John Gunnell’s purpose is to let flies out of the social science bottle, specifically the buzzing irritants of mentalism and realism. The larger source of noise arises from the self-inflicted need to seek legitimating foundations in philosophy and the natural sciences. In developing some of his earlier essays, one of the most distinguished analysts of American political science provides a fresh, powerful, and coherent whole.

The central contention is that the linguistic conventions through which concepts are developed and deployed constitute the social scientist’s only reality: conventional realism. Appealing to a posited realm beyond it, such as the mind, thought, or brute fact, is delusory. This is not, I think, to argue that there is nothing beyond language (p. 44). As Hume held of extreme skepticism, we cannot live as if it were true. Or as Gunnell cites John Searle’s variation on Dr Johnson’s kicking a stone to refute Bishop Berkeley’s idealism: if you don’t believe cliffs are real, you’ll be convinced when you fall off one (p. 129). Gunnell’s argument rather is to insist that the limits of language are the limits of the intelligible world. Anything posited as beyond it has to be an empty classifier, or as couched in words, “ontologies are embedded in languages” (p. 44). This may not be an adequate response to Searle, or it may be to restate what in Conceptual Idealism (1973) Nicholas Rescher called a post-Kantian truism—that it is only through our conceptual/linguistic operations that we understand or explain anything. Gunnell’s argument illustrates a general propensity in philosophy, reaching back at least to Hobbes and commonly called the “linguistic turn.” The activity is one of conceptual therapy, but without the overarching authority often accorded it since Plato.

Gunnell’s position, however, is explicitly Wittgensteinian, and he claims that political and much other social science have been characterized by the superficial use of philosophy and philosophical accounts of science (rarely science itself) to provide spurious and unnecessary legitimating foundations. Perhaps it goes without saying that to designate x from beyond a form of discourse as foundational for it is to identify x by its function within that discourse. As Quine similarly argued (The Ways of Paradox, 1976), we cannot work beyond all conceptual schemes, and from within any given one, it is meaningless to question what it posits as real. Thus, if Gunnell eschews philosophical or scientific foundationalism, he wants a closer attention to the philosophy that can be used in the social sciences. Wittgenstein becomes a philosophical authority for curtailing the scope of philosophy (pp. 141–42). Although occasionally Gunnell’s account sounds a little like discipular infighting, it is lucid, insightful, and careful, valuably ranging across the whole of Wittgenstein’s oeuvre.

Additionally, he enlists Gilbert Ryle and J. L. Austin, albeit in a less systematic manner, as providing compatible arguments. Ryle and Wittgenstein were friendly, and there was certainly cross-fertilization between them, though as Gunnell notes, Ryle published first on common themes (pp. 73–74). The relationship is more indeterminant with Austin who, in the Hobbesian idiom, was apt to be bracingly rude about most other philosophers, including Wittgenstein (p. 84). The three of them constitute the principal artillery for the action.

Chapter 1 provides an overview. Chapter 2 discusses doctrines of the mind and the misrepresentation of concepts. Chapter 3 deals specifically with the problem of realism in international relations. Chapter 4 outlines the attack on representational philosophy provided by Ryle, Austin, and Wittgenstein. Thereafter Wittgenstein takes center stage as Gunnell confronts in chapter 5 some of those who draw on him, such as Wilfrid Sellars, Donald Davidson, Hilary Putnam, John McDowell, and Daniel Dennett. Chapter 6 most explicitly treats the informing distinction between presentation and representation. Chapter 7 explicates conventional realism, and chapter 8 responds to the fear that it leads to relativism, prohibiting secure transcultural judgments. Such judgments are usually held to be the province of philosophy as an arbiter of rationality. Gunnell thinks not: there is hardly a philosophical position of relativism to attack (p. 17).

Important to his argument is the distinction between first- and second-order activities, though he also flirts with the possibility of additional orders (p. 15). The postulation of discursive orders underlines the difference between presentation and representation, the confusions between which blight much social science analysis. First-order activities are substantive human practices that present the world (p. 15), inviting the reflection of second-order inquiries as modes of representation. These elucidate but neither legislate for nor participate in first-order discourses without creating confusion, such as when academic political theory is mistaken for a form of political action. The example indicates that Gunnell’s argument is at once a little schematic and philosophically corrective. Nevertheless, on the basis of distinguishing orders and presentation from representation, he is able to assert rightly that the confusion of representational-theoretic language with its subject matter—the reification of concept terms (Ryle’s category mistakes)—is pervasively damaging in social inquiry (p. 45).

Only in chapter 3 does he deal extensively with the content of a social science second-order practice: international relations. Specifically, the focus is on the ubiquity of realism. It embraces both the (self-serving) distinction between realism and idealism that E. H. Carr projected as central to all international relations theory and what Gunnell calls the “new realism,” reliant on legitimating images drawn from the natural sciences and evidencing an ongoing theoretical insecurity (p. 55). From his succinct and telling analysis, the insecurity is likely to be ongoing.

I am in sympathy with most of what Gunnell argues, though conventional realism seems occasionally to be expressed implausibly. We have no good reason to doubt the existence of a world independent of us, and that is not, as Gunnell suggests, a matter of faith rather than philosophy (pp. 129, 112): it is a condition for living and a philosophical limit to skepticism. The correlative, if tautological, insistence on the impossibility of conceiving externalities independently of linguistic and semiotic convention is nevertheless sufficient to support his principal contention: accepting conventional realism helps dispel the persistent issues of how speech, thought, and reality are related (p. 149). More substantive illustration, as with international relations, would have been useful, especially from social psychology, a field particularly susceptible to the issues Gunnell raises.

He is clearly aware of the endemic problem of academic time lag; it has a bearing on his claims that social scientific appeals to philosophy and natural science are frequently superficial and outmoded (pp. 62–63). His own case provides passing illustration with an outdated image of early modern natural philosophy freeing itself from religion (p. 57). But when examining the substantive practices of social sciences, it may be that what might seem to be bad philosophy or misunderstood science is what has been adapted with functions distinct from its origins, thereby making philosophy’s corrective interventions less than decisive. The history of political thought is surely technically misnamed, but most now working in the field would (I hope) recognize that they are dealing only with language. Thomas Kuhn’s paradigms have a vibrant life beyond the history of science in which they are now largely extinct. As Gunnell rightly insists, language games are not static. More substantive illustration would have helped make a firmer point about philosophy’s limitations, but this is a carping comment on a valuable and stimulating work.