Click to Visit Click to Visit
Nature Japan
Subscribe  
Register  
 


 

 

 

 

 


   


The evolution of syntactic communication

Nature 404, 495 - 498 (2000)

Japanese summary (日本語要約)
進化:統語的コミュニケーションの進化

動物のコミュニケーションは普通、非統語的であり、信号は状況全体をさすことになる。ヒトの言語は統語的であり、信号はそれ独自の意味をもつ別々の要素からなる。統語法は、組合せ論の長所を利用する、つまり「限定された意味を限定されない方法で使う」ために必要である。ヒトの言語がもつきわめて高い表現能力は、統語法なしには不可能だろうし、非統語的コミュニケーションから統語的コミュニケーションへの遷移は、ヒト言語の進化に不可欠な段階だった。我々は、この遷移の進化的な動態を解明し、また自然選択がこれをどのように導きうるかを解析することをめざしている。今回、言語進化の個体群動態のモデルを提示し、言葉の基本的増殖速度を明らかにし、語彙の最大の大きさを算定した。統語法によって、レパートリーは拡大され、事前に学習しなかった情報を定式化する可能性が高まる。それでも、我々のモデルでは、もし必要な信号の数が閾値を超えた場合、自然選択は統語法の出現に有利にしか働きえない。この結果から、なぜヒトだけが統語的コミュニケーションすなわち複雑な言語を進化させたかを説明づけられるかもしれない。

MARTIN A. NOWAK*, JOSHUA B. PLOTKIN* & VINCENT A. A. JANSEN-

* Institute for Advanced Study, Princeton, New Jersey 08540, USA
- School of Biological Sciences, Royal Holloway, University of London, Egham Surrey, TW20 0EX UK

Correspondence and requests for materials should be addressed to M.A.N. (e-mail: nowak@ias.edu).


Animal communication is typically non-syntactic, which means that signals refer to whole situations1-7. Human language is syntactic, and signals consist of discrete components that have their own meaning8. Syntax is a prerequisite for taking advantage of combinatorics, that is, "making infinite use of finite means"9-11. The vast expressive power of human language would be impossible without syntax, and the transition from non-syntactic to syntactic communication was an essential step in the evolution of human language12-16. We aim to understand the evolutionary dynamics of this transition and to analyse how natural selection can guide it. Here we present a model for the population dynamics of language evolution, define the basic reproductive ratio of words and calculate the maximum size of a lexicon. Syntax allows larger repertoires and the possibility to formulate messages that have not been learned beforehand. Nevertheless, according to our model natural selection can only favour the emergence of syntax if the number of required signals exceeds a threshold value. This result might explain why only humans evolved syntactic communication and hence complex language.

The uniqueness of language has been compared to that of the elephant's trunk13. Human language is as different from animal communication as the elephant's trunk is from other animals' nostrils. Yet few biologists worry about the evolution of the elephant's trunk (which is a most complex organ that consists of about 6,000 individual muscles and that can perform an unparalleled variety of mechanical tasks), whereas many philosophers, linguists and biologists have great difficulties in imagining how language could have arisen by darwinian evolution17-21.

A challenge for evolutionary biology, therefore, is to provide a detailed mathematical account of how natural selection can enable the emergence of human language from animal communication. Animal communication is based on three basic designs: a finite repertoire of calls (territorial calls or warning of predators); a continuous analogue signal (for example, the dance of bees); and a series of random variations on a theme (such as the song of birds). All natural animal communication appears to be non-syntactic; some caution, however, seems appropriate as the final verdict on complex vocalization patterns of certain primate species or dolphins has not been reached. In contrast, human language is clearly syntactic: messages consist of components that have their own meaning. We compare non-syntactic and syntactic communication and evaluate their relative performance in an evolutionary setting.

First, we formulate a mathematical model for the population dynamics of language evolution. Suppose a language contains n words. Each individual is born not knowing any of the words, but can acquire words by learning from other individuals. Individuals are characterized by the subset of words that they know. The general equations for the resulting evolutionary dynamics are complicated (see Methods), but an analytical approach is possible if we describe the process in terms of independent, elementary steps on the basis of two assumptions: first, in any one interaction between two individuals only a single new word can be learned; second, words are memorized independently of each other. With these assumptions, we obtain for the population dynamics of xi, which is the relative abundance of individuals who know word Wi

where i = 1,..., n. The abundance of word Wi spreads by the interaction of people who know the word with people who do not know the word; hence its rate of increase is proportional to the product xi(1 - xi). The rate constant, Ri = bqphi i, is the basic reproductive ratio of word W i. This is the average number of individuals who acquire word Wi from one individual who knows it. The parameter b is the total number of word-learning events per individual per lifetime. The parameter q is the probability of memorizing a single word after one encounter, and phii is the frequency of occurrence of word Wi in the (spoken) language. The term - xi denotes a constant death rate, setting the average lifetime of each individual as one time unit.

For a word to be maintained in the lexicon of the population, its basic reproductive ratio must exceed one, which implies that phi i > 1/(bq). Suppose Wi is the least frequent word. We certainly know that phi i is less than 1/n, which is the frequency of a word if all words have the same frequency. Thus the maximum number of words is nmax = bq. Note that this number is always less than the total number of words, b, that are presented to a learning individual. Hence, the lexicon of the population cannot exceed the total number of word-learning events for each individual.

Assuming that word frequency distributions follow Zipf's law22, 23, phii = C/i, where C is a constant, we find that the maximum number of words is roughly given by the equation nmaxln(n max) = bq (Fig. 1).

Figure 1 How many word-learning events per individual are required for a population to maintain a certain number of words in its combined lexicon assuming that word frequencies follow Zipf's law?
High resolution image and legend (7k)

We now use this mathematical framework to analyse how natural selection can guide the transition from non-syntactic to syntactic communication. Imagine a group of individuals who communicate about events in the world around them. Events are combinations of objects, places, times and actions. (We use 'object' and 'action' in a general way to represent everything that can be referred to by nouns and verbs of current human languages.) For notational simplicity, suppose that each event consists of one object and one action. Thus, event Eij consists of object i and action j. Non-syntactic communication uses words for events, whereas syntactic communication uses words for objects and actions (Fig. 2). Events occur at different rates, which are specified by the entries of an 'event rate matrix', Gamma.

Figure 2 To understand the essence of the evolution of syntax, we imagine a world where each event consists of one object and one action.
High resolution image and legend (8k)

For natural selection to operate on language design, language must confer fitness. A plausible assumption is that correct communication about events provides a fitness advantage to the interacting individuals. In terms of our model, the fitness contribution of a language can be formulated as the probability that two individuals know the correct word for a given event summed over all events and weighted with the rate of occurrence of these events.

The population dynamics of non-syntactic communication are again given by equation (1) with word Wij referring to event Eij. As before, the maximum number of words that can be maintained in the population is limited by bq. We calculate the fitness of individuals using non-syntactic communication (see Methods).

We now turn to syntactic communication. Noun Ni refers to object i and verb Vj refers to action j, hence the event Eij is described by the sentence NiVj. For the basic reproductive ratios we obtain R(Ni) = (b/2)qsphi(N i) and R(V j) = (b/2)qsphi( Vj). Here phi(N i) and phi(Vj) denote the frequency of occurrence of noun Ni and verb V j, respectively. The factor 1/2 appears because either the noun or the verb is learned in any one of the b learning events. The probability of memorizing a noun or a verb is given by qs. We expect qs to be smaller than q, which simply means that it is a more difficult task to learn a syntactic signal than a non-syntactic signal. For both signals, the (arbitrary) meaning has to be memorized; for a syntactic signal one must also memorize its relation to other signals (whether it is a noun or a verb, for example).

For noun Ni to be maintained in the lexicon of the population, its basic reproductive ratio must exceed one, implying that phi(Ni) > 2/(bq s). Similarly, for verb Vj we find phi(Vj) > 2/(bq s). This means that the total number of nouns plus verbs is limited by bqs, which is always less than b. The maximum number of grammatical sentences, however, which consist of one noun and one verb, is given by (bqs)2/4. Hence syntax makes it possible to maintain more sentences than the total number of sentences, b, encountered by a learning individual. All words have to be learned, therefore, but syntactic signals enable the formulation of 'new' sentences that have not been learned beforehand.

For calculating the fitness of syntactic communication, note that two randomly chosen individuals can communicate about event Eij if they both know noun Ni and verb V j. If we denote the relative abundance of individuals who know Ni and Vj by x(N iVj), we obtain

The abundance of individuals who know noun Ni and verb Vj increases if someone who knows N i meets someone who knows Vj but not Ni. Similarly, the abundance increases if someone who knows verb Vj meets someone who knows N i but not Vj. We calculate the equilibrium abundances and thence the fitness of individuals using syntactic communication (see Methods). Figure 3 shows the fitness of syntactic and non-syntactic communication as a function of b for different examples of the event rate matrix, Gamma.

Figure 3 The fitness of non-syntactic and syntactic communication, F n and Fs, as function of the total number of word learning events per individual, b, for three different choices of the event rate matrix, Gamma.
High resolution image and legend (24k)

When does syntactic communication lead to a higher fitness than non-syntactic communication? Suppose there are n objects and m actions. Suppose a fraction, p, of these mn events occur (all at the same frequency), and the other events do not occur. In this case, R( Wij) = bq/( pmn) for those events that occur. Making the (somewhat rough) assumption that all nouns and all verbs, respectively, occur on average at the same frequency, we obtain R(N i) = bqs/(2n) and R(Vj) = bqs/(2m). If all involved basic reproductive ratios are well above one, the fitness of syntactic communication exceeds the fitness of non-syntactic communication provided (m 2n + mn2)/ (m2 + mn + n 2) > (2q)/(pqs) (see Methods). If this inequality holds then syntactic communication will be favoured by natural selection; otherwise non-syntactic communication will win. Observe that m 3 and n 3 are necessary conditions for the evolution of syntax. For m = n, the relevant condition is

Hence, the size of the system has to exceed a critical value for syntactic communication to evolve. For example, if it is twice as hard to memorize a syntactic signal than a non-syntactic signal, q/q s = 1/2, and if a fraction p = 1/3 of all noun verb combinations describe meaningful events, then at least an 18 times 18 system is required for syntactic communication to have any chance of evolving. Figure 4 shows the excellent agreement between our approximative analytical results and exact numerical computations.

Figure 4 Numerical validation of the approximate threshold condition given by equation (3).
High resolution image and legend (10k)

The parameter p quantifies to what extent the perceived world has a compositional structure. A small p means that events often consist of unique pairings of objects and actions. The smaller the value of p, the harder it is for syntactic communication to evolve (the critical n in equation (3) is large).

Our results suggest that the crucial step that guided the transition from non-syntactic to syntactic communication was an increase in the number of relevant events that could be referred to. 'Relevant event' means there is a fitness contribution for communication about this event. As the number of such 'relevant communication topics' increased, natural selection could begin to favour syntactic communication and thereby lead to a language design where messages could be formulated that were not learned beforehand. Syntactic messages can encode new ideas or refer to extremely rare but important events. Our theory, however, does not suggest that syntactic communication is always at an advantage. Many animal species probably have a syntactic understanding of the world, but natural selection did not produce a syntactic communication system for these species because the number of relevant signals was below the threshold illustrated by equation (3) . Presumably the increase in the number of relevant communication topics was caused by changes in the social structure24 and interaction of those human ancestors who evolved syntactic communication.

Methods
Suppose there are n words. Individuals are characterized by the subset of words they know. There are 2n possibilities for the internal lexicon of an individual. Internal lexica are defined by bit strings: 1 means that the corresponding word is known; 0 means it is not. Let us enumerate them by I = 0,..., nu where nu = 2n - 1. The number I is the integer representation of the corresponding bit string. (For example, I = 6 represents the string 000...0110.) Denote by x I the abundance of individuals with internal lexicon I. The population dynamics can be formulated as

where I = 0,..., nu. We have delta 0 = 1 and deltaI = 0 otherwise; thus all individuals are born not knowing any of the words. Individuals die at a constant rate, which we set to 1, thereby defining a time scale. The quantities QIJK denote the probabilities that individual I learning from J will become K. Equation (4) is a general framework for the population dynamics of the lexical aspects of language. Assuming that in any one interaction between two individuals only a single new word can be acquired and that words are memorized independently of each other, we obtain the specific system described by equation (1).

Let us now assume that the world is made up of events E ij consisting of objects i and actions j. The 'event rate matrix', Gamma has the entries gammaij which specify the relative rate of occurrence of event Eij. Denote by phiij the frequency of occurrence of event Eij. We have phiij = gammaij/Sigmaijgamma ij.

Non-syntactic communication uses words, Wij, for events Eij. The basic reproductive ratio of W ij is given by R(W ij) = bqphiij. If R(Wij) > 1, the word Wij will persist in the population, and at equilibrium the relative abundance of individuals who know this word is given by x*(Wij) = 1 - 1/R(Wij) .

The fitness contribution of a language can be formulated as the probability that two individuals know the correct word for a given event summed over all events and weighted with the rate of occurrence of these events. Hence, at equilibrium, the fitness of individuals using non-syntactic communication is given by

For syntactic communication, we assume the event Eij is described by the sentence NiVj . The population dynamics of individuals knowing both N i and Vj are described by equation (2). The basic reproductive ratios are given by R(Ni) = (b/2) qsphi(Ni) and R(Vj) = (b/2)qsphi(V i). The frequencies of occurrence are phi( Ni) = Sigmajphi ij and phi(Vj) = Sigmaiphiij. If the basic reproductive ratios, R(Ni) and R(Vj), are greater than one, the equilibrium frequency of individuals who know both Ni and Vj is given by

At equilibrium, the fitness of syntactic communication is given by

Assuming there are n objects and m actions that give rise to pnm meaningful events that all occur at the same frequency, we obtain R(Wij) = bq /(pnm). If we also assume that the combinations of objects and actions are arranged in a way that all nouns and all verbs, respectively, have about the same frequency, we can write R( Ni) = bq2/(2 n) and R(V j) = bqs/(2m) . For the fitness values, we obtain Fn = pnm[1 - 1/R(W ij)]2 and Fs = pnm[1 - 1/R(N i)]2[1 - 1/R(V j)]2/[1 - 1/(R(N i) + R(Vj))] 2. Assuming that all involved basic reproductive ratios are well above one, we obtain for Fs < F n the condition (m2 n + mn2)/(m2 + mn + n2) > (2q)/( pqs). Defining the ratio, alpha = m/n , we can rewrite this condition as

Hence the size of the system has to exceed a critical threshold for syntax to be favoured by natural selection.

Received 1 November 1999;accepted 26 January 2000

------------------

References
1. Von Frisch, K. The Dance Language and Orientation of Bees (Harvard Univ. Press, Cambridge, Massachusetts, 1967).
2. Marler, P. Birdsong and speech development: could there be parallels? Am. Sci. 58, 669-673 (1970). | PubMed | ISI |
3. Wilson, E. O. Animal communication. Sci. Am. 227, 52-60 (1972). | PubMed | ISI |
4. Gould, J. L. & Marler, P. Learning by instinct. Sci. Am. 256, 74-85 (1987). | ISI |
5. Burling, R. Primate calls, human language, and nonverbal communication. Curr. Anthropol. 34, 25-53 (1989).
6. Cheney, D. L. & Seyfarth, R. M. How Monkeys See the World: Inside the Mind of Another Species (Chicago Univ. Press, 1990).
7. Hauser, M. The Evolution of Communication (MIT Press, Cambridge, Massachusetts, 1996).
8. Bickerton, D. Species and Language (Chicago Univ. Press, Chicago, 1990).
9. von Humboldt, W. Linguistic Variability and Intellectual Development (Pennsylvania Univ. Press, Philadelphia, 1972).
10. Chomsky, N. Aspects of the Theory of Syntax (MIT Press, Cambridge, Massachusetts, 1965).
11. Jackendoff, R. The Architecture of the Language Faculty (MIT Press, Cambridge, Massachusetts, 1997).
12. Pinker, S. & Bloom, P. Natural language and natural selection. Behav. Brain Sci. 13, 707-784 (1990). | ISI |
13. Pinker, S. The Language Instinct. (Harper Collins, New York, 1994).
14. Maynard Smith, J. & Szathmary, E. The Major Transitions in Evolution (Freeman, Oxford, 1995).
15. Hurford, J. R., Studdert-Kennedy, M. & Knight, C. Approaches to the Evolution of Language (Cambridge Univ. Press, Cambridge, UK, 1998).
16. Nowak, M. A. & Krakauer, D. C. The evolution of language. Proc. Natl Acad. Sci. USA 96, 8028-8033 (1999). | Article | PubMed | ISI |
17. Chomsky, N. Language and Mind (Harcourt Brace Jovanovich, New York, 1972).
18. Chomsky, N. Language and Problems of Knowledge: The Managua Lectures (MIT Press, Cambridge, Masachusetts, 1988).
19. Premack, D. Gavagai! or the future history of the animal language controversy. Cognition 19, 207-296 (1985). | PubMed | ISI |
20. Lieberman, P. The Biology and Evolution of Language (Harvard Univ. Press, Cambridge, Massachusetts, 1984).
21. Bates, E., Thal, D. & Marchman, V. in Biological and Behavioural Determinants of Language Development (eds Krasnegor et al.) (Erlbaum, Mahwah, NJ, 1991).
22. Zipf, G. K. The Psychobiology of Language (Houghton-Mifflin, Boston, 1935).
23. Miller, G. A. & Chomsky, N. Handbook of Mathematical Psychology. Vol. 2 (eds Luce, R. D., Bush, R. & Galauter, E.) 419-491 (Wiley, New York, 1963).
24. Dunbar, R. Grooming, Gossip and the Evolution of Language (Harvard Univ. Press, Cambridge, Massachusetts, 1996).

Acknowledgements. This work was supported by the Leon Levy and Shelby White Initiatives Fund, the Florence Gould Foundation, the J. Seward Johnson Sr Charitable Trusts, the Ambrose Monell Foundation, the National Science Foundation, the Wellcome Trust and the Alfred P. Sloan Foundation.

c Nature Japan K.K. 2002  掲載記事の無断転載を禁じます。

TITLE:Nature Japan
DATE:2002/09/12 21:49
URL:http://www.natureasia.com/japan/nature/chinosouzou3/13nowak/paper.php