下载掌阅APP，畅读海量书库
立即打开
How language emerged uniquely in our species is a central issue toward understanding the basis of our humanity. Giving the issue a name, such as‘language organ’, and attributing it vaguely to some genetic mutation is not productive. Rather, the issue should be examined from the perspective of evolution theory. Here I suggest that the first phase transition, the trajectory toward language, started when we first assumed bipedal posture. This first phase transition occurred with the Australopithecine over 3,000,000 years ago.The second phase transition occurred with the emergence of our genus Homo over 2,000,000 years ago, when our ancestors exhibited symbolic behavior by producing and maintaining a variety of stone tools. The third occurred when primary communication changed from gestures and prosodies to sequences of syllables made up of vowels and consonants, which provided an efficient signal space; this occurred some 20,000 years ago with the emergence of our species Homo sapiens. The fourth phase transition was the invention of writing some 6,000 years ago, with numerous far reaching consequences.
Language evolution
When I saw the above cartoon in New Yorker magazine some years ago, the issue of language evolution immediately came to mind. Many linguists were then mesmerized by the belief that languages are homogeneous systems that can be explained in simple, algebra-like rules. However, the more language data one looked at, the more the rules became complicated, abstract, and often implausible. Terms like ‘language organ’,‘language instinct’, and ‘language bioprogram’ were used, much as ‘step two’depicted in the cartoon.
Instead of exploring in depth the abilities which enable infants to master language so effortlessly, terms like ‘language acquisition device’ and ‘universal grammar’ were advocated, as if giving the problem a rhetoric can substitute for solving it. When the late Roger Brown pioneered empirical methods with corpora and experiments to study how children actually construct their language from the sparse data in their environment, his project at Harvard was irresponsibly and rudely criticized as “the biggest waste of time in the history of science”. 1
The difficulty with the ‘language organ’ approach was remarked upon in a recent book by the psychologist Corballis:“The idea that the basis for language emerged in a single step in a single individual is remarkable, and smacks of the miraculous.” (2011: 24) Several years earlier, the neuroscientist Ramachandran made the same point about this approach; he further noted a similarity of this approach to the dilemma that famously divided the two discoverers of the theory of evolution, i.e., Charles Darwin and Alfred Russell Wallace. As Ramachandran(2004) put it:
But how could an extraordinarily complex mechanism like language with so many interlocking components have evolved through the blind workings of chance — through natural selection. ... Alfred Russell Wallace said the mechanism is so complicated it couldn’t have evolved through natural selection at all and must have resulted from divine intervention. ... Chomsky said something quite similar, although he didn’t invoke God. ... He almost says it’s a miracle. Unfortunately, neither Wallace’s nor Chomsky’s theory can be tested.
The critical ingredient missing in the ‘language organ’ approach is the dimension of evolutionary time . Going back to Corballis, the salient observation is that evolution is gradual and takes time on a grand scale:
... language can be understood to have evolved gradually, rather than having emerged suddenly in some comparatively recent individual on the family tree, called Prometheus. (2011: 34)
Prometheus existed only in Greek mythology, and unfortunately is not here to make ‘step two’ in the cartoon more explicit for us. Several decades earlier, the linguist Hockett also linked the ‘language organ’ approach to Greek mythology.He titled his 1978 discussion “In Search of Jove’s Brow”, recalling the myth of Minerva suddenly emerging from Jove. Following in Hockett’s tracks several years later, I contrasted such a Minerva Theory with what I called a Mosaic Theory, which I characterized in [ 1982] 1991 as follows:
[Language] evolved in a mosaic fashion, with the emergence of semantics, phonology, morphology and syntax all at different times and according to different schedules ... language is regarded as a kind of‘interface’ among a variety of more basic abilities . These abilities underlie nonlinguistic processes as well, and involve the perception of patterns in the frequency and temporal domains, the coding and storage of events and objects at different levels of memory, the manipulation of various hierarchical mental structures.
Continuing the metaphor in Greek mythology in 1984, I specifically compared the ‘language organ’ to the gods that would suddenly spring out of nowhere in classical Greek tragedies to resolve impossible dilemmas. The literary device is called Deus ex machina in Latin, where the machine in the phrase refers to some mechanical contraption to help bring a god onto the stage. Although the emergence of language is a very hard problem, resorting to miracles or to such literary devices is no help. Rather, we need to bring together all pieces of relevant information that may eventually contribute to its solution. To my mind, the following thoughts from Ramachandran are consistent with the Mosaic theory and point in the right direction:
Early hominids were very good at tool use, especially what is known as the sub-assembly technique ... There is a close operational analogy between this function and the embedding of noun clauses within longer sentences. So perhaps what originally evolved for tool use in the hand area is now exapted and assimilated in the Broca’s area to be used in aspects of syntax such as hierarchical embedding.
Each of these effects is a small bias, but acting in conjunction they may have paved the way for the emergence of sophisticated language. This is very different from Steve Pinker’s idea that language is a specific adaptation which evolved step by step for the sole purpose of communication. I suggest,instead, that it is the fortuitous synergistic combination of a number of mechanisms which evolved for other purposes initially that later became assimilated into the mechanism that we call language.
I have highlighted the word ‘exapted’ above because of its importance for the theme of the discussion here. This new term was introduced by the biologists Gould and Vrba in 1982 to refer to new functions served by old structures, which is something often observed in the life sciences. The concept is an elaboration of what the geneticist Jacob advocated in his famous 1977 paper, Evolution and Tinkering. Nature hardly ever introduces organs de novo , but typically makes use of old structures to serve new functions. As Jacob put it:
Living organisms are historical structures, literally creations of history. They represent not a perfect product of engineering, but a patchwork of odd sets pieced together when and where opportunities arose.
In the context of linguistics, exaptation or tinkering is a concept that was known much earlier. In his brilliant book on language, 1921, Sapir used the word‘overlaid’ to refer to the same phenomenon:
Physiologically, speech is an overlaid function, or, to be more precise,a group of overlaid functions. It gets what service it can out of organs and functions, nervous and muscular, that have come into being and are maintained for very different ends than its own. (p.9)
For producing speech, the situation is easy to see. We cannot speak without an outgoing stream of air provided by the more fundamental function of respiration. However, we should note that many respiratory refinements have additionally evolved over past millennia for much more precise control of this outgoing air stream so that it can synchronize with the varying phrases and stresses in the speech. Even more importantly, syllables are overlaid on the fundamental function of mastication. The rhythmic movements of opening and closing the jaw when we chew food, and the minute movements of the tongue as it transports the food to be processed by different dental structures,provided the sensori-motor skills exapted for producing our vowels, consonants,and syllables.
Going back to Ramachandran’s observation quoted above, what he called‘the sub-assembly technique’ lies at the heart of hierarchical structures, which we find ubiquitously in the cognitive behaviors of higher animals. The utility of hierarchic patterns and processes was charmingly explained by Herbert Simon in his classic paper of 1962 on the architecture of complexity; he gave a parable of two watch-makers, the one using sub-assembly winning over the one who does not. The idea of hierarchical structures has been expressed by a variety of notations, including nested parentheses linearly as well as tree diagrams.
In linguistics, the idea of sub-assembly or hierarchic processes has long been used when sentences were parsed in language textbooks. In mid-20 th century,the idea was discussed in depth by Wells (1947) for syntax and by Pike & Pike(1947) for phonology, under the label ‘immediate constituents’ or IC; the idea was further extended to ‘discontinuous constituents’ when a constituent is interrupted,such as the underlined morphemes of the auxiliary separated by the main verb in ‘he was dream ing .’ The terminology of ‘immediate constituents’ later changed to ‘constituent structure’ and ‘phrase structure’, though the basic idea remains largely intact. Indeed, it is the iterative and recurrent use of linguistic materials that underlies the open ended nature of language, both in phonology and in syntax.
Another observation Sapir made in his book on the age of language is also worth quoting in the present context:
The universality and the diversity of speech lead to a significant inference. We are forced to believe that language is an immensely ancient heritage of the human race, whether or not all forms of speech are the historical outgrowth of a single pristine form . It is doubtful if any other cultural asset of man, be it the art of drilling for fire or of chipping stone,may lay claim to a greater age. I am inclined to believe that it antedated even the lowliest developments of material culture, that these developments,in fact, were not strictly possible until language, the tool of significant expression, had itself taken shape. (p.23)
Let us first comment briefly on his remark on whether all languages come from a single source — ‘ a single pristine form ’, which is the hypothesis of monogenesis. Reasoning from probability theory and basing on demographic data available for prehistoric times, Freedman and Wang (1996) gave various reasons to argue for the opposite hypothesis of polygenesis, namely that the languages of the world descend from many sources. Polygenesis admittedly makes for a messier evolutionary scenario; but it is more realistic.
Sapir’s main conjecture in the quote was that language emerged earlier than ‘ the lowliest form of material culture .’ The basis for this conjecture is that language was the mental tool that our ancestors used to build a rudimentary culture some millions of years ago that was beyond the capacity of all other species at that time. As the diagram below suggests, the launching pad for the trajectory toward language and culture was the transition to erect posture and bipedal movement. This took place with a pre- Homo genus called Australopithecus over 3 million years ago.
Over 2 million years ago, the genus Homo appeared, which systematically made the first stone tools. Sapir’s conjecture was that some form of primitive language must have been in place in their minds which enabled them to achieve this rudimentary culture. As discussed above, Ramachandran similarly linked tool use to the evolution of language. Whereas other species evolved primarily by biological evolution , through transmission of genes from generation to generation, the evolution of Homo was driven in addition by cultural evolution ,through much more complex modes of transmission of teaching, learning, and numerous forms of social interaction. It is this powerful combination of dual evolution , the biological plus the cultural that has created the unique endowments that characterize our species, Homo sapiens .
Over the millions of years, many species have become extinct. Our closest relatives today are the chimpanzees, whose lineage diverged from ours some 6 million years ago. Sapir, writing in 1921, was not in a position to go the next step and compare our culture with that of the chimpanzee, since most of what we know about the chimpanzee was learned in recent decades, as de Waal recently summarized (2005). Call and Tomasello (2008) is a more advanced discussion of the chimpanzee’s mental world; Herrmanm et al (2007) report on cognitive differences among the chimpanzee, the orang utan, and the human child (Senghas Kita and Ozyurek 2004).
Nor could Sapir know much about the earliest material culture of our ancestors; he passed away in 1939, just about when Louis Leakey was discovering the stone tools at Olduvai Gorge in Tanzania. Now we have much more data for such comparisons, including in addition better knowledge of the biological equipment that makes language possible.
The remainder of this essay will attempt an update on these issues, almost a century after Sapir’s conjecture. The term ‘phase transition’ in the title of this essay is borrowed from physics, where it refers to qualitative changes brought about by nonlinear development. A common example is the gradual addition of a common amount of heat to H 2 O, changing it successively from solid ice to liquid water and to gaseous steam. We will discuss four such phase transitions in the evolution of language.
The first phase transition took place over 3 million years ago. 1974 was a good year for paleoanthropology, for the discovery of one of the most famous set of fossils in human evolution in Ethiopia. The remarkably complete set was the remains of a young girl, known to science as Lucy, a member of a species named Australopithecus afarensis . The importance of these fossils for language has been discussed by Johanson, one of their discoverers (Johanson and Blake 2006).Detailed anatomical studies of Lucy and of many other fossils discovered later in nearby regions of Ethiopia have convinced scientists that these creatures walked upright.
Unlike all other primates, Lucy was bipedal, standing erect with her head balanced vertically upon her vertebral column. The other great apes, i.e.,chimpanzees, gorillas and orangutans, may walk bipedally occasionally, even carrying things in their arms over short distances sometimes; they may also use ‘knuckle walking’ complementing the two legs by leaning on an extended hand. Nonetheless, Lucy’s species was the first species in the primate order that assumed an erect stance much as humans do now.
This major ‘tinkering’ that evolution has done with our bodies has significantly restructured our anatomy, especially the head and neck areas. As a reminder of our quadrupedal past, the nerve pathway between the brain and the larynx has been greatly lengthened because it needs to descend below the aorta and then double back up again to control the laryngeal muscles. The larynx itself descends as well during early infancy, creating a vertical volume in the throat approximately perpendicular to the mouth. The resulting anatomy, resembling a bent tube, is shared by food and drink, as well as by the air we breathe and speak with. While such an arrangement allows us to speak a greater diversity of sounds, the negative side is that it makes us more vulnerable to choking, as noted early by Darwin:
The strange fact that every particle of food and drink which we swallow has to pass over the orifice of the trachea with some risk of falling into the lungs, notwithstanding the beautiful contrivance by which the glottis is closed. (1859: 191)
Below is a very informative diagram from Vallender et al (2008) which shows some of the vital differences between us and our primate relatives.
Comparing humans with the chimpanzee is particularly instructive. Note that while we are only some 1.2% different genetically, our brain is more than 4 times larger than the chimpanzee brain. Measurement of the many skulls of Australopithecus afarensis that have been unearthed since Lucy shows a volume similar to that of the other great apes. Such an explosive growth of the brain over these 3 million years — from a mean of 450 cc to 1350 cc - is remarkable in biological evolution.
Nonetheless, there is still much cognitive that we share with the chimpanzee due to our long common lineage. The chimpanzees also have socio-political structure living in small groups, and their cultures vary from region to region.They are the only primate other than man that has been shown to have self recognition. They use twigs to fish for termites. They use stones to crack nuts;mothers have even been observed to shape the hands of young ones in teaching them to do this. They coordinate with each other when trapping monkeys for food. While there is still quite a gap between our cognition and theirs, a good case can be made that we lie on the same continuum of development.
On the other hand, efforts to teach the other great apes language or language like systems have stopped at a relatively low ceiling. The most sustained studies so far are those of Kanzi the chimpanzee (Savage-Rumbaugh and Lewin 1994),and of Koko the gorilla (Patterson and Matevia 2001). Whether the medium of instruction is the American Sign Language, or plastic chips on a magnetic board,or lexigrams on a computer screen, the language attained by the ape students has been compared to human five-year-olds at best. Herrmann et al (2007) report interesting experiments comparing human infants, chimpanzees and orangutans,proposing that humans have a greater advantage in ‘cultural intelligence’. Call and Tomasello (2008) provide insightful discussions on the mental differences between us and the chimpanzee with respect to ‘theory of mind’, i.e., the ability to understand and/or empathize the thoughts and/or feelings of another individual.
An obvious consequence of bipedal posture is the freeing of the hands from the heavy task of pounding the earth during walking and running, allowing them to develop finer and finer skills. While the primate hand in general has achieved some flexible use from climbing trees and various simple chores, it was in the human lineage that the hand developed the greatest dexterity. This remarkable achievement reaches its peak in our times in musical performances,as exemplified in the rapid and precise movements of the pianist’s fingers as they fly across the keyboard. A similar kind of finger dexterity is displayed as modern teenagers play various video games on their cell phones.
Increasing demands on the hand select for more brain tissues for controlling its many parts; witness the large areas in the sensory and motor cortices devoted to the thumb and the fingers. Conversely, finer neural control from the brain enables more diverse activities for the hands to execute. In general, control and execution reinforce each other mutually, as is often the case of co-evolution of structure and function. This interaction between hand and brain, plus all the associated infrastructure of vision, kinesthesia, etc., must have been one of the major engines pushing the explosive growth of the brain.
Bipedal posture provided the foundation for many aspects of behavior that are uniquely human, including linguistic behavior but not exclusively so. The second phase transition upon which language was launched was the concept of symbolization — that one thing could represent something else. Deacon (1997) has aptly named us The Symbolic Species . It was the mental manipulation of geometric relations that led the genus Homo to initiate making a variety of stone tools, for pounding, for cutting, for drilling, and so on. These mark the ‘ lowliest developments of material culture ’ that Sapir spoke of, and the technique of sub assembly mentioned by Ramachandran, quoted earlier. Bipedal posture provided us with a biology with immense potential. Symbolization marked a new phase when cultural evolution began to replace biological evolution as the major engine of change.
The earliest stone tools were made in Africa over 2 million years ago.Shortly after that date, wave after wave of early Homo left Africa for many parts of Asia. For instance, a recent report in Science magazine describes a rare complete skull found in Dmanisi in the Republic of Georgia, dated to 1.8 million years ago (Lordkipanidze et al., 2013). The more famous representatives of our genus include: Peking Man, whose fossils were found in the outskirts of modern Beijing and whose dates center around 500,000 years ago; the Neanderthals that ranged over Europe and Western Asia, whose remains date to 600,000 through 30,000 years ago; and the diminutive Homo floresiensis of Indonesia, who may have survived until as recently as 12,000 years ago.
A recent overview of the origin of our species is Stringer (2012), who is an articulate advocate of the single origin version of the Out-of-African hypothesis.In contrast, many scholars prefer a multiregional hypothesis, which posits significant interbreeding of Homo sapiens with the earlier Homo species. There is as yet no consensus on the details of our evolutionary past, giving a consistent account that accommodates all the data from paleoanthropology, population genetics, and evolutionary linguistics. Relevant data are constantly emerging in the form of unearthed fossils or newly discovered genetic materials. There can be no doubt that impressive advances have been made in recent decades since the various disciplines have begun to work together integrating each other’s results toward a coherent consensus.
Taking clue from the modern behavior of the other primates, early Homo must have communicated with a combination of gestures and prosodies. Gestures require less shared information to communicate intent — extending the palm to beg, raising the fist to strike, baring the teeth to bite. It should not take much to learn when your neighbor bares his teeth that he means harm. Prosodies in the form of growls, grunts, hoots, screams, etc. to express pleasure, anger, and other emotions easily become conventionalized as signals, and typically may co-occur with gestures. (Baring the teeth is a lot more threatening when paired with a growl!) A famous case is the use of prosody for alarm calls in vervet monkeys, as reported by Seyfarth et al (1980).
The coupling of gestures with vocalization continues to this day. Our facial expressions change and we motion with our hands and shoulders even when our addressee is not within sight, such as when we are speaking on a phone. McNeill(2005) and Goldin-Meadow et al (2008) discuss gesture and thought from a modern perspective; Corballis (2011) connects gesture with vocalization; and Arbib (2013) integrates these concepts with the exciting discoveries of the mirror neuron system.
Gestures and prosodies persist in modern languages though in roles which are much reduced in importance; see Corballis (2002). They vary in form and content according to the cultural context. Movements of the head to show agreement or disagreement, of the hand for beckoning, protrusion of the tongue as a sign of respect or disgust, etc. all depend critically on social norms. Similarly,prosodies in the form of intonations and contrasting stress vary from language to language. On the whole, however, these two communicative devices are limited in their semantic range, and do not combine sequentially with each other easily.They function on continua in analog fashion rather than digitally as in segmental phonology.
The expressive power of language was greatly enhanced by the introduction of phonetic segments, i.e., consonants and vowels, as the basic building blocks of speech, organized around the syllable, stress group, and breath group according to the sub-assembly technique. Segments require much less energy to produce,when compared with bodily gestures or complete intonations, and take a great deal less time — in the order of tenths of a second. As a simple example, a phonology of 5 vowels and 20 consonants can generative 100 distinct syllables of the shape CV, and many more syllables of the shape CVC. When these syllables are combined they generate ample phonetic space for a vocabulary of many thousands of words.
The generative power is further enhanced at the level of syntax, where the words are combined, again making use of the sub-assembly technique(Schoenemann and Wang, 1996; Schoenemann, 2005; Vogt, 2005). Given the time efficiency of segmental phonology, a working memory with a chunk size of several seconds can therefore easily take in sentences of moderate length for multiple analyses. It is likely that the third phase transition occurred shortly after modern Homo sapiens appeared in Africa, and the new form of language played a critical role in enabling our ancestors to settle in all parts of the world. Recent research in genetics has greatly increased our understanding of the biological bases of these linguistic abilities; see Chow (2005), Lieberman (2013), Sia, Clem and Huganir (2013).
Writing was invented much more recently, the earliest samples dating back some 6,000 years ago in some civilizations. Daniels and Bright (1996) give a comprehensive survey of the history of various systems of writing; Wang and Tsai (2011) compared alphabetic writing, derived from the Phoenician script, with sinograms, invented in Ancient China.
Writing first emerged as an instrument of privileged groups, such as royalty or priests. Mass literacy among the common people is a very recent phenomenon,within the last century or so; see Olson and Torrance (2009). Because of its relative recency in human evolution, our brain has not yet co-adapted with writing as it has co-adapted with speech; see Wang (2012). As a consequence,the acquisition of written language does not come effortlessly to all children as spoken language does. Instead, there is a minority of children in all cultures who have difficulty in learning to read and write; Dehaene (2009) is a general discussion of such issues from the viewpoint of cognitive neuroscience.
Whereas spoken language evaporates the instant it is produced, written language allows for information to be transmitted across space as well as to accumulate across time. The sharing and accumulation of information, made possible by writing, has grown exponentially, mostly in the form of books during past centuries. Now this immense amount of information, and still growing ever faster, is in the form of electronic files, accessible on the Internet and other resources.
Before the advent of language information about how to do a particular task could only be transmitted by actual demonstration by the teacher in the presence of the student. With the emergence of phonology, this information can be presented in spoken form. With the invention of writing, the information can be externalized as graphic patterns on paper or as polarizations on various magnetic or optic media. In all these cases, the information to be transmitted originates from the sender’s brain. The processes of externalization and transmission have certain constraints as well as possibilities of error.
It would be foolhardy to predict at this point what the fifth phase transition for language will be. But extrapolating from the line of reasoning given above,it is not impossible that the time will come when we can monitor the sender’s brain waves with enough sensitivity and fidelity, and transmit these directly to the brains of the receivers, without any intervening media. Science fiction writers have already depicted such scenarios, sometimes suggesting that information can be transplanted into the receivers’ brain by implanting chips containing integrated circuits. Fiction has been known to predict the future, so perhaps the fifth phase transition will embody some of these scenarios.
Let us now recapitulate the four phase transitions that led to language which we have reviewed above. [1] The first phase transition occurred when Australopithecus changed from quadrupedat 2 posture to a bipedal one, over 3 million years ago, causing a fundamental re-structuring of our anatomy. This freed the hands for forming gestures, and provided the mouth and throat with greater phonetic possibilities. [2] The second phase transition occurred with the emergence of our genus Homo , over 2 million years ago. The production and maintenance of a variety of stone tools give indirect evidence that these individuals had achieved symbolic behavior. [3] The third phase transition occurred when phonology emerged, with the appearance of our species Homo sapiens over 100,000 years ago, built on hierarchies of vowels, consonants,syllables and prosody. Information of much greater complexity can be transmitted within the narrow time window of working memory than is possible with gestures and prosody. [4] The fourth phase transition occurred with the invention of writing, over 5,000 years ago, when words were represented by graphic symbols.This overcame the limitations of time and space, and ushered in a much more powerful mode of thinking.
On this last point, it is useful to ponder a response from the physicist Richard Feynman to historian Charles Weiner, reported in Gleick (1993: 409). Clark(2008: xxv) retold it in this way:
Weiner, encountering with a historian’s glee a batch of Feynman’s original notes and sketches, remarked that the materials represented ‘a record of [Feynman’s] day-to-day work’. But instead of simply acknowledging this historic value, Feynman reacted with unexpected sharpness:
“I actually did the work on the paper,” he said.
“Well,” Weiner said, “the work was done in your head, but the record of it is still here.”
“No, it’s not a record, not really. It’s working. You have to work on paper and this is the paper. Okay?”
As I understand this famous exchange, Feynman was insisting on the point that writing it down is itself an important part of the creative process. Indeed,although language is our heads in its entirety, but the creation of a complex piece of work, whether it be in science or in literature, is not simply downloading something from mind to paper, or to a computer file. Rather the process is very much an interactive process going back and forth between mind and paper.
Here is an example to illustrate the point. A sentence with one degree of center embedding like “ The black cat his dog chased killed the big rat ”would normally present no problem for comprehension. On the other hand,a sentence with three degrees of center embedding like “ The black cat his dog the horse my uncle bought kicked chased killed the big rat ” 3 would not be intelligible without writing it down, even though its constituent sentences are all simple. The situation is not unlike not being able to multiply large numbers in our head, even when we are perfectly well versed in the rules of multiplication. Neither very complex sentences nor multiplication of large numbers would have been possible without writing. Thus writing does much more than record language for messages to be sent across space and preserved across time; rather it extends language by enabling it to form ever more complex and intricate messages.
1.Quote attributed to Chomsky by S. Pinker, Cognition 1998: 206. For detailed critiques of Chomsky’s work, see Postal 2009 and references therein.
2. Quadruped is to be distinguished from Tetrapod . The latter is a taxonomic unit in evolutionary biology which includes all vertebrates with quadrupedal ancestors, including mammals, reptiles, amphibians, and birds.
3.The five constituent sentences are: The black cat killed the big rate; his dog chased the black cat; the horse kicked the dog ; and my uncle bought the horse .
ARBIB, Michael A. 2013. How the Brain got Language: The mirror System Hypothesis . New York: Oxford University Press.
CALL, Josep and Michael Tomasello. 2008. Does the chimpanzee have a theory of mind? 30 years later. Trends in Cognitive Sciences 12: 187—192.
CHOW, King L. 2005. Speech and language — a human trait defined by molecular genetics. In Language Acquisition, Change and Emergence: Essays in Evolutionary Linguistics , ed. J. W.Minett and W.S.-Y. Wang, 21—45. Hong Kong: City University of Hong Kong Press.
CLARK, Andy. 2008. Supersizing the Mind: Embodiment, Action, and Cognitive Extension .Oxford: Oxford University Press.
CORBALLIS, Michael C. 2002. From Hand to Mouth: The Origins of Language . Princeton:Princeton University Press.
______. 2011. The Recursive Mind: The Origins of Human Language, Thought, and Civilization . Princeton: Princeton University Press.
DANIELS, Peter T. and William Bright, eds. 1996. The World’s Writing Systems . New York:Oxford University Press.
DARWIN, Charles. 1859. On the Origin of Species by Means of Natural Selection or The preservation of Favored Races in the Struggle for Life . London: John Murray.
De WAAL, Frans B. M. 2005. A century of getting to know the chimpanzee. Nature 437: 56—59.
DEACON, Terrence W. 1997. The Symbolic Species: the Co-evolution of Language and the Brain . New York: W.W. Norton.
DEHAENE, Stanislas. 2009. Reading in the Brain: The Science and Evolution of a Human Invention . New York: Viking.
FREEDMAN, D.A. and W.S-Y. Wang. 1996. Language polygenesis: A probabilistic model. Anthropological Science 104.2: 131—138.
GLEICK, James. 1993. Genius: The Life and Science of Richard Feynman . New York: Vintage.
GOLDIN-MEADOW, Susan, Wing Chee So, Asli Ozyurek and Carolyn Mylander. 2008. The natural order of events: How speakers of different languages represent events nonverbally. PNAS 105: 9163—9168.
GOULD, Stephen Jay and Elizabeth S. Vrba. 1982. Exaptation — a missing term in the science of form. Paleobiology 8(1): 4—15.
HERRMANN, E., J. Call, M.V. Hernández-Lloreda, B. Hare and M.Tomasello. 2007. Humans have evolved specialized skills of social cognition: The cultural intelligence hypothesis. Science 317: 1360—1366.
HOCKETT, C.F. 1978. In search of Jove’s brow. American Speech 53: 243—313.
JACOB, Frarnçois. 1977. Evolution and tinkering. Science 196: 1161—1166.
JOHANSON, Donald and Edgar Blake. 2006. From Lucy to Language . New York: Simon &Schuster.
LIEBERMAN, Philip. 2013. Synapses, Language, and Being Human. Science 342: 944—945.
LORDKIPANIDZE, David, Marcia S. Ponce de León, Ann Margvelashvili, Yoel Rak, G. Philip Rightmire, Abesalom Vekua, and Christoph P. E. Zollikofer. 2013. A complete skull from Dmanisi, Georgia, and the evolutionary biology of early homo. Science 342: 326—331.
McNEILL, David. 2005. Gesture and Thought . Chicago: University of Chicago Press.
MINETT, J.W. and W.S-Y. Wang, eds. 2005. Language Acquisition, Change and Emergence: Essays in Evolutionary Linguistics . Hong Kong: City University of Hong Kong Press.
OLSON, David and N. Torrance, eds. 2009. Cambridge Handbook on Literacy . Cambridge,N.Y.: Cambridge University Press.
PATTERSON, F. G. P. and M. L. Matevia. 2001. Twenty-seven Years of Project Koko and Michael. In All Apes Great and Small . Vol. 1 African Apes , ed. B.M.F. Galdikas, N.E. Briggs,L.K. Sheeran, G.L. Shapiro and J. Goodall, 165—76. New York: Kluwer Academic Publishers.
PIKE, Kenneth L. and Eunice Victoria Pike. 1947. Immediate Constituents of Mazateco Syllables. International Journal of American Linguistics 13: 78—91.
PINKER, Steven. 1998. Obituary: Roger Brown. Cognition 66: 199—213.
POSTAL, Paul M. 2004. Skeptical Linguistic Essays . New York: Oxford University Press.
______. 2009. The incoherence of Chomsky’s ‘Biolinguistic’ ontology. Biolinguistics 3: 104—123.
RAMACHANDRAN, V.S. 2004. A Brief Tour of Human Consciousness: From Impostor Poodles to Purple Numbers . New York: Pi Press.
SAPIR, Edward. 1921. Language: An Introduction to the Study of Speech . New York: Harcourt.
SAVAGE-RUMBAUGH, Sue and Roger Lewin. 1994. Kanzi: The Ape at the Brink of the Human Mind . New York: John Wiley & Sons.
SCHOENEMANN, P. Thomas. 2005. Conceptual complexity and the brain: Understanding language origins. In Language acquisition, change and emergence: Essays in evolutionary linguistics , ed. J. W. Minett and W.S.Y. Wang, 47—94. Hong Kong: City University of Hong Kong Press.
SCHOENEMANN, P.T. and W. S-Y. Wang. 1996. Evolutionary principles and the emergence of syntax- Commentary on Müller: Innateness, autonomy, universality. Behavioral and Brain Sciences 19(4): 646—647.
SENGHAS, Ann, Sotaro Kita and Aslı Ozyurek. 2004. Children creating core properties of language: Evidence from an emerging sign language in Nicaragua. Science 305: 1779—1782.
SEYFARTH, R. M., D. L. Cheney and P. Marler. 1980. Monkey responses to three different alarm calls: Evidence of predator classification and semantic communication. Science 210: 801—803.
SIA, G. M., R. L. Clem and R. L. Huganir. 2013. The Human language-associated gene SRPX2 regulates synapse formation and vocalization in mice. Science 342: 987—991.
SIMON, Herbert. 1962. The architecture of complexity. Proc. Amer. Philos. Soc . 106: 467—482.
STRINGER, Chris B. 2012. The Origin of Our Species . London: Penguin.
V ALLENDER, Eric J., Nitzan Mekel-Bobrov and Bruce T. Lahn. 2008. Genetic basis of human brain evolution. Trends in Neurosciences 31(12): 637—644.
VOGT, Paul. 2005. On the acquisition and evolution of compositional languages: Sparse input and the productive creativity of children. Adaptive Behavior 13: 325—46.
WANG, W.S-Y. [1982]1991. Explorations in Language Evolution . Osmania papers in linguistics. Supplement; v. 8 Hyderabad: Osmania University Press. Reprint, Taipei: Pyramid Press, in Explorations in Language , 105—131.
______. 1984. Organum ex machina. Behavioral and Brain Sciences 7(2): 210—1.
______. 2012. Language learning and the brain: An evolutionary perspective. In Breaking Down the Barriers: Interdisciplinary Studies in Chinese Linguistics and Beyond , ed. G. Cao, H. Chappell,R. Djamouri and T. Wiebusch, 21—48. Taipei: Institute of Linguistics.
WANG, William S-Y. and Yaching Tsai. 2011. The alphabet and the sinogram: Setting the stage for a look across orthographies. In Dyslexia Across Languages: Orthography and the Brain Gene-Behavior Link , ed. P. McCardle, J.R. Lee, B. Miller and O. Tzeng, 1—16. Baltimore,Md.: Paul H. Brookes Pub. Co.
WELLS, Rulon S. 1947. Immediate constituents. Language 23: 81—117.
要想理解人之所以为人的基础何在，那么探索语言如何独一无二地在人类涌现是个核心议题。把这个议题取个像「语言器官」（language organ）一类的名字，并把它含糊地归因于某种基因突变，是无济于事的作法。这个议题理应从演化论的观点加以检视。在此我主张，语言涌现的轨迹，始于我们首次采取双脚直立的姿势，这比科学分类上我们「人属」（ Homo ）的出现还要早。第二个相变出现在当我们的祖先制造各类石器工具展现了象征行为时。促成语言涌现的第三个相变，出现在当口语沟通从手势和韵律的模式转变为主要依靠元音、辅音构成音节串时，这种转变为人类提供了有效的信号空间。第四个相变则是文字的发明，也造成了若干深远的影响。
语言演化