Slaves who serve their master well

When Gordon Baker and Peter Hacker embarked on their project of writing a detailed analytic commentary on Wittgenstein's Philosophical Investigations in 1976, no one could have foreseen that it would take 20 years and four volumes to complete - the first two co-authored and the later pair written by Hacker alone. But there is no doubt that the result justifies the heroic effort. It is as good a commentary on the Investigations as seems humanly possible. Given the complexity and perennial fascination of the book, others will no doubt shed new light on it, and approach it from as yet unpredictable angles. But this will remain the definitive starting point for the foreseeable future. Indeed, it must rank alongside the greatest contributions to philosophical scholarship (such as Ross on Aristotle or Vaihinger on Kant), since it combines, on a momentous scale, authoritative textual exegesis, philosophical insight, encyclopedic knowledge of the historical background and lucidity of expression.

Despite its undeniable excellence, the Commentary has had a mixed reception. Some are annoyed by the suggestion that a book published not yet 50 years ago should provoke a commentary of this size and complexity. Others jeer at the suggestion that a book published almost half a century ago should not have been decisively superseded. Both reactions are misplaced. It is true and ironical, that Wittgenstein pursued his professed aim of conceptual clarity in a fashion that is frequently obscure. He often condensed his insights to the point of impenetrability and failed to spell out the arguments behind them. Doing so would "spoil their beauty", he once maintained, to which Russell trenchantly replied that he should acquire a slave to take over this task. But while this can well be held against the "master", it is all the more reason to applaud the philosophical slave labour of the Commentary, which teases out insights, arguments and structure through incisive reconstruction controlled by close attention to the text and careful detective work on its emergence in the manuscript volumes of Wittgenstein's Nachlass. Few philosophical works since Kant's Critique of Pure Reason require such labours more than the Investigations, but none of them repays the effort more handsomely.

Equally, it is true that the Commentary seeks to defend Wittgenstein's work against popular objections wherever possible. But this is only good hermeneutic practice and far from holding that it is the last word in philosophical matters. If philosophy can make definite progress, as the objection rightly assumes, it can also objectively stagnate or retrogress. Naturally, Baker and Hacker's suggestion that much of the alleged progress really falls into these categories is controversial. But even if not all of their arguments are compelling, most are ingenious and noteworthy. The four volumes of the Commentary display the originality, resourcefulness and scope of Wittgenstein's work. They should lay to rest the idea, surprisingly popular in some quarters, that it consists of footnotes to Frege or a single insight about rule-following.

This final volume - Wittgenstein: Mind and Will - is devoted to the last 265 paragraphs of part one of the Investigations, and follows the pattern of its predecessors. It divides the remarks into chapters, continuous stretches of text devoted to a specific topic. The essays illuminate the background of Wittgenstein's discussion (in this case mainly James, Russell and the Tractatus), scrutinise the arguments and indicate their relevance to contemporary debates. Finally, the exegesis analyses each remark, traces its ancestry in the Nachlass, and comments on specific interpretative puzzles.

The editors report that Wittgenstein intended to suppress a good deal of these final remarks, and to work into its place material from part two. But Hacker argues persuasively that part two is not part of the same book, since Wittgenstein never started this revision; given the nature of the material, it is also difficult to see how it could have progressed. The final stretch of part one has attracted less attention than earlier sections, partly because it poses even greater challenges to the reader than the earlier ones. The structure and general point of the discussion is often even less obvious, and many of the remarks have been condensed even more than usual. Despite these obstacles, Hacker succeeds brilliantly in showing that these passages are essential to the discussion of language and linguistic meaning that is the leitmotif of the Investigations.

This holds in particular for the chapter on intentionality, which Hacker shows to contain Wittgenstein's deepest criticism of the picture theory, as well as a fascinating but little understood investigation of the logical problems of intentionality. The Tractatus is fuelled by the puzzle of how propositions can "reach right up to reality". If the proposition that p is true, what it says must be what is the case, namely that p; but if it is false, what it says cannot be what is the case; yet its content is the same in both cases. This requirement rules out the idea that the content or sense of a true proposition is a thing that "stands between" the proposition and the fact which verifies it - an abstract entity (Frege) or mental phenomenon (Russell). One might add that it also counts against Frege's more plausible suggestion that the sense of a proposition is the "mode of presenting" its referent, that is, truth-value. While a single referent can be presented in different modes, different referents cannot be presented in the same mode, with the unpalatable consequence that the sense of a proposition depends on whether its referent is the true or the false.

The picture theory of the Tractatus explains how a proposition can reach right up to reality by holding that its sense is neither a mediating entity nor a mode of presentation, but a possibility - a possible "state of affairs" - that is actualised if the proposition is true, and not actualised if it is false. Wittgenstein's later account of intentionality holds on to the idea that the relation between a thought and what verifies it is an internal one: the content of the thought that p is that p, and nothing else. Hacker shows, however, that he abandoned not just the atomistic ontology of the picture theory, but also the idea that the internal relation between a proposition and the fact that verifies it is the result of a logico-metaphysical isomorphism between language and reality. Instead, it derives from an intralinguistic convention according to which the expression of a thought involves a clause ("that p") which also occurs in the statement of the corresponding fact.

The chapter on mental states and processes provides an excellent interpretation and defence of Wittgenstein's approach to philosophical psychology, in particular of his denial that our psychological concepts constitute a "folk psychology" that must be replaced by a more scientific alternative. Similarly, the chapter on the will is the most authoritative discussion yet of Wittgenstein's fiendishly difficult treatment of that topic. But the most exciting idea that begins to emerge from the Commentary is the link between meaning and intentionality on the one hand, and volitional concepts on the other. Wittgenstein's claim that meaning is determined by social conventions is often contrasted with a mentalist account according to which meaning is determined by the intentions of the speaker. But Wittgenstein insists that the meaning of an utterance depends not just on the constituents and structure of the type-sentence, but on how a token of that type is used on a particular occasion, and thereby on the intention of the speaker. The difference is that these intentions are neither of a nonsemantic kind, as for Paul Grice, nor constituted by mental accompaniments of the utterance, as for mentalism. Rather, they are evident from how the speaker explains the utterance when challenged. Meaning something by an utterance is not a "phenomenon", a mental or physical accompaniment of the word, because such phenomena do not have the same normative consequences as meaning something. That my utterance "Napoleon was a bad general" counts as a claim about Napoleon III rather than Napoleon I does not depend on what mental images cross my mind (perhaps talking about the former always conjures up mental images of the latter). Only my avowal or explanation that I meant Napoleon III commits me to such a claim. It is the ability of speakers to avow and explain authoritatively what they mean that underlies meaning something and thereby intentionality. Wittgenstein here takes up a Kantian idea that threatens both materialist and mentalist conceptions of the mind: meaning, intending or thinking something are not states or processes in which we feature passively, since they presuppose our volitional capacities as rational human agents.

Part one of the Investigations ends abruptly. Hacker's intention was to conclude his Commentary with a synoptic essay. This, however, he has not done. Instead he has written an epilogue to the Commentary - a separate book entitled Wittgenstein's Place in Twentieth-Century Analytic Philosophy. It is a magisterial treatment of this topic. Like the Commentary, it shows how scholarship, the history of ideas and analytic philosophy can be combined to great effect.

Hans-Johann Glock is reader in philosophy, University of Reading.