The Bertrand Russell Society Quarterly - November 2005/February 2006

appreciating the varieties of analysis:
a reply to ongley

Michael Beaney

In ‘What is Analysis?’ (2005), John Ongley reviews my entry on analysis in the Stanford Encyclopedia of Philosophy. On the whole, he gives a fair summary of the survey of conceptions of analysis in the history of philosophy that I offered, and his criticisms raise important issues. However, he fails to do justice to my account in one fundamental respect, and this gives those criticisms an inappropriate edge. As I state explicitly at the beginning of my entry, one of my main aims was to give a sense of the varieties of analysis that can be found in the history of philosophy. It was not my aim to pigeonhole philosophers into particular categories, which is what many of Ongley’s criticisms seem to suggest. Of course, some kind of conceptual framework must be developed to elucidate the various forms of analysis and their interconnections, but it was not my intention to impose a rigid taxonomy. Analytic methodology in the history of philosophy is a dense and tangled forest, and it has too often been assumed that the trees are more or less the same. In recent years there have been fine studies of individual philosophers’ conceptions and practices of analysis, but few attempts to see the wood as a whole. Ongley makes pertinent points in relation to individual philosophers, but in offering them up as criticisms of my account, mischaracterizes my project.

In section 1 of his review, Ongley notes correctly that I distinguish three main modes of analysis – decompositional, regressive and interpretive. But he then remarks that “In general, [Beaney] says, analysis breaks a concept or proposition down into elements that are used in synthesis to justify or explain it” (p. 33). This is not well expressed, and is not what I say; at best, it just reflects the decompositional conception. One of my aims in writing about analysis has been to try to break the stranglehold that the decompositional conception has had on philosophical methodology in the modern period, and in discussions of twentieth-century analytic philosophy, in particular. What I call ‘regressive’ analysis, understood as the process of working back to first principles (by means of which something can then be justified or explained in a corresponding process of ‘synthesis’), was the dominant conception in the pre-modern period, and is still influential today. (Such a conception is illustrated, for example, in Russell’s 1907 paper, ‘The Regressive Method of Discovering the Premises of Mathematics’.) Interpretive analysis, too, I argue, is an important mode of analysis, which came to prominence in early analytic philosophy in the emphasis placed on translating propositions into ‘correct’ logical form, but which also has been implicitly involved in practices of analysis throughout the history of philosophy and science. Although he recognizes these three modes, Ongley fails to appreciate that the assumption that ‘analysis’ essentially means conceptual decomposition is what most needs to be questioned in understanding the nature of analytic philosophy (in my view). In his final section, he talks of a ‘revisionist turn’ in the recent history of early analytic philosophy movement, a turn which my work has helped foster. But it is my attack on this assumption that I would want to single out as fundamental in my work.

This is not to say, however, that the decompositional conception is not important, or even central, in many projects of analysis. Rather, when we look at actual practices of analysis, we must recognize that other conceptions may also be involved. Ongley notes this, too, in the first section of his review and in the first paragraph of the second section (p. 34). But he then seems to forget it in the rest of his essay. In the light of his criticisms, I can see now that I should have stressed it more throughout my entry, but as I have said, my main aim was to clarify some of the key forms of analysis and not to do justice to any individual philosopher’s conception or practice. For example, in my discussion of Kant (which I admit is far too brief), I was mainly concerned to illustrate the decompositional conception that reached a highpoint in the Leibnizian/Kantian conception of an ‘analytic’ truth as one in which the predicate is ‘contained’ in the subject. I had not meant to imply that this was the only conception of analysis in Kant’s philosophy. Indeed, on the contrary, I have elsewhere indicated some of the complexities involved in Kant’s actual talk of ‘analysis’ and the ‘analytic’ method (Beaney 2002). As Ongley quite rightly says (p. 41), Kant also has a regressive conception of analysis. I also agree with Ongley (p. 42) that sorting out the sense in which the Critique of Pure Reason is a ‘synthesis’ and the Prolegomena is an ‘analysis’ is a key question for Kant scholarship.

Another issue that Ongley raises in his discussion of Kant is that of whether any analytic method is apriori or not. In fact, my failure to address this issue is the main complaint that he makes in his review. In section 2 he writes:

Beaney does not claim that any philosophical method is apriori—in fact, he does not consider whether they are apriori or not. What he does, for the most part, is describe various instances of analysis as ‘regressive’, ‘decompositional’, or ‘interpretive’. But by simply attaching one of these labels to a method of analysis, we do not learn the details of how the method works, and it is the details that will tell us such things as whether it is empirical or apriori, that is, whether or not empirical propositions must be assumed in order to analyze some concept or proposition. With his own approach to analysis, Beaney cannot answer such questions. This is the major limitation of his approach. (p. 36)

Ongley is right that I do not adequately address the issue of the apriority of analytic methodology, and he has persuaded me that I need to say more about it in my subsequent work. But part of my target in attacking the assumption that analysis is essentially conceptual decomposition is indeed the idea that analysis consists in uncovering the meanings of terms by some apriori method. Ongley comments on the issue at various points in his review, and I found his remarks pertinent and helpful.

Nevertheless, this concession aside, Ongley is bizarrely uncharitable in the passage just cited. For the impression is given that my ‘approach’ is simply to label different instances of analysis as ‘regressive’, ‘decompositional’ or ‘interpretive’. This is a caricature of the crudest kind, which is reflected elsewhere in Ongley’s review. In opening section 4, for example, he writes (p. 39): “Kant’s method of analysis is likewise ‘decompositional’ according to Beaney. I hope it is becoming apparent how limited the use of these metaphorical labels to describe types of analysis is.” Fortunately, however, this caricature is contradicted by Ongley’s own summary of my survey, a summary which provides at least some details of specific methods of analysis; and many more details are provided in my survey itself. Ongley makes use of my terminology, too, in pointing out (correctly, as just noted) that Kant has a regressive as well as a decompositional conception of analysis: expressing it like this neatly encapsulates a feature of Kant’s work which has not been sufficiently recognized. Of course, ‘regressive’, ‘decompositional’ and ‘interpretive’ are only terms that represent the first step in going beyond simple talk of ‘analysis’, and one needs to look at the details of how any given method works to understand it properly. I find it baffling that someone could have read my entry on analysis and thought that all I was doing was offering a tripartite taxonomy, not least because of my emphasis on the way that all three modes are typically implicated in any actual practice of analysis. The conceptions of analysis I distinguish are intended as tools to open up our thinking about analysis, and not as a classificatory device to block further understanding.

In fact, in elaborating my account, I draw all sorts of other distinctions (which can be found in the literature) – between whole-part (decompositional) analysis and function-argument analysis, between ‘logical’ or ‘same-level’ analysis and ‘metaphysical’ or ‘new-level’ analysis, between ‘analysis’ and ‘quasi-analysis’, between reductive analysis and connective analysis, and so on. I also discuss related conceptions such as that of Plato’s method of division and Carnap’s notion of explication, and issues such as the paradox of analysis and Ryle’s idea of a ‘category-mistake’. Ongley mentions some of this (pp. 47, 50-1), which makes it even more surprising that he should think that I am essentially engaged in a pigeonholing exercise. Ongley and I share an interest in the history of early analytic philosophy, and it is here, in particular, that Ongley’s assumption that I am essentially pigeonholing distorts his discussion of my account, and motivates some unwarranted complaints. In section 5 of his review, he writes:

Examining the twentieth century, Beaney begins with a general characterization of 20th century philosophical analysis. “What characterizes analytic philosophy as it was founded by Frege and Russell,” he says, “is the role played by logical analysis, which depended on the development of modern logic. Although other and subsequent forms of analysis, such as linguistic analysis, were less wedded to systems of formal analysis, the central insight motivating logical analysis remained.” Beaney admits that this characterization does not fit Moore or one strand of analytic philosophy, but thinks that the tradition founded by Russell and Frege is analytic philosophy’s central strand. (p. 44)

At the level of conversational implicature, this is misleading. For it makes it look as if I am offering a general definition, but then finding myself forced to admit an important exception. What I actually do in beginning my section on conceptions of analysis in analytic philosophy (note the use of the plural here) is criticize the assumption that decompositional analysis is what characterizes analytic philosophy (since decompositional analysis was around long before analytic philosophy emerged). I remark that “This might be true of Moore’s early work, and of one strand within analytic philosophy; but it is not generally true”. It is at this point that I then say what Ongley quotes me as saying. The “as it was founded by Frege and Russell” makes clear that I am just referring to one – albeit central – strand in analytic philosophy, and not to analytic philosophy as a whole. Ongley gets the dialectic of my argument wrong. I am not forced to ‘admit’ that my characterization does not fit Moore. It was never intended to do so in the first place.

As I said above, one of my targets in writing about analysis is the view that philosophical analysis is essentially conceptual decomposition, and that this is therefore what characterizes ‘analytic’ philosophy. But this view does no justice at all to the actual methodologies employed by those who are generally regarded as analytic philosophers (understood as including Frege and Russell, as well as later philosophers such as Wittgenstein, Carnap, Ryle, etc.). So in focusing on logical analysis, and the Frege-Russell strand, my aim is to correct this mistaken view. Ongley seems to think that I am merely replacing one crude definition of analysis in analytic philosophy with another, whereas my main concern is to show just what a rich variety of conceptions of analysis there are even within analytic philosophy. In fact, we have only to consider the Frege-Russell strand itself to see that there are important differences here, too. As I point out in my entry, and have argued in more detail elsewhere (2003b, §6), for Frege function-argument analysis is fundamental, whereas for Russell decompositional analysis remains at the core of his thinking. (Cf. also Levine 2002; Hylton 2005b; Griffin forthcoming.) The case of Russell is instructive here. For it shows just how complex a particular philosopher’s practice or conception of analysis can be. Russell may engage in logical analysis, in showing, for example, how definite descriptions can be ‘analysed away’ when sentences in which they appear are recast into their ‘correct’ logical form. But decompositional analysis is still assumed to be required in identifying the ultimate constituents of a proposition.

Ongley’s failure to appreciate all this leads him to make some quite unjustified criticisms of my account. He writes, for example:

Beaney finds G. E. Moore’s notion of analysis to be of a traditional decompositional sort, where complex concepts are analyzed into their constituents. This puzzles Beaney: while he admits that Moore influenced conceptions of analysis among analytic philosophers, Beaney does not address the fact that this means that his theory that 20th c. analysis as Fregean/Russellian logical analysis does not seem to work even for the major analysts. He simply ignores this problem and goes on to Wittgenstein. (p. 50)

This is a travesty of my account. There is much to be puzzled about in Moore’s philosophy. (Indeed, Moore would hardly approve if one did not feel puzzlement.) But I am not puzzled that he had a decompositional conception of analysis. I say it is “surprisingly traditional”, given his status as one of the founders of ‘analytic’ philosophy, but that just shows that the use of decompositional analysis cannot be the hallmark of ‘analytic’ philosophy. More importantly, I do not have a ‘theory’ that twentieth-century analysis is Fregean/ Russellian logical analysis, and so do not feel flummoxed that Moore does not fit this straitjacket. On the contrary, I pointed out from the very start that Moore represents one genuine strand in analytic philosophy. So there is no problem that I ignore and quickly cover up by turning to Wittgenstein.

Am I just being overly sensitive to the rhetorical flourishes of Ongley’s exposition? As I said at the beginning of this reply, Ongley gives a fair summary of the main elements of my survey. However, it is to some of his connecting critical patter that I object. The impression is given at numerous points that I am simply pigeonholing philosophers and offering a crude generalization as to what ‘analytic’ philosophy is, which does not do justice to my aim of showing the variety of conceptions of analysis in the history of philosophy. In concluding his account of my survey of twentieth-century analytic philosophy, Ongley remarks: “it should be obvious even from this brief description of Beaney’s survey of the 20th c. that his model of 20th c. analysis as based on logical analysis does not fair well even on his own terms. In the end, Beaney changes tack and defines analytic philosophy as being a set of interlocking subtraditions unified by a shared repertoire of conceptions of analysis that different philosophers drew on in different ways.” (p. 51) I do indeed suggest that analytic philosophy should be seen in this latter way (but not ‘defined’ like this, which is not how I put it). I am not changing tack, however, since I was never in the game of offering a ‘theory’ (or ‘definition’) of analytic philosophy. As I have stressed, I was concerned all along to indicate the richness and complexity of conceptions of analysis throughout the history of philosophy, and not least, within analytic philosophy itself.

Let me end, though, by thanking John Ongley for his detailed review. As he notes at the beginning of his essay, I am currently writing a book on analysis, and my entry in the Stanford Encyclopedia was a first report on the work I have been doing. The hypertext format of the Stanford Encyclopedia, and the fact that entries can be updated in the light of criticism and further research, made writing such an entry the ideal way to proceed. I could offer an outline of conceptions of analysis in the history of philosophy in the main document, while reserving further details for the linked subsections. I could also make available the extensive bibliography I had been compiling, to help and encourage others to explore the topic of analysis. Of course, even with the subsections, attempting to cover twenty-six centuries of history of philosophy in just one entry is asking for trouble, and as Ongley notes at various points, there are significant gaps (not least concerning conceptions of analysis in the nineteenth century), some of which I am hoping to fill in soon. But I am grateful for the generous remarks Ongley makes in the concluding section of his review. I have concentrated in this reply on the main (and only real) grumble that I have with Ongley’s review, but as indicated above, I accept his key criticism, about the need to address the issue of the apriority of analytic methodology. Ongley also makes other, more specific comments in his review, such as those concerning Kant mentioned above. I know that these, too, will be helpful to me both in revising my Stanford Encyclopedia entry and in completing my forthcoming book on analysis.

REFERENCES