We will soon reach the centenary of one of the key texts of modern cultural history - Sigmund Freud's The Interpretation of Dreams - if among a somewhat Gadarene rush to discredit the conceptual shift that psychoanalysis gave to humanity's view of itself. Yet whether Freud was right or wrong about dreams, it is primarily due to him that they are now seen as important.

Anthony Stevens is among those who take a generally dim view of Freud's contribution - but since he is an ardent follower of C. G. Jung, he would, wouldn't he? On the other hand, having been also trained in experimental psychology and medicine, Stevens is no narrowly focused analyst. Apart from an odd "polysemy" and "atman", the writing here is mercifully jargon-free and accessible to those on both sides of his "epistemological divide" - the Two Cultures, in fact. His primary concern is to build bridges between disciplines, most particularly between analytical interpretations and the scientific data of neurophysiology or ethology; since dreams are "psychobiological" events, it is no longer appropriate to discuss them as if they were either one kind of phenomenon or the other. Starting at these very divergent poles, Stevens shows much skill in producing a theoretical amalgam that is both erudite and humane.

Modern research has demonstrated that all mammals dream and that human infants spend much of their time in dream sleep. In Stevens's view, this disposes of most of Freud's hypotheses about dreams, notably that dreams are disguised expressions of repressed wishes or that their primary function is to preserve sleep. Dreams seem more like natural products of the psyche, with a biological purpose of promoting adaptation, as Jung proposed - a positive view, rather than a negative one. This raises the question, though, of how much the unconscious might be structured by man's evolutionary origins, and how much dreams might reflect that structure.

On this Jungian basis, the meaning of dreams is not so much disguised as "formulated in a pictorial language", which needs to be interpreted by combining understanding of the dreaming individual with that of humanity's collective symbols. As John Kerr recently pointed out, this hermeneutic approach meant Jung "saw the manifestations of the unconscious as symbolic communications from a better, wiser, deeper self". Even some psychoanalysts have now abandoned Freud's original view for such reformulations as Charles Rycroft's that dreams are metaphorical messages to the conscious ego from another part of the self.

ÁñÁ«ÊÓƵ

Neurophysiologically, the key discovery was in 1953 of a connection between dreaming and episodes of rapid eye movement (REM) sleep, which occupy about 25 per cent of each night, though the mind is never entirely at rest and some sort of dreams may also occur in the non-REM intervals. Stevens quite reasonably suggests that if sleeping and dreaming (about six years out of a 75-year life) do not perform vital biological functions, they must represent a colossal waste of time - a point ignored by some experimental scientists. A more specific question, though, is the function of REM sleep, which does seem to play an important role in the development of the infant brain, encouraging the emergence of genetically programmed patterns of behaviour. In REM periods, a regular rhythm (theta) can be recorded electrically; this suggests activity in phylogenetically ancient structures of the midbrain concerned with the storage of memory, but Stevens also sees them as a "self-energising system, programmed to seek psychosocial goals".

These same areas were proposed by Jung as the substrate for the archetypal repertoire of the collective unconscious that, converted by the more modern forebrain into linguistic forms, would guide individuals through the life-cycle. It would also connect them with the earliest concerns of humanity - the "two-million-year-old self". Stevens concludes that in dreaming sleep, strategies for survival are being updated in the light of both individual experience and that of the species, since it is only at these times that the brain is free enough of external preoccupations to be able to carry out this function. Like phobias when one is awake, nightmares might be warnings to avoid situations that were dangerous to early man. One of the pioneers of scientific sleep studies, William Dement, claimed that "only the dream can allow us to experience a future alternative as if it were real", but with the advent of virtual reality, that may no longer be true - a possibility not examined here.

ÁñÁ«ÊÓƵ

To appreciate "the incredible subtlety and ingenuity of the dreaming self", Stevens says it is necessary to make a detailed examination of specimen dreams, set in their context. He describes some of his own, some of his patients' dreams, some from Jungian history, creative dreams (Niels Bohr's, S. T. Coleridge's), and other famous dreams (Hitler's, Descartes's). Many are impressive - prophetic or problem-solving - but they have the inescapable drawbacks of anecdotal evidence. Can we be sure they were really dreamt, rather than retrospectively fantasised, and how can these few be compared with the innumerable meaningless, trivial or totally misleading dream experiences that practically everyone has had? Aristotle's shrewd observation - that out of all the dreams people have, at least some must resemble later events - is quoted but ignored.

It is very likely that a night's sleep, like a country walk, can help to resolve intellectual puzzles, though the mechanism by which this happens is uncertain. But there is little convincing evidence here that, apart from some exceptional cases, dreams can add much to mere incubation. Stevens suggests that in western societies dreams seem to be getting less serious and more violent, which, if true, may well reflect the barrage of stimuli most people now experience, particularly from television and radio.

In a most general sense, the importance of dreams may be uncontroversial, but are they, as Stevens suggests, "the only natural oases of spiritual vitality left to us?" From the Jungian viewpoint, dreams have much in common with myths and rituals, in that they are all seen as manifestations of the collective unconscious, but contemporary modernised societies have largely abandoned rites and ceremonies, while public myths survive only in technologically packaged forms. The lost rituals, in Stevens's view, are "some of the most crucial mechanisms of survival in our repertoire", so that their absence could threaten the future of civilisation. Maybe so, but rituals only work if they are believed in and practised unselfconsciously; Stevens does not confront the problem of how these could survive and remain useful in present-day societies, which is a matter of particular concern in Britain.

Yet what is really so remarkable is not how much contemporary science has revealed about dreams, but rather how little. As in so many areas of brain functioning, such as the physical basis of psychiatric disorders, we remain stuck at the inferential stage, with no actual proof that dreams have any one function - or indeed that they have any functions at all. Like Jung himself, Stevens has been tempted too far into a dogmatism that, as a scientist, he should have avoided. However intuitively attractive such ideas may seem, there is in fact no way of knowing that "people who make a practice of remembering their dreams do not suffer", that "living symbols have a . . . life-enhancing effect", or that "in dreaming sleep . . . the human animal is updating strategies for survival". And the claim that Jung proposed "nothing less than the fundamental concept on which the whole science of psychology could be built" is simply not credible.

ÁñÁ«ÊÓƵ

Eventually, Private Myths: Dreams and Dreaming becomes unable to bear its weight of speculative theory and of scientific beginnings pushed far beyond their limits. But it is a brave attempt, which by encouraging new research may help to advance us towards some real understanding of the still largely unsolved mystery that dreams represent.

Hugh Freeman was editor of the British Journal of Psychiatry, and is an honorary visiting fellow, Green College, Oxford.