Recently I’ve been reading up on various aspects of the history of biology, and I noted some similarities between biology and linguistics that I found quite amazing. Maybe historians of science will dispute my interpretations, but I cannot resist the temptation to draw some parallels between what I call “morphists” (scholars who emphasize pure “form”) and adaptationists in both biology and linguistics.

The alleged contrast between “formalists” and “functionalists” is well-known to most linguists (cf. Newmeyer 1998), but I never really understood it, and I don’t normally use the term “formalist”. (After all, everyone recognizes that languages have forms that need to be described – though it is true that some linguists seem to be completely oblivious of the often striking match between functions and forms.)

However, it’s clear that some linguists are interested in explaining the forms of languages with reference to their functions, and others tend to downplay or ignore the functions of grammatical patterns. So it’s interesting to see that in 19th century biology (before Darwin), there were two main approaches to understanding the similarities observed in comparative biology: what I call here morphism (the idea that pure form somehow determines what animals and plants look like), and adaptationism (the idea that the shapes of animals and plants are adapted to their environment, or “conditions of existence”).

Adaptationism was always a widespread view (since Plato and Aristotle), as many animals wear their useful features on their sleeves. It involves explaining biological features with respect to their “final (telic) causes”, or teleologically. This sort of explanation is straightforward of one assumed an Intelligent Designer, so it was closely associated with Natural Theology (of which William Paley (d. 1805) was the most widely mentioned representative). Another famous adaptationist was Georges Cuvier (d. 1832).

By contrast, the idea of pure form as a constraining factor is less evident, but in the 19th century (especially before Darwin), it had some prominent adherents. In Germany, the famous poet Johann Wolfgang von Goethe (d. 1832) was actually also an enthusiastic naturalist, and he pursued the idea of an Urpflanze and an Urtier (archetypal plant and archetypal animal). Goethe was followed by other Naturphilosophen such as Lorenz Oken (d. 1851), and two even more famous morphists were Étienne Geoffroy Saint-Hilaire (d. 1844) in France and Richard Owen (d. 1892) in Britain. Owen coined the term homology for similarities between animals and plants that could not be attributed to adaptation and that had to be explained by some independent principle of form. Other terms that the morphists used were unity of composition (Geoffroy Saint-Hilaire), and unity of type (the latter term was used by Darwin to refer to the morphist proposals).

Here is how a nice summary of the 19th century debates published by the UC Museum of Paleontology puts it:

“The zoologist and historian of science E. S. Russell summed up the great biological controversy of the first half of the nineteenth century: “Is function the mechanical result of form, or is form merely the manifestation of function or activity? What is the essence of life — organization or activity?” While Cuvier founded the “functionalist” school of organismal biology, with his insistence on animals as functionally integrated wholes, Geoffroy continued the more “formalist” tradition of biology that had started with Buffon and was being continued by Goethe, Lamarck, and others.”

I invented the general term “morphism” for the approaches that emphasized the constraints of pure form (or organization, or composition, or “type”), because the alternative term transcendental anatomist that was current at the time sounds opaque and cannot be used for other disciplines. And the term typology/typologist (used by Amundson 1998) is completely confusing in a linguistics context (I’m not sure if anyone ever used it in biology).

(A side note: The term morphology was originally coined by Goethe in 1790 for the morphist approach in biology. It was introduced into linguistics in 1859 by August Schleicher, a Jena linguist who had a strong interest in biology; but Schleicher was firmly in the historicist camp of linguistics, and became the first to highlight the similarities between linguistic and Darwinian biological evolution.)

As I noted earlier, linguistics also has its adaptationist (“functionalist”) and morphist (“formalist”) approaches. Interestingly, this contrast appeared in linguistics only in the second half of the 20th century, while it largely disappeared in biology after the second half of the 19th century. What might explain this difference between linguistics and biology?

I would like to venture the following hypothesis: Adaptation is easy to see in living organisms, but hard to see in languages. By contrast historical development is easy to see in languages, but hard to see in living organisms.

It is true that it look linguists quite some time to realize that similarities in word forms of neighbouring languages are often due common descent. But by the 1820s, it was well-established that the Romance languages derived from a proto-language (Latin) through descent with modification, that the Germanic languages derived from a Proto-Germanic language, and so on (including the even more ambitious idea of an Indo-European proto-language). By the 1870s, when biologists were still struggling to comprehend the significance of the Darwinian reframing of the issues, linguists had made a large number of discoveries based on the idea of descent with modifcation – and these results were universally accepted (and remain valid until this day).

The historical view of biological species as interconnected in a single Tree of Life (Darwin 1859; 1872) was biology’s way of reconciling the opposing approaches: Darwin noted that the obvious adaptations that earlier authors had sometimes attributed to a higher purpose (or teleology) could be explained by variation and natural selection, and that the “unity of type” could be explained by common descent. Both analogies (= adaptations) and homologies (= retentions from a common ancestor) were thus explained, and in a way, both the adaptationists and the morphists turned out to be right.

Likewise in linguistics, the historical approach of the 19th century explained the many homologies that were obvious even to a casual observer: Adjacent languages (and dialects) often had similar words with similar meanings (e.g. French la main ‘hand’, Spanish la mano, and so on), and by invoking limited sound shifts and meaning shifts, proto-forms could be reconstructed from which the attested forms could be derived by descent with modification. Thus, there was apparently no need to invoke a morphist principle of “pure form” to explain similarities between language forms.

This changed, however, in the second half of the 20th century, when syntactic patterns came under closer scrutiny, and the similarities across languages that were observed in this domain did not have a good historical explanation. Ideas such as a universal deep structure (Chomsky 1965), a universal X-bar schema (Jackendoff 1977), and a rich set of innate parameters (Baker 2001; Huang & Roberts 2016) can be seen as the modern-day counterpart of the morphist approaches of the 19th century. The “transcendental anatomists” may not have had a good causal story for the constraints of pure form that they invoked, while modern nativist morphists are clear that they derive their claims about pure form from humans’ genetic endowment (e.g. Pinker 1994), but otherwise the similarities are striking: Functionalists are telling an adaptationist story (similar to the teleological stories from Aristotle to Cuvier), but modern linguistic morphists find these ideas “too easy” and are instead exploring a wide range of “pure form” ideas (various formal frameworks, from Relational Grammar through Cartography, OT and Nanosyntax).

Again, the UC Museum of Paleontology’s characterization of the 19th century debate will sound familiar to many linguists – note the use of the terms “transformation” and “derivation”:

“While Cuvier founded the “functionalist” school of organismal biology, with his insistence on animals as functionally integrated wholes, Geoffroy continued the more “formalist” tradition of biology that had started with Buffon and was being continued by Goethe, Lamarck, and others. In his 1818 book Philosophie anatomique, Geoffroy asked the question: “Can the organization of vertebrated animals be referred to one uniform type?” The answer for Geoffroy was yes: he saw all vertebrates as modifications of a single archetype, a single form. Vestigial organs and embryonic transformations might serve no functional purpose, but they indicated the common derivation of an animal from its archetype.”

Geoffroy was obviously on to something – animals were clearly more similar to each other than could be explained by adaptation, because homologies (to use Owen’s later term) were often functionally diverse but showed a “unity of type” that could not be explained by function/adaptation/teleology (e.g. the fact that the arms of primates, the wings of bats, and the front flippers of whales are formally similar despite their different functions). Likewise, in linguistics, not everything can be explained by adaptation – the double articulation (duality of patterning) seen in languages, or the structure-dependence of displacements, for example. But morphist linguists have sometimes stretched their ideas in extremely implausible ways, and they were preceded in this by Geoffroy Saint-Hilaire – as noted by the UCMP:

“in hindsight, Geoffroy stretched many examples of homology, or “Unity of Type,” farther than was warranted by the evidence available. One of his more infamous theories was that the segmented external skeleton and jointed legs of arthropods such as insects were equivalent to the internal vertebrae and ribs of vertebrates; insects literally live inside their own vertebrae and walk on their ribs. He is said to have stated, “There is, philosophically speaking, only a single animal.” “

The idea that there is in reality “only a single animal” sounds rather similar to the idea that there is underlyingly only a single language. It is unclear how Geoffroy’s theory was empirically constrained, because there was apparently no independent way of testing the idea that the insect skeleton and the vertebrate vertebrae represent the same type. In 1830, there was a famous debate in Paris between the adaptationist Cuvier and the morphist Geoffroy (they had been colleagues at the Muséum national d’Histoire Naturelle, in Paris’s Jardin des Plantes), which is well-known to historians of biology  (the German Wikipedia has an even more detailed account of the debate, where it is noted that 80-year old Goethe was an excited observer from Weimar, happy that Geoffroy was presenting his morphist views with so much vigor; and that Geoffroy was associated with more modern, liberal views in the context of the 1830 revolution). In the end, it seems that Cuvier was thought to have “won” the debate because he presented more detailed data, while Geoffroy’s position was thought to be more speculative.

Linguists are not in the happy position of having resolved their disputes – we seem to be still waiting for our Darwin to come up with a new perspective that (hopefully) shows both sides to be right (in some sense). But our “Darwin” cannot be a historicist Darwin, because we found the historical solution to many of the observed homologies a long tome ago (well before Darwin).

In linguistics, we don’t (seem to) have a single Tree of Languages, so the historical explanation does not go all the way. Our “homologies” (for example, similar word order patterns across families and continents, similar sound-segment inventories, and similar limits on question-word fronting) surely give evidence of a striking “unity of type”, of the sort that Goethe and Geoffroy St. Hilaire would have admired had they known about them. But these “homologies” cannot be distinguished from “analogies”. In fact, after Darwin explained (Goethe’s, Geoffroy’s, and Owen’s) homologies as retentions from a common ancestor, the term “homology” changed its meaning, and came to denote similarities that can be explained by ancestry.

Thus, our typological generalizations are not homologies, but analogies – and analogies are explained by adaptation (or convergent evolution) in biology.

So why is there so much reluctance to explain cross-linguistic similarities of language structures (“unity of type”) by adaptation? I would venture the hypothesis that it is because there is simply far less comparative linguistics than there has been comparative biology since the 18th century. Linnaeus, Buffon, Goethe, Cuvier, Geoffroy, Owen, Darwin – all the naturalists of the 18th and 19th century (and of course long before, going back to Aristotle and Theophrastus) examined a large number of different plants and animals and compared them, in one way or another. It would not have occurred to any of the naturalists of earlier times to draw significant conclusions about nature after examining just one or two species. But this has been the state of linguistics since the 20th century: While Saussure, Trubetzkoy and Bloomfield knew many languages and were thus comparative linguists (in a broad sense), Jakobson was happy to write papers with bigger theoretical points in which he dealt with only a single language (e.g. his 1936 paper on the Russian case system), and this spirit was adopted later by other linguists in the Boston area (and far beyond). It also coincided with a drastic drop of prestige of the older languages (which all 19th century linguists were intimately familiar with), and with the rise of English as the only world language.

But adaptation is hard to recognize in languages – one doesn’t necessarily see the obvious adaptive role of, say, differential object marking, or of zero-coding of inalienable possessive relations, when one only studies a single language. Collectively, of course, linguists are now studying more and more languages (and fieldwork on small, out-of-the-way languages is very prestigious in the 21st century), but the number of linguists who are taking the broadly comparative approach of Cuvier, Geoffroy and Darwin is very small. Thus, until comparative linguistics not only gets more prestigious (it already is – typology is widely paid attention to), but also more widely practiced, it may be that adaptationist approaches will continue to be unduly in the background of theoretical linguistics, instead of taking center stage (where they belong).

References

Martin Haspelmath

More Posts