Rorty’s Philosophy and the Mirror of Nature suffers from conceptual distortions fatal to the project he sets out to complete. Through these distortions the underlying issue the book addresses is rendered obscure and unsolvable, such that addressing this issue using the tools offered is like trying to rebuild an accurate reflection of oneself in a fun house mirror, an odd mix of delight and exasperation that leaves one worse off than plain and simple ignorance of what one really looks like.
First and foremost, his main argument suffers from a serious self-referentiality problem. This is not the churlish accusation of self-reference, like the liar’s paradox, or the accusation that the statement “all truths are relative” applies equally to itself, therefore making that “truth” both absolute and relative, or rather self-contradictory. It’s not a self-referentiality problem of formal logic. Rather it is an existential one of execution.
Rorty devotes the entire third part of the book to developing and extending Kuhn’s notions of normal versus revolutionary science in terms of the incommensurability of paradigms, with a special emphasis on the role of the familiar and unfamiliar to investigators. For him, inquiry is a matter of a hermeneutics that respects the abnormal and unfamiliar, as opposed to an epistemology that seeks to justify knowledge claims using terms familiar to its own paradigm, always with an eye to what it calls ‘the true source of knowledge’, to wit, the framework for permanent justification. To abbreviate the argument here, Rorty says that competing paradigms making claims over the same issue are incommensurable, in that there is no permanent, neutral, non-contextual matrix for assessing the knowledge claims between them. Instead of being adjudicable, they mark descriptions and re-descriptions of varying usefulness, “usefulness” presumably being defined as an admixture of prediction, control, familiarity, and comfort. This incommensurability represents the culmination of the book, on which he bases his claim that the arts and the sciences belong on equal footing as just so many alternatives in the effort to describe and redescribe ourselves and the world we live in.
Except that he doesn’t put this theory into practice in the second part of the book. In fact, in the second part he violates both the letter and the spirit of what he concludes in the third. For how else can the history of philosophy from Descartes to Russell and Husserl be seen together as representing one continuous attempt to do epistemology, i.e. to mirror nature through an analysis of mind seeking to justify knowledge in privileged representations? Conceptually, don’t Descartes and the rationalists belong to a different “paradigm” than the empiricists who followed them (or were sometimes their contemporaries)? So too with Kant and the idealists who drew from him. Don’t Husserl and Russell too belong to different paradigms, both between the two of them and with respect to the rest? One can parse the paradigms in different ways without disagreeing that some parsing into paradigms is necessary, and it simply won’t do to say they all belong to the same paradigm because they all seek to explain knowledge through a representational theory of mind. In the first place, they don’t, and in the second place, even if they did, by that logic there would be no scientific paradigms either because science since Newton science has always tried to explain the world with mathematical models. In the final analysis, wherever the battle lines are drawn regarding the paradigms—and they must be drawn somewhere—all of them are for Rorty commensurable; they are adjudicated—not against a permanent neutral framework as the disclosure of truth and being, for sure (Rorty doesn’t make the mistake he accuses epistemology of making). But nevertheless they are adjudicated. Against what? Against what background does Rorty commensurate philosophies as diverse as modern rationalism and empiricism, transcendental idealism, contemporary transcendental phenomenology, and analytic philosophy a la Russell?
In practice it is the premise that they are all doing the same kind of epistemology, a premise argued using the assumptions, methods, and puzzles of the paradigm familiar to him, the “normal science” of analytic philosophy. This is evident not only in his explicit use of Sellars and Quine to dismantle two central claims of epistemology in part 2; it is also evident in his claims that Locke confuses causal explanation for epistemic justification and that Kant confuses synthesis with predication. Confuses? Who says? A causal explanation can be a justification, as when an employer asks why an employee is late for work, and the employee offers an explanation of what caused him to be late. Whether that explanation is also a justification is a question for the two of them to decide on some other ground. Maybe it is a good one, maybe it isn’t, but one surely begs the question at issue if one says, upfront, offering any explanation confuses explanation with justification. Maybe no explanation at all is necessary. Locke certainly thought his causal explanation of the operations of the understanding offered epistemic justification for parsing out competing knowledge claims, so who says it doesn’t just because it’s an explanation? A similar reasoning applies to Kant’s “confusion” of synthesis and predication. For Kant, predication in knowing requires a synthesis of representations, so who says the way he uses “predication” and “synthesis” is invalid?
Some analytic philosopher, apparently, not as an issue of fact that Locke’s particular explanation fails, or that Kant’s particular use of apriori synthesis yielding universally valid propositions doesn’t cut the mustard, but as a matter of principle that these things must fail: explanations and justifications cannot be the same; predication and synthesis must be differentiated. But why is not analytic philosophy its own paradigm, with its own normal operations, its own problem sets, concepts, and principles? Elsewhere Rorty insists it is. So why (and how) are these principles suddenly dispositive, in that they are suddenly “commensurable” with and thus able to “pass judgment” on Cartesian metaphysics, Lockean empiricism and Kantian transcendental idealism, all philosophies well over two centuries old, developed in entirely different social, economic—in general—historical settings? If the position developed in Part Three is valid—and recall that the previous arguments in the earlier parts of the book are attempts to say that it is—then how is Rorty immune to the charge of what’s good for the goose is good for the gander: judging Descartes, Locke and Kant (much less Husserl and Russell) against analytic philosophical distinctions using analytic philosophical tools should be out of bounds; they should be judged in their own terms, specifically where their claims don’t live up to their own assumptions and arguments, where those assumptions and arguments lead to paradoxes or aporias that can’t be explained away even as they must be explained in accord with their own stated aims. To do otherwise is to end up begging the question, as he does, just as anyone inevitably does when they judge distinctions and concepts from one “paradigm” using distinctions and concepts from a competing one. Why then is Rorty not in the same pickle as Galileo and Bellarmine, with Galileo relying on a telescope and Bellarmine relying on scripture, when ostensibly arguing over the ‘same thing’? Rorty has no answer. In fact, he doesn’t even see the problem, and this blindness is rather fatal to the utility of the distinction he draws. Not only does it leave one going nowhere new; it can’t even get itself off the ground to get there.
Another dimension of the distortion caused by Rorty’s self-reference problem can be seen in a pair of related false (read useless) dichotomies he offers, the first of which pertains to his bete noire—namely, science as the premier model of valid knowledge—and the second of which ultimately grounds the first—namely, the distinction between epistemology and hermeneutics.
As an antidote to science’s frequent claim to be a privileged form of valid knowing, one that sometimes claims jurisdiction over other forms, Rorty claims that natural science, as a description of “ourselves” or “reality” is “on par with the various alternative descriptions offered by poets, novelists, depth psychologists, sculptors, anthropologists, and mystics.” To this end he says “normal scientific discourse can always be seen in two different ways—as the successful search for objective truth, or one discourse among others, one among many projects we engage in.”
The dichotomy, of course, is silly; science can be—and in fact is—both. That is, science can be a discourse, one project among others that we engage in, and still be one that is a successful search for “objective truth” (read: reliable knowledge). Perhaps the particular form of knowledge yielded by science shouldn’t be seen as an arbiter of all other forms of knowing, but the only way that it couldn’t be “a way of reliable knowing” because it is also “one description among others” is if one exports a specific assumption from a “normal paradigm” (in this case hermeneutics) into the dichotomy—an assumption that says a discourse like science either 1) has to be a source of objective truth (which it can’t be because hermeneutics stipulates there is no such thing), or 2) has to be what hermeneutics says it is (which Rorty claims is true because of his arguments for hermeneutics). But why even offer a dichotomy like that, since it amounts to nothing more than silly, circular declaration stating that “x is the way it is because the assumption stipulates that that’s the way it is.” All arguments might ultimately boil down to similar reaffirmations of an assumed premise, but that just shows how limited arguments are for clarifying subject matters with sound distinctions, something any pragmatist should minimally do. Yet Rorty relies on an argument like this throughout the entire book. In fact, this larger argument contextualizes all the specific arguments he makes against other related distinctions. In any case, as it stands, Rorty’s way of looking at “science” and “alternative descriptions” simply begs the question entirely with a dichotomy that exploits an assertion that is trivially true of any description making any kind of claim with regard to ‘objective knowledge’ relative to competing claims—in other words, it’s useless. As an anal-lytic philosopher-convert to pragmatic hermeneutics, he should reject the dichotomy himself simply for failing to re-describe anything worthwhile. That’s the pragmatic way of saying it is not just silly and useless but also false.
The second and most pervasive false dichotomy (again, useless) in PMN has already been approached to some extent above: it is the very distinction Rorty makes between epistemology and hermeneutics, the one he relies on implicitly throughout his entire argument against epistemology and for hermeneutics. For in it Rorty boils down the choices for adjudicating competing knowledge claims in different paradigms to two: either one has a permanent, neutral, non-contextual framework for adjudicating them (epistemology), or they cannot be adjudicated at all (hermeneutics). In short, either there is “a permanent framework for inquiry” (which for him there is not), or no inquiry can be judged more reliable, more accurate, more believable—in a nutshell, more true or false—than another (which for him they can’t). This dichotomy pervades, indeed is, the entire argument of the book.
Now it has already been suggested that Rorty in practice doesn’t hold himself to his own dichotomy as he adjudicates “epistemology” across 350 years of various philosophical paradigms through the single lens of mirroring nature with privileged representations in an attempt to justify knowledge. After that adjudication he rejects this epistemology as—what? Inadequate? False? Misguided? Useless? To some extent, all of these, but is the move justified? On what grounds can one be persuaded to make it? To repeat the point, in order to get to the moment of pragmatic decision between epistemology and hermeneutics off the ground, Rorty has been adjudicating all along. What matrix, then, could he possibly have relied on, since by his own stipulation adjudication is an exclusive either/or proposition: one either has a permanent neutral framework, or one doesn’t. Since his critiques of Lock and Kant and just about everyone else presuppose the normal paradigm of analytic philosophy, is that the framework? In execution for Rorty it looks to be—and therein lays problem: it can’t be because those arguments merely belong to one “paradigm” among others—unless, one claims, they have independent validity, that they represent real progress in real, “objective” knowledge. But again, Rorty can’t claim that; he denies such a thing exists (except, apparently, for the tools he’s familiar with). If he can’t claim that, then what matrix allows him to arbitrate? What force can that arbitration have to persuade one to choose hermeneutics over epistemology? Whence the persuasion? The circle just goes on and on and the question gets begged over and over again. One never even gets off the ground to a point where the dichotomy can be used as a guide for making an informed decision. That’s the pragmatist’s way of saying it’s not just useless but also false.
Now is not asserted here that cross-paradigm assessment cannot be done; pragmatism asserts that it can. It is only said that Rorty cannot execute his enterprise of paradigm adjudication using the paradigm of analytic philosophy (which he does) and be consistent with his own theoretical assumptions and conclusions (which he isn’t). This failure in practice makes the dichotomy he relies on fatal for his own effort; it is useless, and therefore, equivalently, false. Rorty apparently doesn’t see this as a problem, and this blindness calls into serious question the utility of the arguments he makes. For in so far as they remain stuck in the paradigm in which he is comfortable working, their force never really gets one to the pragmatic point of decision to which Rorty wants to bring his readers. As stated before: not only does his hermeneutics leave one going nowhere new; it doesn’t even get one off the ground to get there, merely begging as it does the very question it poses. Instead of laying out here how arbitration between “paradigms” is in fact possible, it is only asserted that common sense shows that arbitration between competing knowledge claims occurs all the time, both within and across paradigms, without recourse to the permanent matrix of epistemology. That arbitration is pragmatism. The problem is, by starting with the dichotomy between absolute adjudication or justification versus no adjudication or justification, Rorty abandons pragmatism before he even starts.
It’s a shame that Philosophy and the Mirror of Nature employs such distortions. The underlying issue it addresses—commensurability versus incommensurability in an arena of inherently contingent but often competing knowledge claims—is an important one. It’s just that the mirror Rorty holds up to philosophy from Descartes to Russell and Husserl fails to reflect its predilections faithfully on the whole, even if some parts are shown back with only minor distortions. By his own stipulations, it can’t even perform this reflection, and his attempt to do so anyway distorts the underlying issue so much that the problem becomes unsolvable—and this is a failure worse than posing the question correctly and offering a wrong or partial solution. At the end of the day one is left more frustrated than delighted by the effort of working through Philosophy and the Mirror of Nature, which doesn’t say much for any promise of edification the book might contain. The fun-house mirror of nature is best addressed by being avoided in the first place, if only to save one the frustration of a useless formulation of important problem.
 Remember, this is Rorty’s own logic of incommensurable paradigms talking here, not the anus’. The question by the logic of Rorty’s own Part Three commits him to the burden of establishing some kind of independent validity if he is going to arbitrate ‘across’ paradigms, as he does.
 A prime example of begging the question is his claim that Kant’s assumption that ‘a manifold is given but a unity is made’ is “unjustified.” It is, however, only unjustified if psychological nominalism is true. Now psychological nominalism is a theory specific to analytic philosophy; it has no traction in non-analytic philosophical studies of language and perception. Setting aside the fact that there is no evidence offered for it, that it remains a stipulation (one that is proven false by everything known about infant and primate cognition), its exclusive use within analytic philosophy alone makes it, pragmatically speaking, part and parcel of a paradigm. As such, it can have no independent validity, and therefore it cannot bear the weight Rorty puts on it (which is considerable) without begging the question.
 This is taken up in some detail in E-piss-temological Behaviorism.