What a load of cratylisme

Translation is not always the happiest fate to befall a book, especially when the book in question is well past its sell-by date. It is now nearly 20 years since Mimologiques was published in Paris and the passage of time has not been kind to it. Belatedly rendered into English, Gerard Genette's ponderously "structuralist" exercise in linguistic historiography conjures up images of a stranded whale from which the tide has cruelly receded.

The original subtitle (inexplicably omitted from this translation) was Voyage en Cratylie. The allusion is to Plato's Cratylus, in which Genette detects a certain linguistic "doctrine" that he dubs cratylisme, survivals of which he claims to identify in various forms in the western tradition. The book purports to offer a conspectus of "typological" variants of this doctrine, arranged in chronological order. Each chapter is in fact an essay on some particular author(s) or text(s) regarded by Genette as cratyliste. These range from John Wallis to Leibniz and from Mallarme to Proust.

The essays themselves are uneven in quality. Several could have been cut or omitted without loss. But surgery alone would not have saved the book, which is flawed at a more basic level. The initial mistake lies in the misidentification of the so-called doctrine of cratylisme itself.

The source of this misidentification is fairly clear. As with many French intellectuals of his generation, Genette's basic linguistic notions are Saussurean. Unfortunately, in spite of constant references to the Cours de linguistique generale, Genette's reading of that text seems to have been very superficial. It is by retrojecting a simplistic Saussureanism back into history that Genette manages to foist onto the Cratylus of Plato's dialogue a position which is diametrically opposed to Saussure's famous "first principle", ie the arbitrariness of the linguistic sign. The opposition thus established falsifies at a single stroke both Plato's linguistics and Saussure's. (Genette's exegesis of Plato has Socrates deliberately cheating in order to put down Cratylus' opponent, Hermogenes. Why Socrates should stoop to this deception Genette never explains.) Having committed that initial error, Genette compounds it by equating cratylisme with mimologisme and explicitly treats both terms as synonymous. This is another howler. In the western tradition, mimetic approximation is only one possible form of representation, and the extent to which it is a "natural" form of representation is highly controversial. But worse is yet to come. A corollary of the equation cratylisme = mimologisme is, for Genette, the corresponding equation arbitraire linguistique = conventionalisme. This distorts Saussure beyond recognition, as well as begging the question of what a convention is. We thus end up with a ludicrous chart of metalinguistic distinctions, in which Cratylus is the representative of mimologisme absolu, Saussure the champion of conventionalisme absolu, while in between come various less extreme positions, with Socrates standing for mimologisme secondaire and Leibniz for conventionalisme secondaire. To cap it all, Genette evidently views mimesis as yielding a transitive relation. Thus if x is an imitation of y, and y of z, then x is an imitation of z as well.

What can be salvaged from such a thorough muddle? Very little. If the chapters on de Brosses and Court de Gebelin are still worth reading, that is in large measure because they are authors who continue to be neglected. This translation may yet serve to draw the attention of English readers to their existence. On the other hand, it takes a Frenchman to find a morbid interest in exhuming the decently buried intellectual corpse of Charles Nodier. To this worthy a long and unforgivably tedious discussion is devoted. He could hardly be left out of Genette's chronicle, if only because it was Nodier who coined the key word mimologisme. The occasional shrewdness of Genette's comments on his chosen mimologists hardly compensates for the fact that treating the entire gamut of mimological claims over 2,000 years as variants on a single view of language involves a gross misreading of the history of ideas.

The translator has done her best with Genette's fussy prose, but the extent to which many of the points depend on the specific form of a French word makes frequent recourse to footnotes or explanatory parentheses inevitable.

Roy Harris is emeritus professor of general linguistics, University of Oxford.