Noam Chomsky Between the Human and Natural Sciences
University of California at Berkeley
The problem of the relationships between the human and natural sciences, if there is a problem, pertains to method, not object. Some people don’t like the word "object" because they believe that it must be something like a stone or a table. But in ordinary English, insofar as I understand it, an object is what we know, talk or write about and any such object may be approached from different points of view and studied by various methods. The human animal is an object of natural as well as social and human sciences. It follows that the problem about the human and natural sciences is not only about method but about classification. It is concerned with the question whether there are methods that are unique to natural or human sciences. Since this question pertains to methodology, perhaps an unprofitable topic of discussion, I shall focus attention on a contemporary scientist whose work exhibits all the features we need to consider. The work of Noam Chomsky, the topic of my final section, will be put in perspective by introductory sections on communication and knowledge.
The term science, the only noun in my title, has become controversial. Many people, even scientists who use the expression "modern science" in other contexts as a kind of yardstick, maintain at the same time that the concept of science reflects a particular culture. I shall not discuss that problem here but believe that it can be resolved by studying the history of science cross-culturally as Joseph Needham has done in his monumental Science and Civilization in China. These eighteen volumes demonstrate that the notion of science is universally valid even if particular sciences and the classification of sciences are not. As it happens, the Euro-American classification of sciences is not only non-universal but has led to pernicious confusions in our own backyard.
Paradoxically, contemporary English usage adopts a nineteenth century German distinction that does not exist in any other European language or any other language as far as I know. Its source lies in the confused ideas of the nineteenth century German philosopher Wilhelm Dilthey, who was primarily a historian and student of civilization but knew little or nothing of mathematics or physics (Staal 1989, Ch.29). Dilthey introduced the distinction between natural sciences (Naturwissenschaften) and sciences of culture (Kulturwissenschaften). C.P.Snow, a physicist and successful author, who did not know anything about human sciences, did the rest. The result is that contemporary English not only restricts the term science to the natural or exact sciences, but has ceased to regard the humanities as sciences and balks at the notion of human sciences. By adopting an outmoded German distinction, English science has become different from French science (spelled the same but pronounced differently), German Wissenschaft, Dutch wetenschap, and Japanese gaku, terms and concepts that refer to all forms of systematic knowledge of ourselves and the universe in which we live.
If we go back further into European history we are in a position to find earlier causes that account for another confusion: that between human sciences and humanities, sometimes referred to as letters, arts or liberal arts. The "seven liberal arts," Greek in origin, were included in the medieval Trivium of grammar, rhetoric and logic and the Quadrivium of arithmetic, geometry, astronomy and music. Picking and choosing from these ancient traditions, many contemporary trends in the humanities have become entangled in non-scientific issues, ranging from educational policy to relativistic fashion, rather than concentrating on the search for reliable knowledge that is the hallmark of science. The present essay hopes to celebrate the human sciences, free them from the grip of the noble but trendy humanities and bring them back into the open forum of science.
The first part of this article deals with communication. The second is concerned with knowledge which, apart from being the primary aim of science, is also widely found among animals. But while nonhuman animals use sounds for communication, they do not employ them for the expression of knowledge. That discussion will take us to the question of the role of language in the acquisition of knowledge and the enduring debate between Chomsky and his critics on the design of language. The work of Chomsky and his followers has reminded us that linguistics is a human science that is exact. By adding the category of exact sciences to our pair of human and natural sciences we add complexity and in so doing pave the way for a solution.
I begin with a verbal expression from Vedic or early Sanskrit, spoken around 1,000 BC but increasingly rare in later Sanskrit and the modern languages of India: the optative. Its morphology is as rich and intricate as are its uses, functions and meanings. The Indian grammarian Panini formulated in his Sanskrit grammar of the fifth century BC the following rule about the semantics of this mode of the verb:
The optative means injunction, invitation, permission, respectful command, deliberation, or request (Panini 3.3.161).Examples are given by commentators:
(a) injunction: "You should go to the town";
(b) invitation: "Please eat here";
(c) permission: "You may sit here";
(d) respectful command: "(Please) teach my son";
(e) deliberation: "Should I study the Vedas or logic?"
(f) request: "May I get a meal?"
Later grammarians discuss simplifications, improvements or replacements of this classification.
Some suggest to reduce the first four subdivisions to one; others include in that reduction also the fifth, etc. (Gune 1978:17).
One simple feature shared by all these cases is that they only involve "I" and "you," the first and second persons, the only personal pronouns that are in a position to directly communicate. Formal and semantic complexity, associated with the communicative functions of the first and second persons, is found in a wide range of early, classical and tribal languages. Many languages possess three persons in the singular and plural, and incorporate (as the third person does in English) different genders. Arabic distinguishes between masculine and feminine also in the second person. Modes of address depend in many languages on the person addressed and this is not confined to pronouns. In Javanese, ceremonial and non-ceremonial languages are different from each other. Here is how Clifford Geertz (1968:248) introduces the subject:
In Javanese it is nearly impossible to say anything without indicating the social relationship between the speaker and the listener in terms of status and familiarity. Status is determined by many things – wealth, descent, education, occupation, age, kinship, and nationality, among others, but the important point is that the choice of linguistic forms as well as speech style is in every case partly determined by the relative status (or familiarity) of the conversers.English "you" is exceptionally convenient which helps to account for the popularity of English among people that wish to break away from barriers of caste and social stratification. English has influenced other languages in this respect. When I buy bread in an Amsterdam store, I am now addressed with the same pronoun as when I was a child but it was used then only when talking to children. Some languages that we regard as exotic, possess a wide range of pronouns reflecting complex kinship systems. Lardil and other Australian languages reflect a system that distinguishes between harmonic and disharmonic relations with kinsmen which Kenneth Hale defines as follows: A person is harmonic with respect to members of his own generation and all even-numbered generations counting away from his own (e.g., his grandparent’s generation, his grandchildren's generation, etc.). He is disharmonic with respect to members of all odd-numbered generations (e.g., that of his parents, that of his children, that of his great-grandparents, etc.). These languages possess different sets of pronouns: if the persons referred to are harmonic with respect to one another, a pronoun from one set is used; if they are disharmonic, a pronoun from the other (1966:319).All these distinctions exemplify how human language is used for communication and the expression of social relations between the members of a linguistic community. While the formal complexities have been studied by linguists, the communicative functions are emphasized by anthropologists. After pointing out that war is unknown among the South African San, M. Konner writes: Conflicts within the group are resolved by talking, sometimes half or all the night, for nights, weeks on end. After two years with the San, I came to think of the Pleistocene epoch of human history (the three million years during which we evolved) as one interminable marathon encounter group. When we slept in a grass hut in one of their villages, there were many nights when its flimsy walls leaked charged exchanges from the circle around the fire, frank expressions of feeling and contention beginning when the dusk fires were lit and running on until dawn (1982:6).Eibl-Eibesfeldt (1989:525-6) writes about the !Kung and Eipo: In both cultures, a great deal of talk centered on food… A larger part of the food conversations are concerned with the social aspects of nutrition. Topics include who gives what to whom, and criticisms of those who do not share their food. Three quarters of all words of the Mek-speaking In are based upon giving and taking.I don't know what is the probative value of these accounts or how much their authors know of the languages about which they write. I quote them because of the communicative flavor they depict. They certainly show what could not fail to strike any nonhuman observer: humans like to talk and talk a great deal.
When studying the commentators’ examples of Panini’s rule about the semantic features of the optative in Sanskrit, it dawned upon me that our nonhuman cousins and relatives engage in at least some of the kinds of behavior of "injunction, invitation, permission, respectful command, deliberation, or request" without using language. There are nonhuman antecedents for the communicative complexity of many features of Panini’s analysis and the classics of ethology describe them. Konrad Lorenz (1963) provides examples of non-linguistic expressions of rank and status, quarrels about ranking order or appeasement gestures. Eibl-Eibesfeldt (1971) mentions forms of submission, cups begging for food or contact, etc. Of course, it had long been known that animals use sounds in mating calls as well as shouts of aggression, territorial, distress or rain calls. There are now comprehensive surveys of the vast literature on animal communication, e.g., Marc Hauser’s Evolution of Communication of 1996. These nonhuman antecedents show that a good part of human communication is about the same things other animals communicate about. That is perhaps not surprising but I regard it as an important observation for the human sciences.
I now come to knowledge which takes us back to the beginning of the animal kingdom. The media devote much attention to new information technologies and communication devices that are claimed to transform our world, but they do not clarify the difference between communication – often no more than a commercial for a new gadget – and what pertains to knowledge or its more fashionable near-synonyms, cognition and the processing of information. Knowledge and communication are very different things and my talk does not make much sense if we do not distinguish between the two. We shall see that communication generally depends on knowledge, not vice versa. Let me begin with a typical case of animal knowledge that does not involve communication. It comes from the amphibian class of vertebrates and is due to a celebrated classic: Lettvin, Maturana, McCulloch and Pitts’s 1959 article "What the Frog's Eye Tells the Frog's Brain." These authors summarize some of their findings as follows: The frog does not seem to see or, at any rate, is not concerned with the detail of stationary parts of the world around him. He will starve to death surrounded by food if it is not moving. His choice of food is determined only by size and movement. He will leap to capture any object the size of an insect or worm, providing it moves like one. He can be fooled easily not only by a bit of dangled meat but by any moving small object. His sex life is conducted by sound and touch. His choice of paths in escaping enemies does not seem to be governed by anything more devious than leaping to where it is darker. Since he is equally at home in water and on land, why should it matter where he lights after jumping or what particular direction he takes? He does remember a moving thing provided it stays within his field of vision and he is not distracted.
Lettvin and his colleagues do not use the term "knowledge" but refer to "concern" and "remembering" which imply knowledge. In telling us what the frog's eye tells the frog's brain, they tell us what he knows. Frogs know much else (e.g., they recognize calls and localize the source of a sound) but I shall confine myself to this knowledge which they derive from vision and utilize to feed. That knowledge is accurate: the frog knows the precise location and speed of movement of a fly since otherwise his long tongue could not catch it. That knowledge serves no other purpose and exists without talk or communication.
What kind of knowledge does the frog possess? Not conscious knowledge like some of ours. The frog needs to know whether there is a fly, not what is a fly, the kind of knowledge a human zoologist seeks. An amphibian Aristotle or Heidegger could declare: "we are concerned only with the existence of flies: THAT THEY ARE, not with their essence: WHAT THEY ARE."
Since the natural history of knowledge starts early in the evolution of life, I cannot avoid a brief mention of the basic question, whether biological evolution involves a kind of progress – not an uncontroversial notion. Connie Barlow’s 1994 anthology Evolution Extended: Biological Debates on the Meaning of Life opens with a general discussion of the differing views of "Prophets of Progress" and "Naysayers." All agree that different animals possess different kinds of knowledge. It is common for Naysayers to restrict knowledge to the ecology of a single species. It is common for Prophets to emphasize that an important feature of the evolution of knowledge is that knowledge is increasingly independent of a species' environment. One of the prophets of progress in the Barlow volume, Francisco J. Ayala, accepts that knowledge is not always increasing:
Paramecium, a single-celled organism. . . does not seek its food or a favorable environment, but simply avoids unsuitable conditions. Euglena, also a single-celled organism, exhibits a somewhat greater ability to process information about the environment. . . it not only avoids unsuitable environments but it actively seeks suitable ones . . . An increase in the ability to gather and process information about the environment is not a general characteristic of the evolution of life. Progress has occurred in certain evolutionary lines but not in others. Today's bacteria are not more progressive by this criterion than their ancestors of three billion years ago.Along with this assessment, Ayala recognizes a form of progress that is not inconsistent with it: When organisms are measured by their ability to process and obtain information about the environment, mankind is, indeed, the most progressive organism on earth.There are species of animal where even mere comparison seems to be fraught with difficulties. There is the well known case of bats' echo-locating which seems to be incompatible with our customary modes of perception even though those same techniques are now applied by humans in contexts of war and medical diagnosis. Richard Dawkins (1986:77-81, etc.) has put it in simple terms: bats see the world with their ears. But what they hear and what we see is the same world and that is relevant to the discussion of human and natural sciences: for even those who assert that Chinese, Indian and American concepts of astronomy are different, do not claim that there are different Chinese, Indian and American suns or moons. In mythology, Yes; in truth, No. The Chinese, Indian and American terms for "sun" and "moon," if they refer at all, refer to two referents that describe their orbits in the universe outside that globe on which our discussion takes place. In logic, the expression "pure reference" is used for such kinds of reference and biologists use it in their discussion of the limits of meaning in animal calls.1 What is the meaning of that term itself?
We have seen that with respect to knowledge, humans are not so very different from nonhuman animals. But humans possess an apparently more distinct faculty: language. I have mostly approached language from the angle of communication but it is also concerned with knowledge, often "referential knowledge" which presupposes "pure reference." Both these concepts are concerned with expressions of language: "pure reference" with words such as nouns or noun phrases such as "this chair" that refer to particulars such as chairs; "referential knowledge" with assertions or affirmative sentences. Since sentences consist of words, "referential knowledge" and "pure reference" form a pair. But there is an important difference that was pointed out by Aristotle: affirmative sentences are true or false; but uncombined expressions such as "man," "white," "runs" or "wins" are not true or false.
I use the term "sentences" (though "affirmative sentences" would be better and one might use "statements," assertions" or "propositions") because it was used by Alfred Tarski, the logician who, in the footsteps of Aristotle, has done most to clarify the concept of "truth" (Tarski 1935, 1969).2
Tarski writes: "The problem of the definition of truth obtains a precise meaning and can be solved in a rigorous way only for those languages whose structure has been exactly specified." This clearly refers to artificial languages, but Tarski has thrown light on natural language as well. Although it is part of the same system, Tarski's favorite illustration of a true sentence, the sentence "snow is white," is different from the sentences I discussed in our sample of intricate morphology from the verbal systems of moods, voices, tenses and aspects of early Indo-European and other classical or tribal languages. The language of truth uses only a tiny fragment of all that communicative complexity that is so characteristic of natural language.
True sentences are different from optatives or questions. They are not like the more fashionable performatives that make things true: they are true (or false). They use only one person (the third), no gender, one mood (the indicative), no aspect, and one tense (the present). The third person is precisely the one used least in contexts of communication. One philosopher who paid attention to these facts was Leibniz: according to him, Latin would be a better language for the expression of scientific truth if all distinctions of gender, number, tense, person and mood were omitted from it. Leibniz would have preferred Japanese San to the English distinction between Mr. and Mrs. He quotes evidence in support of his view: a Dominican priest from Persia he had met in Paris spoke with great fluency a broken Latin in which such distinctions were entirely neglected, but was completely intelligible to anyone who took the trouble to listen to him. He supported his view further with a comment on the alleged intelligence of mathematicians: when dealing with things other than mathematics, mathematicians are not smarter than anyone else. Leibniz concluded that their apparent intelligence is due not to themselves but to the mathematical language they have learnt (Leibniz in: Mates 1986:179).
I am not a biologist but I have the impression that nonhuman animals are not concerned with the truth of sentences. Biologists do not appear to have pondered truth either although the closely related concept of "pure reference" has become increasingly apparent in a series of studies by Seyfarth, Cheney and others. Cheney and Seyfarth showed that the vocabularies of nonhuman primates are larger, more specific and less dependent on context than scientists had initially assumed. Playback experiments demonstrated that vervet monkeys utter different alarm calls when they see eagles, leopards and snakes; they do not give eagle alarm calls when they have seen leopards or vice versa. The alarm call, presented on its own with the help of a tape-recording, elicits the same response as would the predator itself. Cheney and Seyfarth concluded that alarm calls functioned as representational or semantic signals, something very similar to "pure reference" (Cheney and Seyfarth 1985, 1990, 1992).
Cheney and Seyfarth's conclusion has not been without its critics. Charles Snowdon noted that the ability of animals to form inferences from playback studies, that appear to be referential, does not mean that the communicator produces utterances at the same cognitive level. Cheney and Seyfarth accepted this. Snowdon expressed his objection later in clearer terms:
Playback studies are frequently used to argue that calls are representational or symbolic, but there is a logical problem with playback studies. By focussing on the behavior of the recipients, the studies show us the nature of the inferences formed by the recipients but do not inform us that the communicator is communicating representationally. If I shriek with joy in my office, a listener might infer that I have just had a paper accepted or a grant funded, but if I give the same shriek at home, a listener will form a very different inference. In neither case, though, have I specifically designated the reference of my shriek (Snowdon 1990, 1992).The same conclusion applies to linguistic utterances such as "Fantastic!" Some are on the border of language: "Wow!" They are, like shrieks, not referential but may be expressive, for example, of surprise or joy. And yet, the continuation of the biological debate made it clear, that the earlier assumption, that these alarm calls express emotional or affective states of the animals rather than references to external referents, should not be easily discarded. We may express the difference by making use – as we do in any case – of human language. In no case does a leopard alarm call made by vervets correspond to the utterance: "Leopard!" It may mean: "I warn you that there is a leopard!" But it could also express an alarm or urgency of response that leads the hearers (along with the caller) to run into trees. The eagle alarm, similarly, could mean: "I warn you that there is an eagle!" but it might simply induce hearers (along with the caller) to take cover in underbrush.
The evolutionary reasons were elucidated later when Macedonia and Evans (1993) studied primates and rodents and paid attention to sound production as well as perception. These studies led to the conclusion that animal signals form a continuum ranging from reflections of motivational states, such as distress or contentment, to those that respond to particular stimulus categories and are relatively unaffected by affectionate responses. Macedonia-Evans do not mention reference, but they proceeded to show that responses to predator threats are different among different species.
Vervets and ringtailed-lemurs are highly terrestrial compared with their close relatives the ruffed lemurs; but they are able to escape into trees and may therefore respond to predators on two different planes: on the ground or in the trees. The ruffed lemur lives mostly in trees and is therefore unlikely to have to face a carnivore on the ground. Let us denote the calls uttered by the ruffed lemur as A and B, and the calls that vervets and ringtailed lemurs use as C and D (See figure 1):
A may mean: "Monster!" or "I warn you that there is a monster!"OR: "Escape into the trees!"
B may mean: "Eagle!" or "I warn you that there is an eagle!"
C may mean: "Monster!" or "I warn you that there is a monster!"
D may mean: "Eagle!" or "I warn you that there is an eagle!"OR: "Take cover in the underbrush!"
The importance of conflicting escape tactics in the case of the two-plane animals could have led to the evolution of functionally different referential alarm calls. There was no need for it to develop through natural selection among the highly-arboreal one-plane ruffed lemurs. Language evolved because we descended from the trees for these reasons in addition to the more familiar consequences such as standing up, thus freeing hands and therefore mouths, which could now be used for speech.
Are "referential alarm calls" referential expressions in the sense of modern logic? They are not for the leopard alarm call never refers in the manner the word or term "leopard" sometimes refers. The expression biologists use to characterize these calls is apt: they are "functionally referential."
Hauser and Marler (1993) used the term "functionally referential" in their study of food-associated calls in rhesus macaques but concluded that "we do not know whether these calls represent labels for food or command or request others to come and eat the food."
It is clear from these few illustrations that contemporary research on animal communication has reached a high level of sophistication. So far, no simple and definite conclusions appear to have been reached but it would seem fair and safe to conclude that if reference to outside referents exists among nonhuman animals at all, it is not "pure" but "functional" reference which remains firmly embedded in the context of communication. That conclusion suggests that human language itself may also have been selected primarily for communication.
3. Noam Chomsky and the So-Called Exact Sciences
Was language selected for knowledge or communication? To answer that question we must take into account several features of language, natural selection and evolution that I have not so far mentioned. The first is that language and therefore linguistics consists of several components and to ask whether it was selected for "A" or "B" is an oversimplification. Following Jacob Grimm, we may refer to this feature as God’s Teeth.
A century and a half ago, a debate was raging about the question whether God spoke Hebrew. Jacob Grimm, one of the brothers to whom we owe the collection of fairy tales, pointed out gently that if God spoke language, any language, we must assume that he had teeth, but since teeth were not created for speech but for eating, we must assume that he also ate, and this leads to so many other undesirable assumptions that we better abandon the idea altogether (1851, reprint 1958:28).
Although it was eight years prior to 1859 when The Origin of Species was published, Grimm’s observation is consistent with Darwin’s natural selection in that it shows, that any feature or organ selected for one reason may be used for another. It holds for the mouth as well as the ear, that exquisite instrument that enables us not only to listen to symphonies but identify dentals such as t and d, produced by the teeth, velars such as k and g, produced by the throat, etc. But the hearing apparatus within the ear was not selected for any such reason. It is a result of the growth of the skull in early mammals that caused a bone of the reptilian jaw to migrate to the ear.
Let us reduce the number of linguistic components as much as we can and start from the general idea that language uses sounds to express meanings. It is a tautological statement if that is how we define language. Whatever it is, it implies that there must be at least two components of linguistics: phonology and semantics. A third, syntax, relates them to each other but, as it happens, in a very particular, roundabout and unexpected manner. The first major contribution of Noam Chomsky was to show that syntax is independent from phonology and semantics. This led to major advances in linguistics but the analysis continued to be phrased in terms of rules, which had been employed for more than two and a half millennia (they go back to Panini’s Vedic predecessors). Then Chomsky made another radical change: he showed that the formal properties of syntax may be derived from more abstract principles. Syntax and phonology, then, are essentially different (Sylvain Bromberger and Morris Halle 1991). The evolutionary antecedents must also be different, but how does this affect the oversimplified question with which I started and for which my two earlier sections paved the way, namely: Was language selected for knowledge or communication? Chomsky, the greatest linguist since Panini, has always emphasized that language is not primarily for communication but serves other functions such as the expression of knowledge, thought or ideas. What do we mean, he asks (1980:230), by "communication" in the absence of an audience, or with an audience assumed to be completely unresponsive, or with no intention to convey information or modify belief or attitude? It seems that either we must deprive the notion "communication" of all significance, or else we must reject the view that the purpose of language is communication. Chomsky's emphasis on language as a cognitive system and means for the expression of truth continues a European tradition in which Wilhelm von Humboldt played the key role (see, e.g., Chomsky 1964; 1966). Humboldt (1836) defended the view that no language should be regarded merely, or primarily, as a means of communication. According to him, the instrumental use of language (its use for achieving concrete aims) is derivative and subsidiary. Chomsky's development of these ideas within a contemporary scientific framework has shaped modern linguistics and psychology and led to new disciplines such as the cognitive sciences.
Chomsky has not written much about the origin of language and has expressed skepticism about the relevance of natural selection. He has often been attacked on these grounds; and in that context, misrepresented and quoted out of context. I shall therefore quote in extenso three paragraphs from his 1986 Managua Lectures that address the problem of the evolution of language:
Can the problem be addressed today? In fact, little is known about these matters. Evolutionary theory is informative about many things, but it has little to say, as of now, about questions of this nature. The answers may well lie not so much in the theory of natural selection as in molecular biology, in the study of what kind of physical systems can develop under the conditions of life on earth and why, ultimately because of physical principles. It surely cannot be assumed that every trait is specifically selected. In the case of such systems as language or wings, for example, it is not easy even to imagine a course of selection that might have given rise to them. A rudimentary wing, for example, is not "useful" for motion but is more of an impediment. Why then should the organ develop in the early stages of its evolution?I shall return to counting later and first take up the principle of cumulative selection that evolutionary biologists adopt. It’s "God’s Teeth" once again. In single-step selection, entities selected are sorted once and for all. In cumulative selection, they "reproduce": the results of one sieving process are fed into subsequent sieving, which is fed into. . . and so on. Although the human eye could not have arisen directly and in a single step by natural selection from no eye at all, it could have arisen from something slightly different from itself, call it X. If we allow ourselves a sufficiently large continuous series of X's, that is, sufficient time, we have no difficulty in connecting the human eye by means of such a series to a state with no eye at all (Dawkins 1986:45, 77-81, 302 etc.).
In some cases it seems that organs develop to serve one purpose and, when they have reached a certain form in the evolutionary process, become available for different purposes, at which point the process of natural selection may refine them further for these purposes. It has been suggested that the development of insect wings follow this pattern. Insects have the problem of heat exchange, and rudimentary wings can serve this function. When they reach a certain size, they become less useful for this purpose but begin to be useful for flight, at which point they evolve into wings. Possibly human mental capacities have in some cases evolved in a similar way.
Take the human number faculty. Children have the capacity to acquire the number system. They can learn to count and somehow know that it is possible to add one indefinitely. They can also readily acquire the technique of arithmetical calculation. If a child did not already know that it is possible to add one indefinitely, it could never learn this fact. Rather, taught the numerals 1, 2, 3, etc., up to some number n, it would assume that that is the end of the story. It seems that this capacity, like the capacity for language, lies beyond the intellectual range of otherwise intelligent apes. It was, incidentally, thought for a time that certain birds could be taught to count. Thus it was shown that some birds could be taught that if they are presented with four dots, then they can find food in the fourth container in a linear array. The task could be performed up to about seven items, leading to the conclusion that birds can count. But the conclusion is incorrect. The most elementary property of the number system is that the series of numbers goes on indefinitely; you can always add one more. Birds may have certain limited capacities to match arrays of not too many items, but that has nothing to do with the faculty of number. The ability to count is not "more of the same" but entirely different in character (Chomsky 1988:167).
Cumulative selection accounts for the fact that original design and current function need not be the same. Dawkins shows that for an explanation in terms of natural selection to work, the complexity must be analyzed and a series of events must be shown to have been adaptive, it doesn't matter for what. In cases such as the eye or language, many evolutionary trajectories must have come together and we must not only account for all the steps, but demonstrate why the endpoint is adaptive. In the case of the eye, the answer is straightforward: good vision – and that may be a matter of life and death. For a fast-flying bird, a properly focused lense can make all the difference between catching a fly and smashing into a cliff.4
I am concerned with the endpoint and the pre-final steps of language evolution, when various evolutionary trends happened to come together. At that point in time, nonhuman animals had communicated by means of sound for a very long time. Cumulative selection explains why language could have been selected for communication then, long before it was used for other ends. That is precisely what I suspect to be the case. To demonstrate it we have to construct chains of events and apply some other general principles that I have not yet mentioned. These principles are also concerned with the explanation of human nature and have to be taken into account by the human sciences. Adaptations take place in a given environment, but different species exist in different environments and environments change. Adaptations, therefore, reflect the past, and given the pace of evolution, it is a distant past. An evolutionary account of the origin of language is subject to these constraints and the same holds for human nature in general: "The brain/mind mechanisms that constitute human nature were shaped by selection over vast periods of time in environments different in many important respects from our own, and it is to these ancient environments that human nature is adapted" (Donald Symons 1992:138).
If language was selected for knowledge, it must be knowledge derived from the perceptual system which, in the words of Roger Shepard (1982), "over the eons of vertebrate evolution, internalized the most important invariants and constraints in the external world." If that is so, our language reflects knowledge of the world not as it is but as it was or appeared, many millions of years ago, to lowly creatures from whom we happen to descend. In the mean time, the world itself has also changed. We may, in some cases, be able to perceive it as it is; but our perceptual apparatus and language both predispose us to see it as it is no more. We are in no position to determine whether our perception or knowledge of the world is adequate unless scientific analysis has cleared up the matter.
The implications are clear. It is an undeniable fact that language has been used for the expression of knowledge as well as communication for thousands of years. But if it had been originally selected for knowledge, it could not have been adaptive, since the only knowledge it was able to deliver was out of date by millions of years. These features of biological evolution support the view that language was primarily selected for communication, not knowledge.
Both knowledge and communication presuppose meaning and therefore semantics. Animals have semantics: whatever their calls convey, they convey something. Their kind of meaning is restricted to basic communication and not richly structured. I mentioned already that it is likely, in view of cumulative selection, that semantics, syntax and phonology have different evolutionary antecedents. I believe that the evolutionary antecedents of syntax and phonology were combined first and that natural selection refined the resulting combination further by gradually introducing greater structural complexity into existing sound systems. An example is birdsong to which I return in a moment and, among humans, mantras and Lieder ohne Worte, "songs without words." The crucial step in the origination of language was the use of these richly structured sound systems for the expression of meaning. It is not a surprising development since animals used sound already for the expression of communicative meaning and such a development would certainly be adaptive because it greatly improved communication — so much so that it also began to be used for the expression of knowledge, an animal faculty that had never before in biological evolution been expressed by sound. These were great steps forward, but that emergence would also explain that the resulting language was not a perfect instrument for communication and especially imperfect as a means for the expression of knowledge. What held it back was that syntax came from elsewhere, and was not designed for these tasks. What enabled it at the same time to open up new avenues is the feature of syntax that Chomsky refers to as digitalinfinity. Chomsky had shown, as Panini and von Humboldt had already suggested, that language is a recursive procedure that generates an infinity of expressions. Chomsky could show this by deriving formal properties of syntax from deeper and more abstract principles, something like the equations of physics. He concluded that "the language faculty has properties of simplicity and elegance that are not characteristic of complex organic systems, just as its infinite digital character seems biologically rather isolated" (Chomsky 1995:29). It is not surprising that Chomsky expressed, in such contexts, his skepticism about natural selection by emphasizing that that type of evolutionary explanation functions within a structured space of possibilities, determined by physical-chemical laws, historical contingencies, etc.
I shall take up the two parts of digital infinity one by one. The richest systems of digital expression among nonhuman animals come from the study of birdsong. As for humans who do not descend from birds, we too do a lot of singing, chanting and reciting. Darwin wrote that our ancestors probably used voice largely for singing in sexual selection "as does one of the gibbon-apes at the present day" (Darwin 1871:56 quoted by Hauser 1996:66).
The study of the syntax of birdsong developed relatively late because ornithologists analyzed songs by means of sound spectrograms which depict continuity. For syntax we need discrete or digital elements. About two decades ago, ornithologists began to use the human concept of syntax to distinguish the order of notes or syllables within a song. It is a purely serial syntax and the difference between the New York and Minnesota dialects of swamp sparrows is syntactic in precisely this sense. The most interesting system of avian syntax with which I am familiar is not serial but hierarchical like the syntax of mantras or language. It is exhibited by the calls of the Black-Capped Chick-a-dee tit (Hailman, Ficken and Ficken: 1985, 1987). In these calls, there are four "note-types" that are combined to produce hundreds of different "call-types." As in the case of Capuchin Monkeys that were studied by J.G. Robinson (1984), these notes appear almost always in fixed order (permutations are less than 0.3 %). In any given call, one note type may be omitted, given once or repeated an indefinite number of times. Each sequence resembles a small Markov chain and by referring to these chains, Hailman and his colleagues remind their readers of another early discovery of Chomsky, viz., that human language is not analyzable in terms of such mechanisms.
If the four notes are referred to as A, B, C and D, and parentheses indicate indefinite repetition, the most common call-types are: (A)(D) and ((B)(C)(D)), generating, for example:
AAAADDD and BCCCDBCCCD. There are fifteen different note-types:(A) (A)(C) (A)(B)(C)
(B) (A)(D) (A)(B)(D)
(C) (B)(C) (A)(C)(D)
(D) (B)(D) (B)(C)(D)
(A)(B) (C)(D) (A)(B)(C)(D).
There are several constraints: A may be repeated almost indefinitely; B generally not more than three times; as the number of B grows, C is likely to be omitted and the call may skip to D or omit both C and D and come to an end. If there are many C's, there are likely to be fewer or no D's. If the string is long the number of D's decreases. Once a sequence of D's is engaged in, indefinitely many D's may follow. 86 % of all calls contain at least one D.
It is not surprising to find indefinite continuation in nonhuman animals for their activities are in general, as E.O. Wilson (in Pinker 1994:340) put it, "repetitious to the point of inanity." The Chick-a-dee system is different from human syntax in that there seems to be an upper limit to repetition. Whatever it is, the relevant difference is not what Hauser says it is in his book The Evolution of Communication:
"Although there is some intuition that each note may mean something different, no formal analyses have been conducted. Without any sense of this system’s semantics, however, there can be no real sense of its syntax" (Hauser 1996:632).Hauser refers here to a true feature of the relationship between syntax and semantics in human language. Human language combines the meanings of words into larger units by means of syntax. The result is a powerful system for the expression of meaning, but not one without its problems as we shall see later (and as was already alluded to at the end of note 2). There is no evidence that nonhuman animal semantics works in this manner. Moreover, mantras, Lieder ohne Worte and music itself all illustrate that complex syntactic structures do not stand in need of any meaning. Chomsky showed that the mechanisms of human syntax are independent from semantics. That very independence is likely to be due to a non-semantic evolutionary antecedent of syntax that is purely formal. Chick-a-dee syntax is interesting because it may be such a system and therefore analogous to the antecedent of human syntax before it combined with semantics.
One question remains. What is the adaptive value of these songs?
I alluded already to Darwin’s answer and Hauser goes into much greater detail elsewhere in his book when he discusses female preference and runaway selection (pp.364 ff). Darwin had accepted female preference as given and the question has been much debated in recent biology.
The peacock’s tail, the grotesquely enlarged claw of the male fiddler crab and the machismo of members of the human species are all exaggerated features that may cause injury to individuals that display them but attract females. Since male ornaments and female choice reinforce each other, they have evolved together, leading increasingly to more extreme traits and preferences. Female platyfish (related to swordfish) prefer to their "normal" males, who have no swords, males with virtually any bizarre appendage, including artificially attached, colored ersatz swords that exaggerate their body size and coloration. Birds have simpler tastes: females prefer males when their songs are complex and numerous. Even in humans, braininess is sexually desirable — or at least such abilities as to remember the steps of a long and complicated ritual dance. All of this (Fischer 1930; Dawkins 1986; Basolo 1990; Searcy 1992; Ryan and Rand 1993; Reeve and Sherman 1993) shows in passing that there is no Cartesian dualism in natural selection. It adds fuel to the thesis that rites and mantras paved the way for the emergence of syntax (Staal 1985, 1989, etc.).
It is time to take on infinity, return to Chomsky’s third paragraph on counting and his contention that birds cannot count. I am sure he is right, but I am not sure that it may not also hold for at least some humans. So let us return once more to the Vedas with which we started when we studied the Vedic optative.
The Yajurveda, thousands of mantras and observations on ritual, altar construction, mythology and other topics, contains a few long hymns that consists entirely of numbers (Taittiriya Samhita VII.2). The first begins: "Hail to One! Hail to Two! Hail to Three!" etc. including 4 through 20, then 29, 39, 49, through 99 ending with 100 and 200. It may well be an abbreviation of the series of natural numbers from 1 to 200 (Hayashi, forthcoming).
The second enumeration begins: "Hail to One, Hail to Three" and continues with most but not all odd numbers up to 99 and adding 100. The third list takes up multiples of four, and the remaining lists multiples of 5, 10, 20 and 50. Afterwards some larger numbers are added: 1000, 10,000, 100,000 up to parardha or or one thousand billion, ten to the power 12. Later Jaina and Buddhist sages went up to much larger numbers, the largest perhaps the Jaina sirsa prahelika which is 8,400,000 to the power 28, a number of 194 digits (Joseph 1990:242). In all these cases, the numbers are larger than in birds where it was seven; but that does not prevent their enumeration from being finite and different from "the human number faculty" in Chomsky’s sense.
What is the survival value of these statements? Human interest or utility and perhaps female preference. The composers of the Vedas were great sages. There are stories in Indian mythology and folklore about women who desire to give birth to a son that has been fathered by such a sage. Tempting a sage is difficult, especially when he is an ascetic. Since ascetic power and female choice reinforce each other, they might have evolved together, leading to increasingly extreme traits and preferences. But Alf Hiltebeitel has reminded me that these stories are told by men. What would be a perfect argument for runaway selection may be no more than a male fantasy.
We don’t know at present what happened and whether language was selected for knowledge or communication in the manner I have suggested. Chomsky is right that "There is a rich record of the unhappy fate of highly plausible stories about what might have happened, once something was learned about what did happen" (Chomsky, forthcoming). But some facts are available and while we are waiting for others that may clinch the matter, we need not sit still and refrain from using logic. I’ll be satisfied if the reader regards my story as plausible. It is consistent with what I found in an earlier study: meaningless mantras are in important respects closer to bird song than to human language5 (Staal 1985; 1989 Ch.23). We seem to find again that singing and reciting sages are closer to birds than to contemporary intellectuals.6 The majority of humanity, apparently, hovers in between."
We are back in human company and return to the human sciences. Chomsky has gone further than any other student of language in discovering, establishing and formulating abstract principles from which the formal properties of language can be derived. Though he has not yet discovered the Schroedinger equation of syntax, it is largely due to his discoveries and impact that a contemporary study on theoretical linguistics looks like a treatise on algebra. That’s not just a fashionable trend. It shows that not only the rest of the universe to which they belong, but human language and humans themselves are profoundly different from what they appear to be on the surface and from what can be understood and expressed about them by means of that natural language itself.
It might be objected that not all scientists derive the facts they encounter from abstract principles. Human scientists feel that it is an attractive feature of the human sciences that one may study the rich detail of cultural diversity by concentrating on a single historical person or event, a linguistic structure such as the optative or a poem. But that holds true for all sciences.
Schroedinger was in a position to discover his equation because of his braininess and that of countless experimenters who had measured spectral lines of individual elements or studied the properties of compounds under special circumstances of temperature or pressure. A linguist and a zoologist may work in Nepal side by side, studying respectively a rare language and an almost extinct species. In all these cases, degrees of abstraction vary but "to complain of bare bones is like criticizing the physicist for failing to capture the richness of the rain forest" (Quine 1981:186).
Let’s take one more step and draw a conclusion that is significant in our present context. More than that, it is the hinge upon which my talk turns. The emergence of formal, artificial languages demonstrates that natural language is unfit for the expression of truth about what lies behind the appearances and suggests that it was not designed for that purpose. The languages of classical science – Greek, Latin, Sanskrit, Chinese and Arabic – were formal, but not formal enough to trigger a global scientific revolution. They contributed to the origin of algebra which introduced a new step of language evolution but in an entirely unprecedented direction (Staal 1995). That’s what Galileo had in mind when he declared that the book of the universe is written in the language of mathematics. The so-called European scientific revolution depended on the artificial language of algebra which was imported from the Arabs who developed it from what they had inherited from Mesopotamia and discovered in India and China (Graham 1973). That same so-called European scientific revolution would not have spread as wide and fast as it did, if mathematicians like Euler had not translated Newton’s cumbersome Latin, obsolete Euclidian methods and unwieldy formalizations into simple algebraic equations that everybody could learn to understand (Truesdell 1968:167 etc.; Park 1988:210,249).
The spectacular results of the so-called exact sciences are largely due to mathematization. The adoption of an artificial language for the expression of a fundamental truth, as in the Schroedinger equation, is not an abbreviation of what may be expressed through a natural language; it possesses a structure that may be called linguistic or syntactic but is totally different from the syntactic structure of natural language. It preserves the property of digital infinity and is still a sentence in Tarski’s sense: it must be true or false. But it does not consist of a subject and predicate – hence the inadequacy of Aristotelian logic. It lacks gender, mood, aspect or tense and does not refer to "I" or "you." And yet, artificial languages developed from the syntax of the natural, incorporating other symbolic systems such as expressions for numbers. The functional expression f(x), for example, may in some simple cases be interpreted as "x is f", which may in turn be read as a subject-predicate sentence.
The formal languages of mathematics have replaced natural language in all the so-called exact sciences. The chief purpose of this third and final section of my talk has been to show that the successes, not only of the natural sciences but also of linguistics, are due to formalization. Though syntax paved the way, natural language is unable to express such results adequately. That is relevant to our context because language is the chief characteristic of the human animal and its study is a human science if anything is — yet it is almost as formalized as physics.
Is all of this the result of the work of a single individual? Of course not. Although Chomsky seems to have gone further in deriving the formal properties of natural language from abstract principles than anyone else so far, there is a prehistory to it. In the European tradition, the formalization of natural language was undertaken by logicians starting from Aristotle.
Formalization in linguistics on a grand scale was undertaken by Chomsky’s own teacher, Zellig Harris. Leonard Bloomfield paved the way in one respect by the use of ordered rules in one of his publications – Menomini Morphonemics of 1939. Whatever the explanation (Chomsky suggests "schizophrenia"), Bloomfield did not attach importance to that work, did not exploit its formal structure or introduce its methods into modern linguistics. De Saussure was more perceptive: he did not do the actual work but noted in 1894, more or less in passing, that the expressions of linguistics "will be algebraic or will not be" (quoted in Koerner 1976:704).
The history of linguistics takes us back to our point of departure because the greatest precursors of formalization in the human sciences were the Sanskrit grammarians. Panini was inspired by the Vedic science of ritual and a tradition of recitation which inspired in turn other human sciences that analyze rule-governed activities such as poetics, literary criticism and the law. By using the case endings of Sanskrit metalinguistically as well as other metarules, rule-order, recursiveness and other formal properties of rules, Panini developed not just artificial expressions, but an artificial language.
Panini’s Sanskrit grammar is a derivational system, the only parallel to Euclid's system of geometry in any civilization with which I am familiar (Staal 1965, 1988: 143-60). His derivations seem to lack the degree of generality of Euclid's proofs; but the distinction is one of object, not structure. We are entitled to generalize and characterize early Indian science as primarily concerned with a treatment in derivational form of rule-governed human activities. It centered on linguistics but went beyond it just like Roman Jakobson’s work on poetry, literature and other features of civilization. If it is anywhere, it is in such cases that we are justified to use the expression "human science."
I have spent time on Panini and Chomsky not because their contributions are unique or uniquely valuable, but because they illustrate that science is universal and that there is no intrinsic difference in method between the human and other sciences. The mere existence of a Panini and a Chomsky demonstrates that linguistics is an "exact science" and if that is so, the distinction between human and natural sciences falls to the ground. We can go one step further and note that there is no such thing as a single or unique scientific method. Scientific knowledge, like all reliable knowledge, springs from intuitions steeped in facts, sharpened by logic, and continuously tested by both. Such knowledge is never final not because anything else is, but because it is likely to be replaced, sooner or later, by something more probable and more nearly true.
My conclusion that there is no significant difference in method between natural and human sciences should not surprise anyone who has followed, even from a distance, the progress of scientific knowledge during the past millennium. Ancient religions, philosophies and sciences had separated heaven and earth. The development of the sciences picked up speed after it had been discovered that the celestial world is not essentially different from ours. We now know that we live in one world, appropriately called the universe. Since that insight took root, progress has been rapid. The transition from physical to chemical was understood roughly a century ago. That from inorganic to organic can now be simulated in a flask. We are able to see that the origin of life is an unlikely event; but given the vast expanses of time, that event is bound to happen if the surroundings are suitable which means: a few elements are available and the temperature is approximately right. The appearance of intelligent life is again a matter of time. We may continue to draw boundaries in terms of specific concepts such as consciousness, but we also know that we share 98% of our genes with our closest nonhuman relatives, and very high percentages with a large number of other animal species. Why should we need special or even unique methods to acquire a particular knowledge?
I have tried to bring the human sciences out in the open and must now face up to the fact, that I have not only been unkind to the humanities, but have spent much of your valuable time on the execution of an ignoble task. The humanities deal with what is noble in men, women and children, their aspirations and dreams, their artistic creations and spiritual gleanings. I too believe in the need of such a civilizing education, in Greek paideia provided it is not confined to the Greek. Presently that need may be as great as the need for scientific education. We cannot act responsibly unless we know the differences between violence and non-violence, tolerance and intolerance, love and hatred. But we should never forget that these acts and dispositions are not restricted to the human animal. All I have been implying about the humanities is that they are not concerned with knowledge or scientific insight. They continue to be constrained by a particular European cultural background. Like much else, the humanities themselves can only flourish and get rid of their trendiness – another unfriendly expression in which I have allowed myself to indulge – if they stand on the firm foundation of the human sciences in the widest context possible: that of the universe. In the final analysis, the two faces of this Janus look in the same direction.7
1. The evolutionary account of knowledge is not so much inconclusive as beset with problems of terminology. According to many biologists, the reasons for applying terms such as "knowledge" to animals in certain situations are just as good or bad as applying them to humans in similar situations. This is supported by studies that relate the evolutionary biology of primates and apes to child development and psychology and has become even more apparent in another group of partly overlapping disciplines such as animal cognition, cognitive ethology and animal psychology. Donald Griffin, who coined the term "echolocation" in the course of his research on bats, is one of the founders of the first two of these disciplines and has gone furthest in applying "human" terms and concepts to other animals. He regards information processing, for example, simply as thinking (Griffin 1981, 1992). I don't know whether Griffin is thinking of thinking in relation to Paramecium and other single-cell organisms and it is not my task to attempt clarification in this area. I have suggested that there is a natural history of knowledge, using one term to refer to all the different types of knowledge that different animals possess. Perhaps I should have been less generous and confine knowledge only to higher animals whose knowledge is more specific, detailed or extensive than that of single-celled organisms. Whatever the terminology, there exists a cognitive continuum and animals possess kinds of knowledge or cognition without the ability to communicate it to others. Our use of language should express these facts as clearly as it can.
2. Aristotle's doctrine of truth is best known in its medieval form which defines truth as "correspondence between thing and intellect." Tarski went further in his study of the concept of truth in artificial languages but began by making this notion more precise. According to Tarski, any adequate definition of truth for a sentence such as "snow is white" must imply the following equivalence:
The sentence "snow is white" is true if, and only if, snow is white.
This is neither useless gibberish nor tautologous. It related a sentence of language (the name of which occurs between the quotes) to a state of affairs in reality. It is concerned with the basic relation between language and the world, including language itself. (For such an equivalence to be true, it is not necessary that the structure of the sentence mirrors the structure of reality; in fact, it is not as Bertrand Russell, Wittgenstein and many other philosophers have explained).
3. Evans and Marler called "functionally referential" those vocalizations that are sufficient to elicit appropriate behavior from conspecifics even in the absence of contextual information, such as that which might normally be gleaned from the nonvocal behavior of the sender. Such behavior is an act of communication, not a form of knowledge. Functionally-relevant vocalizations have the following properties: they are production-specific, possess a discrete structure and are context-independent. Having laid out the conceptual framework, the aim of another 1993 study of Hauser and Marler was to find out whether alarm calls are functionally referent in a familiar animal: the chicken. The result was that they are, but there was some discussion of the possibility of alternative interpretations. I shall not try to evaluate the details: the authors admit that the balance of evidence is not striking though it seems to tilt to interpretations in terms of functional reference. Another possibility is that the alarm calls make information available not about external referents but about the sender's subsequent behaviour. That discussion did not lead to a definite conclusion either and the article ends: "This is an initial step in a programme of experiments designed to determine how, and to what degree, information about eliciting events, subsequent behavior and affective state are encoded in animal signals."
4. The need for a Darwinian account of the origin and development of language was stated most perspicuously and persuasively by Steven Pinker and Paul Bloom in their opening statement and final response to a collection of papers entitled "Natural language and natural selection" by linguists, biologists, psychologists, cognitive scientists, and anthropologists in the December 1990 issue of the journal Behavioral and Brain Sciences. Pinker and Bloom state repeatedly that "language is a system of complex design and the only known explanation of such systems in life is natural selection." If we replace "the only known" by "one" this is a truism in the life sciences. Pinker and Bloom also give a brief "argument for design in language" (1990: 712-15) where they try to show that language enables humans to refer to the world. They start with nouns, verbs, adjectives etc. which "are exploited to distinguish basic ontological categories such as things, events or states, and qualities." They assume that language corresponds to the world although one of the main achievements of logic and philosophy has been to show that it does not (cf. note 2).
5. In important semantic respects: mantras and birdsongs do not express meaning systematically and may be used on all kinds of occasions. In syntactic respects, mantras are situated between language and birdsong: their syntax is hierarchical like language and not serial like the majority of birdsongs.
6. Writers and actors express these things in their own insightful manner. J.M. Coetzee describes in Disgrace a former professor of modern languages who has become, "since Classics and Modern Languages were closed down as part of the great rationalization, adjunct professor of communications." He finds preposterous the first premise of his new discipline, as set down in the college handbook: "Human society has created language in order that we may communicate our thoughts, feelings and intentions to each other" - believing that "the origin of speech lie in song, and the origins of song in the need to fill out with sound the overlarge and rather empty human soul" (Banville 2000:23). When asked by Charlie Rose what makes great literature, Patrick Stewart answered "Language" and added: "There must be something extra in the language that makes it sing."
7. I am grateful to Noam Chomsky for his comments on an earlier draft of this essay and for giving me access to a pre-publication copy of Chomsky (forthcoming).
Banville, John (2000), "Endgame," The New York Review of Books 47/4 (March 9).
Barkow, Jerome H., Leda Cosmides and John Tooby (1992), The Adapted Mind: Evolutionary Psychology and the Generation of Culture, New York - Oxford: Oxford University Press.
Barlow, Connie (1994), Evolution Extended: Biological Debates on the Meaning of Life, Cambridge,Mass and London,England: The MIT Press.
Basolo, A.L. (1990), "Female Preference Predates the Evolution of the Swordtail Fish," Science 250:808-10.
Bloomfield, L. (1939), Menomini Morphophonemics, Travaux du cercle linguistique de Prague, 8:105-115.
Bromberger, Sylvain and Morris Halle (1991), "Why Phonology is Different," in: Kasher (ed.): 56-77.
Cheney, Dorothy L. and Seyfarth, Robert M. (1985), "Vervet Monkey Alarm Calls: Manipulation through Shared Information?" Behaviour 93:150-166.
Cheney, Dorothy L. and Seyfarth, Robert M. (1990), How Monkeys See the World: Inside the Mind of Another Species, Chicago: The University of Chicago Press.
Cheney, Dorothy L. and Seyfarth, Robert M. (1992), "Meaning, Reference, and Intentionality Vocalizations of Monkeys," in Nishida et al.: 315-328.
Cheney, Dorothy L. and Seyfarth, Robert M. (eds.) (1992), "How Monkeys See the World," Behavior and Brain Sciences 15:135-182.
Chomsky, Noam (1966), Cartesian Linguistics. A Chapter in the History of Rationalist Thought, New York: Harper & Row.
Chomsky, Noam (1975), Reflections on Language, Pantheon
Chomsky, Noam (1980), Rules and Representations, New York: Columbia University Press.
Chomsky, Noam (1986), Knowledge of Language: Its Nature, Origin, and Use, New York, Wesport, Conn. and London: Praege.
Chomsky, Noam (1988), Language and the Problems of Knowledge,The Managua Lectures, Cambridge and London: M.I.T.
Chomsky, Noam (1995), The Minimalist Program, Cambridge, Mass. and London, England: The MIT Press.
Chomsky, Noam (forthcoming), "Language and the Brain" (Address at the European Conference on Cognitive Science, 27th-30th October, 1999, Siena).
Darwin, Charles (1859), On the Origin of Species, London: John Murray.
Darwin, Charles (1871), The Descent of Man and Selection in Relation to Sex, London: John Murray.
Dawkins, Richard (1986), The Blind Watchmaker. Why the evidence of evolution reveals a universe without design, New York - London: W.W.Norton & Co.
Eibl-Eibesfeldt, Irenaus(1971), Love and Hate. The Natural History of Behavior Patterns, New York, Chicago & San Francisco: Holt, Rinehart and Winston.
Eibl-Eibesfeldt, Irenaus (1989), Human Ethology, Aldine de Gruyter.
Evans, Christopher S., Evans, Linda and Marler, Peter (1993), "On the Meaning of Alarm Calls: Functional Reference in an Avian Vocal System," Animal Behaviour 46:23-38.
Fischer, R.A. (1930),The Genetic Theory of Natural Selection, New York: Dover Publications Inc.
Geertz, Clifford (1960), The Religion of Java, Illinois: The Free Press of Glencoe.
Graham, A.C. (1973), "China, Europe, and the Origins of Modern Science: Needham's The Grand Titration," in: Nakayama and Sivin: 45-69. Reprinted in Graham (1992):137-58.
Graham, A. C. (1992), Unreason within Reason: Essays on the Outskirts of Rationality, Lasalle,Ill: Open Court.
Griffin, Donald (1981), The Question of Animal Awareness. Evolutionary Continuity of Mental Experience, New York: Rockefeller University Press.
Griffin, Donald R. (1992), Animal Minds, Chicago and London: The University of Chicago Press.
Grimm, Jacob (1851), Ueber den Ursprung der Sprache. Reprint 1958.
Gune, Jayashree A. (1978), The Meaning of Tenses and Moods. The Text of Kaundabhatta's Lakararthanirnaya with Introduction, English Translation and Notes, Pune: Deccan College.
Hailman, J.P., M.S. Ficken and R.W. Ficken (1985), "The ‘chick-a-dee’ calls of Parus atricapillus: A recombinant system of animal communication compared with written English," Semiotica 56:191-224.
Hailman, J.P., M.S. Ficken and R.W. Ficken (1987), "Constraints on the structure of combinatorial ‘chick-a-dee’ calls," Ethology 75:62-80.
Hale, Kenneth L. (1966), "Kinship Reflections in Syntax: Some Australian Languages," Word 22:318-324.
Hauser, Marc D. (1996), The Evolution of Communication, Cambridge,Mass: The MIT Press.
Hauser, Marc D. and Marler, Peter (1993), "Food-associated Calls in Rhesus Macaques (Macaca mulatta)," I & II, Behavioral Ecology 4/2:194-212.
Hayashi, Takao (forthcoming), "Mathematics in the Vedic Texts," Enciclopedia Italiana Treccani.
von Humboldt, Wilhelm (1836), Uber die Verschiedenheit des menschlichen Sprachbaues und ihren Einfluss auf die geistige Entwicklung des Menschengeschlechts, Berlin: F. Dummler. Reprint 1935, Berlin: Lambert Schneider. English translation by Peter Heath (1988), Cambridge University Press.
Joseph, George Gheverghese (1990), The Crest of the Peacock: Non-European Roots of Mathematics, Penguin Books.
Kasher, Ada (ed.) (1991), The Chomskyan Turn, Cambridge Mass. And Oxford: Basil Blackwell.
Koerner, E. F. K. (1976), "Towards a Historiography of Linguistics," in Parret: 685-718.
Lettvin, J.Y., Maturana, H.R., McCulloch, W.S. and Pitts, W.H. (1959), "What the Frog's Eye tells the Frog's Brain," Proceedings of the IRE, 47 (No.11):1940-1959. Reprinted in: McCulloch 1965: 230-255.
Lorenz, Konrad (1963), On Aggression, San Diego, New York & London: Harcourt Brace Jovanovich.
Macedonia, Joseph M. and Evans, Christopher S. (1993), "Variation among Mammalian Alarm Call Systems and the Problem of Meaning in Animal Signals," Ethology 93:177-197.
Marler, Peter, Evans, Christopher S. and Hauser, Marc D. (1992), "Animal Signals: Motivational, Referential, or Both?" in Papousek: 66-86.
Mates, Benson (1986), The Philosophy of Leibniz. Metaphysics and Language, New York - Oxford: Oxford University Press.
McCulloch Warren S. (1965), Embodiments of Mind, Cambridge, Mass: The M.I.T. Press.
Nakayama, Shigeru and Sivin, Nathan (1973), Chinese Science. Explorations of an Ancient Tradition, Cambridge, Mass: The M.I.T. Press.
Nishida, Toshisada, McGrew, William C., Marler, Peter, Pickford, Martin and de Waal, Frans B.M. (1992), Topics in Primatology, I: Human Origins, Tokyo: University of Tokyo Press.
Papousek, Hanus, Jurgens, Uwe and Papousek, Machthikd (1992), Nonverbal Vocal Communication, Cambridge: The University Press and Paris: Editions du centre national de recherche scientifique.
Park, David (1988), The How and the Why. An Essay on the Origins and Development of Physical Theory, Princeton: Princeton University Press.
Parret, Herman (1976), History of Linguistic Thought and Contemporary Linguistics, Berlin: Walter de Gruyter.
Pinker, Steven (1994), The Language Instinct. How the Mind creates Language, New York: William Morrow & Co.
Pinker, S. and Bloom, P., eds. (1990), "Natural Language and Natural Selection," Behavioral and Brain Sciences 13:707-784. Reprinted in: Barkow, Cosmides and Tooby: 451-93.
Quine, W. V. (1981), Theories and Things, Cambridge, Mass.:Harvard University Press.
Reeve, Hudson Kern and Paul W. Sherman (1993), Adaptation and the Goals of Evolutionary Research," The Quarterly Review of Biology 8:1-32.
Robinson, J.G. (1984), "Syntactic structures in the vocalization of wedge-capped capuchin monkeys, Cebus nigrivittatus," Behaviour 90:46-79.
Ryan, Michael J. and A. Stanley Rand (1993), "Sexual Selection and Signal Evolution: The Ghost of Biases," Proceedings of the Royal Society, B 340:187-95.
Searcy, W.A. (1992), "Song Repertoire and Mate Choice in Birds," The American Zoologist 32:71-80.
Shepard, Roger (1982), "Perceptual and Analogical Bases of Cognition," in: Mehler, Garrett and Walker (eds.).
Snowdon, Charles T. (1990), "Language Capacities of Nonhuman Animals," Yearbook of Physical Anthropology 33:215-243.
Staal, J. F. (1965), "Euclid and Panini," Philosophy East and West 15: 99-116. Reprinted in: Staal 1988:143-60.
Staal, Frits (1985), "Mantras and Bird Songs," Journal of the American Oriental Society 105:549-58.
Staal, Frits (1988), Universals. Studies in Indian Logic and Linguistics, Chicago - London: The University of Chicago Press.
Staal, Frits (1989, 1993), Rules without Meaning. Ritual, Mantras and the Human Sciences, New York etc.: Peter Lang.
Staal, Frits (1995), "The Sanskrit of Science," Journal of Indian Philosophy 23: 73 – 127.
Symons, Donald (1992), "On the Use and Misuse of Darwinism in the Study of Human Behavior," in Barkow, Cosmides and Tooby: 137-159.
Tarski, Alfred (1935), "Der Wahrheitsbegriff in den formalisierten Sprachen," Studia philosophica 1:261-405.
Tarski, Alfred (1969), "Truth and Proof," Scientific American 220: 63-77.
Truesdell, C. (1968), Essays in the History of Mechanics, New York: Springer.
Rules Without Meaning : Ritual, Mantras and the Human Sciences
by Frits Staal
The Fidelity of Oral Tradition and the Origins of Science
by Frits Staal
Universals : Studies in Indian Logic and Linguistics
by Frits Staal
Lingua ex Machina: Reconciling Darwin and Chomsky With the Human Brain
by William H. Calvin, Derek Bickerton
New Horizons in the Study of Language and Mind
by Noam Chomsky, Neil Smith
Aspects of the Theory of Syntax
by Noam Chomsky
Language and Thought
by Noam Chomsky
On Language : Chomsky's Classic Works Language and Responsibility and Reflections on Language in One Volume
by Noam Chomsky