The child likes to play, but he likes also to break his toys so as to see what is inside them. —Antoine Thomas, Coup d’œil sur l’histoire et la méthode de la science étymologique (1904).Isidore stands in a tradition. Nutnut etymologies had been interrogated in Plato's Cratylus—a sceptical, beautiful, and very funny dialogue about the origin of language and the relation of word to world. In around 45 BC, Varro wrote a treatise on the Latin language, of whose 25 books we possess six, full to the 10/6 hatbrim of precarious etymologies, much like Isidore's. The early Fathers indulged in the same sort of thing. But Isidore was the limit. Where could one go after the Origines? No further on the same road, certainly. So Isidore's material would be subsumed into the whelm of mediaeval lore, surfacing when needed in the course of an argument.
For Howard Bloch, Isidore embodies the pre-modern attempt to establish place, firmness, solidity: 'Like Cicero’s notion of topos or Varro’s concept of locus, Isidore seeks the places where language comes to a standstill, where meaning becomes intrinsic'. Bloch is referring to this etymology in Varro: 'Loqui (to talk), comes from locus (place), because. . . he talks, who with understanding puts each word in its own place'. Isidore constructs a system of language around his 'uncompromising reverence for etymology', such that the explicatory words have no need of further explanation. Lucus needs to be analysed; lucere does not. (Isidore had evidently learnt nothing from the work of Jacques Derrida.) But the decision to stop at lucere is arbitrary, or at least there is no reason given for it. The foundation was not yet firm; and so it had been in Varro, and so it would remain through the Middle Ages and the Renaissance.
It wasn't until the late eighteenth century that a new wave of Isidoreans appeared, announcing a firm foundation of 'first principles'—the start of a great age of etymology-cranks. Rather than just repeating Varro and Isidore, they had something new to offer. They all sought to reduce, in the way that science reduces the chaos of the world to a system of regular relations. The material in this post is urbane, rather than sublime; it may seem dry, but I hope that it will stand in contrast with what came before—Isidore—as well as with what follows.
The great object of the etymological art is not to give an account of the origin of all words without exception—and I daresay this would be a frivolous aim—rather, the art is principally valuable for the materials and observations with which it can furnish philosophy, so as to raise the great edifice of a general theory of Languages.So writes (in my translation) the statesman Anne-Robert Turgot, in his article on etymology for the 1751 Encyclopédie. An attempt to explain all words without exception would be quixotic, for so many words are impossible to analyse, being arbitrary in origin, or having origins too far lost in the dusts of time. No etymological analysis should be an end in itself, as it had been for Isidore—who had begun anew with each word—but rather should contribute a single fragment towards the 'great edifice' of a linguistic theory. Turgot's article is thus a guide to the critical and cautious use of etymology as a philological tool, and it very much embodies the Lockean empiricism of its period:
It results from all that we have said in this article, that an etymology is a supposition; that it receives its character of truth and certitude only by comparison with the known facts; from the number of circumstances of these facts that it explains; from the probabilities of its results, as the critic appreciates them.I first came across the first quotation used as an epigraph for The Diversions of Purley, an eccentric work on etymology written in two parts (1786, 1805) by the English politician John Horne Tooke. Only here it is attributed to one 'de Brosses'—ie. Charles de Brosses (1709-1777), a man with an oar in every pie, and well-known for various literary endeavours throughout the second half of the century. (Wikipedia is not bad on him.) De Brosses' masterpiece is his 1765 Traité de la formation méchanique des langues et des principes physiques de l'étymologie, which I read in its 1801 edition, its colophon dated 'An IX'.
This Traité, very much in the same tone as the 1751 Turgot article, and full of calm, rational thinking on the development of languages, is also the first great modern monument of bambashka etymology. This apparent disparity is due to a simple fact: while Turgot's strictures are very sensible, both for science in general and for etymology in particular, they are not enough. The sober etymologists of the next century demanded more than plausibility and coherence: they demanded evidence from the regularity of historically-determined laws of language evolution. De Brosses did not. Instead, he returned to the sort of quasi-empirical speculation found in the Cratylus. Near the end of that dialogue Socrates progs at the waspnest of sound-symbolism—the notion that each vowel and consonant has a basic significance, and that words are fadged up from clusters of these sound-meanings:
Now the letter rho [ρ], as I was saying, appeared to the imposer of names an excellent instrument for the expression of motion; and he frequently uses the letter for this purpose: for example, in the actual words ρειν [to flow] and ροη [torrent] he represents motion by rho.Plato's concept is a primitive version of what we now call the linguistic root; this latter notion, in a good approximation of its modern form, would be imported from classical Sanskrit grammars into Western linguistics in the 19th century. But de Brosses and his successors, like Plato, could only gesture towards this. De Brosses' hunt for the symbolism of sound was a hunt for fundamental causes. Where Isidore had been concerned merely with transforming the structure of the world into a structure of words, and with ascribing an aetiology to each word, de Brosses goes further: he wants to derive all words from a much smaller group of meanings embodied in discrete sounds. In this way his interests parallel the rise of atomism and naturalist classification in the science of his century. Physics had become explicitly empiricist, and so de Brosses advocates an empirical survey of language to determine the meanings of its underlying sounds. Early on in his book he claims
Thus did the legislator, reducing all things into letters and syllables, and impressing on them names and signs, and out of them by imitation compounding other signs.
That the system of the first fabric of human language, and of the imposition of names on things, is therefore not arbitrary and conventional, as we have been accustomed to picture it, but rather a true system of necessity, determined by two causes. The first is the construction of the vocal organs, which can only produce certain sounds, analogous to their structure; the other is the nature and propriety of the real things that one would name—this obliges one to apply to their names the sounds that depict them, establishing between thing and word a connection by which the word can excite an idea of the thing.Lockean epistemology is evident in the last sentence: if, as Locke had said, words exist so as to communicate the ideas of things between men, the linguist had to explain the mechanism by which a word 'excited' an idea. It was natural to assume some sort of actual similarity or homology between the idea and its verbal sound. Thus the root FR is said to denote breaking or cutting—Hebrew PHouR, Latin FRango and FRio, French BRiser and BRoyer, Latin FuRFuR, FaRina—'After the farine (wheat) is cooked in a FuRnace, the bread, necessary aliment, is the principal foodstuff with which one FuRnishes (fournir) one's house; but we have since generalised the word fournir'—FRustrum, FRaus, FRagmenta, FRuctus, FuRor, FRigo, French eFFRoyable, FeRveur, FoRt, and so on. It is hardly surprising that de Brosses would be praised by his 1801 editor for his 'facility for grasping the very real connections between objects that seemed the most opposed'.
But de Brosses goes further than sound-symbolism, and he is not averse to traditional Isidorean speculation: for example, this digression on the celestial origins of the words for many abstract concepts—
To admire is to observe the Sun; mirari, from Mihr, ie. sol. To contemplate is to observe the Sky; contempleri, from Templum, ie. coelum [sky], aether. To consider is to observe the stars, and to desire is to lose them from view; considerare, desiderare, from Sidera [stars]. Admonition is the view of the moon; Moneo, from moun, ie. Luna.The corollary to this type of investigation is a pronounced scepticism towards the enquiry, so common in the previous century, into the first language among men, specifically that spoken at Babel before the confusion of tongues. (Most, naturally, had said Hebrew. Becanus, notoriously, had said Brabantic.) De Brosses argues, in almost a modern tone, that even the language at the origin of several modern languages could be 'a mixture derived from many even more ancient'—for no language 'is formed all in one go'.
The results of de Brosses' theories are not so different from the results of Isidore's, but their attitudes and aims are poles apart. De Brosses still 'seeks the places where language comes to a standstill', but he seeks this in pure sound, which in turn can be mapped onomatopoeically onto the sounds of natural phenomena. Language thus evolves directly out of the natural world. Isidore's text acts as a poetic exploration of the possibilities inherent in language, and as a mnemotechnic for humanists learning about their world. De Brosses, like Turgot, regards his own activity as scientific—as seeking first causes by empirical enquiry and rational abstraction. He unlike Isidore, is thus after the 'one true hidden meaning', only he sees the pursuit in terms not of hermeneutics, but rather of mechanics and physiognomy.
When Tooke quotes Turgot, it is entirely appropriate, for he too wants to use etymology as the means to a 'general theory of Language'. For Hans Aarsleff—the only scholar to have written seriously about English linguistics at the turn of the century—John Horne Tooke (1736-1812) is the man who ruined English linguistics at the turn of the century. (The Aarsleff story is that good linguistics left England when Condillac brought Lockeanism to France, stayed there until the Germans nabbed it, and finally came back to England with its tail between its legs.) Tooke's Diversions of Purley, a strange quasi-political treatise on the origin of English words, in the form of two long dialogues at a gentleman's country home, did not participate in the Lockean ideals still motivating the French linguists of the period.
Locke had argued that words come from ideas, which in turn come from things. Language was at the bottom of the hierarchy, neither fully public nor fully private, and the most susceptible of the three strata to confusion. For Tooke, on the contrary, words come directly from things, and thoughts from words. Language is thus an active process of the soul, shaping the mind as it evolves—a Romantic way of thinking, later revived by Heidegger and his postmodern epigones, as well as by Sapir-Whorf. The history of words, for Tooke, is the history of thought: and hence we can discover the true essence of a thought by learning the original meaning of its word. We are, of course, back with Isidore. And in the second volume of the Diversions, matters get distinctly political—Tooke, the progressive, wants to deny the natural or rational basis for English law, and so he remarks that law is only that which has been laid down by men, and that what is right (rectus) is that which is ordered.
Tooke's methods were considerably more elaborate than his predecessors, with an unusual use of historical English—he quotes Chaucer and Gower, and the great prototype etymological dictionaries of Minsheu (1627) and Skinner (1678)—as well as its supposed antecedents, notably Gothic, which he cites in Wulfila's alphabet, much to my amusement. But without scientific constraints, even the insufficient dicta of Turgot, Tooke's serious ideas and linguistic gymnastics produce fanciful results; and my readers will be aware that I use the word impejoratively. Tooke reduces the parts of speech to two—the noun and the verb—and proceeds to reveal a noun or a verb in all conjunctions and prepositions. If, for instance, is traced to give(n): 'if you go to town' is made equivalent to 'given that you go to town'.
Tooke's Diversions were hugely popular in England and America, and his imitators and followers were legion. (He also spawned a number of critics, one of whom went so far as to produce a book-length rebuttal, the Anti-Tooke, full of etymologies contradicting Tooke but no more plausible.) The most notable of these was Walter Whiter (1758-1832), the author of at least three literary curiosities: a dissertation on suspended animation, a fragment-commentary on Shakespeare's verbal free-associations, and a huge tome (in two versions) on etymological metaphysics. The first version of this last is the Etymologicon Magnum: or Universal Etymological Dictionary, on a new plan (1800); it was revised in 1822 as the Etymologicon Universale, with a subtitle of truly Elizabethan length. He writes in the 1822 introduction:
Among the Etymologists, no idea of submitting a race of words to a general law had ever been adopted. One word was supposed to be derived from another single word; not was there any attempt to discover an abstract or Universal Principle, to which these various separate instances might be referred, and by which they might all be connected with each other.And this is Whiter's great plan, and great contribution to the science of nutty etymology. The structure of etymology is for him like the structure of a monist metaphysics: a search not just for first causes, but for the first cause—a construction of the manifold out of variations on one basic principle. Thus he starts by seeking the underlying unity of those linguistic elements that seem most diverse, which he identifies with those 'most familiarly employed' and therefore 'so perpetually liable to change from frequent use'. It is, we might now say, like looking for connections between asmi, sum, eimi, am. But Whiter's big idea is that every word in every language has an original sense relating to the earth—or as he prefers it, THE EARTH—and its cultivation. (Thoreau would be a keen reader, in fact.) His reasoning?
Where can we find, or where can we expect to find an agent sufficiently potent and predominating for a purpose like this, but in that great object, which is ever present with us, at all times, and on all occasions, on which all other objects, capable of being seen or felt, either actually exist, or exhibit their force and influence?He adds, rather wishfully, that there is no one 'who will not instantly grant' the self-evidence of such a proposition. In the 1800 work, lacking the grand projet, he had already connected EARTH to g-EARTH or GARDEN, and to HARD, Latin DUR-us and TER-ra, Greek STER-eos [solid], ORCHARD, GUARD, WARD, YARD, Hebrew ARETZ [earth], and Greek ARATOS [plough]. Now, in 1822, he identifies soil, clod, chalk, slate, sludge, silex, calx, clay, heel, clear and clean (ie. that from which clay has been removed—the old antiphrasis again). He connects write with verto, which was originally 'to turn the earth'. And he goes further with patronyms: the Celts were originally 'workers in Clay', ie. those who made bricks; they are the 'illustrious nation, I had almost said, the only nation of the Globe', and their name is cognate with those of the Gaels, Gauls, Welsh, Waldenses, Belgae, Scoloti (Scythians), Scots, Goths, Chaldaeans, Galileans, Gaetuli, Atlantidae, Itali, Latins, Lusitanians, Cilicians, Kalmucks, Moguls, Sclavonians, Caledonians, Castilians, Catalonians, Andalusians, Angles, Lacones, Albanians and Albioni, Hellenes, Pelasgi, Philistines, Poles, Pehlevi and Volsci.
Whiter thought that 'consonants are alone to be regarded in discovering the affinities between words, and the vowels are to be wholly rejected', which puts us in mind of the famous bon mot ascribed to Voltaire, that 'etymology is a study in which consonants count for very little, and vowels for nothing at all'. (This discussion thinks not Voltaire, but at least Max Müller.) But Whiter's dictum of course made it very easy for him to spin etymologies as he pleases, and for that we are grateful. He also disregards spelling conventions, pointing out that Greek γγ corresponds to Latin ng, Greek σ to Latin ns, and so on—and therefore that the letter-groups can be substituted willy-nilly. It is a 'ghoti' logic.
Took up Plato's Cratylus. Full of comic and absurd etymologies of the kind that raged till about a century ago. But also half-conscious essays about semantics of a very interesting kind. The miracle is that soon after 400 B.C. there was a community with disinterested mental leisure to ponder over such nonutilitarian matters.I have, I confess, gone on too long. And I have not even mentioned the most bizarre of all the speculators—Alexander Murray, whose 1823 History of the European Languages begins with the assertion that all words in all languages derive from nine basic roots denoting movement (ag, bag, dwag, gwag, lag, mag, nag, rag, swag), and goes on to trace the development of these roots with the most fantastical elaboration over a thousand-odd pages. (This work, unlike the others mentioned here, can happily be bought in an Elibron reprint.) There are countless more.
— Bernard Berenson's diary, 21 March, 1942.
The new Isidoreans had similar (albeit more sophisticated) results, different self-presentations, and ultimately the same overall goal as Isidore himself. All of them, to a man, were creating their world with words, as if the hierarchy were not thing—idea—word but idea—word—thing; as if language, guided by an overarching theory, were recreating the real world. But there was no mysticism among this group, as there would be for the occultists of the later nineteenth century; De Brosses, Tooke and Whiter were all staunch materialists, and all understood their own project to be a naturalist one—pure common sense—however silly, or however magical, it might seem to us.