What began as a compelling coup against authority has hardened, he suggests, into a new form of authority, a new kind of conventional wisdom. The rebels have become priests. "When psychoanalysts spend too much time with each other, they start believing in psychoanalysis. They begin to talk knowingly, like members of a religious cult." Instead of emphasising and accommodating strangeness, modern analysis imposes a kind of conformity: the professional elite confronts unhappy people with a glib checklist of probable causes.
In a way it seems right and proper that serious Freudians should revolt, like children turning on their parents, against their intellectual father- figure. But Phillips is careful to train his guns on the priests, not the prophet. "Most psychoanalytic theory now," he writes, "is a contemporary version of the etiquette book."
Analysts, he declares, were supposed to be our liberators; instead they have set themselves up as "the masters of our suffering." What he urges them to renounce is any notion of themselves as experts - expertise being merely a new form of authority - posing as superior knowledge and protected by jargon - the surest sign there is of a defence mechanism going full blast.
Phillips tells his story with the help of hundreds of quotations from Freud and Fereneczi, but also with select phrases from writers as various as Chesterton and Tobias Wolff. The star of his show is a perfect Kafka parable, called "Leopards in the Temple". It is a one-sentence short story: "Leopards break into the temple and drink to the dregs what is in the sacrificial pitchers; this is repeated over and over again; finally it can be calculated in advance, and it becomes part of the ceremony." It is a gorgeous idea, and for Phillips it illustrates what has happened to Freud's ideas. They have been brought into the ceremony by the "experts" and tamed.
Where the plot has been lost, Phillips suggests, is in the idea that psychoanalysis is curative or therapeutic. To him this should be merely a desirable side-effect. "The aim of psychoanalysis," he writes, "is not to cure people of their conflicts but to find ways of living them more keenly."
The leopards have to remain at large, he feels; they can't be turned into pussy cats (everyone knows, thanks to Freud, what happens when we pretend that terrors are not terrifying). It is a robust argument which reveals that Phillips wants Freudian theory to be less like science, and more like literature. Pyschoanalysis, he suggests, is a form of fiction. It tells "persuasive stories about where misery comes from".
It is thoroughly refreshing to be reintroduced to Freud the storyteller rather than Freud the categorical ruler of our inner lives. Phillips quotes Freud himself admitting that there is nothing in pyschoanalysis that isn't in literature; Sophocles invented Oedipus long before he became the tragic hero of our unacknowledged childhood. If analysis does have a therapeutic function, it is only that it "makes pain bearable by making it interesting". In this view, Freud was a literary critic, and Phillips makes much of this notion in his chapter on dreams. The interpretation of dreams, he writes, has the same relation to dreams that criticism has to literature: it can debate the meaning, but cannot relive the experience.
Phillips decorates this pleasing argument with a startling array of engaging insights and quips. He even succeeds in his paradoxical ambition, which is to be both genuinely self-deprecating about the limits of what psychoanalysis can presume to do; and also extremely impressive. It might even be that the very excellence of Phillips own work undermines his argument. He pleads for an acceptance of variety and strangeness, for a sceptical sense that everything is fallible; but his own style is anything but equivocal.
He likes unqualified, gnomic assertions ("The idea of knowing oneself makes a fetish out of memory... "pain makes us believe that other people have something we need"). In a masterly example of a sentence which expresses the exact opposite of what it says, he snaps: "Too much definition leaves too much out." And he is strikingly fond of the words "never" and "always". "Defences are always defences against the provisional," he writes. "We always know too much and too little." He writes, in other words, like an expert, like someone who knows what he's talking about. But it doesn't seem forbidding or oppressive. And so long as we accept that expertise is a relative term, it's okay to acknowledge that he has plenty of it.