Herman Cappelen's aim in this absorbing book is to argue that, despite what they may profess about their methodology, professional philosophers do not place any important reliance upon intuitions in support of their philosophical theories. He takes as his stalking horse the assumption of ‘Centrality’: the claim that ‘contemporary analytic philosophers rely on intuitions as evidence (or as a source of evidence) for philosophical theories’ (3). His aim is to show that Centrality is false. The claim is intended descriptively, not normatively: Cappelen is concerned with what our actual philosophical practice is with regard to intuitions, not with how it should be conducted. In discussing the merits and demerits of Centrality, he criticizes the view that intuitions are essentially linked to thought-experiments in a manner that reflects our conceptual competence about a given subject area.

Thought experiments work, Cappelen claims in his chapter 9, by drawing our attention to philosophically interesting features of the world; they do not point to basic intuitions underlying our judgments about a given topic. Philosophers do not in general seek conceptual truths, or a priori knowledge. Rather, they raise questions of various different kinds, which lead a deeper understanding of the issues involved. The resulting meta-philosophical picture is one in which different methodological issues arise depending upon the specific sub-field of philosophy under inquiry. There is no one general methodology that unifies the whole discipline.

In the first half of the book, Cappelen covers some useful ground, arguing against attempts to defend Centrality by appeal to philosophers' use of the term ‘intuition’ and its cognates. He does a good job in showing how widely these terms are used, and that many of these uses do not function to pick out anything philosophically significant. Three interpretative strategies, which he calls ‘hedge strategies’, are described in support of his view that philosophers do not make use of the term ‘intuitive’ in any significant sense. Some forms of intuition talk are like a verbal tick – the point can be equally well conveyed if the references to ‘intuitively’ and the like are simply removed. Others indicate that a snap claim is being made with little reflection. More problematic is a third kind of case, where he claims that to say that a given proposition p is ‘intuitive’ means roughly the same as that p is ‘pre-theoretic’: p is a judgment that has been or can be justified without taking a stand on the question under discussion; it is accepted in the context as ‘common-ground’. Cappelen concludes Part One by claiming that the actual philosophical uses of ‘intuition’ and the like do not support Centrality.

However, while interpretation of intuition talk as ‘pre-theoretic’ may be a fair assessment in respect of many contexts, it is still arguable that in some situations the application of the term ‘intuition’ has a rather different import – it refers to a claim that is accepted without justification from perception or inference, and hence does indicate that something important is going on. Some of the points that Kripke makes about the necessary properties of individuals and kinds appear to rely heavily on intuitions understood in this sense. While Kripke does not claim that intuitively held beliefs are to be regarded as decisive, they can still carry strong justificatory weight. Moral philosophy is another area where the appeal to intuitions arguable plays an important role. Again, the claim would not be that such intuitions cannot be contested, rather that they are often used as entry points to debates about the appropriateness of an action in a given set of circumstances.

In Part Two, Cappelen rejects the idea that Centrality can be defended by appeal to philosophers' actual practice. The argument in this part is less than straightforward, largely because of the protean interpretations given to the notion of an intuition. Cappelen attacks what he calls ‘Methodological Rationalism’, a view his opponents take to be at the core of philosophical inquiry. The idea he criticizes is that we can make progress in philosophy by describing actual and possible situations in order to assess the intuitions we form about them. After much elaboration of possible variants, he eventually sets out a ‘watered down picture’ corresponding to a weakened form of Methodological Rationalism. He summarizes his diagnosis of the three complex features that are supposed to be the symptomatic of appeals to the intuitive:

This characterization leads on to the central debate about the role of intuitions that comprises chapter 8, where Cappelen considers a number of case studies, involving examples raised in many of the thoughts experiments that have been the subject of discussion in recent analytic philosophy. This chapter – and especially the discussion of the second thesis, Rock, that it contains – forms the heart of the book. It includes a discussion of the cases cited by Perry in his paper, ‘The Essential Indexical’, Burge's arthritis case, Judith Thomson's violinist example, Trolley problem cases, Chalmers's appeal to zombie examples, and others. Several notable thought experiments which might be assumed to manifest strong support of Centrality are not considered. In particular, although Twin-earth cases are mentioned in other places in the book, there is no extended analysis of what is shown by the discussion of the kinds of thought experiments raised by Putnam, Block and others; nor is there a full discussion of the Gettier examples. Cappelen argues, reasonably enough given the amount of space devoted to other stages of his overall argument, that it would not be feasible for him to consider further cases. It might, however, perhaps have helped the overall argument to be more effective had he compressed some of the material in Part I, allowing greater room for a more detailed examination of some of the examples mentioned. He suggests that his analysis demonstrates a pattern in the study of cases approach, which is that ‘actual instances of philosophical argumentation hardly ever exhibit signs of relying upon intuitions’. He expresses some sympathy for the views of Timothy Williamson about the nature of philosophy, while also noting differences between their respective views, and is sceptical of the notion that philosophy is concerned with conceptual analysis. In the final main chapter, experimental philosophy is dismissed as a ‘big mistake’.

There is a problem, however, in that Cappelen never really tells us what is going on when we construct thought experiments, and consider how they relate to the categories and classifications we find ourselves naturally inclined to adopt, beyond claiming that philosophy proceeds by our looking at the facts that are relevant to a philosophical issue, and at all ‘sources of evidence’. He argues that in situations where we might appear to be claiming support for a certain position on the basis of a particular intuition, in fact that intuition is responsive to a prior accepted theory about the area under discussion. However, it is arguable that in many cases the direction of justification is the other way round, and that it is our intuitions that form an important part of the ‘evidence’. The appeal to intuitions is often relevant to the question of whether a key classification, which we have hitherto assumed to pick out a unitary notion, is in point of fact more complex. Thus a consideration of the twin earth cases shows that what we take to be a unified notion of ‘sense’ in fact is a conflation of two distinct ideas: one aspect of the traditional idea of sense, connected with what is subjectively available to the subject, has to do with what a person understands in referring to an individual or kind; the other aspect of sense has to do with the mechanism of reference, and involves externalist considerations. The intuitions we have about the actual and possible cases are what show us that these two notions of sense come apart. In a contrasting type of case, a thought experiment can show that where there appeared to be two different categories of entity, there is in fact a deeper unity. Such may (or may not) be the case when we analyse what it is to be a sentient being. Some thought experiments put forward in debates over functionalism encourage the intuition that, after all, a suitably constructed functioning metal robot may be fully conscious in the way that normal human beings are; other thought experiments point to the opposed conclusion.

Cappelen is right to say that when a philosopher uses intuition talk, this does not necessarily mean that she is in fact appealing to an intuitively held belief that plays some privileged role in an argument. But this principle cuts both ways. A philosopher may be appealing to intuitions in a significant way, even though she makes no use of the term at all. So far as I can tell, Cappelen hardly considers such a possibility. But it is arguable that the implicit appeal to our intuitions often does figure in philosophical argument, and constitutes evidence for Centrality. To take one example, in his discussion of veridical hallucination and non-standard causal chains, David Lewis is plausibly read as appealing to intuitions about which causal processes issue in hallucinations, when he considers cases involving a ‘nonstandard causal process’ linking an object before the subject's eyes to a subsequent visual experience. Our intuitions about these cases are taken by Lewis to support the need to add a further sufficient condition into the analysis of seeing. More could be said also about the method of reflective equilibrium used in assessing normative questions. An important passage from Nelson Goodman's Fact Fiction and Forecast – one which is cited by Cappelen, but not then further explored in any detail – contains the key proposal: ‘A rule is amended if it yields an inference we are unwilling to accept; an inference is rejected if it violates a rule we are unwilling to amend’ (italics in the original). I do not know how to read the uses of ‘unwilling’ in this passage except as connected with appeals to our intuitions, when reasoned argument reaches an end.

Despite the reservations I have about some features of Cappelen's overall argument, I find this an interesting work about the nature of philosophical methodology, focusing on matters that are deserving of careful study. While there is some repetition of points and occasional backtracking, the cases he considers generate many points of interest, and on the whole the careful approach is to be commended. The enterprise is an original one, and the book raises important issues, about which we need to develop a deeper understanding.