October 5, 2010
Three stages in the evolution of language
A debate has been raging for years among linguists as to whether the development of language was gradual and early or sudden and more recent. In this section of my book, Towards a Democracy of Consciousness, I argue for three stages of language evolution: (1) mimetic language beginning as long as four million years ago; (2) protolanguage emerging around 300,000 years ago (which would also have been spoken by the Neanderthals); and (3) modern language which may have begun to emerge about 100,000 years ago but probably only achieved fully modern syntax around the time of the Upper Paleolithic revolution around 40,000 years ago.
The co-evolution of language and the pfc
Imagine the world of our hominid ancestors over the four million years between Ardi and the emergence of homo sapiens. As we’ve discussed, it was a mimetic, highly social world, where increasingly complex group dynamics were developing. Communication in this world was probably a combination of touching, gestures, facial expressions, and complex vocalizations, including the many kinds of grunts, growls, shrieks and laughs that we still make to this day. Terrence Deacon believes that it was in this world that “the first use of symbolic reference by some distant ancestors changed how natural selection processes have affected hominid brain evolution ever since.” He argues that “the remarkable expansion of the brain that took place in human evolution, and indirectly produced prefrontal expansion, was not the cause of symbolic language but a consequence of it.”
The findings from Kuhl may help to explain how prefrontal expansion could have been a consequence of symbolic language. If we apply what we learned about how a patterning instinct leads the brain to shape itself based on the patterns it perceives, then we can imagine how a pre-human growing up in mimetic society would hear, see and feel the complex communication going on around him, and how his pfc would shape itself accordingly. Those infants whose pfcs were able to make the best connections would be more successful at realizing how the complex mélange of grunts, rhythms, gestures and expressions they were hearing and seeing patterned themselves into meaning. As they grew up, they would be better integrated within their community and, as such, tend to be healthier and more attractive as mates, passing on their genes for enhanced pfc connectivity to the next generation. It was no longer the biggest, fastest or strongest pre-humans that were most successful, but the ones with the most enhanced pfcs. In Deacon’s words, “symbol use selected for greater prefrontalization” in an ever-increasing cycle, whereby “each assimilated change enabled even more complex symbol systems to be acquired and used, and in turn selected for greater prefrontalization, and so on.”
Deacon’s view of this positive feedback loop between language and evolution is shared by others, including linguist Nicholas Evans who describes it as “a coevolutionary intertwining of biological evolution, in the form of increased neurological capacity to handle language, and cultural evolution, in the form of increased complexity in the language(s) used by early hominids. Both evolutionary tracks thus urge each other on by positive feedback, as upgraded neurological capacity allows more complex and diversified language systems to evolve, which in turn select for more sophisticated neurological platforms.” While this view represents some of the most advanced thinking in the field, it’s interesting to see that it has a solid pedigree – as far back, in fact, as Charles Darwin himself, who wrote in 1871:
If it be maintained that certain powers, such as self-consciousness, abstraction etc., are peculiar to man, it may well be that these are the incidental results of other highly advanced intellectual faculties, and these again are mainly the result of the continued use of a highly developed language.
Perhaps now, armed with this new perspective on our patterning instinct and the coevolution of language and the pfc, we are finally equipped to tackle the conundrum of when and how language actually evolved. Surely it was early and gradual after all, if Deacon and Evans are to be believed? It’s difficult to conceive how it could be anything else. But what, then, accounts for the Upper Paleolithic Great Leap Forward? How could we have had language for millions of years and not produced anything else with symbolic qualities until forty thousand years ago?
Three stages of language evolution
There is, in fact, a possible resolution to this conundrum that permits both the “gradual and early” and the “sudden and recent” camps to both be right. This approach views language as evolving in different stages, with major transitions occurring between each state, and was first proposed as a solution by linguist Ray Jackendoff in a paper entitled “Possible stages in the evolution of the language capacity.” Jackendoff was well aware that his proposal could “help defuse a long-running dispute,” and he pointed out that if this approach to language evolution becomes widely accepted, it will no longer be meaningful to ask whether one or another hominid “had language,” but rather “what elements of a language capacity” did a particular hominid have.
Jackendoff proposed nine different stages of language development, but for the sake of simplicity, I’d like to suggest three clearly demarcated stages. Additionally, I think these three stages can be closely correlated to the different stages of tool technology found in the archaeological record, so an approximate timeframe can also be applied to each stage. Importantly, the last of these stages, the transition to modern language, would be contemporaneous with the Upper Paleolithic revolution, and would therefore solve the conundrum posed by the “sudden and recent” camp. If we consider the analogy of language as a “net of symbols,” then we can visualize each stage as a different kind of net: the first stage may be visualized as a small net that you might use to catch a single fish in a pond; the second stage would be analogous to a series of those small nets tied together; and the third stage could be seen as the vast kind of net that a modern trawler uses to catch fish on an industrial scale. I’ll describe each stage in turn.
Stage 1: Mimetic language. This stage may have begun as early as Ardi, over four million years ago, and continued until slightly before the advent of modern homo sapiens, around 200,000-300,000 years ago. It would have been concurrent with what Merlin Donald calls the mimetic stage of human development, as discussed in Chapter 1. It would have involved single words that began to be used in different contexts, thus differentiating them from the vervet calls discussed earlier which only have meaning in a specific context. Examples of these words could be shhh, yes, no or hot. Jackendoff gives the example of a little child first learning language, who has learned to say the single word kitty “to draw attention to a cat, to inquire about the whereabouts of the cat, to summon the cat, to remark that something resembles a cat, and so forth.” If you were to imagine a campfire a million years ago, a hominid may have pointed to a stone next to the fire and said “hot!” and then might later have caused his friends to laugh by using the same word to describe how he felt after running on a sunny day.
The correlative level of technology would have been the Oldowan and Acheulean stone tools that, as described in Chapter 2, changed very little over millions of years. Interestingly, a recent study employed brain scanning technology to analyze what parts of the brain people use when they make these kinds of stone tools. Oldowan tool-making showed no pfc activity at all, while Acheulean tools required some limited use of the pfc, activating a particular area used for “the coordination of ongoing, hierarchically organized action sequences.”
Stage 2: Protolanguage. This stage, which has also been proposed by linguist Derek Bickerton, may have gradually emerged around 300,000 years ago (the period when Aiello and Dunbar believe that language “crossed the Rubicon”) and remained predominant until the timeframe suggested by Noble & Davidson for the emergence of modern language, roughly 70,000-100,000 years ago. It would have involved chains of words linked together in a simple sentence, but without modern syntax. If we go back to our Stone Age campfire, imagine that the fire’s gone out, but an early human wants to tell his friends that the stones from the fire are still hot. He might point to the area and say “stone hot fire” or alternatively “fire hot stone” or even “hot fire stone.” The breakthrough from mimetic language is that different concepts are now being placed together to create a far more valuable emergent meaning, but the words are still chained together without the magic weave of syntax.
Interestingly, it was around 300,000 years ago that new advances were being made in stone tool technology, leaving behind the old Acheulean stagnation. As described by anthropologist Stanley Ambrose, “regional stylistic and technological variants are clearly identifiable, suggesting the emergence of true cultural traditions and culture areas.” The new techniques, known as Levallois technology from the place in France where they were first discovered, represent according to Ambrose “an order-of-magnitude increase in technological complexity that may be analogous to the difference between primate vocalizations and human speech.” Ambrose believes that this type of “composite tool manufacture” requires the kind of complex problem solving, planning and coordination that is mediated by the pfc, and may even have “influenced the evolution of the frontal lobe.”
Stage 3: Modern language. This stage may have begun to emerge around 100,000 years ago, but possibly only achieved the magical weave of full syntax around the time of the Upper Paleolithic revolution, about 40,000 years ago. By this time, our fully human ancestor could have told his friend: “I put the stone that you gave me in the fire and now it’s hot,” with full syntax and recursion. The correlative level of technology would be the sophisticated tools associated with the Upper Paleolithic revolution, including grinding and pounding tools, spear throwers, bows, nets and boomerangs. The same brains that could handle syntax and recursion could also handle the complex planning and hierarchy of the activities required to conceptualize and make these tools.
But as we’ve seen, the Great Leap Forward involved more than sophisticated tools. It also delivered the first evidence of human behavior that’s not just purely functional but also has ritual or symbolic significance. For the first time, humans are creating art, consistently decorating their bodies, constructing musical instruments and ritual artifacts. I suggest that these innovations in the material realm are correlated to one particular aspect of language that may also have emerged at this time: the use of metaphor.
 Deacon, op. cit., 321-2.
 Ibid., 340.
 Evans 2003, op. cit.
 Darwin (1871). The descent of man, and selection in relation to sex. Cited by Bickerton (2009) op. cit., 5.
 Jackendoff, R. (1999). “Possible stages in the evolution of the language capacity.” Trends in Cognitive Sciences, 3(7:July 1999), 272-279.
 Stout, D., Toth, N., Schick, K., and Chaminade, T. (2008). “Neural correlates of Early Stone Age toolmaking: technology, language and cognition in human evolution.” Phil. Trans. R. Soc. Lond. B, 363, 1939-1949.
 Bickerton, D. (1990) Language and Species, Chicago: University of Chicago Press.
 Ambrose, S. H. (2001). “Paleolithic Technology and Human Evolution.” Science, 291(2 March 2001), 1748-1753.