Oxford University Press's
Academic Insights for the Thinking World

Languages, Species, and Biological Parallels

Languages: A Very Short Introduction

By Stephen R. Anderson

Human languages are not biological organisms, despite the temptation to talk about them as “being born,” “dying,” “competing with one another,” and the like. Nonetheless, the parallels between languages and biological species are rich and wonderful. Sometimes, in fact, they are downright eerie. Recently, for example, it was shown that the areas of the world that are richest in their biodiversity are also among the most diverse in the number of languages spoken by their indigenous people. Absolutely no logical reason exists for such a correlation: there is only one species (Homo sapiens) that has language, but the members of that species that live in places they share with many other species tend to have similarly rich diversity of language.

We think we know reasonably well what makes a species in the biological sense, but the notion is surprisingly hard to make precise, as a vast and contentious literature attests. We ought to be able to identify members of different species by their characteristic appearance, patterns of behavior, etc. But if that is the case, it is hard to understand why we call chihuahuas, pit bulls, and Bernese mountain dogs all “dogs” and assign them to one species (Canis lupus familiaris, a subspecies of the grey wolf, Canis lupus), while leopards (Panthera pardus), snow leopards (Uncia uncia), and clouded leopards (Neofelis nebulosa) are assigned to three different species — indeed, in some taxonomies, to three different genera. Compare this with the fact that we think of the English spoken across North America, Britain, Australia, New Zealand, South Africa, and India as all forms of the same language, despite considerable and quite obvious differences, while we treat the speech of Maastricht and Aachen as two distinct languages (‘Dutch’ and ‘German’ respectively).

Biological species change over time, members of a species differentiate from one another and new species come into existence, through the accumulation of small changes in their genetic constitution.  Similarly, languages change over time, their local forms diverge and new ones appear through the accumulation of small changes (especially sound changes, but also changes in aspects of their grammar).  Darwin was already struck by this parallel, and in The Descent of Man observed that “[t]he formation of different languages and of distinct species, and the proofs that both have been developed through a gradual process, are curiously the same.” The notion of ‘descent with modification’ is a fundamental explanatory concept in both evolutionary biology and historical linguistics.

One way biologists have tried to make the notion of ‘species’ more precise is by suggesting that organisms belong to the same or different species depending on whether they can produce (fertile) offspring. This approach does not get very far in the plant world, where hybridization between members of distinct species is extremely common, and even among animals it provides an imperfect definition. See the ability of mallard ducks (Anas platyrhynchos) to virtually wipe out other Anas species through rampant inter-breeding. The linguistic parallel is the notion that two forms of speech constitute separate languages if they are not mutually comprehensible. Again the problems with this seemingly obvious criterion are clear. Speakers of Danish can largely understand Swedish and Norwegian (but not always vice versa), and speakers of Slavey, a Canadian Athabaskan language, are reported to have responded “they’re speaking our language!” upon seeing a television program about the Navajo (whose quite distinct language is historically related to theirs).  On the other hand, speakers of ‘Italian’ from Venice and Naples are almost certain to have to resort to something other than their native language to make themselves understood to one another.

The notion of distinct species could surely be made precise in terms of distinct genetic makeup, of differences in the structure of the DNA of the organisms in question, but different individuals that we surely want to treat as members of the same species will also differ in significant ways in their DNA. Some such differences ‘count’ for species identity while others simply represent variation within the species; but the treatment of any particular difference as one way or the other cannot be decided in a non-arbitrary manner without knowing in advance how we want the answer to come out.

The linguistic analog here is the notion that different languages not only have different words, but also different grammars, so we ought to be able to identify distinct languages by looking for differences in grammatical structure. Again, though, we find very considerable differences among the grammars of people who would otherwise be seen as speaking the same language. In the American South, for example, many speakers can combine two modal verbs in the same sentence, saying something like “he might could do that, but he shouldn’t oughtta” (a construction which is quite impossible for most other speakers of ‘English’).

A particularly distressing parallel between biological species and languages that has attracted a good deal of attention in recent years is the fact that both are disappearing at alarming rates as a consequence of developments in modern society. A language is inexorably moribund when it ceases to be learned by children, and linguists estimate that as many as 50% or even 90% of the languages currently spoken, however we choose to enumerate these, will be in this situation by the end of the present century. This is a rate that is much greater than anything we find in the extinctions taking place in the plant and animal worlds. The situations are not exactly parallel, but the loss of diversity in both domains is deplorable and something that will leave our world much poorer if allowed to proceed unchecked.

Languages and biological species are much more nearly comparable concepts than their obviously different bases might suggest. Why this should be so is quite unclear.  Perhaps it shows us something very profound about the categorizing faculty of the human mind, but in any case it surely contributes to the sense that human language is something deeply rooted in our biology.

Stephen R. Anderson is Professor of Linguistics at Yale University. His most recent book, Languages: A Very Short Introduction was published this month.

Subscribe to the OUPblog via email or RSS.
Subscribe to only VSI articles on the OUPblog via email or RSS.
View more about this book on the  

Recent Comments

  1. Kniffler

    The analogy is not so very mysterious. Things we recognise as ‘species’ arise because of ‘descent with modification’. Languages, like the genetic recipes that define biological species, are the result of slightly-inaccurate transmission of the genes/memes over time.

    Selection pressures act directly on languages (e.g. favouring simpler grammars, easier-to-replicate phonemes, aggressive education policy), as well as on the communities of speakers (usually conquest), and so some languages adapt and survive and others die or are killed.

    This also goes some way to explaining why language diversity and biodiversity correlate – in both cases, speciation occurs when cross-pollination is rare or difficult, where travel is hard, and where domination of a community by an invading species is difficult.

  2. Stephen R. Anderson

    I try to develop analogies of the sort you suggest in the book: they’re not mysterious, but still surprising, given the difference in type between languages and species.

    On the other hand, evidence is essentially lacking in the case of languages for selection pressures of the sort you suggest. Languages are constantly getting more complicated, not less: wierd structures develop (and persist) where none were before. The mechanisms of linguistic change are to a significant degree quite different from the influences on biological evolution, at least to the extent we understand them (and the effort to do so actually antedates corresponding work in biology).

  3. Kniffler

    Good point, I was only guessing at pressures that might exist, but of course there don’t need to be any particular selection pressures on language itself. Simple memetic drift will suffice to cause speciation, given some barriers to panmixia.

    It would be interesting to see if there are any useful parallels between language and other things in biology which are believed to be products of sexual selection mechanisms. After all, *how* we speak is a social indicator more than anything.

Comments are closed.