Biochemical cascade? Yes, I remember it well

Memory, the authors of this book tell us in their opening pages, is "the most amazing phenomenon in nature". Well, maybe. Having spent most of my research life studying the brain mechanisms involved in learning and memory (neuroscientists have an unfortunate habit of eliding the two terms), I certainly agree that it is one of the most fascinating.

But is it "the key to consciousness" the subtitle claims? Well only if, as authors Richard Thompson and Stephen Madigan do, you reduce the richness of our conscious experience to mere awareness - and even then their book does not begin to unpick how memory might indeed be such a key. What they have written, though, is a useful and cheerfully accessible account of current ideas on the phenomenology of, and neurophysiological processes associated with, memory formation.

Over the past 30 years, memory has become one of the more heavily researched topics within contemporary neuroscience, and Thompson, a neurophysiologist, is a distinguished veteran of the field who has spent many years tracing the network of cells, primarily in the cerebellum, that are necessary to the learning of a simple reflex: that which associates a flash of light or the sound of a buzzer with the imminent delivery of a puff of air to the eye. His co-author, Madigan, is a psychologist with an interest in mental imagery and visual memory. The commission to write the book came from the US National Academy of Sciences, apparently as part of its public understanding of science programme. This may perhaps account for the slightly uneven tone of the book, in which the reader is chummily addressed as "you" and "real-life" anecdotes appear in boxes alongside more technical material.

The sequence is straightforward enough. It begins with the now-standard taxonomy of memory (declarative, procedural and so on), followed by the ontogenetic story of memory development in infants. Then come forgetting, amnesia, false memory, a somewhat surprising excursion into language acquisition, the now-obligatory references to the molecular mechanisms required for memory formation, and a nod at the end towards the future: will genetic engineering enhance memory; will computers substitute for human memory?

Most of the account is based on human studies, but there is much reference to animal work. I am not sure how a readership more sensitised these days to some of the ethical problems of animal experimentation will respond to the unquestioning reference to lesion experiments in monkeys, though few are likely to fret much over the sufferings of sea slugs. My own view is that such experiments need to be carefully contexted if they are to be justifiable, and that failure to do so in a "popular" science book is a mistake. Apart from any ethical doubts, we animal experimenters really do need to be able to convince our colleagues and others that when we observe an adaptive change in the behaviour of an animal as a result of experience (or, still more reduced, a change in the electrical responsiveness of a slice of brain tissue), then this is an analogue to my remembering where I parked my car last night.

Of course, such animal experiments have yielded a great deal of data about the molecular and cellular processes that underlie memory formation and consolidation - that is, the processes that are inferred to occur during a learning experience and in the minutes to hours following, during which it is assumed that a biochemical cascade takes place in particular nerve cells that results in a modification in their connections (synapses) with their neighbours. The modified pattern of connections is then assumed to be the "store" or "code" in the brain of the memory for the new experience.

Recall, then, involves scanning for and reactivating this circuit.

That such a cascade occurs in species as different as fruit flies, baby chicks and mice is certain. That the same drugs that interfere with these processes in animals and enhance or block memory formation do so in humans is also largely true. This is how both the present and the future generations of drugs to treat such conditions of memory loss as Alzheimer's disease have been developed. But there remain huge gaps in our understanding of just how the coding occurs, and how such specific memories can be retrieved almost at will. There are indeed some distinguished neuropsychologists who doubt that there is any such fixed store at all, that memories are evanescent and highly dynamic, remade each time we recall some past experience.

Regrettably, such theoretical problems are not permitted to raise their heads in Thompson and Madigan's clear and didactic presentation. Maybe they thought them to be too troublesome to introduce in a book aimed, I suppose, as much at first-year psychology students as the general public. A pity.

Nonetheless, as an introduction to "the facts of the case" as perceived by mainstream contemporary, primarily US-based neuropsychology, the book works well.

Steven Rose is professor of biology, Open University.