470

56. Wiessner, P.W. (2014) Embers of society: firelight talk amongthe Ju/’huansi Bushmen. Proc. Natl Acad. Sci. U. S. A. 111,14027–14035

57. Harari, Y.N. (2014)Sapiens: A Brief History of Humankind,HarvillSecker

58. Grice, H.P. (1989) Studies in the Ways of Words, CambridgeUniversity Press

59. Sperber, D. and Wilson, D. (2002) Pragmatics, modularity andmind-reading. Mind Lang. 17, 3–23

60. Scott-Phillips, T. (2015) Speaking Our Minds: Why HumanCommunication is Different, and How Language Evolved toMake it Special, Palgrave Macmillan

61. Premack, D. and Woodruff, G. (1978) Does the chimpanzeehave a theory of mind? Behav. Brain Sci. 4, 515–526

62. Penn, D.C. et al. (2008) Darwin’s mistake: explaining the dis-continuity between human and nonhuman minds. Behav. BrainSci. 31, 108–178

63. Call, J. and Tomasello, M. (2008) Does the chimpanzee have atheory of mind? 30 years later. Trends Cogn. Sci. 12, 187–192

64. Krupenye, C. et al. (2016) Great apes anticipate that otherindividuals will act according to false beliefs. Science 354,110–116

65. Wimmer, H. and Perner, J. (1983) Beliefs about beliefs: repre-sentation and constraining function of wrong beliefs in youngchildren’s understanding of deception. Cognition 13, 103–128

66. Rosati, A.G. and Santos, L.R. (2016) Spontaneous metacogni-tion in rhesus monkeys. Psychol. Sci. 27, 1181–1191

Trends in Cognitive Sciences, April 2017, Vol. 21, No. 4 235

67. Peer, M. et al. (2015) Brain system for mental orientation inspace, time, and person. Proc. Natl Acad. Sci. U. S. A. 112,11072–11077

68. Buckner, R.L. et al. (2008) The brain’s default network: anatomy,function, and relevance to disease. Ann. N. Y. Acad. Sci. 1124,1–38

69. Vincent, J.L. et al. (2007) Intrinsic functional architecture in theanaesthetized monkey brain. Nature 447, 83–86

70. Lu, H. et al. (2012) Rat brains also have a default mode network.Proc. Natl Acad. Sci. U. S. A. 109, 3979–3984

71. Darwin, C. (1871) The Descent of Man and Selection in Relationto Sex. (2nd edn), Appleton

72. Savage-Rumbaugh, S. et al. (1998) Apes, Language and theHuman Mind, Oxford University Press

73. Kaminski, J. et al. (2004) Word learning in the domestic dog:evidence for ‘fast mapping’. Science 304, 1682–1683

74. Pilley, J.W. and Reid, A.K. (2011) Border collie comprehendsobject names as verbal referents. Behav. Proc. 86, 184–195

75. Falk, D. (2016) Evolution of brain and culture: the neurologicaland cognitive journey fromAustralopithecus to Albert Einstein. J.Anthropol. Sci. 94, 99–111

76. Everaert, M.B.H. et al. (2015) Structures, not strings: linguisticsas part of the cognitive sciences. Trends Cogn. Sci. 19, 729–743

77. Pfenning, A.R. et al. (2014) Convergent transcriptional special-izations in the brains of humans and song-learning birds. Sci-ence 346, 1256846

78. Emmorey, K. (2002) Language, Cognition, and Brain: Insightsfrom Sign Language Research, Erlbaum

79. de Condillac, E.B. (1971) An Essay on the Origin of HumanKnowledge: Being a Supplement to Mr. Locke’s Essay on theHuman Understanding, A Facsimile Reproduction of the 1756Translation by T. Nugent of Condillac’s 1747 Essay, Scholars’Facsimiles and Reprints

80. Rousseau, J.-J. (1782) Essai sur l’Origine des Langues. Geneva

81. Vico, G. (1744/1953) La Scienza Nova, Laterza

82. Hewes, G.W. (1973) Primate communication and the gesturalorigins of language. Curr. Anthropol. 14, 5–24

83. Kimura, D. and Archibald, Y. (1974) Motor functions of the lefthemisphere. Brain 97, 337–350

84. Armstrong, D.F. et al. (1995) Gesture and the Nature of Lan-guage, Cambridge University Press

85. Rizzolatti, G. and Arbib, M.A. (1998) Language within our grasp.Trends Neurosci. 21, 188–194

86. Corballis, M.C. (2002) From Hand to Mouth: The Origins ofLanguage, Princeton University Press

87. Rizzolatti, G. and Sinigaglia, C. (2006) Mirrors in the Brain: HowOur Minds Share Actions and Emotions,Oxford University Press

88. Pollick, A.S. and de Waal, F.B.M. (2007) Ape gestures andlanguage evolution. Proc. Natl Acad. Sci. U. S. A. 104, 81–89

89. Tomasello, M. (2008) The Origins of Human Communication,MIT Press

90. Arbib, M.A. (2012) How the Brain Got Language: The MirrorSystem Hypothesis, Oxford University Press

91. Gardner, R.A. and Gardner, B.T. (1969) Teaching sign languageto a chimpanzee. Science 165, 664–672

92. Hobaiter, C. and Byrne, R.W. (2011) Serial gesturing by wildchimpanzees: its nature and function for communication. Anim.Cogn. 14, 827–838

236 Trends in Cognitive Sciences, April 2017, Vol. 21, No. 4

93. Kimura, D. (1973) Manual activity during speaking.1. Right-handers. Neuropsychologia 11, 45–50

94. Goldin-Meadow, S. and Brentari, D. (2015) Gesture, sign andlanguage: the coming of age of sign language and gesturestudies. Behav. Brain Sci. Published online October 5, 2015.http://dx.doi.org/10.1017/S0140525X15001247

95. Hilverman, C. et al. (2016) Hippocampal declarative memorysupports gesture production: evidence from amnesia. Cortex85, 25–36

96. Kendon, A. (2011) Vocalisation, speech, gesture, and thelanguage origins debate. Gesture 13, 349–370

97. McNeill, D. (2012) How Language Began: Gesture and Speechin Human Evolution, Cambridge University Press

98. Corballis, M.C. (2014) The word according to Adam: The role ofgesture in language evolution. In From Gesture in Conversationto Visible Action as utterance: Essays in Honor of Adam Kendon(Seyfeddinifur, M. and Gullberg, M., eds), pp. 177–197, JohnBenjamins

99. Kohler, E. et al. (2002) Hearing sounds, understandingactions: action representation in mirror neurons. Science 297,846–848

100. Haberling, I.S. et al. (2016) Language, gesture, and handed-ness: evidence for independent lateralized networks. Cortex 82,72–85

101. Donald, M. (1991) Origins of the Modern Mind, Harvard Univer-sity Press

102. Burling, R. (2005) The Talking Ape, Oxford University Press

103. MacNeilage, P.F. (2008) The Origin of Speech,Oxford UniversityPress

104. Burling, R. (1999) Motivation, conventionalization, and arbitrari-ness in the origin of language. In The Origins of Language: WhatHuman Primates can Tell Us (King, B.J., ed.), pp. 307–350,School of American Research Press

105. Studdert-Kennedy, M. (2005) How did language go discrete? InLanguage Origins: Perspectives on Evolution (Tallerman, M.,ed.), pp. 48–67, Oxford University Press

106. Bass, A.H. and Chagnaud, B.P. (2012) Shared developmentaland evolutionary origins for neural basis of vocal–acoustic andpectoral–gestural signaling. Proc. Natl Acad. Sci. U. S. A. 109,10677–10684

107. Gentilucci, M. et al. (2001) Grasp with hand and mouth: akinematic study on healthy subjects. J. Neurophysiol. 86,1685–1699

108. McNeill, D. (1985) So you think gestures are nonverbal? Psy-chol. Rev. 92, 350–371

109. Petrides, M. and Pandya, D.N. (2009) Distinct parietal andtemporal pathways to the homologue of Broca’s area in themonkey. PLoS One 7, e1000170

110. Corballis, M.C. (2012) How language evolved from manualgestures. Gesture 12, 200–226

111. McGurk, H. and MacDonald, J. (1976) Hearing lips and seeingvoices. Nature 264, 746–748

112. Ramachandran, V.S. and Hubbard, E.M. (2001) Synaesthesia –

a window into perception, thought and language. J. Conscious.Stud. 8, 3–34

113. Blasia, D.E. et al. (2016) Sound–meaning association biasesevidenced across thousands of languages.Proc. Natl Acad. Sci.U. S. A. 103, 10818–10923

114. Corballis, M.C. (2004) The origins of modernity: was autono-mous speech the critical factor? Psychol. Rev. 111, 543–552

Box 1. Structure Connecting Sound and Meaning

Language is structured at all levels: phonology (sound structure), morphology (word structure), and syntax(phrase structure). The examples below show how structures built by the computational system aresystematically mapped onto sound and meaning.

The two ways of pronouncing the string ‘white board eraser’, also reflected in spelling, are paired with adifference in their meaning: (i) whíteboard eraser: an eraser for whiteboards; and (ii) white bóard eraser: aneraser that is white.

The eraser in (i) itself could have any color, but in (ii) it has to be white. Both meanings and stress patterns aresystematically derived from the structural patterns given in Figure IA,B, respectively.

In Figure IA, the adjective ‘white’ first merges with the noun ‘board’ and constructs the nominal compound‘whiteboard’, which, as a unit, is merged with the noun ‘eraser’, yielding a bigger compound, an eraser,erasing what is written on whiteboards. When pronounced, the structure gives rise to the stress pattern[whíteboard eraser].

In Figure IB, ‘board’ first merges with ‘eraser’, producing the nominal compound ‘board eraser’, which, as aunit, is merged, in syntax, with the adjective ‘white’, yielding a nominal phrase, a board eraser that has awhite color. When articulated (in speech), the structure gives rise to the stress pattern [white bóard eraser].

(A)

N

Ns Nweraser

As Nwwhite board

(B)

LetterWhat is Languageand How Could itHave Evolved?Martin B.H. Everaert,1

Marinus A.C. Huybregts,1

Robert C. Berwick,2,3

Noam Chomsky,4

Ian Tattersall,5 Andrea Moro,6

and Johan J. Bolhuis7,8,*

Unraveling the evolution of human lan-guage is no small enterprise. One couldstart digging somewhere in the largelyunobservable past, working forwards tothe present, hoping to surface in the rightspot. Alternatively, one could start withthe currently observed and well-estab-lished properties of human language,the phenotype of language, and workbackwards, with these ‘knowns’ guidingthe search for otherwise speculative his-torical ‘unknowns’. In a recent issue ofTrends in Cognitive Sciences, Corballis[1] appears confident that only the firststrategy will serve. Evolutionary explana-tions necessarily are historical, but fewevolutionary biologists faced with sucha paucity of historical evidence wouldforge ahead without first defining what,exactly, the phenotype is that ultimatelyevolved [2]. Yet, Corballis criticizes whatwe actually know about the human lan-guage phenotype, because it does notconform to his speculations [3]. Webelieve that Corballis’ odd research inver-sion suffers from misconceptions regard-ing what we know about both languageand evolution.

NP

Aw Nswhite

Ns Nwboard eraser

Figure I. Nominal compound (A) versus nominal phrase (B). Abbreviations: A, adjective; N, noun; s,strong; w, weak.

The Nature of the HumanLanguage Phenotype (Is NotCommunication)There is no denying that language issometimes used to communicate, likethis Letter. However, this should not leadto the apparently common fallacy that thedesign of language can be inferred

reverse-engineering style from this singlefunctional perspective. Artful kinestheticsof human skeletal structures in motion,aka ‘dance’, also communicates, butone would be hard pressed to derivethe evolutionary history of the human tibia

Trend

from watching Swan Lake. In the sameway, linguists now know with near surgi-cal-precision how the sentence ‘skeletalstructure’ generally fixes meaning. Weknow how ‘communication’ in the senseof transfer of propositional meaning is

s in Cognitive Sciences, August 2017, Vol. 21, No. 8 569

facilitated by language: our manipulationof meaning is systematic and relies “on aningrained ability to recognize structure inlanguage.” [4]. Meaning builds on acomputational system that is sensitiveto structural factors leading to hierarchicalstructure (Box 1). All this has been over-looked by Corballis.

‘Language’ reduced to a mode of com-munication becomes vacuous as anexplanatory motivation. Bees communi-cate with flowers, your router communi-cates with your computer, wecommunicate via oil on canvas, and soon [5]. The Corballian world of communi-cation is so diffuse that it becomes all themore puzzling why “the emergence of anorgan as complex as language” wouldapparently be limited to humans.

Structured ThoughtCorballis claims thought is structured,and that the “nature and structure ofthought have a long and gradual evolu-tionary history”, suggesting that, on thispoint, he is in agreement with the positionadvocated by Berwick and Chomsky [3].However, all examples we know point tothought structured by syntax, not thereverse. The hierarchical structure builtby the brain when processing sentencesfeeds our conceptual apparatus. Conse-quently, language is basically a thought-expressing tool [3,6]. For Corballis,‘thought’ permeates the entire animalkingdom. That may be so, but he failsto give any clue as to what he might meanby this expansive notion of ‘thought’, howwe could find out how ants, or songbirds‘think’, and why, in this view, “expressivelanguage” would take millions of years toappear.

RecursionCorballis suggests that “generative gram-mar may [our emphasis] depend on thegenerative nature of spatiotemporal imag-ination, rather than on any propertyunique to language itself.” However, this

570 Trends in Cognitive Sciences, August 2017, Vol. 21, N

is merely idle verbiage. There is not even aremote connection between the two. Ifsomeone were to say that the waggledance of the honeybee ‘may’ dependon the laws of motion, no one wouldpay attention. A recent comparative studyof cross-species generative systemsasserts that nonhuman animals havenothing resembling human recursive syn-tax [7]. While many animal species recog-nize statistical-probabilistic sequences,linear associations, or even algebraicrules, only humans appear capable ofinternalizing generative algorithms. In linewith his view on ‘spatiotemporal imagina-tion’, Corballis appears to assume thatlanguage inherits these sorts of propertyfrom similarly structured actions. Such anapproach, linking recursion to observa-tions from interactive language use, how-ever, fails [8]. Moro has shown that thesuperficial parallels here between actionsequences and sentences are misguided,again essentially backwards [9,10]. Self-reference, a defining property of recur-sion, appears to be absent from thedomain of motor action and spatiotempo-ral imagination of nested maps [7], yet arich part of human language.

The Nature of EvolutionWe are surprised that Corballis sub-scribes to the view that anything otherthan an ancient and gradual origin forlanguage “is counter to the theory of evo-lution”, for he is clearly aware that ourunderstanding of evolutionary mecha-nisms has been refined considerably overthe past 150 years. Certainly, evolutionarychange requires transitions from one via-ble state to another, but this does notentail that phenotypic steps are necessar-ily the tiny and incremental ones he favors.It is also unclear why he believes thatchanges in gene regulation cannot “addcomplexity”, especially when it is almostcertainly the case that modifications atthis level gave rise to the anatomicallydistinctive species Homo sapiens [11].Furthermore, the archeological recordcontains no evidence of behaviors com-patible with the style of linguistic

o. 8

information processing until after anatom-ical Homo sapiens had come on thescene some 200 000 years ago: a factthat Corballis mentions, but whose rele-vance is left undiscussed.

Mental Time TravelCorballis also refers to, but regrettablydoes not discuss, the position that “Noother organism, instantaneously andeffortlessly extricates from the environ-ment language-relevant data, and in arather comparable way quickly attainsrich linguistic competence, again a featutterly beyond other organisms even in itsrudimentary aspects.” [12]. Corballisattributes these achievements to mentaltime travel (MTT) and Theory of Mind(ToM), although, as frequently noted, bothof these competences are often dissoci-ated from language ability: “autistic chil-dren highly defective in theory of mind[ . . . ] can acquire rich linguistic compe-tence (and in fact a great deal of languageacquisition proceeds before a childshows any sign of having attained theoryof mind).” [12]. Significantly, again, nomechanism is suggested to lead fromToM to the specific structures of lan-guage. Worse, there is no discussion ofwhat the mechanisms of recursivethought in MTT or ToM are, or how theylead to the feat that has to be explained.Shifting the burden from recursive lan-guage to recursive thought in MTT andToM appears to us to leave the problemexactly where it was, adding nothing.

1Utrecht Institute of Linguistics, Utrecht University, 3512

JK Utrecht, The Netherlands2Department of Electrical Engineering and Computer

Science, Massachusetts Institute of Technology,

Cambridge, MA 02139, USA3Department of Brain and Cognitive Sciences,

Massachusetts Institute of Technology, Cambridge, MA

02139, USA4Department of Linguistics and Philosophy,

Massachusetts Institute of Technology, Cambridge, MA

02139, USA5American Museum of Natural History, New York, NY

10024-5192, USA6School of Advanced Studies - Instituto Universitario di

Studi Superiori, Pavia 27100, Italy7Cognitive Neurobiology and Helmholtz Institute,

Department of Psychology, Utrecht University, 3584 CH

Utrecht, The Netherlands

8Department of Zoology and St Catharine’s College,

University of Cambridge, Cambridge CB2 3EJ, UK

*Correspondence: j.j.bolhuis@uu.nl (J.J. Bolhuis).

http://dx.doi.org/10.1016/j.tics.2017.05.007

References1. Corballis, M.C. (2017) Language evolution: a changing

perspective. Trends Cogn. Sci. 21, 229–236

2. Everaert, M.B.H. et al. (2015) Structures, not strings: lin-guistics as part of the cognitive sciences. Trends Cogn.Sci. 19, 729–743

3. Berwick, R.C. and Chomsky, N. (2016) Why Only Us:Language and Evolution, MIT Press

4. Winter, Y. (2016) Elements of Formal Semantics: An Intro-duction to the Mathematical Theory of Meaning in NaturalLanguage, Edinburgh University Press

5. Moro, A. (2016) Impossible Languages, MIT Press

6. Chomsky, N. (2010) Some simple evo-devo theses: howtrue might they be for language? In The Evolution of HumanLanguage (Larson, R.K., ed.), pp. 45–62, Cambridge Uni-versity Press

7. Hauser, M.D. and Watumull, J. (2016) The Universal Gen-erative Faculty: the source of our expressive power inlanguage, mathematics, morality, and music. J. Neuroling.Published online November, 9, 2016. http://dx.doi.org/10.1016/j.jneuroling.2016.10.005

8. Legate, J.A. et al. (2014) Recursive misrepresentations:reply to Levinson (2013). Language 90, 515–528

9. Moro, A. (2014) On the similarity between syntax andactions. Trends Cogn. Sci. 18, 109–110

10. Moro, A. (2014) Response to Pulvermueller: the syntax ofactions and other metaphors. Trends Cogn. Sci. 18, 221

11. Tattersall, I. (2012) Masters of the Planet, PalgraveMacmillan

12. Chomsky, N. (2015) Some core contested concepts. J.Psycholing. Res. 44, 91–104

LetterLeaps of Faith: AReply to Everaertet al.Michael C. Corballis1,*

In their comments on my Opinion in TiCS[1], Everaert and colleagues [2] make sev-eral leaps concerning the evolution of lan-guage, including the idea that languageimplies a mental structure unique tohumans, and that this came aboutthrough an evolutionary ‘great leap for-ward’ within our limited time span.Against this backdrop, perspectives onlanguage evolution have changed overthe past few decades, with languageincreasingly regarded as a device forcommunicating thought rather than asthought itself, with the structure of

thought emerging gradually accordingto neo-Darwinian principles.

Everaert et al. suggest there ‘isn’t even aremote connection’ between spatiotem-poral imagination and grammar. There isnevertheless a growing understandingthat language evolved primarily as ameans of communicating about the non-present, with the property of displace-ment providing ‘the road into language’[3]. Neurophysiology shows that the ento-rhinal-hippocampal complex even in therat can compute information about eventsin space and time with language-like gen-erativity, and includes incidental episodicinformation [4]. These internal events arenot language, and the human system hascertainly gained in complexity [5], but lan-guage may well have evolved primarily asa means to express them.

Theory of mind is also critical to the com-municative aspect of language. As Chom-sky once wrote, ‘Communication relies onlargely shared cognoscitive [sic] powers. . . ’ [6]. Gilles Fauconnier elaborates:

When we engage in any languageactivity, we draw unconsciously onvast cognitive and cultural resources,call up models and frames, set upmultiple connections, coordinate largearrays of information, and engage increative mappings, transfers, andelaborations [7].

Such resources go well beyond ‘I-lan-guage’, and it is difficult to believe theysurfaced in a single step late in the evo-lution of our species.

On Erasers and SticksWhat of the distinction between a white-board eraser and a white board eraser?Analogous distinctions do appear to arisein nonhuman animals. New Caledoniancrows can use a short stick to extract alonger stick from a barred cage, and usethis longer stick to extract food from along tube, and do so immediately andspontaneously. Therefore, they might besaid to know the distinction between along-stick extractor and a long stick

Trend

extractor [8]. Hierarchical structuresmay well be part of the cognitive appara-tus by which humans and animals parsethe world, and can exist in the absence ofany means to communicate them. Seek-ing analogs of human cognition does notimply digging into the past and movingforwards, ‘hoping to land in the rightspot’, as Everaert et al. suggest [2]; ratherit takes what we know about languageand human cognition, and seeks likelyprecursors.

Everaert et al. complain that I do not giveevidence as to how nonhuman animalsthink, and ‘why . . . “expressive lan-guage” would take millions of years toappear’ [2]. The extraction techniquesused by the crows provide an exampleof a window into the nonhuman mind.There are others, a few of which are men-tioned in my article. Moreover, the actionsof the crows might themselves be con-sidered communicative, perhaps to becopied by a watching bird. Expressionmight then begin with overt behavior,gradually becoming decoupled and con-ventionalized, especially in social specieswhere survival depends on effective com-munication transcending time and space.Thus, the emergence of expressive com-munication can itself be considered agradual process, as in the gestural theoryof language origins I outlined [1].

On EvolutionAs for evolution, even Tattersall seemsbemused by the idea that the mind couldhave changed so dramatically in such ashort window of time. In a book cited byEveraert et al., he wrote:

‘It is a qualitative leap in cognitive stateunparalleled in history. Indeed . . .the only reason we have for believingthat such a leap could ever have beenmade, is that it was made. And itseems to have been made well afterthe acquisition by our species of itsdistinctive modern form’ [9].

This sounds more like fiat than fact. WasHomo sapiens was really so distinctive?Evidence of mating between early sapiens

s in Cognitive Sciences, August 2017, Vol. 21, No. 8 571

and both Neanderthals and Denisovanseven raises the question of whether sapi-ens was a species distinct from theseother large-brained hominins, uniquelycapable of language [10]. Perspectivesare indeed changing. As two recent com-mentators put it:

‘myth of a “modern human revolution”is now rejected by archaeologists,although it lingers on in linguisticcircles, as illustrated, for example, byChomsky (2010). The myth dissolvesas soon as one considers the archae-ological record of the whole Old World,and especially of Africa, where a grad-ual, piece-meal process of culturalaccretion took hundreds of thousandsof years’ [11]. Rather than appearingde novo in Homo sapiens, languagewas more likely part of that ‘gradual,piece-meal process of culturalaccretion’.

The Modern Synthesis does offer a moreflexible account of evolution than simpleDarwinian theory allows, but the idea thatit can really account for the one-stepemergence of a faculty as complex aslanguage has been seriously questioned(e.g., [12]). However, I would be glad tohear of any case of a system as complexas the digestive system or the immunesystem, which Chomsky claimed to becomparable to language, emerging denovo in a single step.

Nevertheless, such a case would stillsmack of special pleading. Everaert andcolleagues appear secure in the belief thatthe nonhuman mind is unstructured.However, absence of evidence is notevidence of absence, and there are nowsigns from behavior or neurophysiologythat the animal mind is indeed structured.Moreover, our common ancestry with ournearest nonhuman great-ape relativesdates back some 6 million years. That isplenty of time for a gradual increase incognitivecomplexity, and includes the emergence,in Homo, of features such as bipedalismand a threefold increase in brain size thatappear well adapted to more highly struc-tured thought, action and communication,

572 Trends in Cognitive Sciences, August 2017, Vol. 21, N

leading, as Darwin put it, to differences ofdegree, but not of kind.

1School of Psychology, University of Auckland, Private

Bag 92019, Auckland 1142, New Zealand

*Correspondence:

m.corballis@auckland.ac.nz (M.C. Corballis).

http://dx.doi.org/10.1016/j.tics.2017.05.006

References1. Corballis, M.C. (2017) Language evolution: a changing

perspective. Trends Cogn. Sci. 21, 229–236

2. Everaert, M.B.H. et al. (2017) What is language and howcould it have evolved? Trends Cogn. Sci. this issue

3. Bickerton, D. (2014) More than Nature Needs: Language,Mind and Evolution, Harvard University Press

4. Moser, M.B. et al. (2015) Place cells, grid cells, and mem-ory. Cold Spring Harb. Perspect. Biol. 7, a021808

5. Nadasdya, Z. et al. (2017) Context-dependent spatiallyperiodic activity in the human entorhinal cortex.Proc. Natl. Acad. Sci. U. S. A. 114, E3516–E3525

6. Chomsky, N. (2007) Biolinguistic explorations: design,development, evolution. Int. Philos. Stud. 15, 1–21

7. Fauconnier, G. (2003) Cognitive linguistics. In Encyclope-dia of Cognitive Science (vol. 1) (Nadel, L., ed.), In pp. 539–543, Nature Publishing Group

8. Taylor, A.H. et al. (2007) Spontaneous metatool use byNew Caledonian crows. Curr. Biol. 17, 1504–1507

9. Tattersall, I. (2012) Masters of the Planet, PalgraveMacmillan

10. Johansson, S. (2013) The talking Neandertals: what dofossils, genetics, and archeology say? Biolinguistics 7,35–74

11. Dediu, D. and Levinson, S.C. (2013) On the antiquity oflanguage: the reinterpretation of Neandertal linguisticcapacities and its consequences. Front. Psychol. 4, 397

12. Christiansen, M.H. and Chater, N. (2008) Language asshaped by the brain. Behav. Brain Sci. 31, 489–558

SpotlightHow Does SocialNetwork PositionInfluence ProsocialBehavior?Oriel FeldmanHall1,*Consider the sheer number of people youinteract with on any given day. How doyou keep up with all the complex bonds,relationships, and hierarchies betweenpeople in your social community? Beingable to track the quality of your connec-tions to individuals within your socialgroup, and to subsequently track theirrelationships to others, reflects the topol-ogies of our social networks [1].

o. 8

A new study by Parkinson and colleagues[12] is one of the first to shed light on howthe brain encodes the dense structure ofits social network. Whereas previouswork focused primarily on indexing pop-ularity [2], this study used an innovativeapproach for measuring multiple addi-tional dimensions in a large sample, effec-tively capturing a more holistic picture ofhow humans map their social groups.These dimensions included: (i) the num-ber of connections between individuals(social distance); (ii) how well connectedindividuals are to other well-connectedindividuals (a robust measure of socialstatus, known as eigenvector centrality);and (iii) the degree to which individuals areconnected to other unconnected individ-uals (to pass on relevant social informa-tion, known as brokerage). A subset ofparticipants then partook in an imagingstudy, in which they passively viewed vid-eos of different members of their groupwho varied in social distance, social sta-tus, and brokerage.

Merging representational similarity analy-sis (RSA) with a dense topographicalsocial network structure allowed Parkin-son and colleagues [12] to not only locatewhere BOLD activity was occurring in thebrain, but to also characterize how theseneuronal populations were encodingsocial information, a more effectivemethod for inferring mental states. Thisis because RSA capitalizes on fine-grained spatial pattern differences, ratherthan on the overall activation of specificbrain regions. What the authors observedwas that a distributed network of separateregions encodes each dimension of thesocial network: social distance wasindexed by the inferior parietal lobule(IPL), social status by the medial prefron-tal cortex (mPFC), and brokerage bysuperior temporal cortex (STC). Thesefindings highlight that social informationused in our everyday interactions isencoded in brain regions known to trackdomain-general spatial navigation andpsychological distance (IPL), and