Big game flushed out by beaters

Developmental Science , now in its third year, was founded to publish "the best of contemporary scientific developmental psychology". No journal should admit to less, and so it is of greater interest to read the list of topics that the editors consider represent contemporary scientific developmental psychology. These are "comparative and biological perspectives, connectionist and computational perspectives, and dynamical systems theory". Cognitive, affective and social aspects of development get relegated to the "in-addition-we-shall-also-consider" category.

This clear ranking of disciplines gives an insight into the journal: it sees the future of scientific developmental psychology as lying in the biological foundations of development and in studies of how the developing brain responds to its environment. If not actually dispensed with, the traditional and rather more touchy-feely studies of subjects such as visual cliffs (do babies have an innate tendency to avoid cliffs? - they do), terry-cloth mothers (do babies suffer when suckled by artificial mothers? - they do) and studies of social learning (can babies imitate complex motor behaviours? - they can), are now seen as something of an anachronism, perhaps.

Important changes to research in developmental psychology began about 15 years ago with some of the first well-controlled investigations of "cross-modal" matching. Locke famously wondered in a 17th-century gedanken whether a man blind from birth could, if his sight were restored, recognise by sight alone objects with which he was familiar by touch. Thus,could the newly sighted man recognise a ball without touching it? It is a good question and one that modern developmental psychologists have been asking, in a number of experimental guises, of infants. Infants demonstrate some elements of cross-modal matching when they successfully post round objects through round holes and triangular ones through triangular holes. But a well-controlled experiment was required to answer Locke's query: put a marble in a baby's mouth then show it pictures of marbles and cubes. The baby's patterns of visual attention to these stimuli shows that it discriminates which one is in its mouth.

Experiments in cross-modal matching have given way to demonstrations that infants can add, that they prefer billiard balls to roll in the right direction when struck, and that they have an expectation that gravity pulls objects straight down. Developmental Science has published some of this important work. Taken together, the baby experiments show that human infants come into the world with a far wider and more sophisticated set of abilities than previously suspected. The issues are of deep philosophical implication. Kant, confronted by Hume's terrible conclusions that inductive knowledge was impossible, was forced to make the somewhat desperate assertion that humans simply possess synthetic truths of the world - things that we just know to be true, such as causation, without prior experience. Two hundred and fifty years on, developmental psychologists are showing that Kant was right.

Older theories of development - such as Piaget's stages - can be easily accommodated and re-shaped into the new approaches. "Critical periods" of development fit nicely into a view of the developing organism as a facultative agent, equipped with certain abilities, but whose final form is moulded by the environment at key points. Evolution gives critical periods a certain plausibility: if the individuals of a species typically evolve in a predictable range of environments, then the agency of natural selection can shape development to take advantage of those features at the right time. Here journals such as Developmental Science must lead the way as our understanding of critical periods (if they exist) has important implications for how educational and social systems are structured. Recent stories of a boy who lived in a troop of monkeys between the ages of two and four or the horrible stories of some Eastern European orphanages reinforce the belief that critical periods of development exist, and if missed, can have irreversible consequences.

Encouragingly, the authors of two recent books on the role of early learning experiences in humans are among the editors of Developmental Science and are regular contributors. It would be wrong to imply that it is only the experiments of a Kantian nature that fill the journal's pages. Equally good investigations in language learning, concept development, and acquisition of spatial and temporal abilities appear regularly. These are complemented in each issue by "peer commentaries and response", in which a point of view is expressed, followed by reactions from interested researchers. It is too early to tell if this feature of Developmental Science will be a success. But it is a good idea and one that, if tightly controlled by the editors to prevent self-aggrandising and carping, could rival the similar and successful features in journals such as Behavioural and Brain Sciences.

After many years of increasing specialisation, the practice of science looks to be changing course. The University of California at Berkeley is spending $500 million on a multidisciplinary programme uniting biologists, geneticists, physicists, mathematicians and computer scientists with the goal of investigating human disease. Other wealthy institutions will follow with similar programmes. There is an emerging view that the new challenges in human sciences will require teams that can combine knowledge and methods from related fields. Traditional single-discipline research may be about to take a back seat to more synthetic approaches.

Developmental psychologists and their journals cannot afford to miss this juncture. For example, the findings on infants' in-born abilities cry out for embryological and neurological explanation. And yet, there still seems to be a naivety among many developmental psychologists about the links between psychological development, biology and genetics. One recent contributor to Developmental Science uses the analogy of cells differentiating into skin, teeth, eyes and other specialised cells to argue against the idea that children may have innate knowledge structures. This contributor's notion is that the embryological development of specialised cells proceeds not from a pre-written blueprint in the genes but from chemical gradients and so forth that arise spontaneously as development proceeds. In an analogous manner, children can acquire knowledge structures solely from experience, and without any pre-existing blueprint. But this argument would surprise developmental biologists who see it as their sine qua non to understand how the chemical gradients that guide embryological development arise not from nowhere, but precisely from the genes that ultimately produce those chemicals, and the spatial relationships among cells.

If the first years are a reliable guide, Developmental Science looks to have positioned itself as an outlet for high-quality research of both an empirical and theoretical nature. It is consistent from issue to issue, conveying a serious and confident image of an active field. As a quarterly,it naturally acquires some prestige from being a scarce commodity. Academic developmental psychologists will wish to visit its pages regularly, and cognitive psychologists, linguists, philosophers and developmental biologists should keep an eye on it. Advancing its position will require vigilance and imagination. The editors of Developmental Science need continually to beat the bushes looking for scientists who are taking the next steps and not just fine-tuning paradigms. A key next stage for developmental psychologists will be to marry their impressive and hard-won facts about what infants bring to the world to the advances in embryology, neurology, genetics, and biological development. Computational neural network and connectionist models may also contribute, but their successes seem to be more in mimicking what we know of perceptual systems than in telling us how they are structured.

Achieving these next steps may require scientific structures such as those being assembled to investigate human diseases. Journals such as Developmental Science can fruitfully encourage them by seeking to be interdisciplinary about what they publish. To stay in front, behavioural scientists and their journals must increasingly embrace the biological revolution.

Mark Pagel is in the school of animal and microbial sciences, University of Reading.