Sebastian Rödl is, before anything else, a decidedly Kantian philosopher. And like his Konigsberg captain, Rödl is intrigued by philosophy’s ability to discover truths about the world by examining the structure of pure thought. Rödl is, like Kant before him, concerned that an analyst of the world were he to proceed without first having completed an analysis of the thought through whose forms he conducted the world’s analysis, would not be able to distinguish the object of study from the instrument used to study it. If such an analysis were to have already begun by the time the analyst had recognized his methodological mistake, one would, by necessity, have to start all over again if he was still committed to his initial pursuit of truth. Refusing to begin again from the beginning is not a mandatory activity, however, for those unconcerned to untangle the forms of thought from the forms of that which is thought, but anything less than an entire overhaul such as the one Kant recommended and carried out in the first Critique would no longer be worthy of the name of ‘analysis.’
Hence, Rodl writes “[a]nalytic philosophy…for the most part abstains from logical analysis of the statements that represent the object of investigation. (This is a shame, considering Analytic philosophy’s proud name” (CT, Rodl, pg. 11) here, highlighting the irony inherent in naming a practice ‘analytic’ when it is unable (and perhaps unwilling) to properly analyze the object of its study, which is, in this case, the form of thought. This is of course, as much due to confusion as it is to disavowal – most analytic philosophers do not realize that they have not been doing analytic philosophy – but one gets the sense, reading Rodl, that he believes there to be a great deal of disingenuousness implicit in the work of a number of more than a view ‘big names’. Rodl makes his point in this way:
Epistemology defines its subject matter by the formula S knows that p, where “p” stands for a sentence and “S” for the name of a particular subject. This presupposes that we can think about what it means how it is possible to know something without investigating the logical categories under which the object and the subject of knowledge fall. In consequence, analytic epistemology is largely, and largely without being aware of it, skeptical.” (CT, Rodl pg. 11)
That is, epistemology takes it for granted that we know what the subject and object of knowledge are and proceeds to interrogate only the ‘that’ portion of the general proposition S knows that P. Rodl believes that it is this confusion which continues to stultify theories of knowledge which, taking a page out of the natural sciences, think of knowledge as a process of simple induction, or evidence accrual, a process appropriate to fields of study in which the form of their object has already been well defined and well understood. The problem is not induction per se but the fact that epistemology is “grasping clouds of error”, to borrow Hegel’s line, since it has not properly deduced what it is in fact it is trying to induce in statements represented in the form S knows that p (or its variants, S believes that p, S desires that p, etc). For until the form of knowledge is properly analyzed, instantiations of knowledge, i.e. particular judgments or propositions, have no clear content. Lacking clarity and inspired by empiricist methodologies, epistemology takes knowledge to be given, and sets to work not on explaining the possibility of judgment, but on the evidential warrant separating good judgments from poor ones, which is the subtopic known as justification. In this way, failing to give an account of the general form of judgment itself, epistemology dooms itself to the “careless talk of inductive justification, which Hume demolished so effectively,” (CT, Rodl, ph. 12) preventing any talk of knowledge as knowledge, instead limiting itself to the cataloging and refereeing of inferential regularities and constant conjunctions. What empiricism calls general knowledge (or what Quine called the ‘observation categorical’), represented for example in the statement ‘mooses molt in spring’, is supposed to be that which is backed up by a sufficient number of corroborating particular judgments, e.g. Quinean observation statements. But compiling a multi-volume encyclopedia of particular judgments and emboldening certain entries which are corroborated by a satisfactory number of others will never tell us what a judgment is and whether or not such an act is possible. Without a formal account of what judgment is we will never be able to know if, why and how one judgment is closer to truth than another nor what the criterion of justification should be (why, for example, regularity should increase rather than decrease a proposition’s warrant). For Rodl, pragmatism always leads to empiricism, which in turn, results in skepticism. Therefore, Rodl insists that empiricism and pragmatism (which, on this view is just the ‘Casual Friday’ version of the former) must be abandoned as foundations for epistemology.
Before we get onto examining Rodl’s positive thesis which is intended to tell us both what knowledge is and how it is possible, we might do ourselves well by turning the pages back a few chapters and seeing what others have done in their own way to steer the Titantic of theory from the iceberg of incoherence. As mentioned above, there are theorists who would rather deny that there is a problem here in the first place than to seek a means to solve it. Being told that you’ve been more or less building a ship in a bottle when it was meant to sail the sea can be devastating news, especially when your ship has been touted as unsinkable by the very best engineers. Nevertheless, it is important to note that according to the charge being volleyed here, the state of epistemology is actually as bad as this mixed metaphor makes it out to be: it’s not merely a case of comparing the sea-worthiness of one ship to the next, nor is epistemology simply the process by which an already sailing ship stays afloat by patching up its leaky bits faster than it sinks. To be sure, learning (or Bildung if you prefer) will certainly involve reformative discourse in these ways. Having the ability to route out rotten concepts and incorporate new and supple replacements certainly contributes to the buoyancy of any theory of knowledge. But unless philosophy can give an account of what it means to have a concept in the first place and for that concept to contribute to the production of knowing and not just the conduction of knowledge, epistemologists will still be arguing about who can carve the best canoe while the scientist passes over the problem completely, so to speak, by passing underneath it, watching the whole affair by periscope perhaps. Science can do this because its business is, strictly speaking, neither knowledge nor truth but prediction and manipulation. Even mathematics brackets out the question of knowledge and studies formal entailment, which does not concern itself with the human’s ability to think such entailment. Similarly, to paraphrase Ray Brassier, cognitive-neuroscience and psychology investigate the ‘what’ of the mind and the ‘how’ of its relation to the brain without first interrogating what ‘what’ means, what the meaning of ‘means’ is and what ‘is’ means, nor ‘how’ we know these things (Brassier, “Concepts and Objects,”, Speculative Turn, pg. 57).
We have neither the time nor the inclination to rehearse the entire history of western metaphysics and epistemology so I will here limit us to a peek at what I take to be some of the reigning theories of knowledge at present. This puts us on a plane to Pittsburgh, where we can hopefully pencil in a spot with Sellars, Brandom and McDowell, the so-called “Neo-Hegelians”.
Now, what makes all three of the above thinkers “Hegelian” is their rejection of ‘the Given’ which is seen by many to be a gesture analogous to Hegel’s rejection of what Kant called the-thing-in-itself, the world apart from the way it is accessed via concepts provided by the human mind, i.e., the world beyond thought. Famously, Hegel declared that the rational is the actual, and the actual, rational, leaving no space for an absolute beyond of thought, a non-conceptual ‘Given,’ insisting instead that we must conceive of thinking not only as something that subjects do in the world, but also as something that the world does to itself (or in-itself) thereby giving us the equation of substance and subject, matter and mind, thinking and being. One good reason for adhering to this theory – which we might understandably taken to be arguing for a complete demolition of the wall erected between the world and our ideas about the world, (with epistemology then becoming a branch of ontology) – is its affinity with the picture of reality given to us by the natural sciences that teach of humanity’s birth and evolution out of the womb of nature. Before Darwin persuaded the world of this origin story empirically, Spinoza persuaded the philosophical world by sheer logical aplomb of man’s necessary and essential dependence upon the cosmos of which man and by extension his actions, are mere accidents (in both senses of the term). If God or Nature is genuinely the first and only prime-mover, then man’s actions, including his thinking, will be, regardless of how they appear to us, determinations of Nature herself who is self-determining. On this Spinozistic view, thinking, along with the flowing of rivers, the accretion of matter, and the expansion of space-time, becomes just another way that Nature natures itself.
The difference between Hegel and Spinoza is found however, in the former’s attempt to place man on the crest of the cosmic wave of being rather than simply inside of it. For since only in man is Nature self-conscious, God only knows Himself as Himself in man’s mind, and not just any mind, but the philosopher’s mind, who has translated the religious, artistic and historical premonitions of this fact into the pure and formally truth conferring system of logical certainty. Demonstrating how Hegel goes about doing this involves, as might be expected, a great deal of scholarship and careful analysis of his dialectical method but suffice it to say that, despite centuries of being mistaken for the Emperor of Idealism , Hegel does not think that epistemology and ontology are identical nor that mind and world can simply be collapsed into one another.
Notoriously, Schopenhauer accused Hegel of being a sophist and a charlatan and insisted that German Idealism was a “spider’s philosophy” that succeeded not in grasping truth but only in catching and liquefying generations of bright minds by assigning them the fool’s task of deciphering an encrypted system whose original message would turn out to be sheer nonsense, footnotes to footnotes to footnotes, more or less. Schopenhauer, followed later by Russel and Moore, soon smelled the spider coming down the line and quickly unspun themselves from the Hegelian web of (in)coherentism. Schopenhauer, good Kantian that he was, thought that it was impossible to find a foundation for knowledge given that the law of logical entailment, the principle of sufficient reason itself simply could not be applied to the thing-in-itself, since logic was a necessary element of conceptual thinking only, and could not be known to reflect the nature of the world beyond human thought. That something is conceptually necessary does not mean that it is actually necessary. Thus, Hegel’s edifice was built upon quicksand. But where to turn? Examining the possibility that the laws of the mind might serendipitously just happen to reflect the laws of the in-itself, Schopenhauer rejected this Romantic fideism for what he thought was the much more rigorously evidenced and argued position coming out of the natural sciences and the ancient Eastern wisdom of the Upanishads. Both, in their own ways, portrayed man, from head to toe, as a soulless beast whose most noble faculties, namely his good will and his rationality, were simply higher levels of base functions shared by all animals, designed to impel the individual to survive and procreate on behalf of the species, to keep the Wheel of Ixion spinning, so to speak. On this cynical view, even man’s selfishness was not his own, working more or less as a means through which the mindless drive of the Will – Schopenhauer’s name for Nature’s formless vital energy – discharges itself. Diving from Hegel’s web, Schopenhauer did not even look for solid ground on which to fall; his stoical philosophy is not concerned to explain to man his fall from grace, but only to teach him how to fall gracefully. Indeed, Schopenhauer’s pessimistic denial of meaning as purposiveness inspired Nietzsche to later reject meaning as truth itself. This was a rejection that simultaneously cut the ties with an entire tradition of thought beginning before Socrates which was predicated on the presupposition of Truth as both real and knowable and cut the ribbon for a full century of philosophy that would be performed without the pretensions of Absolute Truth and oriented no longer by the latter’s guiding compass of logic and mathematical a priority but led instead by the careful observation of power relations and phenomenological unconcealment.
Meanwhile, Russel and Moore, sought solid ground upon the tried and true rocks of logic and mathematics and were largely successful in their endeavours, along with Frege and the Vienna Circle, but only for a time. In order to build a sturdy foundation that would be able to ground all knowledge philosophy simply had to figure out a way to derive the truths of mathematics from the rules of logic, thereby securing the certainty of both, and then show how our concepts can be derived or constructed from a base of physical, mathematical and logical simples. Metaphysics may just be a matter of mixing primaries. The picture we would paint would be a neat, tidy and complete correspondence theory of truth without an inconvenient remainder. A grand system, no doubt, that would be both in-and-for-itself like Hegel’s, but this time around built on solid foundations grounded in elegant physico-mathematical principles instead of the strange logic of negation and the mysterious, movement of the ‘Concept’ [Begriffe]. Unfortunately, Godel, Wittgenstein and Quine came around and gave very convincing arguments that called into serious doubt the completeness and consistency of mathematics, the “word-world” correspondence theory of meaning, and the ability to ground our logical concepts in anything but hypotheses whose terms are set to accommodate practical but never actual or natural necessities. On the continent, Derridian deconstruction was using internal inconsistencies and contradictions within systems of semiotic signification or ‘texts’ to overturn structures of knowledge and meaning in similarly devastating ways. Theories hawking an absolute grounding for any kind of mind-independent reality were everywhere losing ground. Clearly the mind and the world were correlated in some fundamental way but how and to what extent? It seemed to many, both in and outside the philosophical circles, that philosophy had come to the end of the road having failed to present a theory of knowledge that was adequate to the theoretical rigor of the sciences and true to the lived reality of human experience. As Foucault observed, by fleeing Hegel, philosophers on both continents found themselves running right into his arms in a way that was eerily exemplative of the dialectical process they had originally rejected. Was fear of error really fear of truth? We will see shortly that, in a very important way, the truth was the fear itself. Self-doubt is the first wrung on the ladder of reason. Understanding this will require us to familiarize ourselves with the Pittsburgh Neo-Hegelians, Wilfrid Sellars and his successors, Robert Brandom and John McDowell. Though the philosophies presented by these three thinkers are by no means identical, they do share enough of a family resemblance to justify the group moniker. They are all certainly committed wholeheartedly to rejections of those epistemologies which are based on seeking foundations for knowledge in non-conceptual Givens. Of the three, McDowell is the black sheep in the family since he does present a theory of knowledge that can be called Foundationalist, though one that does not plant its roots in any acrid and barren non-conceptual ground. His contemporary, Brandom, on the other hand, follows Sellars in his proposal of a broadly inferentialist epistemic program which we will now briefly examine.
It’s a Sellars Market:
We recall that we began our current discussion by entertaining the idea, proposed by Sebastian Rodl, that philosophy has been unable to explain how knowledge is possible because it has not yet clarified the nature of the object of which a knower, or rational subject, is supposed to have knowledge. Let us call this the What Gives Problem (or WGP for short). Philosophers prior and posterior to Sellars who refuse to entertain this possibility or who believe that the WGP has been solved tend to take the object of knowledge to be a discrete physical entity (or combination thereof) existing in the external world independently of the mind which knows it and whose essential properties belong to the object in question irrespective of whatever properties the mind of the knower may attribute to it in the form of the concept under which it is judged. On this view, the direction of fit goes from the Given object to the concept with the properties of the former determining the content of the latter, so that the concept is under constraint as to what kinds of objects it can be properly applied to. This is a representationalist view wherein the concept is said to represent not only the object’s existence but also its essence. Any discrepancy between the object and the concept will be due either to the rational subject’s misapplication of the latter or else it will be due to the inadequacy of the concept’s capacity to accurately and completely represent its object’s essential nature. Interestingly, on this view, where the direction of fit stretches from the world to the word (or from the object to the concept), every concept necessarily succeeds in representing the existence of its object, notwithstanding the fact that it may fail to represent with any degree of fidelity the way in which that object exists, i.e., its determinant properties. On this view, every act of conceptualization captures with perfect adequacy, the existence of the object that is Given. The mind catches everything that is Given to it, though it may not ‘grasp’ everything it catches. Thus, even if the object is brought under the concept of a fictional entity, say a faerie, then one can have absolute knowledge that the object so conceptualized exists but whose form of existence, qualities, properties, essence, etc., are not knowable with the same certainty. This thing I conceive as a faerie, exists absolutely, though it’s faerie-ness may be simply part of the added content of my concept. Perhaps a more accurate conceptualization of the object would be to conceive of it as a dragon fly, a hologram beamed into the garden by my neighbors, or, if I am especially lacking in confidence, perhaps I will conceive of it simply as a batch of particularly effervescent neurons firing faerie-wise. The key point to recognize about this representationalist view of the Given is that every single judgment or conceptualization will, by formal requirement, presuppose the existence of the object so adjudicated. Existence predication (i.e. existential quantification) is the invisible first premise in any judgment, be it syllogism or enthymeme. In order to ascribe a predicate to any object the object itself must first be predicated as predicable. This formal requirement obviates any recourse a rational defendant may take in shoring up his judgment by citing the non-conceptual content that precipitated it. Since the same non-conceptual, purely existential content is used to ground all judgments, it will be the very same content grounding any particular judgment. Since non-conceptual existential quantification is the ground of judgment in general, no judgment citing the ground as a premise will be able to get off the ground long enough to assert its conclusion. In the realm of judgment, existence of the object is always already posited as Given. Even when I negate the object I say “there is no such object” but I am only negating the conceptual content compressed inside the ‘such’; the ‘there is’ remains untouched.
Wilfrid Sellars in his paper ‘Empiricism and the Philosophy of Mind’ diagnosed the above-mentioned What Gives Problem by claiming that it was a symptom of a larger illness plaguing epistemology in general, an illness which he famously classified as the ‘Myth of the Given’. The pejorative use of the word ‘myth’ here is designed to draw our attention to the possibility that idea of ‘the Given’ – that non-conceptual content, whether physical or non-physical, can be used as a premise in a judgment somehow justifying subsequent premises or conclusions which of course are themselves wholly conceptual – is an idea which fundamentally misunderstands what the act of judgment is. The representationalist view conceives of judgment (and knowledge) under the model of description whereby an already (though non-conceptually) determined object is examined and analyzed, squeezed for information and then copied in the form of a concept. Sellars saw the difficulties inherent in such a view, difficulties, such as those we rehearsed earlier, regarding the seeming inability to make heads or tails of how a non-conceptual medium interacts with a conceptual one or how to explain in a non-question-begging way our criteria for justification. Sellars thus offered a theory of knowledge that presented the act of judgment as the exercise of a prescriptive power rather than a merely descriptive and receptive one. When I say of object X that it is red, I am not taking in its property of redness like a camera takes in light (regardless of whether or not this is an accurate picture of what my eyes are doing), but rather ascribing to the object certain inferential entailments and constraints that are logically and materially determined by all the other properties I can now infer about this object in virtue of its being classified as red. I commit myself to treating the object in a way that I treat other objects that I take to be red and I entitle myself to infer properties about this object that I can infer about other red objects, for example that they are not green, are surfaced, reflect light and so on. Thus the relationship described in the formula ‘S’ knows that ‘p’ is not one between the world of concepts and the world of objects, but is rather a relationship between concept users and the concepts used, both of which together make up the Space of Reasons which, though perhaps supervenient on the space of physical causality, is nevertheless functionally independent of it. Though Sellars makes no use of the following analogy, we might find it helpful to think of the Space of Reasons as an economy of logical entailment, whose currency is the Reason (as opposed to the Dollar or the Yuan, for example) whose ‘purchasing power’ can be thought of in terms of the use value of inferential warrant. When we say of someone that she is right, or that she has reason (elle a raison) what we are saying is that the reasons she has given entitle her to the inferences she has claimed (or ‘bought) for herself in her judgment. Thus, giving and asking for reasons is, by analogy, similar to the activity of bartering. When we inquire as to the price of a commodity we inquire as to its value relative to the cost of other commodities for sale in the marketplace. We do not, at least not anymore, look for the value inside of the object apart from its relationship to other objects in a similar basket of goods. To anchor the economic analogy in play here, we can nickname the Space of Reasons ‘the Sellars Market’.
Just as every purchase I make in the economic market expresses what the price of the good purchased is at a given time, it also expresses my assent, approval or endorsement as to what the cost of that good ought to be. By spending n-dollars on object X, I fix its price at that moment and simultaneously reinforce its current pricing. If the transaction is successful, I will have contributed to the valuation of the good (and thoses like it) from that moment on thereby determining the price others will have to pay for it in the future. Similarly, in the Sellars Market, when I make an assertion, or a judgment, I cannot separate my act of description of the object from my act of prescription as to what concepts I think this object should fall under. To designate something as true is to appraise it as true.
Notice that in this inferentialist view, there is no need to talk any longer about the object apart from how it is jointly conceptualized by a given community of rational agents. We will see that this will turn out to be a major point of contention that McDowell will raise regarding this kind of Sellarso-Brandomian inferentialist model but for now, all we need to take notice of is the notion that the Normative Functionalism exemplified in the Space of Reasons can function consistently as a theory of knowledge without referring to its grounding in any non-conceptual foundation.
Interestingly, in the Sellars Market, there is no functional difference between subjects and propositions. A subject and a proposition are different only in degree but not in kind as to their role in the Space of Reasons. To be fair, this is view that, to my knowledge, neither Sellars nor Brandom articulated in their work, but it is an idea I’d like to explore a little bit here since we will see that it helps bring out the nature of the problem that Rodl’s transcendental logic is meant to remedy. In inferentialist epistemologies such as Sellars’ and Brandom’s, we bracket out the notion that there is a physical object represented or referred to behind our knowledge claims or statements of belief. The essence of any judicable object in the Space of Reasons just is whatever inferential commitments its assertion entails and whatever inferential entitlements its assertion licenses. The judicable object can be another rational agent, a deep-space satellite or a bag of cement. It doesn’t matter functionally, that is, according to the internal logic of the system, whether “‘S’ knows that ‘p'” is a statement asserted about a philosopher being able to reliably make inferences about the ontological argument or whether it refers to a thermometre’s ability to reliably track the temperature outside. Brandom parries this Behaviorist characterization of the theory by adding the further stipulation that a judicable object can only be said to ‘know that p’ if it is able, should it be challenged, to recognize and reproduce reasons as to why and how it is able to reliably infer ‘that p’ obtains only when it does. Raising the bar in this way perhaps keeps parrots and pocket watches from possessing knowledge. But it is not clear whether or not this solves the problem. After all, there are many human and non-human ‘S’s’ both animal and mechanical that can reliably reproduce reasonable inferences without necessarily being able to recognize what makes them reasonable (consider young children, chicken-sexers, Turing machines or Searle’s Chinese Room). But the power to recognize what makes a reason a reason is precisely what we are trying to explain without relying on the notion of “reliable differential responsive dispositions” as part of the answer. This would also mean that an inferentialist would have to say more about what makes ‘S’ a rational subject besides simply highlighting its ability to cite a rule or a norm. (Perhaps we should devise a new class of entity and call its members ‘rational objects’?).
We see however that considering once again the Sellars Market helps us think about the Space of Reasons. In basic economic terms the consumer-producer of any commodity is herself a commodity entirely reducible to a dollar value, within the internal logic of the system. Her human rights only enter the economic space as restrictions on the kinds of transactions she is licensed to be engaged in and the minimum price she can be purchased or sold for at any given time. In the knowledge economy, to say ‘S’ knows that ‘p’ is thus to say that S can inferentially ‘afford’ to assert ‘that p’ and afford further to infer her own capacity to make such an inference. Authority to assert ‘that p’ includes a self-reflexive ability to recognize that authority as itself authoritative. In the Sellars Market, notions of knowledge that include first-person experience, non-discursive mood states or phenomenological intentionality have no traction on whether or not something will count as being a rational subject. A man who believed ‘that p’ but could not give a reason why he believed ‘that p’ as opposed to ‘not p’ cannot be said to know ‘that p’ while a Turing machine that could provide both the assertion and the rule would be said to have knowledge in this instance. Sellars writes:
[I]n characterizing an episode or state as that of knowing, we are not giving an empirical description of that episode or state; we are placing it in the logical space of reasons, of justifying and being able to justify what one says. (EPM §36)
Sellars could not be any clearer. This theory of knowledge limits itself to a logical analysis of justificatory relations between propositions. The value of a proposition is its inferential role. It is therefore a deductive logic and an inferential logic that explicates (with admittedly remarkable explanatory power) nothing besides the logical relationship thoughts have to other thoughts. The beliefs you take yourself to be entitled to reveal your reliability as a conductor of good inferences. Ladyman and Ross are just as deflationary about beliefs as Sellars is on this point. The knowledge economy and its rules described here by the Sellars Market is not able to give an account of how reasoners come to know things in general just as the field of economics is unable to explain how consumers and producers come to value commodities in the first place. They do not deny that these preliminary steps must take place, but they do take it for granted that they have already taken place. This does not detract from the ingenuity and utility of such inferentialist theories and their capacity to illuminate the dynamic but logically-trackable patterns of variables in a vector space, be that the Space of Reasons, geometric space, or the global marketplace. By de-psychologizing logic, Sellars and Brandom show how the finite human intellect has access to principles that can be applied with acuity to material realities that are beyond human sensibility. The Sellarso-Brandomian program, if there is such a thing, can certainly be used to succour theories of transcendental philosophy such as Kant’s and Fichte’s, against eliminativist and empiricist denials of the a priori whether in its synthetic or analytic forms. That rational subjects can think such logical systems is an impressive fact about rational subjects. However, Sellars and Brandom do not show whether or how such rational subjects exist. This incapacity is troubling only for philosophers and human beings interested in obtaining self-knowledge. Because it is often thought to be self-evident that we exist and are rational, many philosophers and perhaps most people do not consider this sort of self-examination to be of paramount importance, a luxury of those with the privilege of free-time, perhaps. However, being unequipped with a theory that can distinguish a rational subject from a rational object has led many people to deny that the former exists, or to conclude, for example, that a corporation, which is certainly able to reliably cite reasons and norms, should enjoy the same legal rights as a human individual. This will have grave consequences for how we think of ethics and politics.
Now, finding it alarming that we would be happy with such a theory of knowledge that could not distinguish a subject from a proposition or a rational subject from a rational object, McDowell rejects inferentialist coherence epistemologies out of hand. In their place he reverts to a foundationalist theory of knowledge that does not rely on the Given as its grounding. McDowell’s solution to the problem of knowledge is to get rid of the notion of non-conceptual content almost entirely while retaining the foundationalist’s sourcing of knowledge in perceptual and sensory experience. The price is steep, for McDowell no longer is able to easily distinguish between rational human thought, and say, the thoughts of animals who cannot reason explicitly nor is he able to neatly distinguish the mind from the world since both on his view contain conceptual content, or rather, conceptual content contains both mind and world.
The meat and potatoes of McDowell’s epistemology seems to be his view that perception puts us in a position to see that world is a certain way and to know that it is that way simply by seeing that it is that way.
There is no excessive intellectualism in a conception of a capacity in whose exercise a subject acquires knowledge that is grounded, and known by her to be grounded, in the perceptual presence to her of objective states of affairs. An ordinary adult human being might not put it in those terms, but that is what she would mean if she said something like “I can tell a green thing when I see one”. It does not require much sophistication to be able to claim such capacities. (McDowell, ‘Perception as a Capacity for Knowledge,’ pg. 32-3)
The virtues of this theory of perceptually produced knowledge are first, that the power (here vision) which produces the intuited object for evaluation is also the power which produces its own justificatory authority as a reliable source of knowledge and second, that this ends the regress that plagued other foundationalist theories whose proponents failed to give an account of how a belief could be grounded in a non-conceptual intuition without recourse to another belief and so on ad infinitum. The vices of the theory should be easy to pick out however: how do we explain this power of perceptual knowledge in non-dogmatic terms? That is, how do we know that perception is authoritative? And of course, closely related to this question is the all-too-familiar fact that the way the world seems to us is often different from the way it really is and in non-trivial ways. McDowell’s response may be disappointing to some for its Wittgensteinian quietistic flavour. Essentially, McDowell concedes the skeptic’s point here but urges her to reacquaint herself with the blessings of sound common sense. When perception goes wrong, it is either because the power is being activated in circumstances which render it unwieldy (e.g., when behind the wheel while drunk); otherwise it is because the power itself has been perverted in some way that should, nevertheless, not be taken as evidence that the entire apparatus is faulty. Hallucinations for example, get the world that is given to them more or less right; in such cases there simply is a discrepancy between the part of the world perceived by the hallucinator and the part of the world perceived, say by his psychiatrist, yet both she and her patient have genuine perceptual knowledge. The price paid by endorsing such a theory seems to be that we deflate almost all of the air out of the general concept of the world as a shared space and instead take seriously the possibility that there are as many worlds as their are minds. Despite this being an intriguing thought, McDowell does not want to go that far, and neither do we. Gorilla-gluing the definite article to the concept of the world keeps the otherwise rickety scaffolding of philosophy from collapsing completely, at least long enough for us to re-set its foundations. Demonstrating that the world we inhabit is a shared one will take some work. It will involve establishing the universality of the forms of intuition as well as the universality of the forms of judgment. This means that we must show that despite there being differences in the way an object appears to me from the way it appears to you, we can know that it is the very same, numerically identical object that you and I encounter in our perception. This means we will have to give an account of how pure and synthetic knowledge is possible, pure in the sense of incorruptible and synthetic in the sense of comprising conceptual as well as phenomenological necessity.