The Tower of Babel by Pieter Bruegel the Elder (1563)

Tower of Babel (Book Review)

This article first appeared in The Skeptic. Vol.20, No.1, pp.54-57.

Reproduced with permission

Comment on the linguistic discussion in Pennock R.T. (1999) Tower Of Babel: The Evidence Against The New Creationism, MIT Press, Cambridge (Mass.)

Robert Pennock is a philosopher of science at the University of Texas in Austin. His book constitutes a timely rebuttal of the arguments of the newly arisen cluster of relatively sophisticated (pseudo-)scientific creationists (‘intelligent design theory’ advocates) such as Michael Behe and Phillip Johnson, who have persuaded many that there is after all a scientific case for a literal interpretation of Genesis.

Pennock’s book is thorough, scholarly and insightful, his presentation of the issues is clear, and he is particularly impressive (as one might expect of a philosopher) in drawing fine but crucial distinctions between claims, standpoints and arguments. His book has been challenged (not always very expertly) on various philosophical points and has inevitably received hostile reviews from creationists; but he has rebutted these objections extremely effectively. Here I take the philosophical strengths of the book as given, and focus on a somewhat novel aspect of Pennock’s work: the extensive use which he makes in this context of the facts of human language and in particular of the nature of linguistic change. Pennock’s title suggests a linguistic focus, and in Chapter 3 (pp 117-179), and sporadically elsewhere, he discusses the creationist-evolutionist debate in this specific context rather than in the more familiar context of biological evolution. This, he says, is partly because the issue of language change is less ‘charged’ (which is certainly true), and also partly because Genesis contains the well-known story of the diversification of languages at the Tower of Babel, which most creationists naturally interpret as literally true but which no mainstream linguist would take seriously (for obvious reasons).

The analogy between linguistic change and biological evolution is not, of course, entirely precise, and readers of Pennock must bear this in mind, since this is a point which a casual reader might fail to grasp. Indeed, one of the few criticisms which can reasonably be made of the book is that Pennock – who is, in general terms, well aware of most of the differences between biological and linguistic change – is not always quite explicit enough on such points. And in a few places, notably where he exemplifies adaptive linguistic changes from vocabulary borrowing (pp 140-141), it might be thought that he has chosen examples of a type which suits his case considerably better than some other phenomena would – although in fact I do not believe that this was deliberate. However, these are small blemishes, if such they are, and a careful, well-informed reader will cope with them. Certainly Pennock’s analogising argument from linguistic change, at the level at which it is intended to be understood, is just as persuasive against creationism as are similar arguments involving biological data. And, as we shall see, some of the differences between linguistic and biological evolution are not as great as might at first be imagined.

Nevertheless, it is still of interest to examine these differences between linguistic and biological change. First of all: the distinguishing features of particular languages (grammatical, phonological, lexical etc) are characteristics coded and transmitted culturally rather than genetically, and are thus ‘acquired’ characteristics (= acquired by the organism in its lifetime). Changes involving them are thus evolutionary in this precise sense. With an eye on the history of the theory of evolution, one might be tempted to call such changes ‘Lamarckian’, but as we shall see many of them are not adaptive in the way suggested by the classical Lamarckian account of ‘acquired’ characteristics. The fact of cultural transmission distinguishes such changes from biological changes involving (random) mutation in genetic inheritance. (If anyone doubts this account of linguistic change, it is clearly demonstrated by the fact that – even if, as Chomskyan linguists hold, we inherit the most basic characteristics of language genetically – we do not inherit our specific languages genetically; we acquire the languages/dialects/ accents current where we grow up, and these may be utterly different from those of our parents.)

However, this point does not damage Pennock’s use of the facts of linguistic change against creationism, and he certainly does not try to fudge or conceal it. The precise mechanisms by which changes and diversification occur vary in all domains, and differences in what predominates from domain to domain are not really relevant to the issue. The argument against creationism is at a more general level.

It should be noted that many of the linguists – Chomskyans and such – who are especially interested in the genetic aspects of language acquisition recognise little diversity and hence limited scope for change in the general framework which they believe is inherited. This is true to such an extent that Chomsky himself – though not a creationist or even an ‘intelligent design supporter’ – can be read as equivocal about the possibility of formulating plausible, specific scenarios for the evolution of human language as an entire phenomenon out of pre-human communication systems (and has thus given comfort, perhaps unwittingly, to creationist linguists). But this is not an unavoidable feature of the relevant theories; it may, indeed, be seen as an aberration.

A second major difference between linguistic and biological evolution involves the fact that many short-term, specific linguistic changes do not seem to be adaptive, in that the later forms (eg, in grammar) are no better suited to the speakers’ environment (be it stable or changing) than the earlier forms from which they are derived. They have often been likened to changes in fashion or, more technically and perhaps more helpfully, to ‘cultural evolution’ more generally. Even if motivation for them can be adduced, it does not usually involve accommodation to the environment, still less improvement of the chances of survival/reproduction. There are of course exceptions to this, but they mostly involve less central and less systematic aspects of languages such as vocabulary, eg, the clearly adaptive ‘borrowing’ or coining of new words to deal with new concepts. As noted above, Pennock’s main examples (eg, pp 140-141) are of this latter type, a type which is much more readily understood by non-specialists (and may have loomed large in Pennock’s mind for that reason).

However, this difference is, again, not damaging to Pennock’s approach. As before, the central issue is the well-established facts surrounding the nature of change and diversification, rather than the details of the mechanisms or motivations involved. In any case, there are biological changes too which are hardly ‘adaptive’ in any strong sense.

There are a few more specific differences between linguistic and biological change, but these are mainly matters of rather fine detail which in no way detract from Pennock’s case. For instance, attempts have been made to individuate languages using the criterion of mutually intelligibility; these are parallel with attempts to individuate biological species using the analogous criterion of inter-breeding. Neither of these ideas has held up especially well in strong terms. In linguistics, this was perhaps predictable (at least for socially-aware linguists concerned with variability), given the rather obviously ‘fuzzy’ boundaries of geographically-adjacent, related ‘languages’ (as in a ‘dialect continuum’). Even here, the tendency in developed societies towards standardisation has promoted the concept of separate, countable languages and has discouraged accurate perception of the more usual patterns dominated by complex variation and ‘messy’ geographical boundaries. The limited success of the equivalent notion in biology was perhaps a little less predictable: traditional notions of the species suggested that here the boundaries should be more determinate. But current conceptions of species have had to allow for more ‘fuzziness’ in biology as well – which is, of course, very threatening to creationists with their focus on inherently separate biological ‘kinds’. The upshot is, in fact, that neither species nor languages are ‘water-tight’ entities (which strengthens Pennock’s analogy).

Now within the two sets of arguments there are certain detailed differences between these two cases. For instance, there seems to be nothing in the biological sphere quite analogous to the way in which (for a variety of reasons, some linguistic and some socio-cultural) mutual intelligibility between ‘languages’ is not always equal in the two directions. An example involves Portuguese and Spanish: they are closely related but the latter is structurally somewhat simpler and more perspicuous at most (not all) points where they differ, with the result that untutored speakers of Portuguese typically understand Spanish more readily than vice versa. However, no-one could reasonably argue that the absence of phenomena in biology analogical to something so specific damages Pennock’s general case.

Despite generally careful wording and a rather late excursus (pp 159-163), a harsh critic might suggest that Pennock fails to distinguish adequately between a) the origin of human language as an entire phenomenon and b) the origin (and development) of individual languages, as discussed above. For obvious reasons, rather little is known about the former process (though the volume of published research is growing fast); but it may very well have been more strongly adaptive in character and hence might be deemed ‘evolutionary’ in much the same sense as many aspects of the biological development of the species. (Some linguists actually prefer to restrict the term evolution to processes of type a), because the parallels with biology are closer here and biology is the domain where evolution is most obviously a dominant factor.) A related point involves the fact that all known ‘normal’ languages, ancient or modern, seem to be of roughly the same order of complexity, flexibility etc, and hence presumably represent much the same stage of evolutionary development. More ‘primitive’ languages must surely have existed before human language in its modern form evolved and while it was evolving; but no such languages are available today (all this occurred, on current estimates, at least 50-100,000 years ago, and because writing is a recent invention we have no specific language data more than a small fraction of that age).

On the other hand, there is no reason to suppose that it will ever be possible to draw a really sharp boundary between processes of types a) and b) as described above. Given the nature of evolution more generally, it would in fact be surprising to find such major differences of type or discontinuities in linguistic history. It would surely be unreasonable to require Pennock to base his claims upon distinctions which linguists themselves would struggle to reify.

It is also important for those of us involved in the mainstream of linguistics not to be seduced by the currently fashionable hyper-egalitarian dogma of near-absolute ‘linguistic equality’. Whether one regards additional complexity positively or negatively, and however one seeks to integrate recognition of such differences into theories of the development of human language, there is no doubt that at a detailed level some otherwise similar languages are more complex than others. Indeed, some prominent linguists (of various persuasions) have admitted as much. Icelandic and Afrikaans are both Germanic, but Icelandic grammar (especially the inflectional morphology) is considerably more complex (and more irregular) than that of Afrikaans. Note also cases such as that of Spanish and Portuguese, mentioned earlier. Well-informed, critical outsiders like Pennock can do us linguists a great service by reminding us that we too may have our sacred cows.

Pennock (p 159) likens the origin of human language to that of life itself, and the development of individual languages to that of individual species. This also provides a useful perspective in respect of the mutual-intelligibility criterion, as discussed above: if languages are seen as analogous to inherently ‘fuzzy’ groups/types of individuals such as species, a well-known logical objection to that criterion can perhaps be regarded as less damaging, or at least as damaging only to rather naive conceptualisations of the case. On the other hand, given some of the above points, another analogy (focusing more specifically on the adaptive aspect of evolution) might involve seeing all human languages as rather similar instances of the same species, differing very little in ‘fitness’. Only a few of the factors determining the success or failure of individual languages relate to their own characteristics; the key factors involve, eg, the power and prestige of their users. They still compete with each other in a social and cultural sense; but, in these terms, it is human language as a whole which has competed in a biological sense for the one relevant niche with other, earlier communication systems – and it won its battle completely, long ago. However, these are, of course, only analogies, not competing theories, and must not be pressed too far. The usefulness of one does not obviate that of another.

Pennock’s text contains only a very small number of more specific, problematic statements about language; one of these is simply a slip and the others are venial given that he is not himself a linguist. For instance, he does not seem fully aware (on pp 136, 143) of the highly controversial status (to say the least) of deep-time proto-language reconstructions such as those of Ruhlen, Greenberg and the Nostraticists (even though some of these do tally fairly well with non-linguistic data such as the genetic data correlated by Cavalli-Sforza and his colleagues). But some of his other points about language are very well-made. In addition, he gives a good account of ‘creationist linguistics’ (pp 120-125) – though he might also have referred to the recent (but arguably obscure) work of syntacticians such as Baker who have arrived at some very strange historical linguistic analyses partly on the basis of creationist notions.

But the specific disparities, and the very few problems which I have identified above, do not detract at all from Pennock’s main case, which (as I have said) I find very clear and persuasive indeed. Indeed, Pennock has much to say that is of great interest to philosophically-inclined historical linguists (some of which follows up the important work of Lass) – and indeed to anyone interested in these matters.

Newbrook, Mark. 2000. Review of Pennock, R. 1999 Tower Of Babel: The Evidence Against The New Creationism

Dr Mark Newbrook is a skeptic, linguist and football hooligan. He managed to combine two of these interests to help pioneer the field of Skeptical Linguistics.