Do Analysts and Metaphysicians Disagree?

I have in front of me a terribly obscure paper by Rorty called "Do Analysts and Metaphysicians Disagree?" He wrote it for the 1967 American Catholic Philosophical Association conference to which he attended. I have no idea what the circumstances were that led him to give a paper for this conference, why they should ask an atheist to deliver a paper or whatever. It's lost to time, along with the paper unless you happen to be at a major university with an extensive library that houses most journals. I happen to be at such a university, but that's not how I was led to the paper. Years before I discovered that I could own everything Rorty had ever written for five cents a page by digging up the journals they were originally written in, and photocopying them, I hunted down anthologies at used bookstores. And one day I randomly pulled the Proceedings to the '67 conference off the shelf of one of the many used Madison bookstores I frequent to find, to my utter surprise, an essay by Rorty I'd never read. So I paid six dollars (I talked him down from $8.50, arguing that nobody but me would ever want to buy an ancient conference proceedings by a bunch of nobodys) for an essay that I could've had for 40 cents. A single instance of a long line of such silly financial decisions. But there's also a kinda' funny presidential address by Ernan McMullin, who's kinda' well-known, right? Right?

My disasterous penchant for buying anything with Rorty's name on it is, however, not why I'm writing. This particular paper is, to my mind, very good at bringing together those two disparate parties, pre- and post-linguistic turn philosophers. And, therefore, very good at making sense of how easily Pirsig can be brought into line with contemporary philosophers who, on the surface, seem to have little to do with his pursuits. So I'd like to pedantically summarize and quote extensively (yeah, more pedantic than usual) Rorty's paper so that its insights will be slightly more available.

Rorty opens up his paper very quickly by conceding that, yes, the pigeonholes "analyst" and "metaphysician" are overgeneral, but that "using them is no more unfair to one side than to the other." (39) He says that the metaphilosophical disagreements between the two camps is "usually conceived as centering around a disagreement about what method to pursue. My principal thesis in this paper will be that none of the familiar ways of characterizing such alternative methods are successful." What's more, Rorty says that he will argue that "it would be a mistake to say that analysts and metaphysicians have different subject-matters. It is, to my mind, wildly misleading to say that analytic philosophers take language as their subject-matter, whereas metaphysicians take nature or experience as theirs." The nice thing about this is that it tells us immediately that Rorty is about to sweep out of the way the prima facie objections of metaphysicians who think that analytics are obviously wrong because there is a lot more to life than language (echoes of which we can hear implicitly in Pirsig and explicit cries from most of his disciples) and the naive assumptions of early logical positivists who thought they had brand new, expensive hi-tech weapons to get rid of metaphysics. Both were wrong.

Rorty sets the stage for metaphilosophical confrontation by suggesting that "the only way of describing the subject-matter of either school which does not beg important metaphilosophical questions is to say that both take as their subject the various perplexities and paradoxes found in the philosophical tradition." We can see here Rorty's future position of identifying philosophy as the family romance, or the tradition of writing, from Plato to Descartes to Kant to etc. What's revealing in this old essay (as in his other old essays, where we hear other premature echoes) is the more specific why of the stance: it's to move the conversation along, it's a strategic move in a dialectical encounter, or rather one to produce a dialectical encounter. Here's Rorty at his best:

The aim of both schools is to find a set of truths which will resolve these perplexities and problems--truths which will have a maximum of intuitive plausibility and, taken together as a system, a maximum of theoretical elegance. In my view, the dialectical moves which are made in formulating and defending such truths are much the same on both sides of the fence. [For a further discussion with examples, e.g. Royce and Carnap, see Rorty's early paper "The Limits of Reductionism" (1961).] Among both analysts and metaphysicians, the ultimate criteria are set by a common aim--that of striking a proper balance between effectiveness (at resolving problems), plausibility, and elegance. The apparent differences of subject-matter and method are illusions created by the use of different jargons in which to formulate these dialectical moves. (39-40)

If one, for whatever reason, focuses on arugmentation in philosophy, I think one is led inevitably to a preoccupation with metaphilosophy and, eventually as one reads more and more philosophy, this kind of view. For whatever reason, I find myself drawn to metaphilosophy, probably because I find that the dance of arguments is the beauty of philosophy. Rorty was probably drawn to metaphilosophy because, in lieu of Platonic showing, which he had earlier abandoned as possible, the only way to establish Plato's dream was arguing for it. But it quickly became apparent that that wasn't so easily done either. (Sometimes I think those who are attracted to metaphilosophy are those who never want to lose an argument. Always wanting to win, or at the least never lose, will drive one to being familiar with arguments of all kinds and the various levels of arugments so that one is never caught off guard. This kind of pathology, of which I am no doubt susceptible, leads you to always try and stay one step ahead. Eventually, though, I think you figure out that there are no winners in philosophy qua argument, and so you become an expert at stalemate.)

The bulk of Rorty's paper is taken up by three things: 1) a three-fold division of philosophers into "critical", "speculative", and "empirical", 2) discussion of the distinction between "first-order" and "second-order" questions, and 3) the distinction between "solving" and "dissolving" problems. The first part, of course, is my favorite because I love pigeonholes, even ones that aren't (as Rorty says of this one).

By a "critical" philosopher I mean one who holds that philosophy is, in so far as it is a discipline in which argument and cooperative inquiry are possible, the attempt to dissolve all the traditional philosophical problems and thus to make the discipline of philosophy obsolete.
...
By a "speculative" philosopher I mean one who holds the negative thesis that philosophical inquiry should not be thought of as a pursuit of truths of the sort about which agreement can be reached by argument. On this view, philosophy is much more like poetry than like the sciences. We need not quibble about whether poetry aims at a special sort of truth, as long as we agre that the sort of truth at which it aims is not one about which argument is possible (except in a very diluted and Pickwickian sense of "argument"). On this view, philosophy does not aim at finding solutions to problems. Rather, it aims at finding new ways of seeing things through finding new ways of saying things

It is harder to specify what I intend by the term "empirical philosopher" but the following conditions will have to do: I shall call a philosopher empirical if:

1. He believes that there are procedures by which rational agreement can be reached on solutions to at least some of the traditional philosophical problems.

2. He defends this first belief by saying that (i) philosophy's method is either simply a description of what experience is like, or else a form of the hypothetico-deductive method exemplified by the empirical sciences: (ii) that philosophy is distinguished from these sciences either by (a) being concerned with a different sort of experience than that which supplies the premises for arguments in any other empirical discipline, or (b) having as its task the proposal of hypotheses of a greater generality than those produced by the empirical sciences. (40-1)

Oh, just lovely. So how do we put it to use? Well, Rorty tells us that it is designed to include all traditional metaphilosophical views, save two: rationalism (because nobody nowadays is a rationalist) and transcendentalism (because he's already gerrymandered the definition of "empirial philosophy" to include "whatever Kant is doing in the Transcendental Analytic and whatever Strawson is doing in Individuals."). Rorty suggests that this trichotomy is "a useful heuristic device for spotting difficulties in a philosopher's metaphilosophical views, and disharmonies between these views and his actual practice."

Thus the later Wittgenstein is professedly a critical philosopher, but his practice often makes him seem an empirical philosopher. Austin and Adler are professedly empirical philosophers, but their practice often makes them seem like critical philosophers. Santayana is a professedly speculative philosopher, but his practice often makes him seem an empirical philosopher. Heidegger sometimes talks as if he were an empirical philosopher and sometimes as if he were a speculative philosopher, but his practice is almost impossible to classify. Turning now to an application of these categories to the quarrel between the analysts and the metaphysicians, I think one has to say something like this: Analysts are, by and large, divided into professedly critical philosophers (like Wisdom, the early Carnap, and the early Bergmann) and professedly empirical philosophers (like Austin, Putnam, and Strawson). Nevertheless, one finds in the practice of all these philosophers a bent toward critical philosophy--in the sense that the test of success of the application of their empirical methods turns out to be the dissolution of certain traditional problems. Metaphysicians are, by and large, divided into professedly empirical philosophers and professedly speculative philosophers--into philosophers who have an explicit metaphilosophical view according to which some quasi-scientific procedure directed upon experience will produce philosophical truth, and philosophers who say "don't worry so much about decision-procedures and criteria, just try to find some way of expressing the nature of things, or some way of gaining wisdom." Nevertheless, one finds that a large proportion of the work of both sorts of metaphysicians consists in dissolving problems, in exposing the false presuppositions which have generated traditional problems--in short, in doing precisely the work the critical philosopher wants done. If these remarks accurately describe the contemporary situation, then the difficulty of my title question is clear. For if the analysts and the metaphysicians are, in practice, critical philosophers, and if neither school would whole-heartedly accept this description, then we need to find out why they would not, and, perhaps, wonder if something is wrong with our original trichotomy. (41-2)

One of the things that looks wrong with the distinctions is that "even if such post-positivistic analytic philosophers as Austin, Putnam, and Strawson are (in their explicit professions) neither critical nor speculative philosophers, they cannot be called empirical philosophers, since they deal with language rather than with experience, and pursue linguistic rather than empirical methods." Rorty thinks this looks wrong because "many linguistic philosophers and many critics of linguistic philosophy are still saddling themselves with an awkward and confusing pre-Quinian view" of the distinction between "questions of language" and "questions of fact." Rorty softens up this distinction by noting that "language is part of what we experience. It is not self-evident why consulting language should be opposed to turning to experience." (43) Rorty suggests that this opposition exists partly because analytic philosophers have relied on the analytic-synthetic distinction, but "it seems clear ... that it is a synthetic truth that certain sentences express analytic propositions. In other words, it is an empirical truth that the various expressions of our language exhibit the 'logical' relations to one another that they do." Rorty's point is that languages are grown by us mortal humans and that inspecting how words interact with each other, determining what they mean, is no less an empirical endeavor than seeing how fast a rock falls.

Despite this, and another point, Rorty says it is not "clear what relation such linguistic philosophers take to exist between hypotheses about the meaning ... of expressions on the one hand and solutions to traditional philosophical problems on the other. Prima facie, there is a great gulf fixed between a straightforwardly and explicitly empirical philosopher like Adler and a linguistic philosopher like Austin, and the gulf obviously has something to do with differing conceptions of the relevance of linguistics to philosophy." (43-4) This is where Rorty moves to the heart of what I take to be the irrelevance of a distinction of more than style between pre- and post-linguistic turn philosophers. It is summed up when Rorty wonders "what difference it makes whether we as 'What is the nature of x?' or "How do we use the word "x"?'"

There would be a difference if we were sometimes in a position to claim that we use the word "x" in such a way as to distort our conception of the nature of x, or to claim that more can be known about the nature of x than either language or science reveals. It is natural to suppose that metaphysicians are in such a position--for if they were not, why should they object to their analytic colleagues using the formal mode of speech?

On the utility of the linguistic turn, Rorty says it helps to "remind ourselves that there are no methods except attending to actual or possible linguistic behavior to decide questions about the nature of x, unless these questions can be settled by further data about the behavior of x's," that "the only poin to asking 'How do we use "x"?' is just to remind metaphysicians that they they have nowhere to turn to except to language." The question remains whether metaphysicians have an alternative method to offer. Rorty uses Mortimer Adler (of Hutchins School fame) as an example of a suggestion, that "philosophy is a 'non-investigative' empirical discipline dealing with 'common experience'". (45) Rorty, however, isn't sure the analyst would find anything here to disagree with and that we would need to know if I can "find out anything about x by consulting 'common experience' that I cannot find out by consulting the linguistic behavior of those who use the expression 'x'?" In trying to answer this question, Rorty turns to Adler's "tests of truth": "empirical truth, logical consistency, harmony between principles pertaining to theory and those pertaining to practice, and ability of philosophical principles to resolve questions arising out of scientific inquiry" (compare this set to Pirsig's in Lila, p. 113). But again, who would quarrel with that?

In lieu of a procedure for discovering the inaccuracy or problems with an account of x drawn from a description of the uses of "x", Rorty speculates that the metaphyscian's discomfort might be something like this: "There could be no quarrel, he might say, with the claim that we found the nature of x by finding out how 'x' was used if it were the case that accounts of the use of various expressions were always consistent with each other. But philosophical problems owe their origin precisely to the fact that our usage generates paradoxes, antinomies, and absurdities." Rorty thinks the analyst would reply that either 1) the perplexities don't really exist, but only appear to because of bad accounts of common usage (which better accounts of usage will clear up) or 2) the perplexities are real but we can eliminate them by linguistic reform, and that most will combine them with 3) "we do not yet know whether perplexity can be removed by better accounts of usage, but that if not, linguistic proposals are the only alternative. If these proposals amount to urging the adoption of a full-scale 'Ideal Language"--or, in older language, a metaphysical system--then so be it." (46)

Here is where Rorty puts all his cards on the table and everything becomes very clear.

Here again we seem to be at an impasse. The metaphysician regards the analyst's appropriation of the term "metaphysics" as merely perverse. The analyst asks why regarding metaphysical systems as so many proposals for the reform of language should be so disturbing. If, the analyst may say, you want to consider these proposals for different ways of speaking as discoveries about the world, go ahead. Just do not pretend that you have any criterion for such discoveries save the absence of paradoxes, antinomies, and absurdities from the language you suggest we use.

This, I think, says it all. If one has discarded all the bad junk, those horrible distinctions from the tradition, appearance/reality, subject/object, conditioned/unconditioned, differentiated/undifferentiated, analytic/synthetic, fact/value, then what does it matter if you talk about language or experience? What's more, what could decide in either favor if the non-existence of paradox is the only non-question begging criterion we have?

In Rorty's expansive introduction to his anthology The Linguistic Turn, Rorty puts it in similar, though slightly more aggressive terms. He asks metaphysicians, "If you were not making proposals for such an ideal language, what were you doing? Certainly you were not making empirical inquiries, nor deducing consequences from self-evident truths; so if not this, what?" (7-8)

If there is a single crucial fact which explains the contemporary popularity of linguistic philosophy, it is the inability of its opponents (so far, at any rate) to give a satisfactory answer to this question. It is no good saying that the great philosophers of the past were not interested in anything so piffling as language, but were interested instead in the nature of reality, unless we can get some clear idea of what it was they wanted to know about reality, and of how they would know that they had this knowledge once they had it. If one construes, for example, Spinoza's 'There is only one substance' as a proposal to stop talking about persons and physical objects in the ordinary (roughly, Aristotelian) way, and to start talking about them as dimly-seen aspects of a single atemporal being, a being which is both mental and physical, then one will have some criteria for evaluating his statement (which, unconstrued, strikes one as patently absurd). If one talks Spinozese, one will indeed be unable to state the propositions about minds and bodies which so worried the Cartesians, or the propositions about God's creation of the world which so worried the scholastics. Now it was precisely upon this that Spinoza prided himself--that the mind-body problem and problems about the relation between God and the world could not (or, at least, not very easily) be formulated in his system. It was this fact that made him confident that he had grasped the true nature of things. (8)

I think if we look at Pirsig, we see very similar things. Pirsig prides himself on his ability to dissolve problems of philosophy with his system. He thinks he has given us a new way to better understand situations like the hippies and ghetto violence and things like insanity and religion. But what is this new way if not a new set of categories or concepts, a new set of linguistic tools?

The rest of Rorty's paper consists of him wrestling with Adler a little while longer, and then James Cornman (over "solving" or "dissolving" philosophical problems). While interesting, they don't shed that much extra light over the previous stuff. I shall end with Rorty's conclusion, which is his short, unargued for suggestion about what the difference is between analysts and metaphysicians. And unlike earlier, where you could read metaphysician and analyst as simply pre- and post-linguistic turn pragmatists, like Pirsig and Dewey and Rorty and Putnam, here you'll want to read analyst as pragmatist and metaphysician as Platonist, as holding an appearance/reality distinction.

I shall conclude with an analysis which, though vague, seems to me at least less misleading than those I have been discussing. The difference has to do with wisdom. I think that analytic philosophers have in common the view that the pursuit of wisdom cannot be served by continuing the inquiries traditionally grouped together as "philosophy." Their examination of the history of philosophy leads them to think that philosophy is indeed, a tree from which all, or nearly all, the branches have fallen. They think that the answers to the questions they pose in the formal mode of speech, and the truths which are brought forward to show the falisty of the presuppositions of some or all philosophical problems, are not truths which will give us wisdom. At best, they will give us the sort of wisdom which Socrates had--the wisdom which consists in not being fooled any more. For the sort of wisdom which Plato sought, we must turn to the sciences, the arts, and, perhaps, to speculative philosophy--to philosophy as the articulation of a vision rather than the solution or the dissolution of problems. They tend to conceive of empirical philosophy as a propaedeutic to critical philosophy, and of critical philosophy as a propaedeutic either to speculative philosophy or to a post-philosophical culture.

Metaphysicians, on the other hand, believe that some species of empirical philosophy will bring us to truths which will offer us wisdom in Plato's sense. They believe that the visions of the speculative philosophers can be backed by argument, and that criteria exist for choosing between such visions. Metaphysicians keep the Platonic faith that argument can bring us to truth--truth about the ultimately important things. Analysts have lost this faith. They have not it (pace Adler, Gilson, Heidegger, Blanshard, and many others) because they have adopted a false epistemology or a false metaphysics. They have lost it because they have looked at the history of philosophy and at our culture and arrived at a judgment about how human life may best be enriched. I believe that their judgment is sound. But I do not know how to prove that it is.

This conclusion is an important marker in Rorty's corpus for a number of reasons. I think it brings together for the first time a number of themes that will become very common in Rorty's later writing. 1) I think it formulates for the first time the lines between pragmatists and Platonists/realists. 2) It defers Platonic wisdom to science (for predicting and controlling), the arts (for moral guidance), and to "speculative philosophers", i.e. strong poets. 3) It reconceives the philosopher in terms of Socrates' wisom: "not being fooled anymore". This is a theme that doesn't appear a lot, but one place it does is in "Philosophy in America Today" (in CP) where Rorty reconceives the model of the analytic philosopher, not on the scientist (as before) or the scholar (as in Europe), but on the lawyer. 4) It shows Rorty first coming to terms with the fact that, as he will later make much of, there is no way to argue between the pragmatist and realist (which recurs most reverberatingly in the Introduction to CP, at the end of "Pragmatism, Relativism, and Irrationalism" (also in CP), and in the beginning of CIS). Most interestingly to my mind, though, is seeing Rorty make an unconscious slip of projection. Maybe it was because he was young, merely in his mid-thirties, that caused Rorty to either misjudge or hope for the best of his colleagues. But either way, Rorty is sorely wrong to think that it is anglophone philosophers' "examination of the history of philosophy [that] leads them to think that philosophy is indeed, a tree from which all, or nearly all, the branches have fallen." That his analytic colleagues had "lost faith" because "they have looked at the history of philosophy and at our culture and arrived at a judgment about how human life may best be enriched." My bet is that Rorty, at the time, didn't quite realize at that point the extent of the relative historical illiteracy of his colleagues that he would later diagnose in them. "Lost faith"? That is definitely the Rorty we see in "Trotsky and the Wild Orchids" and in interviews, not the run-of-the-mill analytic philosophy professor. Rorty was definitely projecting his route to pragmatism on his colleagues, but who can blame him? He was young. And on the other hand, it may be a precursor to Rorty's famous "we's", his warm, cozy appellations of what "we pragmatists" or "we liberals" think.