The Evolution of Language

evolution-languageWhat is the link between language and learning? While exploring these issues, fundamental questions arise that have far-reaching effects for the classroom and beyond.

The most fundamental mechanism by which humans share information is language, but does language fall into the category of biologically evolved function or cultural invention? If it is, at least in part, an evolved function, how did language evolve, and what are the mechanisms of the mind that depend upon it? How language came into being has intrigued many great minds, including Charles Darwin, but it is also a question that we can all understand and ponder. In the course of thinking about language and its evolution, we inevitably introspect and examine the very process of thought itself. We grit our teeth in frustration with what the past hides from us. But ultimately, we may experience personal epiphanies about the workings of the mind, triggered by ideas conveyed to us through the language upon which we are reflecting. There is a great deal to read on this topic, and much of it is worthwhile. Here we shall only attempt to whet the appetite.

Evolution and Information

In the scientifically literate world, and even in many parts of the United States, evolution has become mundane fact. Wander through any museum of natural history and view the skeletons of animals from both past and present, and this central theme around which all biology is organized will seem obvious. We recognize the structure of our own hand and pelvis in the articulated forelimb and vestigial pelvis of a whale skeleton. That we share an ancestor with these behemoths of the sea seems clear.

The question arises not whether evolution occurred, but why were people so slow to grasp it? Is evolution like the seemingly unsolvable puzzle that becomes trivial once the answer is revealed? Are we just smarter than people were 150 years ago? And if evolution is obvious, why do we make such a big deal of Darwin? Evolution and the relatedness of animals are in fact very old ideas and common to many cultures, presumably because they are obvious. What is not obvious, however, are the mechanisms by which evolution operates.

The theory of natural selection, Darwin’s great idea, was important because it resolved aspects of evolution that run counter to our intuitive sense about information: information does not arise spontaneously, but is transmitted from a source to a receiver. In a seminal paper in 1948, Claude Shannon formulated these ideas into a precise mathematical theory, known as “information theory”, that allowed for exact quantities of information being transmitted to be calculated (information theory is important in many subfields of neuroscience – for background see Lucents Information Theory Web Page. But in the case of evolution, the source of the information that has led to increasingly complex structures in some organisms is unclear. Darwin’s insight was that the information source is the environment. Random mutations in our DNA do not generate information. Rather it is the selection of individuals that carry these mutations (a euphemism for the relatively higher killing of those who don’t) by the environment that imparts new information to the gene pool.

Many vertebrates, however, do not rely solely on genetics to transmit information to the next generation. Conditions in the mother’s external environment can influence the conditions in the womb, which can in turn greatly influence the developing embryo. After birth, parental nurturing can transfer still more information to the young. In humans, this extra-genetic transmission has become massive through the invention of culture and technology. The written word is certainly a profound cultural invention that has greatly changed the amount of information being transmitted from generation to generation. But is the language upon which writing is built a cultural construct or a biological capacity? This question continues to be one of some debate, but the answer certainly lies somewhere between these two extremes.

What is Language?

Before investigating how language came in to being, we must first decide what language is. When alone in a quiet space, one may think in a continuous stream of internal speech. At such times, language seems to be an integral part of thought. But there is no evidence that language is essential to any particular cognitive operation. Damage to the brain has for some patients resulted in a complete loss of speech, both external and internal, but researchers have been unable to correlate cognitive deficits with this loss. In his book Origins of the Modern Mind, Merlin Donald compares the loss of language in these patients to the loss of a sensory system. The patients have lost a tool that greatly simplifies life in the world, but like a blind or deaf person, there is no diminished intellect or consciousness that accompanies this loss.

Language is thought to be a mechanism for transmitting the information within thoughts. One experiment used to demonstrate this idea requires subjects to listen to a short passage of several sentences. The subjects are then asked to repeat the passage. Most subjects accurately convey the gist of the passage in the sentences they produce, but they do not come close to repeating the sentences verbatim. It appears as if two transformations have occurred. Upon hearing the passage, the subjects convert the language of the passage into a more abstract representation of its meaning, which is more easily stored within memory. In order to recreate the passage, the subject recalls this representation and converts its meaning back into language.

This separation of thought and language is less intuitive than it might be because many people find language to be a powerful tool with which to manipulate their thoughts. It provides a mechanism to internally rehearse, critique, and modify thoughts. Language allows us to apply a common set of faculties to our own ideas and the ideas of others presented through speech. This internal form of communication is a powerful tool for a social animal and could certainly be in part responsible for the strong selective pressures for improved language use.

A Universal Grammar

Linguists are concerned with the structure of language and the rules used to form this structure. These rules constitute a “mental grammar” that is thought to be well-defined, though not identical, in every language. These mental grammars are not the grammars we learn in school, but rather they are the unconscious process by which we decide whether a sentence is well-formed. The process is somewhat independent of meaning. We know that “Bob book read the” is an ill-formed English sentence though we can guess the meaning, whereas Chomsky’s nonsensical “Colorless green ideas sleep furiously” is well-formed.

Chomsky’s insight that there are a limited number of possible grammars came from an intuition about information flow. He believed that children were acquiring language too quickly to be explained by the exposure to language that they were receiving within their environment. He reasoned that the system the children were born with was already highly constrained so that with only relatively few examples, a child could derive the structure of their native language. Chomsky called this set of constraints “the universal grammar.” This set of constraints is like those on a die used to generate a random number. One cannot predict ahead of time whether the result will be 1, 2, 3, 4, 5, or 6, but one would never waste time guessing that the result was 7 or 3.14.

Other evidence for the existence of a universal grammar is the process of “creolization.” When adult humans come together that do not share a common language, they begin to communicate by forming a “pidgin.” A pidgin is not considered to be a true language because though there is a shared vocabulary, there is not a set of grammatical rules from which a rich set of expressive sentences can be generated. Children that are born into a culture speaking a pidgin will speak a language different from that of their parents. They impose a grammar upon their parents’ vocabulary to create a new language called a “creole.”

“Twin speak” is another example in which twins or two children at similar developmental ages will invent a language with each other that no one else can understand. According to Chomsky, invented languages, creoles, and all other human languages (both spoken and gestural) are in part defined by a grammar necessary for generating well-formed sentences, and these grammars share many properties according to the constraints of a universal grammar that all humans, and only humans, carry in their genetic code.

Language and the Brain

Whether all such examples of grammatical analysis hold up to detailed scrutiny is a matter of debate. Phillip Lieberman points out in his book Eve Spoke that no complete description of a grammar of any language has ever been produced, implying that these analyses are somewhat superficial. Lieberman believes that many aspects of language that Chomsky attributes to a specific brain module, “the language organ”, are in fact performed by areas of the brain using more general cognitive and learning capabilities.

Lieberman suggests that just because a skill is uniquely human does not mean that an area of brain evolved to perform that skill. Humans are the only animal that plays chess, but we do not expect to find a “chess module” within the brain. But chess does seem remarkably well-suited for humans. It is an interesting game because it is not too easy, like tic-tac-toe is, and not too hard. But theoretically, chess is not very different from tic-tac-toe. The only difference is that the level of planning required is matched to our brain capacities so that we can not see so far ahead that we know the final outcome, but we can see far enough ahead that we can influence the outcome and do not feel like all is left to chance.

As with chess, our ancestors may have invented a means of communication that matched the cognitive capabilities of the brain. Chomsky’s universal grammar may exist in at least the constraints that these capabilities place on language. Language may have culturally evolved so that it was easy to learn using these capabilities, even with relatively few examples from which to derive its grammatical structure. It may be our desire to communicate with each other that is hardwired into our brains. In the absence of an existing language, this overpowering desire forces us to create a new language. However, because this new language is built upon the same cognitive structures as an existing language would have been, the new language shares common structural themes with existing languages.

The fact that the brain appears to have areas within it that are used for processing distinct aspects of language is thought to be evidence for a language organ. However, this evidence would only be supportive if the language areas appeared before a language was learned. The fact that language areas develop as language is learned, that there is a fair amount of variance in their location, and that their location within the brain can dramatically shift if early damage to the region is incurred, is at least as supportive of the theory that there is not a specific language organ, but rather general cognitive modules that have the potential of becoming language processors.

It seems likely that language was built on top of an existing cognitive structure. However, given the importance of language to human social interaction, including reproduction, it also seems likely that selective pressures would prefer genetic modifications that improved language capabilities. These pressures also existed in the cultural environment, and groups that could create a language that would improve cooperative behavior might have had a distinct survival advantage over other groups. Their language and language capabilities would have effectively isolated them in a speciation process resulting in a new language-based species.

Language and Speech

Language is a complex phenomenon with origins that are difficult to trace. Certain aspects of language, though not without controversy, are better suited to empirical investigation. Speech is one example because it requires some physical properties that can be measured in, or at least partially derived from, the fossil record.

Phillip Lieberman has investigated the origin of speech for many years and has used this research to form hypotheses about the evolution of language. Lieberman suggests that speech improved greatly about 150,000 years ago when the larynx descended into the throat. According to the work of Lieberman and his colleagues, this descension improved the ability of early homonids to make key vowel sounds. Whereas the Neanderthals had a vocal tract similar in many respects to that of a new born baby, the elongated pharynx of a modern adult human is thought to enable production of a more perceptible repertoire of speech sounds. Lieberman suggests that though Neanderthals probably had some form of language, they may have failed to extend this language because they lacked the physical apparatus for producing a more sophisticated set of speech sounds. The theory that the modern human vocal tract is better suited for production of vowels has, however, recently been called into question by Louis-Jean Boe.

evollangBut the strongest part of Lieberman’s argument is evolutionary rather than strictly phonetic. He points out that the descension of the larynx into the throat makes humans much more susceptible to choking than any other mammal. It is unlikely that such a dangerous adaptation would have evolved unless there was some strong selective advantage provided by it. Since the larynx contains the vocal cords and is critical for speech, it seems likely that Lieberman’s hypothesis is correct in that this change somehow improved speech and that improved speech gave a distinct selective advantage, even if his hypothesis about what the exact improvements to speech were turns out to be incorrect.

From Hand to Mouth

Even before the development of speech, there must have first been a desire to communicate. Purposeful communication is very different from animal calls that signal more or less automatically to the whole group. Purposeful communication requires a coordinated transmitter and an attentive receiver. The first form of such communication may have been through hand gestures and facial expressions. Reading facial expressions may have helped individuals to anticipate the coming actions of their peers. Attention to hand movements may have allowed copying of skills requiring manual dexterity such as tool making.

Giacomo Rizzolatti has performed a series of experiments that show a possible neural substrate for such abilities in monkeys. Rizzolatti discovered a cell type in an area called F5 that can respond to a precise hand or mouth movement performed by another animal. A neuron might respond when the experimenter turns his hand clockwise to obtain a food pellet. No other hand motion by the experimenter causes the cell to fire. It might also fire, however, when the monkey itself performs a similar movement with its own hand to reach the pellet. For this reason, the neurons are called “mirror cells.”

Mirror cells are causing quite an uproar in the scientific community. They are exciting because they seem to provide a cornerstone upon which a system of communication can be built. In a concrete usage, the cells might allow an animal to mimic a motion that it sees another animal perform in order to obtain food. In a more abstract usage, these cells might let the animal mimic the hand or mouth movements of another animal to copy a sound or gesture to which that animal had assigned a meaning. In this way, the two can create a shared vocabulary.

Rizzolatti has suggested that F5 corresponds to Broca’s area in the human brain, the area ostensibly responsible for speech initiation and production. An interesting hypothesis about the role such an area may have played in early language evolution has been proposed, though by no means proven, based on this research. According to this hypothesis, dexterous hand movements useful for tool making were adapted for a system of communication based on gestures. Sounds were used to append ancillary information such as emphasis to the gestures. In time, the complex motor control area was adapted to also control the vocal tract, and the primary means of communication shifted to sounds thus freeing up the hands for other work. However, the hands could and still do provide ancillary information such as that previously provided by voice (as Rizzolatti, whose laboratory is in Italy, well knows).

Though highly speculative, this scenario fits well with some other important phenomena. For example, the hypothesis suggests why the sign languages used by the deaf are successful languages. The hand and mouth have always been partners in communication, and one can fill the role when the other cannot.

The hypothesis also fits well with a phenomenon known as the “McGurk Effect” which is a strikingly powerful illusion in which one syllable is pronounced (tape recording), another is seen (silent film), but a third syllable is perceived. This effect demonstrates that we obtain speech information visually as well as aurally. When the two information channels do not agree, we blend the information to perceive a syllable that was not produced. The mirror cells suggest that the visual information channel may serve in language acquisition when we attempt to mimic the speech sounds that we see being made.

Conclusion

The desire to communicate is strong in the human species, and language provides a powerful tool by which to communicate. Although language is not required for complex cognitive thought, it has undoubtedly evolved to aid with an individual’s internal thought. It simplifies symbolic representation, allowing for a host of powerful mental manipulations, such as long-term planning and abstract reasoning.

Natural selection probably played a significant role in the emergence of language. This selection process is effective whether information is transmitted genetically, culturally, or both. Troops of animals that communicated well within their group would have had a distinct advantage in coordinated action over closely related troops that did not. Of all the hominid species, it is ours, presumably the most sophisticated user of language, that alone survives. The emergence of language marked a turning point in human biology when the information that defined the species and was being conveyed from generation to generation was no longer primarily in the genetic code.

Dr. Bret Peterson has been working towards integrating informatics and neuroscience for the past eleven years. He worked most recently as Editor-in-Chief of BrainConnection.com for Scientific Learning Corporation in Berkeley, California. Previously, Dr. Peterson was an associate research scientist at Yale University working on the Human Brain Project. He has also worked as an employee of Apple Computer and as a consultant to BioRad. Dr. Peterson received a bachelor’s degree in Computer Science from the University of California, Berkeley and a master’s degree and a Ph.D. from a joint University of California, Berkeley / University of California at San Francisco Bioengineering program. He also conducted subsequent research at Los Alamos National Laboratory.

Comments are closed.