viernes, 19 de noviembre de 2010

On Concepts and Objects: Can the Gem Destroy the Circle of Correlation?

On Concepts and Objects:

Can the Gem Destroy the Circle of Correlation?


 Ray Brassier’s Concepts and Objects, his contribution to the forthcoming anthology The Speculative Turn (edited by  Nick Srnicek, Levi Bryant, and Graham Harman), advances a programmatic overview of his forthcoming work. In addition to an acerbic assault on Latour and Object Oriented Philosophies which, in a surprising turn, he ends up accusing of being ‘correlationist’, the main thrust of the paper deals with the intricate juncture between metaphysics and epistemology. I will focus here on Brassier’s appropriation of Stove’s indictments against the idealist ‘gems’, which he extends against correlationism. Ray’s basic point of contention, against Meillassoux’s avowal of the strength of the correlationist circle, is that the latter suffers from an ambiguity which, once clarified, makes its purported inescapability seem all the less obvious. In order to show this, Brassier seeks to show how the disambiguation provides two possible claims about the relation between conditions for determination and objects determined by these conditions, one of them trivially true, and the other one speculatively dogmatic or at the very least questionable.

    Case in point, Brassier disambiguates between a) the tautological claim that one cannot posit something without positing it, from b) the less obvious claim which asserts that things cannot exist without being posited. The correlationist can  subscribe to the former without going beyond inanity, but fails to legitimately establish the latter. Now, Brassier stipulates that the correlationist may substitute ‘positing’ with another term, according to the specific brand of determination sought (language, perception, intentional consciousness, culture…). But the basic point is that the correlationist goes on to illegitimately extrapolate the ontological dependence of objects on concepts, of being on thought, from the trivial claim that objects are expressed or postulated through concepts, and that it is thought which gains traction on being.

      Brassier ultimately seeks to rehabilitate the distinction between concepts and objects, in order to reintroduce the epistemological filter on dogmatic metaphysical claims, Latour-inspired OOO included. In this occasion, I would just like to point out some points where I nevertheless think that Brassier’s attempt to extend Stove’s Gem to deflate the correlationist strength is complicated by some considerations about the precise distinction between correlationism and idealism, since it is against the latter which Stove’s reconstruction is ultimately designed to work. I have presented these observations to Ray recently, who seems to think that they are of some interest, and has said he will reply to me in detail at some point. In any case, I share these observations since I myself am a bit confused on the precise force of Stove’s gem. Anyway, on to the ‘things themselves’!

    The point of departure in Brassier’s paper concerns a presentation of Berkeley’s classical polemic against the possibility of pursuing a thought of things independently of their being thought. Let us follow the specifics in Ray’s presentation, in an schematic manner:

1)      After presenting the passage from Berkeley, paragraph (32) states the following:

“From the indubitable premise that ‘One cannot think or perceive something without thinking or perceiving it’, Berkeley goes on to draw the dubious conclusion that ‘Things cannot exist without being thought or perceived.’ Berkeley’s premise is a tautology, since the claim that one cannot think of something without thinking of it is one that no rational being would want to deny.”

     Moments later in the paper, however, Brassier asserts in paragraph (34) that the Berkeleyian Gem “does not assert that there is no mind-independent reality, it merely says it must remain inconceivable”. These two readings seem to be obviously in conflict, and I take it that it is the latter which is representative of Berkeley’s account. That is since he repeatedly emphasizes the absurdity of trying to conceive of a reality independent of us. But from this it certainly doesn’t follow, as Brassier rightly puts earlier, that mind independent reality cannot or does not exist, but simply that it is impossible to prove it does or doesn't, since what we know is relative to the possibilities and to the content given to and mediated by us.

2)           In paragraph (32) Brassier writes again against Berkeley that: “…the metaphysical claim that only minds and their ideata exist is supposed to be the consequence of Berkeley’s argument, not its presupposition”. We know, however, that for Berkeley, ideata are not just properly our concepts or propositional attitudes, but also our imaginings and perceptions. The entire point, I take it, following the basic Lockean/Cartesian problematic of our epistemic relation to quality, is that even pre-conceptual contact with the world is mediated through our faculties (of sense and thought), and so that conception is always anchored on appearances given in and as experience.

      But if this is right then, at least in the passage in question, Berkeley does not seem to be claiming only ideata and minds exist, but solely that there are no epistemic grounds to surmise the existence of beings outside how they are given to us, since everything we know are ideata and these are relative to our faculties. This indicates that to conceive of things independently of givenness implies either knowledge that things exist outside mediation, or knowledge of how they exist independently of the conditions of givenness. But this is precisely what the correlationist  contends we have no warrant in claiming. Meillassoux highlights how the correlationist must deny that knowledge of particular beings obtains in order to assert knowledge of the unreason for things to exist as they appear (Pete Wolfendale highlights this point in his Essay on Transcendental Realism). Since for Berkeley primary/secondary properties are anchored on sense-data, and these prove to be fallible / subject to distortion (thus being specific to our apparatus for mediation), conception as derived from experience cannot be transposed to an independent reality, at a loss for reasons.  At the very least, however, against Brassier’s stipulation of the dogmatist kernel of this move, there is an argument for the claim that all we know are ideata, based on the errancy and deceptiveness of perception, which grounds conception, and thus finally existential affirmation.

    We know that for Sellars (whose naturalism has inspired Brassier’s more recent work) it is precisely the relation between perception and ideation which becomes complicated; but here is where the transcendental realist needs to come in with more than the accusation against the Gem, since the correlationist draws from the empiricist/rationalist attack against the epistemic reliability of perceptual data to substantiate the constriction of our access to the content of phenomena. Thus, when Brassier writes that it “by no means follows that we cannot conceive of things existing independently of concepts, since there is no logical transitivity from the mind-dependence of concepts to that of conceivable objects”, the correlationist claims that it is not the mind-dependence of concepts which logically entails the absurdity of concept/mind independent reality, but the mind-dependence of concepts and perception which empirically puts the hypothesis of mind-independent reality at a loss for justification. The problem is that on the basis of the quoted passage, it seems that Berkeley is to be read as a correlationist rather than an idealist, since in the way of claiming the ‘absurdity’ of the in-itself he appears committed to the unreason of our limited perceptual-conceptual apparatuses (thereby playing a similar role to the one facticity does for the correlationist).

         This is also why we should probably distinguish the Berkeleyian view from the Absolute Idealist, for whom ideation is thoroughly conceptual: the former does nevertheless admit the passive dependence of our phenomenal objective experience on intuition. Having said this, I’m not sure Stove’s attack on the Gem can safely target the weak correlationist agnosticism as well as it can the idealist assimilation of the in-itself. Since the former’s justification is unreason/facticity rather than the formal conflation of being/thought, it’s not clear to me that the circularity proceeds in both cases.

3) In paragraph (34) Brassier claims that proponents of the Gem confuse independence and inaccessibility. That is, they confuse the claim that we cannot access a mind-independent reality (since it must be given through conception) with the claim that there can be no mind-independent reality. But if the prior considerations are correct, it is not so much that the correlationist claims that mind-independence is logically inconceivable, but rather empirically vacuous. ‘Proof’ would then be impossible not because of a performative contradiction resulting from the Gem, but simply because the fallibility/perception-dependence of ideata constrain us to the phenomenal.

Again, it seems that under this reading the correlationist could perfectly admit that a mind-independent reality could exist, and that we can conceive of this possibility. This Meillassoux underlines as the peculiarity of correlationism’s avowal of facticity, against subjective / absolute idealisms. We trivially cannot conceive without an act of conceiving. But more crucially, since all the content of experience is also given within experience, and given that this content is of experience as mediated by our own cognitive-perceptual faculties, reflexivity cannot make this content/knowledge necessary for being outside such constrictions, lest we extend the constriction to things themselves (which Meillassoux does in his absolutizing of contingency).

     The ‘absurdity’ for Berkeley or the correlationist under this reading would thus be to think that what we know provides warrant for proof of the in-itself. Therefore, the conceptual inaccessibility of the in-itself would follow from its vacuity given the facticity of thought/perception. This is just to reassert the ‘correlationist agnosticism’ Meillassoux diagnoses, rather than the alleged logical fallacy of the idealist Gems attacked by Stove.

     In the end, we must not forget the import of the argument against the reliability of perception for knowledge in prefiguring the correlationist agnosticism about the in-itself. It is not just that perception is often deceptive, but that it is quite impossible to think what a reality outside the mediation given by perception and our concepts would be; let alone grounds to prove that such a reality   exists, or even more, that it does in a way that already corresponds to the content targeted by our concepts. This is the point of contention for the realist, who claims that our perceptual/cognitive mechanisms do provide resources to think robustly of the in-itself. More specifically, it is the contention of the scientific realist that science provides the best resources for thought to gain traction on being, while preserving their relative independence.

3)      In paragraphs (40)-(42) Brassier disambiguate between word, concept and referent. He writes:

“It might be objected that we need Saturn to say what Saturn is; that we cannot refer to Saturn or assert that it is without Saturn. But this is false: the first humans who pointed to Saturn did not need to know and were doubtless mistaken about what it is: but they did not need to know in order to point to it. To deny this is to imply that Saturn’s existence—that it is—is a function of what it is—that Saturn is indissociable from Saturn (or whatever else people have believed Saturn to be). But this is already to be a conceptual idealist. Even were the latter to demonstrate that the conditions of sense determine the conditions of reference, this would still not be enough to show that the existence of the referent depends upon the conditions of reference. To do that, one would have to show that ‘to be’ means ‘to be referred to’; an equation tantamount to Berkeley’s equation of ‘to be’ with ‘to be perceived’; yet it would require more than another Gem to dissolve such a fundamentally normative distinction in meaning.”

     Here there is an odd shift in the reading of Berkeley, which is now said to conflate being with being-perceived (whereas earlier it was stated he conflated being with being-conceived). But if the possibility of the in-itself is preserved outside conception or thought in B’s account, then neither reading really suffices, at least on the basis of the quoted paragraph.

        More importantly, however, is the suggestion that ostentation does not require conception. I’m left wondering if this really helps the case against correlationism, since it appears they’d be hostile to the assumption that they have already ‘smuggled in’ the equation between sense and reference. This is because the correlationist claims that even if Saturn is dissociable from Saturn insofar as what we intuit can be conceptualized in a variety of ways, the latter is not dissociable from perception, insofar as any conception is finally anchored on intuition (or the imagination’s reproduction of the same, to use Kant’s language). Since conceptual indices (even primary properties, which are conceptually specified) follow from perception, ostentation does little to move us beyond the domain of mediation. This complicates the following diagnosis of correlationism at large, provided in paragraph (44), characterizing the Gem as a…:

     “…Form of argumentation that slides from the true claim that we need a concept of mind-independent reality in order to make claims about the latter to the false claim that the very concept of mind-independent reality suffices to convert the latter into a concept, which is by definition mind-dependent”

The problem is that for the correlationist it is perception, not just conception, which is also relative to conditions of givenness. And whatever determines perception is said to exhaust the content of our concepts, since conception is anchored on sensing, or in imagining. Thus, for the skeptic/agnostic/idealist/weak correlationist/OOO Saturn would seem to be a function of our sensing, which in turn fails to guarantee knowledge or to provide indices for mind-independent realities, since it is fallible and relative to our very peculiar perceptual mechanisms. Strictly speaking, the correlationist could argue that when humans point at what we call Saturn, they are really pointing at mere sense-data which constitutes the content of the being relative to us. But it no more follows from this that a ‘real Saturn’ exists or doesn’t exist outside our minds than  an unspecified plurality of ‘Saturn-stages’ does, since individuation remains relative to conception, and the latter to sensing. In support of the last claim the correlationist just has to rehearse the usual diatribes about the seemingly endless plurality of meanings and arbitrary ‘partitions’ on the intuited solicited by conception, and thus to some form of primitive ontological indeterminacy. He can thus merrily revert into the agnostic quietism about the existence of the in-itself, since in any case knowledge remains restricted to the straightjacket of mediation. For the realist then, being-in-itself need not just be refractory from conception, but from perception.

This is why I take it Brassier finds special importance in not just a postural realism anchored in real but ultimately unobjectifiable ‘anonymous’ being. Rather, what is needed is the possibility of discrimination between statements which gain traction on being from those which don’t, given a robust conception of reality. This is, I assume, why Sellars’ attempt to draw the grounds for naturalism and realism at the crossroads of conception-perception is particularly significant.

          In any case, it seems that even if Berkeley did endorse a classical Gem, which incurs in the ambiguity in question, other alternative correlationist positions can resist the indictment of having denied the in-itself in the process of conflating the act of thinking with the physical object of thought. They can claim this by saying that although the physical constitution of the being is indeed relative to the act of thought, this doesn’t rule out the impossibility of the in-itself, but merely makes any such content inaccessible and so unknowable. The correlationist can preserve the in-itself in different forms this way, while insisting in that quiddity remains relative to conditions of disclosure. For example, in Heidegger’s account, the Vorhandenheit realm of concrete particulars targeted by scientific rationality remains relative to Dasein’s practical/cognitive comportments. Yet this doesn’t rule out the withdrawal of beings in concealment which is in excess to our comportments, albeit this way the in-itself becomes an increasingly anonymous and unthinkable Otherness, whose presumed givenness is quasi-religious and dubious for reasons Sellars has rightly pointed out.

       However predictable this correlationist strategy may have become, they can always go ahead and associate whatness to the disclosure of beings in ekstatic transcendence for temporal Dasein, eidetic constitution in intentional consciousness, conditions of perception in representation, etc. This finally constricts being-for-us within the factical ‘unreason’ of thought/perception. Considering that the content of thought remains constrained by mediation of sensing-thought, the thesis of the in-itself remains shrouded in darkness or veiled in religious mystery, but never in principle ruled out. This is why I take Meillassoux’s avowal of the force of the circle of correlation resists its wholesale deflation, as attempted by pointing at the fallaciousness of the Gem. Rather, for Meillassoux the way out of this circle is to force the correlationist to accept that the contingency of the being must already be a feature of things in themselves to avoid the idealist appropriation, i.e. in order to say things could be different than they are contingency cannot be merely relative to us, lest we reinstate the idealist coup. But this still means that the knowledge of particular beings gained within the circle of correlation is contingent and uncertain, i.e. what we know is the absolute necessity of the contingency of beings as such.

         The question is finally the obvious one: if we want to say that any conceptual determination does gain traction on the in-itself, and that conception/perception can furthermore already imply the stratification of being itself, then it falls to the realist to show how this is so. And this must be done without invoking a miraculous continuity between conception/description and being, such as in the descriptive metaphysics of OOO which simply solicit all kinds of claims about the in-itself, even incompatible ones. I take it this is also the limitation of a postural realism such as Laruelle’s as advanced by determination-in-the-last-instance; since it ultimately lacks the robustness to subtract from Decision or science anything but the real cause, utterly indifferently to the descriptive register in objectification:

     ”How can we acknowledge that scientific conception tracks the in-itself without resorting to the problematic metaphysical assumption that do so is to conceptually circumscribe the ‘essence’ (or formal reality) of the latter?”

     The stratification of being in relation to the phenomena described by natural science I take to be the decisive import of Sellar’s approach, while the normative criteria for rational deliberation and the withdrawal of authority required for objective truth is expounded in Brandom’s account (and, of course, in the prospective project initiated by Pete Wolfendale).

Kant's Synthetic A Priori

Kant’a Synthetic A Priori 
               Kant's notion of synthesis lies at the center of his philosophy, and of his purported overhaul of classical metaphysics, as well as of Hume’s skepticism and empiricism. He conceived of his project as being a natural extension of the Copernican Revolution in speculative thought, which would place philosophy in the ‘secure path of the sciences’, in continuity with the ideal set by the Enlightenment. He characterized his project in the first Critique on general terms as follows:

       “Up to now it has been assumed that all our cognition must conform to the objects; but all attempts to find out something about them a priori through concepts that would extend our cognition have, on this pre-supposition, come to nothing. Hence let us once try whether we do not get further with the problem of metaphysics by assuming that the objects must conform to our cognition, which would agree better with the requested possibility of an a priori cognition of them, which is to establish something about objects before they are given to us. This would be just like the first thought of Copernicus, who, when he did not make good progress in the explanation of the celestial motions if he assumed that the entire celestial host revolves around the observer, tried to see if it might not have greater success if he made the observer revolve and left the stars at rest” (Preface to Critique of Pure Reason B/XVI)

    The basic idea is the following one: Kant accepts Hume's indictment against Cartesian rationalism in saying that knowledge is anchored in experience. This means that we have no transparent access to the contents of the world as it is 'in itself', i.e. classical metaphysics is nothing but naive realism.

    However, he rejects Hume's skepticism about causality by claiming that empirical knowledge is constituted by a priori subjective faculties. As such, experience is not given to things-in-themselves (noumena), but to how things appear to us through these faculties (phenomena), i.e. knowledge is inherently mediated and so because of that it is representational. The class of the noumenal, however, must be distinguished from the category of object: the latter are functions of the synthesis given to the subject, while the former are the mere limiting concept of a being separable from the way it is given to us through perception in sensibility (CPR, 270-274). This is the sense in saying that noumenal reality is a negative concept. We can define Kant’s basic ontological position thus according to what the philosopher Quentin Meillassoux has called ‘correlationism’: the thesis that we have no access to being-in-itself, but only to the relation between being and the subject. In its ‘weak’ version, which Meillassoux attributes to Kant, the in-itself is not given, but is thinkable (After Finitude, Pgs. 23-27). This is decidedly different from the Absolute Idealist who claims the in-itself is identical to the for-itself of phenomenal thought (as we will see in for example, the case of Hegel).

    Further, since causality is for Kant now said to necessarily derive not from things-in-themselves, but from the phenomena of experience, Kant argues that we can gain certain knowledge of causal necessity about objects, given that we adequately clarify how the subject comes to constitute experience as a result of his own faculties. This is the basic impetus guiding the idea of synthesis: it concerns elucidating those necessary or a priori features which articulate the content of experience into knowledge. Thus synthesis is for Kant an operation which involves both a) affects received passively by the subject in order to form objects, and b) a capacity for actively applying categorical judgments about these objects to discern concrete features which determine the objective content.

     This means that every occasion of knowledge implies synthesis, but it is through the critical enquiry into the principles of knowledge that we understand the a priori features of synthesis, and the faculties they involve.  Correspondingly, if knowledge is to be saved from Hume, it is because we have access to the structure of reason, and thus to the forms of synthesis that articulate experience.

    This critical enquiry into pure reason must seek an answer to the question 'How are a priori synthetic judgments possible?'. This kind of knowledge is for Kant tantamount to judgments about that which simultaneously gives form to the content of experience (and thus synthetic), and yet which is necessarily prior to all actual contingent experience.

     It is these kind of judgments which clarify how the subject organizes the content of experience, and so clarifies the features of those phenomena given to us, i.e. the synthetic a priori is ontological knowledge (Ontologia is Metaphysica Generalis, CPR 845f, B873f, A 247, B 303). Each occasion of experience is thus already determined by this (a priori) synthesis which is the act of the transcendental subject.

    "Knowledge which brings forth the quiddity of the being, i.e. knowledge which unveils the being itself, Kant calls "synthetic". Thus the question concerning the possibility of ontological knowledge becomes the problem of the essence of a priori synthetic judgments" (Heidegger, KPM, Page 9)

      Unlike analytic a priori concepts which concern merely conceptual determinations (matters 'of definition'), the synthetic a priori determines/specifies the content given to experience. It is not therefore abstracted from experience, but applies to it in every case as a matter of necessity. The faculty conformed by a priori principles is what Kant calls 'pure reason', and the enquiry into them is thus the object of the first Critique. Correspondingly, the enquiry into these principles is titled transcendental, insofar as it deals with how the act of the subject constitutes the content it passive receives in affection:

   "I entitle all knowledge transcendental that is occupied in general not so much with objects as with the kind of knowledge we have of objects, insofar as this is possible a priori." (CPR, B25)

     “Transcendental knowledge enquires not into just actual experience, like empiricists purport, but more generally into possible experience. Thus the principles of pure reason Kant also calls 'conditions of possibility' of all knowledge; it is a matter of "transcendental truth, which precedes all empirical truth and makes it possible." (CPR, a 146, B185).

      This enquiry is the most difficult for Kant since "...nothing yet lies at the ground as given except for reason itself, and thus seeks to develop knowledge from its original seeds without seeking the support of any fact." (Prolegomena zu einer jeden kunftingen Metaphysik, $4, Werke, IV, Pg. 23). However, if all objective knowledge is synthetic then the critique must enquire into what exactly is being synthesized. This means that transcendental enquiry concerns just as much the affects, or the general ways in which things are given to us in sensibility, as well as the fundamental operations endowed to the content of sensibility by thinking in the understanding.

- The former enquiry is tethered to what Kant calls the 'Transcendental Aesthetic', which concerns sensibility. This concerns the way in which we receive the matter of experience.

- The latter enquiry is tethered to what Kant calls the 'Transcendental Analytic', which concerns understanding. This concerns the way in which we create or in-form data through concepts in thought.

     We should mention in passing that it is because the transcendental subject needs thinking to supplement the content given to intuition and organize it that Kant can say that the subject is essentially finite. The pathos of finitude, which extends all the way into the 20th Century's obsession with the limits of our knowledge, is doubtlessly anchored and prefigured in Kant's basic thought that we only ever experience phenomenal entities, whose constitution is synthetic insofar as we are subject to receptivity, and need the contribution of the understanding to get 'objects'. Kant was thus, as Alain Badiou remarks, 'without doubt the creator in philosophy of the notion of the object', i.e. the object as something which stands-against a subject, and which is not himself, thus pointing to his limits (that which ob-jects to it). An infinite being, on the other hand would produce the being at the same time it intuits it: thus not needing thinking to organize the content of experience, since it would directly form the being-itself (the noumena) as it thinks it. Heidegger puts it nicely:

     "The question concerning the possibility of a priori synthesis narrows down to this: how can a finite creature, which as such is delivered over to beings and is directed by the taking-in-stride of these same beings, know. intuit prior to all instances of taking the being-in-stride, without being its "creator"? (KPM: Pg 30)

Object, finitude, phenomena, synthesis, sensibility and thinking are thus all to be thought together as constitutive of the act of the transcendental subject.  

     With respect to the Aesthetic, we have already hinted towards its link to perception (aesthesis means perception in Greek). As we indicated, this part of the Critique concerns the enquiry into the faculty of sensibility. Although this faculty is fundamentally receptive insofar as it concerns data given to the subject, its content is already structured in a pre-conceptual manner. Kant clarifies that all active thinking organizes the content received by our sensibility, which is in other words what we passively intuit. However, since intuition is already given-to the subject, it is already conditioned or ‘formed’ by it, irrespective of the concrete contribution of the understanding (which we will review next).

    Therefore, the most general features of this receptive capacity are at the same time the most general conditions of possibility for any thought to synthesize sense-data into objective experience. These conditions of possibility Kant calls the 'pure forms of intuition': space and time. These already constitute representations insofar as they are already given to the subject, mediated by our finite sensibility, as we indicated above.

     It is important to note that these two forms are given in a unified and thus articulated way. Kant writes about space that "Space is represented as an infinite given magnitude" (A 25, B 39). This means that space (just like time) is not an empirical perception, or an object which gives itself within experience among others, discernible through concepts in judgment. Rather, space and time are the general forms or conditions under which any object of experience must be first given to the subject, logically prior to any act of constitution by the understanding. To say space is infinite in this regard simply means to say it does not correspond thus to any empirically given or specific magnitude, all of which are necessarily finite in experience. It is a 'unified whole', an original and 'pure representing' (A32, B48). But what is intuited is by itself not yet an determined object, even if it provides the content for such a determination. in the pure transcendental object Space constitutes in this way the ‘outer sense’ of synthetic activity.

    Time is likewise a pure form, which organizes succession of given data to our minds. It is "the form of inner sense, i.e. of the intuiting of ourselves and our inner has nothing to do with shape or position... Time is the formal a priori condition of all appearances whatsoever" (A33, B49). All appearances/phenomena must happen in time to subsist even as space, i.e. there is only something present insofar as it is present in time; and the same for the past and the future. Thus time has something of a logical precedence over space for Kant. Together, time and space form the pure forms of intuition.

     However, as we mentioned, time-space do not by themselves provide knowledge of any-thing; it has no objective content still. The pure transcendental object is undetermined, and a merely logical pole at the other end of the transcendental subject, but not yet robust. For objectivity to emerge, and thus the specificity of experience, Kant needs the active agency of thought which provides what he calls the 'pure concepts of the understanding', which are basically the basic ways in which the subject synthesizes of intuition-understanding in actual experience. As opposed to the pure forms of intuition, the analytic concerns the pure concepts of the understanding, or what Kant will come to call the 'categories' of the understanding.

    These constitute the basic capacity of judgment, that is to say, the way in which we specify an object X to be such-and-such (A80). Kant thus says famously that "the conditions of the possibility of experience in general are likewise conditions of the possibility of the objects of experience." (A 158) Possible experience is unified in the dyad of intuition-understanding, forms and categories, sensibility and thinking.

Kant describes the categories as 'functions of thought', which have objects as their outputs. They thus relate to all objects generally. For Kant these categorical determinations are essentially twelve, each corresponding to a possible judgment about an object, which we now list schematically:

The Table of Categories / Judgments (A 71)
Unity /Universal
Reality / Affirmative
Inherence-Subsistence / Categorical
Possibility/ Problematical
Plurality / Particular
Negation / Negative
Causality-Dependence/ Hypothetical
Existence / Assertoric
Totality / Singular
Limitation / Infinite
Community/ Disjunctive
Necessity / Apodictic

  Having laid these out, we can understand the crucial distinction between analytic and synthetic judgments. All judgments have a subject-predicate structure, but whereas in analytic judgments the predicate does not add anything to the subject (all bachelors are unmarried), in synthetic judgments they do (My Budweiser is flat). Synthetic a priori judgments are thus simply those judgments which predicate of the object content specified by the categories, i.e. all objects have a cause, no event has a cause... It is interesting that for Kant the statements of science and pure mathematics are actually synthetic a priori. The classical example he provides is that of the simple sum 7+5 = 12. According to Kant, the concept 12 is not intrinsic to 7 or 5, one is said to ‘synthesize’ these by analogy to occasions of perception (one sees one’s fingers…). Of course this conception did not pass the test of time; Peano’s axiomatization of arithmetic shows it is perfectly possible to think of these statements as analytic; while Quine on the other hand launches an attack and questions whether there are any analytic judgments against Kant’s primitive taxonomy (see his Two Dogmas of Empiricism).

        In a certain way, Kant's categories are continuous with the architectonic conception of Aristotle, except he inflects these determinations into the subject rather than locating them within the sphere of objects. Thus subsistence is not a property we know of things in themselves but is phenomenal: "In all change of appearances substance is permanent. This permanence is, however, simply the mode in which we represent to ourselves the existence of things (in the phenomenon)".  As can be expected, the particular layout of the table of categories, as well as the –architectonic view- of the subject in principle, became the source for posterior attacks from other philosophers.

    It is important to note however, that Kant needs both poles for the synthetic a priori to work. Without the understanding, the pure forms of intuition are void of objective specificity. The pure transcendental object stipulated by Kant would thus remain vacuous and helplessly general, one of which we know nothing in its emptiness: "This [transcendental object] cannot be entitled the noumenon; for I know nothing of what it is in itself, and have no concept of it save as merely the object of a sensible intuition in general" (CPR, A 253). Zizek here is illuminating:

    "Transcendental object, that is, the completely indeterminate thought of something in general," has the function of conferring "upon all our empirical concepts in general relation to an object, that is, objective validity" ( CPR, A 109 )... Transcendental object is the form of the object in general by means of a reference to which a priori categories synthesize the multitude of sensible intuitions into the representation of a unified object: it marks the point at which the general form of every possible object reverts to the empty representation of the "object in general." (TWN, Pg. 36)

    However, without intuition, we cannot know what a pure concept ever applies to, and thus we cannot have objective knowledge either:

    “When this condition (of sensible intuition) has been omitted from the pure category, it can contain nothing but the logical function for bringing the manifold together under a concept. By means of this function or form of the concept, thus taken by itself, we cannot in any way know and distinguish what object comes under it, since we have abstracted from the sensible condition  through which alone objects can come under it"

  Kant neatly distributes the act of the subject across these two faculties (CPR 131-136):
1)       We require the forms of intuition for the apprehension of sense data, its sequential-temporal ordering,
2)       We require the categories of understanding for the recognition of the object under the concept.

    However, there is an additionally intermediary ‘faculty’ or power, which is that of the imagination, which reproduces representations and produces combinations of the same (which Kant calls schemas, and which amount to procedural rules obtained from the synthesis of the imagination and time). This ambiguous role of the power of the imagination, somewhere in-between sensibility and the understanding opened the gulf for a series of speculative knots and quarrels with the tradition, of which famously Heidegger made the most of in his Kant and the Problem of Metaphysics (1928).

- It is interesting to note that Kant's transcendental subject introduces a split in the traditional conceptions of subjectivity, including the Cartesian subject. This is because the split between the noumenon/phenomenon goes all the way to the subject, i.e. there is a split between the logical subject which accompanies every 'I think', and the empirical subject or 'person' which can be identified with a substance. This is analogous to the pure transcendental object we discussed earlier to the actual object of synthetic experience.

      This is an important point, since one may ask: what exactly holds the variety of objective syntheses together? This transcendental subject of experience, which logically unifies experience, enacts by Kant what he calls 'pure unity of apperception':

       "There is one single experience in which all perceptions are represented as in thoroughgoing and orderly connection, just as there is only one space and one time in which all modes of appearance and all relations of being or not-being occur. When we speak of different experiences we can refer only to the various perceptions all of which, as such, belong to one and the same general experience. This thoroughgoing synthetic unity of perceptions is indeed the form of experience, it is nothing else than the synthetic unity of appearances in accordance to concepts... The unity of apperception is thus the transcendental ground of the necessary conformity to law of all appearances in one experience" (1929, A111-A127)

    It is the synthetic unity of apperception which provides the basis for a universal cognitive experience. Pure apperception is the 'wellspring' of the synthetic a priori which constitutes the unifying function of the purely logical subject. The unity of apperception guarantees that experience is always of the same subject, no matter the contingent experiences. The subject is thus never, like in Descartes sense, one more substance among others (the hypostasized res cogitans in the chain of being); but it is this pure and empty formalizing agency of unifying synthesis in apperception: morphe - “form ‘informs’ matter”). It thus bridges the empty logical necessity of the analytic a priori, with the contentful empirical contingency of the synthetic a posteriori. This makes Habib’s reading, for example, of the forms of intuition as being ‘in the mind’ problematic (Pg. 363); the transcendental subject is certainly not the substantial mind indicated by the Cartesian cogito, and its unity is merely formal. It thus contains no properties; its synthesizing function is transcendental insofar as the categories are purely logical, i.e. nothing is strictu sensu in the transcendental subject.

    Zizek has remarked, in his Lacanian reading of Kant, that it is precisely this gap between the logical subject of the 'I think' and the empirical self that becomes crucial in understanding Kant's sublation of Descartes. The former is strictly speaking not even an appearance, since it is never intuited, i.e. it is without content. But it is also not noumenal, since it is purely formal and empty; so if were to appear it could be lacking nothing: it would nullify the gap between noumena and phenomena constitutive of finite transcendence: "In the synthetic original unity of apperception, I am conscious of myself, not as I appear to myself, nor as I am in myself, but only that I am" (CPR, B 157)

The following passage clarifies this relationship:

   “This gap which separates the empirical I's self-experience from the I of transcendental apperception coincides with the distinction between existence qua experiential reality and existence qua logical construction, the status of Kant's I of transcendental apperception is that of a necessary and simultaneously impossible logical construction ("impossible" in the precise sense that its notion can never be filled out with intuited experiential reality)" (Zizek, ibid)