Lacan ostensibly pursued three, arguably incompatible, intellectual goals in his psychoanalytic theorizing: (1) a ``return to Freud'', (2) making psychoanalysis into a rigorous, mathematical science, (3) merging psychoanalysis with semiotics, the supposed science of signs. Roustang says little about the first, though he does (e.g. p. 104, n. 34) note that it amounted to little more than rhetorical camouflage, and indeed it is very hard to imagine Uncle Sigmund, with his (sound) mechanism and positivism, recognizing his own ideas in such fantasias as the mirror stage.
As to the second project, turning psychoanalysis into a science, it was ``relentlessly pursued'' (p. 19) over decades. In fact, Lacan set out to put psychoanalysis on the plane of not just any science --- say, mycology --- but the big one, namely physics, while at the same time ``in no way taking truth to be [psychoanalysis's] goal'':
If this seems like some kind of trickery [!], then pause for a just a moment over the actual criteria of truth, and ask yourself what remains, within the vertiginous relativisms of contemporary physics and mathematics, of even the most concrete of these criteria: what has become of simple conviction, the test of all mystical knowledge; of self-evidence, the very basis of philosophical speculation, and even non-contradiction, that most modest requirement of the empirico-rationalist edifice? But more within the scope of our own judgement, can it be said that the scientist asks himself whether a rainbow, for example, is real? What matters to him is that a phenomenon is communicable in a certain language (a condition of the mental order), able to be recorded in a certain form (a condition of the experimental order), and able to be successfully inserted into the chain of symbolic identifications in which its particular science unites the various aspects of its object (a condition of the rational order). [Écrits, p. 79, as quoted by Roustang on p. 21.]Let us pause a moment to savor this passage, from an essay ``Beyond the `Reality Principle' '' (1936), over which hovers the retroactive spirit of Alan Sokal. The suggestion that ``simple conviction'' or ``self-evidence'' had been taken seriously in either mathematics or physics for the last few centuries requires considerable reserves of either disingenuousness or ignorance; the idea that either mathematics or physics has given up relying on non-contradiction argues very strongly that it was simply ignorance, since mathematicians are addicted to proofs by reductio ad absurdum. As to how phenomena are handled in physics (what phenomena are handled in math?), one can only wish that Galileo had shared Lacan's insight. Instead of spending weary hours rolling balls down incline planes, timing them with his pulse, or squinting at Jupiter through a telescope, he could simply have written down whatever he liked,so long as some people thought it made sense. This is a methodology frankly unable to handle cook-books (did Lacan have a cook?) and changing light-bulbs, never mind physics. It was, moreover, written decades after Bernard, Mach and Poincaré, after Russell and Wittgenstein's Tractatus, and was contemporaneous with the great works of Carnap, Popper, Gödel and Tarski. (This is not to say that any of those men had the last word on science, truth and all that, just that a much higher grade of thought was unquestionably available.)
This is how Roustang criticizes the same passage:
In order to avoid being accused of proposing to turn psychology into a second-rate science, Lacan provides himself with the most incontestable model, the one that, through its access to mathematics, has rid itself of any dependence on the qualitative. But for the model to be usable in this instance, its overly sharp edges have to be rounded off. Its universality, based on the fact that it needs only certain kinds of knowledge and no interpretation to be understood, is reduced to communicability in ``some form of language.'' But mathematical language is not just any language --- it is highly specific. Its relation to experimentation, verification, and refutation --- all essential, even if there is some time lag (the Sputnik was the first experiment with Newton's laws) --- is reduced to the simple possibility of being recorded. Many things are recorded in meteorology, yet as far as we know, it is not dependent on mathematical physics. And as for physics' [sic] ability to express its theories and results in algebraic formulae, this supposedly amounts to nothing more than inserting a phenomenon ``into the chain of symbolic identifications in which its particular science unifies the diverse aspects of its object,'' an elegant formulation [!] in which the words ``symbolic identifications'' prepare us for the slippages of meaning to follow, but in which the formal rigidity of sigla stripped of any evocative potential is lost. It is true --- and not without a degree of cunning --- that Lacan chooses the rainbow's appearance in the sky as his example of a physical phenomenon, an example lending itself more to poetic élan than to the austerity of a calculus.Clearly this is better than Lacan's tissue of horrors, but not (as my links indicate) without problems of its own, principally stemming from ignorance of mathematics and physical science (e.g. claiming that math is completely non-qualitative). This is not, I hasten to say, anything in itself to be ashamed of, or even mildly bothered about --- unless you happy to be talking about mathematics or physical science. Lacan, evidently, knew next to nothing about them --- on p. 44 Roustang quotes Lacan as saying that ``Einstein's little formulae ... align inertial mass with a constant and some exponents'', which is so far from actual relativity as to be not even wrong. Roustang makes no such howlers, but still I get the impression of a grimly serious duel conducted with whiffle-bats.[This] procedure can be summed up as follows: After having selected the very best model of science, its most prestigious traits had to be somewhat downgrade, and others selected, equivocal enough to be applied indifferently to both science and psychology [sic!]. [pp. 22--23]
Despite devoting forty pages, i.e. a third of the book, to Lacan's attempts to make an honest science of psychoanalysis, Roustang never mentions the current literature on, precisely, the scientific status of psychoanalysis (e.g. Grünbaum's Foundations of Psychoanalysis, or even Popper). Since this literature almost uniformly damns psychoanalysis, one would think it ideally suited to Roustang's purpose, but it seems that he believes in some form of psychoanalysis himself. In any case, most of his space is devoted to quoting from Lacan, explaining the latter's meaning in slightly less opaque prose, and attempting to refute it. Some of these arguments are cogent, and many others could be made so by additional premises the reader can supply without exhausting effort. Others are not so easily salvaged. I confess that I haven't a clue what Roustang means by an ``algebra'', other than that it is not what mathematicians mean, and when it comes to ``a machine, in order to function, has to be endowed with energy; it is therefore necessary to invoke a form of transcendence, or a God'' (p. 33), words fail me. One peculiar idea he seems to share with Lacan is that ``the subjective'' has no place in science, where ``the subjective'' means something like emotions, feelings, desires, etc. In one sense, this is perfectly true, since scientists are supposed to not be guided by their emotions in their practice, we are supposed to try to be disinterested. Roustang, however, may (I can't quite tell) think that this means science can have nothing to say about emotions, feelings and the like, which of course doesn't follow at all. (Even in everyday life, we often make objective, dispassionate statements about subjective feelings and passions: ``Why's he so snappish today?'' ``He's angry because Robin dumped him over the weekend.'')
On to project number three, the merger of psychoanalysis with semiotics. The guiding aphorism here is ``The Unconscious is structured like a language'' (sometimes even ``is a language''). The closest approach to an argument on behalf of this idea goes as follows: ``Since the psychoanalytic method draws on language alone, and this method provides access to the unconscious, the unconscious is structured like a language'' (p. 57, Roustang's formulation). Roustang is quite properly merciless with this --- ``Just because certain celestial bodies can be studied only by means of a telescope does not mean that celestial bodies have the same nature as a telescope.'' And he does point out that it's not at all clear what it would mean for the unconscious to have the same sort of structure as a language (are we to suppose it's made of words and sentences, or things isomorphic to them? does it have declensions?), let alone the sort of structure which Lacan supposed language to have. But here, more than anywhere else, Roustang is hampered by keeping to approximately the same idiom as Lacan: we're back to the whiffle-bat duel. Lacan pursued through pages and pages of his labyrinth of the Real, the Symbolic, the Imaginary, the Name of the Father, the ``object a,'' the ``sexual non-relation'', etc., with Roustang nipping at his heels, barking happily (or, rather, angrily) at every fallacy, imposition and leap of thought. The option of simply undermining the foundations of the whole maze is not even dismissed. This is a pity, since it would have been much easier on the reader, and it could even have been an elegant demolition, using only three or four precisely shaped charges at the right points: the theory is vacuous, therapeutically ineffective, and predicated on a view of language which is, to say the least, highly contentious, and all but abandoned in linguistics itself. Asking pointed questions about neuropsychology or Darwinian evolution would be overkill.
Chomsky once called Lacan an ``amusing and perfectly self-conscious charlatan''. It is hard not to agree with the charlatan part, and for those in the right frame of mind the whole performance is at once spectacular and amusing: here is a ``prestidigitator of genius'' who proposed to practice sleight-of-hand upon, supposedly, the most learned and skeptical audience in the world, and kept up the act for decades, easily putting Cagliostro to shame. Roustang is not in the right frame of mind to be amused, and no wonder: he was one of those wondering how on Earth Lacan linked those three steel rings. His first chapter asks ``Why Did We Follow Him For So Long?'', and answers, because he set up his school in such a way as to confuse the relation of teacher and student with that of analyst and analysand, and because he insinuated that he was in a position to bestow the key to universal knowledge on us. Well, perhaps --- but for ten or twenty years? And what about non-analysts? (Memeticists' and multi-level marketers' ears will perk up, however, at the observation that Lacan's school was peculiarly focused on creating analysts, especially as patients were encouraged to become analysts themselves.)
It is difficult to find any grounds on which to recommend this book. As a critique it lacks force in crucial places; its description of Lacan's rhetorical strategies is unimpressive; as an account and explanation of the fortunes of Lacanism it is infinitely inferior to Turkle's Psychoanalytic Politics; it does not even aspire to being a memoir, or elegant, or witty. It could, I suppose, serve to explain Lacan's key ideas to an audience which was not familiar with them in detail, is prepared to see them attacked, and is already conversant with post-structural jargon. Where such an audience is to be found, I have no idea.
11 August 1997