Philip Johnson-Laird: How We Reason.
Though simple, the title is breathtakingly audacious. Yet this book is no overly sanguine young gun's effort, but the magnum opus of quite possibly the world's foremost authority on the subject. In it Johnson-Laird has put together the accumulated fruit of a life-time's research into human reasoning. The results do justice to the title.
Mental models are the central concepts that tie together the whole book. Individual models are maximally iconic representations of what would be true in one possible scenario: a possible dinner party seating arrangement, for example. The strengths and weaknesses of human reasoning are explained in terms of the interplay between these models and human cognitive limitations, in particular the 'bottleneck' our very small working memories create. One important weakness is that models represent only what is true in the possibility, and not what is false (112). This is necessary to lower the computational load, but it leads to predictable biases in reasoning. Another weakness is that, even given this simplification, we can hold only a few models in memory at once, making reasoning about disjunctive possibilities particularly difficult.
While this central idea may seem to be almost simplistic, chapter by chapter Johnson-Laird pursues its many implications and applications. One vital implication is that human reasoning cannot be accounted for merely in terms of syntactic relations. Instead, it relies upon content, as our knowledge is used to construct mental models and to modulate interpretations of logical connectives in ways that could not be handled by meaning postulates. This becomes particularly significant for inductive reasoning, which always relies upon knowledge. As a result, Johnson-Laird comes to draw a clear distinction between logical calculus and reasoning: the first captures the implication relations between sentences, while the second is the process of drawing conclusions from premises (171). In effect, the reasoning strategies we pursue are the result of bottom up learning rather than the application of top down principles (272). Another vital implication that mental models have is that counterexamples come to play a crucial role in reasoning. It is by searching for counterexamples among models that we evaluate a line of reasoning. Indeed, Johnson-Laird claims that the ability to grasp the significance of counterexamples is the 'cornerstone' of rationality (214).
True to its comprehensive conception, the book's many chapters go into a plethora of topics that are seen as relevant to reasoning: the cognitive role of emotions, individual and social differences in reasoning, unconscious thought processes, and reasoning about causal and deontic connections, to name just a few. In its thoroughness, it is inevitable that the book has its stronger and weaker parts. One uncharacteristically insubstantial chapter is that on development of reasoning abilities. As Johnson-Laird states, this process is something of a mystery (261), yet the weakness reaches deeper into his account and can perhaps be traced to his reliance upon representation as something of a primitive notion. While this allows him to work at the level of mental models, it makes it impossible to see how this level is built up from more basic elements--particularly in the context of development--as well as why the models represent the very things they do. Johnson-Laird is left stating that development is the result of the interaction of genes and environment, a vague conception that very few would be likely to disagree with these days. It feels somewhat uncharitable to point out this shortcoming, given how much the book does achieve. Yet, the issue is of great significance for him, given that he does not wish to accept the nativist approaches popularised by evolutionary psychology.
It is clear that Johnson-Laird has made every effort to make this book as accessible as possible (even if it does unavoidably overload the short-term memory). All the concepts are, therefore, carefully introduced, making the material relatively easy to understand regardless what kind of background the reader brings to it--a point that is most valuable given the very interdisciplinary nature of current research into cognition. For similar reasons, the book is full of examples that clarify the peculiarities of human thinking at the same time as they allow us a chance to chuckle to ourselves when we realise we share in these foibles. Indeed, the dry humour within the book works very well with its main purpose: the best example perhaps being Johnson-Laird's proof that, despite his best efforts, the book must necessarily include inconsistent statements (339). The one tactic chosen by Johnson-Laird to make the reading easier for non-experts that perhaps detracted from the value of the book was the decision to avoid footnotes. There actually are many pages of highly informative endnotes after the main text, but there is nothing in the text to indicate their presence, so that, in a way, 'the wires' have been hidden. While these might be distracting to some, to the scientific reader they provide a way to map Johnson-Laird's approach onto what they already know as well as directions for where to look next. Indeed, partly due to an apparent desire not to overburden with references, and partly due to its coherence and comprehensiveness, the book at times feels somewhat insular. This is a pity, as the endnotes do make clear the numerous influences upon Johnson-Laird as well as the relevance of his work for others.
While the relevance of Johnson-Laird's research for any philosophical account of mind is discernible throughout the book, the most striking realisation that reading the endnotes leads to is just how significant philosophical thought--C. S. Peirce's pragmatism, in particular--has been for his own approach. Of course, Johnson-Laird's work can be seen as a continuation of Peirce's project to understand logic (in the broad sense that predates formalisation). However, the connection is much deeper, Johnson-Laird's whole approach being profoundly Peircean both in the way it rejects the purely syntactic attempts to understand reason that were popular throughout most of the previous century and in the way it returns to a focus upon the significance of the content of beliefs as well as a bottom up approach to rational strategies. In this context, Johnson-Laird's identification of recognising the force of counterexamples as the cornerstone of rationality can be seen as the descendant of Peirce's point that thought results from doubt. In addition, the connection to Peirce provides a possible solution to Johnson-Laird's difficulty with representation, as Peirce's semiotics should afford him an account that is highly compatible with mental models.
Pragmatism has a history of thinkers who combine psychological and philosophical thought, of course. One contemporary of Johnson-Laird's who cannot help but come to mind is Herbert Simon, whose work was also profoundly shaped by Peirce. Clues to his relevance are sprinkled throughout Johnson-Laird's book, the central idea taken from Simon being that human rationality must be understood in terms of its boundedness. Given the closeness of these two scientists in terms of their intellectual underpinnings as well as some of their methodologies, however, it is surprising how little Johnson-Laird says about Simon. In particular, a discussion of Simon's heuristics--even if it appeared only in the endnotes--would cast significant light on Johnson-Laird's own talk of reasoning strategies and mental processes.
Johnson-Laird ends the book with the very Peircean fallibilist conclusion that chances are that many of his claims are false. Yet, it is clear that for some time to come those who would understand how we reason would do well to start by considering the claims he makes. This book is the best introduction to these ideas. It is sure to become a classic.
Konrad Lorenz Institute for Evolution and Cognition Research
and Marie Curie Sklodowska University
|Printer friendly Cite/link Email Feedback|
|Publication:||Philosophy in Review|
|Date:||Dec 1, 2008|
|Previous Article:||Nancy J. Hudson: Becoming God. The Doctrine of Theosis in Nicholas of Cusa.|
|Next Article:||Michael Krausz: Interpretation and Transformation.|