Monday, May 08, 2023

Max Müller, Chomsky, and linguistics as science

A while ago I posted the following brainteaser on Facebook:
“We do not want to know languages, we want to know language; what language is, how it can form a vehicle or an organ of thought; we want to know its origin, its nature, its laws”. Who wrote this, and when? (No Googling!)
The question elicited a lot of educated guesses. Some associated it with current thinkers in the Chomskyan mould: see for instance Norbert Hornstein's discussion of linguists vs. languists. The word "organ", meanwhile, is reminiscent of Anderson & Lightfoot's 2002 book The Language Organ. Flagging up these similarities was precisely what I was aiming at in asking the question! The actual author of the quotation was Friedrich Max Müller (1823–1900), in his 1861 lectures.

Engraving of Müller by George Perine, pre-1885

Müller, a German-born academic who spent his career in England, is a fascinating figure. He absolutely dominated the linguistics scene in his time, especially in Britain, yet history has not been kind to him. Hornstein, for instance, was candid about the fact that he didn't know who he wasCartesian Linguistics, where Chomsky most famously situates his thought in relation to its predecessors, has no mention of Müller, and Chomsky (p.c.) states that Müller's ideas did not influence him. Historically this is perhaps unsurprising: Müller’s ideas mostly died with him. (Exactly why this happened is an interesting question in its own right, and I may return to it in another post some time.)

Müller died in 1900, and for the most part only historians of ideas have paid attention to his oeuvre since. Lourens van den Bosch's 2002 biography and survey, Friedrich Max Müller: a life devoted to the humanities - henceforth vdB - gives a thorough overview of his life and work. More recently the Publications of the English Goethe Society devoted an issue to Friedrich Max Müller and the Role of Philology in Victorian Thought. Neither of these sources particularly focuses on the parallels between Müller's ideas and today's linguistic thought, however (vdB has a few comments on this, but he is writing from the perspective of a historian of religion). This blog post is intended to draw some of these parallels to the attention of today's reader. In particular, I think there are at least four points where Müller's thought shows a striking similarity to ideas expressed by Noam Chomsky more recently:
  1. Linguistics as science;
  2. The relation between language and thought;
  3. The evolution of language;
  4. (more speculatively on my part) The role of infinity in human exceptionality.
This blog post focuses only on the first point; I'll try to return to the other three points in future posts. 

A note on sources: SL stands for Lectures on the Science of Language, published in two volumes between 1861 and 1864, and is the clearest statement of Müller’s linguistic thought.

Linguistics and natural science

Müller took a position opposed to the Scottish school of thought that held that language was a work of human art (vdB p214 mentions Locke, Adam Smith and Dugald Stewart). Müller’s argument against this was the largely unconscious nature of language change: “it is not in the power of men either to produce or prevent it. We might think as well of changing the laws which control the circulation of our blood, or of adding an inch to our height, as of altering the laws of speech” (SL). Darwin, who read Müller and corresponded with him, took a compromise position in Descent of Man (1871), describing language as “an instinct to acquire an art”.

Although Müller acknowledged the historical approach to language – and in fact spent most of his career focusing on this approach himself – he was also convinced that language could be approached as a physical science, or natural science: “with one foot language stands, no doubt, in the realm of nature, but with the other one in the realm of spirit” (SL). He distinguished philology, the study of languages for the purpose of understanding historical documents, from the science of language (see the brainteaser quote at the start of this post), whose main practical tool was comparative philology.

In labelling the study of language as a physical science, Müller was reacting against the famous 19th-century geologist Whewell’s distinction between Natural History (the physical sciences) and Human History (historical or moral sciences, the study of the contingent), and in turn, according to vdB, this probably derived from Alexander von Humboldt’s division between Naturwissenschaften and Geisteswissenschaften, a toxic dichotomy that continues to shape German and worldwide academia. It’s anachronistic to think that Müller was arguing that language belonged to the study of physics specifically.

The idea of linguistics as a physical science was anathema to the Yale linguist William Dwight Whitney, whose posthumous reputation can only be contrasted with Müller’s: while Müller is largely forgotten, Whitney is lionized, despite Whitney’s scholarly oeuvre being (as far as I can make out) substantially less original and more reactionary, and despite the fact that one of the ideas he is commonly credited with popularizing – uniformitarianism – was actually brought into linguistics by Müller and taken over only later by Whitney. The two were enemies for much of their lives, and Whitney devoted substantial energy to throwing shade at Müller, even suggesting that one of his major scholarly achievements, his edition of the Rigveda, was not by him at all, and getting his students to propagate this view (see vdB p486). Whitney even wrote a whole book, called Max Müller and the Science of Language, with no purpose other than muckraking at Müller’s expense. In his overview of the history of the field, Bloomfield’s Introduction to the Study of Language (1914) praises Whitney to the heavens, but makes no mention of Müller, even while arguing strongly that language can be treated scientifically (calling it a “mental science”) – exactly the opposite of Whitney’s view on the issue: Whitney describes the study of language as a “historical or moral science”, precisely on the other side of Whewell's dichotomy (Morpurgo Davies 1999: 208). Alter's (2005) ironically-named book “William Dwight Whitney and the science of language” has a good chapter on the Whitney-Müller clash, if one that also comes down rather heavily in favour of Whitney.

The Müller-Whitney “debate” is reminiscent of the rationalist-empiricist clashes of today, and it hardly needs to be said that Müller’s stance on linguistics as a science is also Chomsky’s (see Aspects chapter 1, Cartesian Linguistics, and Joseph 2002: ch. 2), though the issue is hardly particularly controversial nowadays. Chomsky’s own writings don’t suggest that this was a recognition he had to fight hard for: we’ve seen that Bloomfield treats language scientifically, Saussure was more psychologistic than he is usually credited with being (see Joseph’s biography), and even behaviourist approaches also viewed language (qua behaviour) as a valid subject of scientific enquiry (contra Whitney). Where Müller and Chomsky differed from these linguists was in the nature of the science. That said, one major distinction between Müller and Chomsky is the role of religion in the distinction: “Physical science deals with the works of God, historical science with the works of man” (SL). But thinking of Müller’s God in a fully present-day Christian way is also a mistake – I’ll hopefully come back to this in a later post.


Why are these similarities found? I'm not really a historian of linguistic ideas, so I can only point to a couple of possibilities. The obvious one is shared influences, of which Wilhelm von Humboldt (one of the heroes of Cartesian Linguistics) is probably crucial, and through him also Herder. There's also the possibility that Müller and Chomsky came across the same ideas independently because they made sense and were a good fit for the evidence; i.e. the ideas are right. (Hey, you can't rule it out!)

But there are also lots of areas in which Müller's and Chomsky's ideas diverge wildly, of course. Müller never paid much attention to syntax at all, and Chomsky has never paid much attention to diachronic linguistics. For Müller, individual lexical items and their historical development (“roots”) were crucial, while for Chomsky the combinatorial mechanism is. And it's a stretch to call Müller's approach psychologistic in any sense; moreover, other things that Müller proposed, like the Turanian macrofamily, turned out to be bunkum. But in any case, it'd be nice to see these similarities (as well as these differences) getting more attention in future, and hopefully this blog post might provoke thoughts along these lines.