We have been fortunate to see, in recent years, an upwelling of serious books about the philosophy of psychiatry and neuroscience, some of which have been written by psychiatrists (e.g. Ghaemi, Reference Ghaemi2003; Brendel, Reference Brendel2006) and others by philosophers (e.g. Bickle, Reference Bickle2003; Murphy, Reference Murphy2006; Bechtel, Reference Bechtel2007). This book is an important addition to this growing literature. The author, Carl Craver, is a philosopher of science with a deep substantive knowledge of the history and workings of neuroscience.
In seven chapters, Craver tries to paint, in a broad canvas, what he sees as the appropriate approach toward the nature of explanation and causation in neuroscience. Three major themes are noteworthy. First, Craver advocates a manipulationist approach to causation following, for example, the work of Woodward (Reference Woodward2003). In this approach, causality is defined in a real-world practical sense that avoids some the metaphysical tangles which the subject of causality can evoke. You know A causes B when you can figure out a way, in the real world, to manipulate A and, holding background factors constant, change B. With receptors, ion channels and long-term potentiation (LTP), neuroscience is chock full of examples of such causal processes.
Second, Craver is a strong advocate of a mechanistic approach toward explanation, advocated for example, in prior writings of Bechtel (Reference Bechtel2005, Reference Bechtel2007; Bechtel and Richardson, Reference Bechtel and Richardson1993). Craver believes that neuroscience will go astray if it predominantly seeks small numbers of deep laws that explain much of the workings of nervous systems. He would argue that this approach, which often dominates typical physics-oriented philosophies of science, is just misplaced in Neuroscience specifically and biologically more generally. Evolution is a tinkerer that builds organisms by trial and error. Its solutions typically work, but are rarely elegant or – when nervous systems are involved – simple. Our universe, by contrast, may have been created in a rather different way – with a few basic laws (quantum theory and relativity would be our current best guesses) – that explain a lot of how things really work.
An advantage of this mechanistic approach is that it steers clear of the old, rather unproductive debates of hard reductionists and emergentists. Mechanistic thinking is inevitably reductionist in the sense that you have to figure out how things really work in the nitty-gritty of the world. But biological systems work in multiple layers. Ions interact with receptors which sit on cells which connect in local networks which form neural systems, etc. So another consequence of the mechanistic approach is that higher-order processes also really matter. Neural systems have top-down control mechanisms that cannot be ignored. That said, Craver warns against what he terms ‘spooky’ emergence – so his idea of higher-order effects is firmly grounded in material processes. One of the most interesting chapters is titled ‘Nonfundamental explanations’ and explores these very questions.
Third, Craver has chosen his subtitle with care. The term ‘mosaic unity’ is one of the leading metaphors of this book and one that is particularly meaningful to psychiatry and psychology. He suggests that by adopting a multi-level mechanistic approach, it becomes conceivable to have an integrated view of neuroscience problems that is neither simple-minded nor hopelessly naive. This is a message that psychiatry might profitably consider given the confusion that currently exists in our field divided as it is into biological, psychosocial and psychodynamic wings. In his final chapter, Craver applies this model of mosaic unity to the problem of learning and long-term potentiation – with rather impressive results.
This book is both descriptive and proscriptive. It both describes the author's view of how neuroscience actually works and how it ought to work. There is alas, almost nothing that is explicitly psychiatric or psychological in this book. Nonetheless, it has many lessons of relevance for our field.
Let me take one of many possible examples. In thinking about the progression of mechanistic models of explanation, Craver draws a very helpful distinction between ‘how possibly’ versus ‘how actually’ mechanisms. Anyone who has read the literature in psychiatry has grown accustomed to fancy models of boxes and arrows or idealized synapses that purport to describe etiologic models for schizophrenia, depression or some other psychiatric syndrome. Clearly, these are (and our field is) at the ‘how possibly’ stage.
This is mostly an ‘easy to read’ philosophy book. ‘Philosophy-speak’ is kept to a minimum. That said, now and again Craver does have bones to pick with other philosophers. He does spend some pages here and there in debates that few who are not professional philosophers are likely to appreciate. But it is not too painful to skim these sections till he gets back into the meat of the issue.
Who should read this book? This book will be of interest if you are curious about what good philosophy of science is doing in a field very close to our own and particularly if you have an itch to see what an integrative vision of neuroscience (and maybe someday psychiatry) might actually look like.