NATURAL LANGUAGE AND NATURAL SELECTION Pinker, S. & Bloom, P. (1990).
Comments from Ivan Zsuzsanna and Toth Eva
*Comment1: Beeing biologists we agree with the authors, that evolution,
and natural selection can be the only explanation for the development
of the language. We think, there is no point in disputing this.
*Comment2: It is not obvious, that the larger brain size is connected
to better communicaton skills. The H. S. neanderthalensis had bigger
brain than the modern Homo Sapiens, and nobody think that they were
better in communication than us.
*Comment3: A question connected to the process of language evolution:
If there is only a man with spontenous mutation, who is using a higher
grammatical step, it is not obviously an advatage, because the other
100 understands him more difficult, although he can express himself
better, and more clear.
*Comment4: We don't agree, and don't understand, why will have a mother
with "m" language rule a baby with "m+1" language rule?
p&b> Many people have argued that the evolution of the human language
p&b> faculty cannot be explained by Darwinian natural selection. Chomsky
p&b> and Gould have suggested that language may have evolved as the
p&b> by-product of selection for other abilities or as a consequence of
p&b> as-yet unknown laws of growth and form.
p&b>
p&b> Reviewing other arguments and data, we conclude that there is every
p&b> reason to believe that a specialization for grammar evolved by a
p&b> conventional neo-Darwinian process.
p&b>
p&b> All human societies have language. As far as we know they always
p&b> did;language was not invented by some groups and spread to others like
p&b> agriculture or the alphabet. All languages are complex computational
p&b> systems employing the same basic kinds of rules and representations,
p&b> with no notable correlation with technological progress: the
p&b> grammars of industrial societies are no more complex than the
p&b> grammars of hunter-gatherers; Modern English is not an advance over
p&b> Old English. Within societies, individual humans are proficient
p&b> language users regardless of intelligence, social status, or
p&b> level of education. Children are fluent speakers of complex
p&b> grammatical sentences by the age of three, without benefit of formal
p&b> instruction.
p&b>
p&b> This list of facts (see Pinker, 1989a) suggests that the ability to
p&b> use a natural language belongs more to the study of human biology than
p&b> human culture.
p&b>
p&b> It would be natural, then, to expect everyone to agree that human
p&b> language is the product of Darwinian natural selection. The only
p&b> successful account of the origin of complex biological structure is
p&b> the theory of natural selection, the view that the differential
p&b> reproductive success associated with heritable variation is the
p&b> primary organizing force in the evolution of organisms.
p&b>
p&b> But surprisingly, this conclusion is contentious. Noam Chomsky,
p&b> the world's best-known linguist, and Stephen Jay Gould, the world's
p&b> best-known evolutionary theorist, have repeatedly suggested that
p&b> language may not be the product of natural selection, but a side effect
p&b> of other evolutionary forces such as an increase in overall
p&b> brain size and constraints of as-yet unknown laws of structure and
p&b> growth.
p&b>
p&b> In this paper we will examine this position in detail, and will come
p&b> to a very different conclusion. We will argue that there is every
p&b> reason to believe that language has been shaped by natural selection
p&b> as it is understood within the orthodox "synthetic" or "neo-Darwinian"
p&b> theory of evolution.
p&b>
p&b> Evolutionary theory is informative about many things, but it has
p&b> little to say, as of now, of questions of this nature [e.g., the
p&b> evolution of language]. The answers may well lie not so much in the
p&b> theory of natural selection as in molecular biology, in the study of
p&b> what kinds of physical systems can develop under the conditions of life
p&b> on earth and why, ultimately because of physical principles.
p&b>
p&b> Since we are impressed both by the synthetic theory of evolution and
p&b> by the theory of generative grammar, we hope that we will not have
p&b> to choose between the two.
p&b>
p&b> In this paper, we first examine arguments from evolutionary biology
p&b> about when it is appropriate to invoke natural selection as an
p&b> explanation for the evolution of some trait. We then apply
p&b> these tests to the case of human language, and conclude that language
p&b> passes. We examine the motivations for the competing nonselectionist
p&b> position, and suggest that they have little to recommend them. In
p&b> the final section, we refute the arguments that have claimed
p&b> that an innate specialization for grammar is incompatible with the
p&b> tenets of a Darwinian account and thus that the two are incompatible.
p&b>
p&b> 2. The Role of Natural Selection in Evolutionary Theory
p&b>
p&b> 2.1. Nonselectionist Mechanisms of Evolutionary Change
p&b>
p&b> The argument is illustrated by an analogy with the mosaics on the
p&b> dome and spandrels of the San Marco basilica in Venice: Spandrels --
p&b> the tapering triangular spaces formed by the intersection of
p&b> two rounded arches at right angles ... are necessary architectural
p&b> by-products of mounting a dome on rounded arches. Each spandrel
p&b> contains a design admirably fitted into its tapering space. ... Anyone
p&b> who tried to argue that the structure [spandrels] exists because
p&b> of [the designs laid upon them] would be inviting the same ridicule
p&b> that Voltaire heaped on Dr. Pangloss: "Things cannot be other than
p&b> they are ... ... Yet evolutionary biologists, in their tendency
p&b> to focus exclusively on immediate adaptation to local
p&b> conditions, do tend to ignore architectural constraints and
p&b> perform just such an inversion of explanation.development. This
p&b> body of research, they suggest, is an antidote to the tendency to
p&b> treat an organism as a bundle of traits or parts, each
p&b> independently shaped by natural selection.
p&b>
p&b> 2.2. Limitations on Nonselectionist Explanations
p&b>
p&b> The key point that blunts the Gould and Lewontin critique of
p&b> adaptationism is that natural selection is the only scientific
p&b> explanation of adaptive complexity. "Adaptive complexity" describes
p&b> any system composed of many interacting parts where the details
p&b> of the parts' structure and arrangement suggest design to fulfill some
p&b> function.
p&b>
p&b> The vertebrate eye is the classic example. Darwin showed how such
p&b> "organs of extreme perfection and complication" could arise from the
p&b> purely physical process of natural selection.
p&b>
p&b> The essential point is that no physical process other than
p&b> natural selection can explain the evolution of an organ like the eye.
p&b> The reason for this is that structures that can do what the
p&b> eye does are extremely low- probability arrangements of matter.
p&b>
p&b> This is also true of the other nonselectionist mechanisms outlined
p&b> by Gould and Lewontin. It is absurdlyimprobable that some general
p&b> law of growth and form could give rise to a functioning
p&b> vertebrate eye as a by-product of some other trend such as an
p&b> increase in size of some other part.
p&b>
p&b> Likewise, one need not consider the possibility that some organ
p&b> that arose as an adaptation to some other task This argument is
p&b> obviously incomplete, as it relies on the somewhat intuitive notion
p&b> of "function" and "design."
p&b>
p&b> What, then, is the proper relation between selectionist
p&b> and nonselectionist explanations in evolution? The least interesting
p&b> case involves spandrels that are not involved in any function
p&b> or behavior, such as the redness of blood, the V-shaped space between a
p&b> pair of fingers, the hollow at the back of a knee, the fact there
p&b> are a prime number of digits on each limb, and so on. The mere presence
p&b> of these epiphenomenal spandrels, that play no direct role in
p&b> the explanation of any species-typical behavior or function, says
p&b> nothing about whether the structures that they are associated with
p&b> were shaped by selection. There are as many of them as there are ways
p&b> of describing an organism that do not correspond to its functional
p&b> parts.
p&b>
p&b> Gould (1987a) describes a kind of wading bird that uses its wings
p&b> primarily to block reflections on the surface of water while looking
p&b> for fish.
p&b> A wing used as a visor is a case where a structure designed for
p&b> a complex engineering task that most arrangements of matter do
p&b> not fulfill, such as controlled flight, is exapted to a simple
p&b> engineering task that many arrangements of matter do fulfill, such as
p&b> screening out. When the reverse happens, such as when a solar
p&b> heat exchanger is retooled as a fully functioning wing in the
p&b> evolution of insects (Kingsolver and Koehl, 1985), natural selection
p&b> must be the cause.
p&b>
p&b> 2.3. Two Issues that are Independent of Selectionism
p&b>
p&b> There are two other issues that Gould includes in his depiction
p&b> of a scientific revolution in evolutionary theory. It is important to
p&b> see that they are largely independent of the role of selection in
p&b> evolutionary change.
p&b>
p&b> 2.3.1. Gradualism
p&b>
p&b> According to the theory of "punctuated equilibrium" most
p&b> evolutionary change does not occur continuously within a
p&b> lineage, but is confined to bursts of change that are relatively brief
p&b> on the geological time scale, generally corresponding to
p&b> speciation events, followed by long periods of stasis.
p&b>
p&b> Once again the explanation of adaptive complexity is the key reason
p&b> why one should reject nongradual change as playing an important
p&b> role within evolution. An important Darwinian insight, reinforced by
p&b> Fisher (1930), is that the only way for complex design to evolve is
p&b> through a sequence of mutations with small effects. However there is a
p&b> clear sense in which such changes are still gradual, since they only
p&b> involve a gross modification or duplication of existing structure, not
p&b> the appearance of a new kind of structure.
p&b>
p&b> 2.3.2. Exaptation
p&b>
p&b> Exaptation is another process that is sometimes discussed as if it
p&b> was incompatible both with adaptationism and with gradualism. People
p&b> often wonder whether each of the "numerous, successive, slight
p&b> modifications" from an ancestor lacking an organ to a modern creature
p&b> enjoying the fully-functioning organ leads to an improvement in
p&b> the function, as it should if the necessary evolutionary sequence is to
p&b> be complete.
p&b>
p&b> Furthermore, it is crucial to understand that exaptation is merely
p&b> one empirical possibility, not a universal law of evolution. Gould is
p&b> often quoted as saying "We avoid the excellent question, What good
p&b> is 5 percent of an eye? Of course no ancestor to humans literally had 5
p&b> percent of a human eye; the expression refers to an eye that has
p&b> 5 percent of the complexity of a modern eye). In response,
p&b> Dawkins (1986: 81) writes: "An ancient animal with 5 per cent of an
p&b> eye might indeed have used it for something other than sight,
p&b> but it seems to me at least as likely that it used it for 5 per cent
p&b> vision. ... Vision that is 5 percent as good as yours or mine is
p&b> very much worth having in comparison with no vision at all.
p&b>
p&b> In sum, the positions of Gould, Lewontin, and Eldredge should not be
p&b> seen as radical revisions of the theory of evolution, but as a
p&b> shift in emphasis within the orthodox neo-Darwinian framework.
p&b>
p&b> 3. Design in Language
p&b>
p&b> 3.1. An Argument for Design in Language
p&b>
p&b> Humans acquire a great deal of information during their lifetimes.
p&b> Since this acquisition process occurs at a rate far exceeding that
p&b> of biological evolution, it is invaluable in dealing with
p&b> causal contingencies of the environment that change within a lifetime,
p&b> and provides a decisive advantage in competition with other species
p&b> that can only defend themselves against new threats in evolutionary
p&b> time.
p&b> There is an obvious advantage in being able to acquire such
p&b> information about the world second-hand: by tapping into the vast
p&b> reservoir of knowledge accumulated by some other individual, one
p&b> can avoid having to duplicate thepossibly time-consuming and dangerous
p&b> trial and error process that won that knowledge. This makes the
p&b> following kinds of contents as worthy of communication among humans.
p&b> We would want to be able to refer to individuals and classes, to
p&b> distinguish among basic ontological categories (things, events,
p&b> places, times, manners, and so on), to talk about events and
p&b> states, distinguishing the participants in the event or state
p&b> according to role (agents, patients, goals), and to talk about the
p&b> intentional states of ourselves and others. Also, we would want
p&b> the ability to express distinctions of truth value, modality
p&b> (necessity, possibility, probability, factivity), to comment on the
p&b> time of an event or state including both its distribution
p&b> over time...
p&b>
p&b> Thus grammars for spoken languages must map propositional structures
p&b> onto a serial channel, minimizing ambiguity in context, under
p&b> the further constraints that the encoding and decoding be done
p&b> rapidly, by creatures with limited short-term memories, according
p&b> to a code that is shared by an entire community of potential
p&b> communicants.
p&b>
p&b> Let us list some uncontroversial facts about substantive
p&b> universals, the building blocks of grammars that all theories of
p&b> universal grammar posit, either as an explicit inventory or as a
p&b> consequence of somewhat more abstract mechanisms. -Grammars are built
p&b> around symbols for major lexical categories (noun, verb,
p&b> adjective, preposition) -Major phrasal categories (noun phrase, verb
p&b> phrase, etc.) start off with a major lexical item -Phrase
p&b> structure rules -Rules of linear order -Case affixes -Verb affixes
p&b> -Auxiliaries
p&b>
p&b> ...And this is only a partial list, focusing on sheer expressive power
p&b> As we write these words, we can hear the
p&b> protests: "Pangloss! Just-so stories!" Haven't we just thought up
p&b> accounts about functions post hoc after examining the structure?
p&b> How do we know that the neural mechanisms were not there for other
p&b> reasons, and that once they were there they were just put to
p&b> various convenient uses by the first language users, who then
p&b> conveyed their invention to subsequent generations?
p&b>
p&b> 3.2. Is the Argument for Language Design a Just-So Story?
p&b>
p&b> Human language is a device capable of communicating exquisitely
p&b> complex and subtle messages, from convoluted soap opera plots to
p&b> theories of the origin of the universe.
p&b>
p&b> Similarly, the facts of grammar make it difficult to argue that
p&b> language shows design for "the expression of thought" in any sense
p&b> that is substantially distinct from "communication." If "expression"
p&b> refers to the mere externalization of thoughts, in some kind of
p&b> monologue or soliloquy, it is an unexplained fact that language
p&b> contains mechanisms that presuppose the existence of a listener...
p&b>
p&b> ...Grammar is a notoriously poor medium for conveying subtle patterns
p&b> of emotion, for example, and facial
p&b> expressions and tones of voice are more informative (Ekman and
p&b> Friesen, 1975; Etcoff, 1986). Although grammars provide devices for
p&b> conveying rough topological information such as connectivity,
p&b> contact, and containment, and coarse metric contrasts such as near/far
p&b> or flat/globular...
p&b>
p&b> For example, even the artificial languages that are focussed on
p&b> very narrow domains of content and that are not meant to be used
p&b> in a natural on-line manner by people, such as computer languages
p&b> or symbolic logic, show certain obvious parallels with aspects of human
p&b> grammar. They have needed means of distinguishing types of symbols,
p&b> predicate argument relations, embedding, scope, quantification, and
p&b> truth relations, and solve these problems with formal syntactic
p&b> systems that specify arbitrary patterns of hierarchical
p&b> concatenation, relative linear order, fixed positions within
p&b> strings, and closed classes of privileged symbols.
p&b>
p&b> 3.3. Language Design and Language Diversity
p&b>
p&b> A more serious challenge to the claim that grammars show evidence of
p&b> good design may come from the diversity of human languages.
p&b> Grammatical devices and expressive functions do not pair up in
p&b> one-to-one fashion.
p&b>
p&b> 3.4. Language Design and Arbitrariness
p&b>
p&b> Piattelli-Palmarini (1989) presents a different kind of argument:
p&b> grammar is not completely predictable as an adaptation to
p&b> communication, therefore it lacks design and did not evolve by
p&b> selection. Frequently cited examples of arbitrary phenomena in
p&b> language include constraints on movement (such as subjacency),
p&b> irregular morphology, and lexical differences in predicate-argument
p&b> structure.
p&b>
p&b> 3.4.1. Inherent Tradeoffs
p&b>
p&b> For example, there is a conflict of interest between speaker and
p&b> hearer. Speakers want to minimize articulatory effort and hence tend
p&b> towards brevity and phonological reduction. Hearers want to
p&b> minimize the effort of understanding and hence desire explicitness
p&b> and clarity. This conflict of interest is inherent to the
p&b> communication process and operates at many levels.
p&b>
p&b> 3.4.3. Arbitrariness and the Relation Between Language Evolution and
p&b> Language
p&b>
p&b> ...When children say breaked and comed, they are using a system that is
p&b> far simpler and more logical than the adult
p&b> combination of a regular rule... There is no deficit in
p&b> comprehensibility; the meaning of comed is perfectly clear. In fact
p&b> the child's system has greater expressive power that the
p&b> adult's. When children say hitted and cutted, they are distinguishing
p&b> between past and nonpast forms in a manner that is
p&b> unavailable to adults, who must use hit and cut across the board.
p&b> Why do children eventually abandon this simple, logical, expressive
p&b> system? They must be programmed so that the mere requirement of
p&b> conformity to the adult code, as subtle and arbitrary as it is, wins
p&b> over other desiderata. The requirement that a communicative code
p&b> have an innate arbitrary foundation ("universal grammar," in the
p&b> case of humans) may have analogues elsewhere in biology.
p&b>
p&b> 4. Arguments for Language Being a Spandrel
p&b>
p&b> 4.1. The Mind as a Multipurpose Learning Device
p&b>
p&b> The main motivation for Gould's specific suggestion that language
p&b> is a spandrel is his frequently-stated position that the mind
p&b> is a single general-purpose computer.
p&b>
p&b> The analogy is somewhat misleading, someone has to reprogram it
p&b> first. Language learning is not programming: We suggest that natural
p&b> selection was the programmer.
p&b>
p&b> The gross facts about the dissociability of language and
p&b> other learned cultural systems, belie the suggestion that language
p&b> is a spandrel of any general cognitive learning ability.
p&b>
p&b> 4.2. Constraints on Possible Forms
p&b>
p&b> The theory that the mind is an all-purpose learning device is of
p&b> course anathema to Chomsky. Chomsky, is in the Continental tradition
p&b> of trying to explain evolution by structural laws constraining
p&b> possible organic forms.
p&b>
p&b> ...perhaps these are simply emergent physical properties of a brain
p&b> that reaches a certain level of complexity under
p&b> the specific conditions of human evolution.
p&b>
p&b> Although Chomsky does not literally argue for any specific
p&b> evolutionary hypothesis, he repeatedly urges us to consider "physical
p&b> laws" as possible alternatives to natural selection. Of course
p&b> human brains obey the laws of physics, and always did, but that does
p&b> not mean that their specific structure can be explained by such
p&b> laws.
p&b>
p&b> Changes in brain quantity could lead to changes in brain quality. But
p&b> mere largeness of brain is neither a necessary nor a sufficient
p&b> condition for language... Nor is there reason to think that if you
p&b> simply pile more and more neurons into a circuit or more and more
p&b> circuits into a brain that computationally interesting abilities
p&b> would just emerge. Neural network modeling efforts have suggested that
p&b> complex computational abilities require either extrinsically imposed
p&b> design or numerous richly structured inputs during learning or
p&b> both...
p&b>
p&b> ...the most likely explanation for the complex structure of the
p&b> language faculty is that it is a design imposed on
p&b> neural circuitry as a response to evolutionary pressures.
p&b>
p&b> 5. The Process of Language Evolution
p&b>
p&b> For universal grammar to have evolved by Darwinian natural selection,
p&b> it is not enough that it be useful in some general sense. There
p&b> must have been genetic variation among individuals in their grammatical
p&b> competence. There must have been a series of steps leading from no
p&b> language at all to language as we now find it, each step small enough
p&b> to have been produced by a random muttion or recombination, and each
p&b> intermediate grammar useful to its possessor. Every detail of
p&b> grammatical competence that we wish to ascribe to selection must have
p&b> conferred a reproductive advantage on its speakers, and this advantage
p&b> must be large enough to have become fixed in the ancestral population.
p&b> And there must be enough evolutionary time and genomic space
p&b> separating our species from nonlinguistic primate ancestors. ...there
p&b> is no reason to think that every aspect of grammar that has a
p&b> genetic basis must be controlled by a single gene.
p&b>
p&b> 5.2. Intermediate Steps
p&b>
p&b> The intermediate links, it has been suggested, would not have been
p&b> viable communication systems. These arguments fall into three classes
p&b>
p&b> 5.2.1. Nonshared Innovations.
p&b>
p&b> ...how a hypothetical "beneficial" grammatical mutation could
p&b> really have benefited its possessor given that none
p&b> of the person's less evolved compatriots could have understood him
p&b> or her. One possible answer is that any such mutation is likely to
p&b> be shared by individuals who are genetically-related. Since much
p&b> communication is among kin, a linguistic mutant will be
p&b> understood by some of his or her relatives and the resulting
p&b> enhancements in information sharing will benefit each one of them
p&b> relative to others who are not related.
p&b>
p&b> 5.2.2. Categorical Rules.
p&b>
p&b> How could such structures evolve in a gradual sequence? ..."5% of an
p&b> eye...
p&b>
p&b> What could it conceivably mean for an organism to possess half a
p&b> symbol, or three quarters of a rule? (p. 3)
p&b> ...monadic symbols, absolute rules and modular systems must be acquired
p&b> as a whole, on a yes-or-no basis -- a
p&b> process that cries out for a Creationist explanation. (p. 30)" No
p&b> single mutation or recombination could have led to an entire
p&b> universal grammar, but it could have led a parent with an n-rule
p&b> grammar to have an offspring with an n+1 rule grammar, or a parent
p&b> with an m-symbol rule to have an offspring with an m+1 symbol rule.
p&b> It could also lead to a parent with no grammatical rules at all and
p&b> just rote associations to have an offspring with a single rule.
p&b>
p&b> 5.3. Reproductive Advantages of Better Grammars
p&b>
p&b> 5.3.2. Grammatical complexity and technology.
p&b>
p&b> Children can learn from a parent that a food is poisonous or a
p&b> particular animal is dangerous; they do not have to observe or
p&b> experience this by themselves.
p&b> ...It makes a difference whether that region has animals that you can
p&b> eat or animals that can eat you...
p&b>
p&b> 5.3.3. Grammatical complexity and social interactions.
p&b>
p&b> Humans everywhere depend on cooperative efforts for survival.
p&b> Conflicts within the group are resolved by talking, sometimes
p&b> half or all the night, for nights, weeks on end.
p&b>
p&b> 5.3.4. Social use of language and evolutionary acceleration.
p&b>
p&b> The social value of complex language probably played a profound role
p&b> in human evolution that is best appreciated by examining the
p&b> dynamics of cooperative interactions among individuals.
p&b>
p&b> But this is only a beginning. Cooperation opens the door to advances
p&b> in the ability of cheaters to fool people into believing that they
p&b> have paid a cost or that they have not taken a benefit. This in
p&b> turn puts pressure on the ability to detect subtle signs of such
p&b> cheating, which puts pressure on the ability to cheat in less
p&b> detectable ways, and so on.
p&b>
p&b> In all cultures human interactions are mediated by attempts at
p&b> persuasion and argument.
p&b>
p&b> 5.4. Phyletic Continuity
p&b>
p&b> Bates et al. (1989), Greenfield (1988), Lieberman (1976, 1984) argue
p&b> that if language evolved in humans by natural selection, it must have
p&b> antecedents in closely-related species such as chimpanzees, which
p&b> share 99% of their genetic material with us and may have diverged from
p&b> a common ancestor as recently as 5-7 million years ago. ...natural
p&b> language represents a discontinuity from other primate abilities
p&b> and so could not have evolved by natural selection.
p&b>
p&b> In fact there is even more scope for design differences than the
p&b> gross amount of nonshared genetic material suggests. The 1% difference
p&b> between chimps and humans represents the fraction of base pairs that
p&b> are different. But genes are long stretches of base pairs and if even
p&b> one pair is different, the entire functioning product of that gene
p&b> could be different. Just as replacing one bit in every byte leads to
p&b> text that is 100% different, not 12.5% different, it is possible for
p&b> the differing base pairs to be apportioned so that 100% of the
p&b> genes of humans and chimps are different in function. Though this
p&b> extreme possibility is, of course, unlikely, it warns us not to draw
p&b> any conclusions about phenotypic similarity from degree of genomic
p&b> overlap.(Note 4)
p&b>
p&b> 6. Conclusion
p&b>
p&b> As we warned, the thrust of this paper has been entirely
p&b> conventional. All we have argued is that human language, like other
p&b> specialized biological systems, evolved by natural selection. Our
p&b> conclusion is based on two facts that we would think would be entirely
p&b> uncontroversial: language shows signs of complex design for the
p&b> communication of propositional structures, and the only explanation for
p&b> the origin of organs with complex design is the process of
p&b> natural selection. Although distinguished scientists from a wide
p&b> variety of fields and ideologies have tried to cast doubt on an
p&b> orthodox Darwinian account of the evolution of a biological
p&b> specialization for grammar, upon close examination none of the
p&b> arguments is compelling.
This archive was generated by hypermail 2b30 : Tue Feb 13 2001 - 16:23:06 GMT