It is difficult these days not to be impressed by the range of technologies and research methods available to study the brain. The 'assault' on the most complex organ in the universe seems to be open on all-fronts, and there isn't a nook or crane that is not being measured, calculated, or interfered with in some manner or other.  From various imaging techniques to ever more sophisticated lesion-studies and cross-method innovations such as the combined use of fMRI and MEG, the field appears to be bursting with promise that great discoveries lie just 'around the corner'. As the editors of Methods in Mind declare in their introduction, "the future of cognitive neuroscience looks exciting" (p. ix), and their book is an invitation for the rest of us "to share in the excitement" (idem) of such promise. Accordingly, what you find here is a collection of 13 chapters, each by a leading authority in their field, explaining the various investigative techniques used nowadays to study human cognition, their strengths and weaknesses.

Let me first of all say that Methods in Mind is indeed "a uniquely valuable resource for the researcher seeking to expand his or her repertoire of investigative techniques" (from back cover). The range of topics covered is comprehensive and illustrative of cutting-edge work in the field, presented in a scrupulous manner that ought to be accessible to graduate-level readers (as a trained-philosopher, I found some of the more technical detail hard to follow, but familiarity with basic notions and concepts in neuroscience permits the main theoretical and practical points to be easily grasped). Also, and more importantly, individual authors have dutifully complied with the Editors' request to highlight possibilities — and current limitations — of cross-method integration and convergence, which results in a book that is not only up-to-date but also an honest and thought-provoking assessment of what the challenges are for current cognitive neuroscience.

One such challenge worth mentioning here is the very own drive towards integration and convergence the Editors are so keen to emphasize. It seems to me that in this respect the chapters collected in Methods in Mind are themselves clear testimony that, despite ever more sophisticated instruments, dreams of a unified cognitive neuroscience are still being hampered by old age conceptual tensions. Hence, whilst all authors praise and recommend the benefits of bringing together different methods, most of them are also careful to emphasize that the task is anything but straightforward.

The point is nicely made by Stewart and Walsh in the first chapter: "…one expects the results obtained using different techniques to converge on explanations of sensory and cognitive function, and that there is, in some sense, a hierarchy of explanation dependent on spatial and temporal specificity. If we examine this somewhat optimistic and simplistic view, it is easily found wanting, at least at our current state of knowledge of cognitive architecture and neural functions." (p.1) The problem here is that different methods rely on different assumptions, and analyze the brain at different levels of description, making it far from clear how such different types of data are to be combined or 'disagreements' dealt with. Hence, if, for example, behavioral deficits tell us one story about hypothetical mechanisms underlying cognitive function and fMRI tell us another, how should we interpret such results? Should we take them as incentive to refine our various techniques to the point where using either method will yield the same basic conclusions; or should we rather try to learn from the differences between them? The answer, again, is anything but straightforward — or even consensual.

More worrying for those waging their bets on the undeniably impressive advances in neuro-imaging techniques is the recalcitrant difficulty of going beyond mere correlations. As many of the authors included in this book note, brain activations can be misleading: what is being measured at the surface of the scalp could arise from a potentially infinite number of intracranial sources, and there is no standard principled way to solve the 'inverse problem' (or the 'non-uniqueness problem', as it is also known). This renders many of the claims to have identified the necessary underlying mechanism(s) to perform tasks x, y, or z particularly vulnerable to criticism.

Does this mean that we are best abandoning hopes of providing "a complete spatiotemporal description of the distribution of task-related activity within the brain" (p.291)? Not at all. If there is one thing that Methods in Mind does well in reminding us is that 'pulling resources together' is the best way to overcome particular limitations and open up more promising avenues of investigation. The 'trick', of course, is not to be seduced by such promises in forgetfulness of our current limitations.

© 2007 Isabel Gois

Isabel Gois has recently gained her Doctorate from King's College London. Her research interests include Philosophy of Mind, Neuropsychology, and Mental Disorder. She has articles published on Emotions, Computationalism, and Consciousness.    

Categories: Philosophical, Psychology