Pragmatism vs Foundationalism

In his recent Spinoza lectures, Brandom also summarizes the context of the 20th century Anglo-American analytic philosophy criticized by Rorty.

“In any case, the broadly Hegelian project Rorty was then recommending as an alternative to the degenerating Kantian research program he saw in analytic philosophy did not look to Europe for its inspiration, but to the substantially distinct tradition of classical American pragmatism.”

“Rorty’s remarkable diagnosis of the ills of analytic philosophy as resulting from an uncritical, so undigested, Kantianism is at least equally radical and surprising as the reimagined, redescribed, and revived pragmatism that he developed as a constructive therapeutic response to it. For Kant emphatically was not a hallowed hero of that tradition. Anglophone analytic philosophers thought that the ‘Kant [or] Hegel?’ question simply didn’t apply to them. After all, Russell had read Kant out of the analytic canon alongside Hegel — believing (I think, correctly as it has turned out) that one couldn’t open the door wide enough to let Kant into the canon without Hegel sliding in alongside him before that door could be slammed shut. Both figures were banished, paraded out of town under a banner of shame labeled ‘idealism,’ whose canonical horrible paradigm was the Bradleyan British Idealism of the Absolute” (Brandom, Pragmatism and Idealism, pp. 4-5).

“The dominant self-conception (combatively made explicit by Carnap) was of analytic philosophy as an up-to-date form of empiricism, a specifically logical empiricism, whose improvement on traditional, pre-Kantian, Early Modern British empiricism consisted in the deployment of much more sophisticated logical tools it deployed to structure and bind together essentially the same atoms of preconceptual sensory experience to which the earlier empiricists had appealed” (p. 5, emphasis in original).

Brandom portrays logical empiricism as foundering on the skeptical “trilemma” (circular argument, infinite regress, or appeal to unjustified justifiers) formulated by the Greek Skeptic Agrippa. Foundationalism in the theory of knowledge typically arises from excessive worries about skepticism.

“Attempts to justify empirical knowledge must either move in a circle, embark on an infinite regress, or end by appeal to unjustified justifiers, which must accordingly supply the foundations on which all cognition rests…. The two sorts of regress-stoppers Rorty saw appealed to by epistemological foundationalists were immediate sensory experiences, as ultimate justifiers of premises, and immediate grasp of the meanings of our terms or the content of our concepts, as ultimate justifiers of inferences. In a telling phrase, he refers to these as two sorts of ‘epistemically privileged representations'” (pp. 5-6).

In order to avoid skepticism, foundationalism makes untenable claims about immediate (noninferential) knowledge, which depend on the assumption that certain representations are specially privileged, so as to be immune to questioning. I have always appreciated Brandom’s exceptional clarity on these issues, and it seems that in this he was preceded by Rorty.

“Rorty takes Kant at his word when Kant says that what he is doing is synthesizing rationalism and empiricism. But Rorty takes it that what logical empiricism made of Kant’s synthesis in the end takes over both sorts of privileged representations: the sensory given from the empiricists, and the rational (logical, inferential, semantic) given from the rationalists…. In this story, Carnap shows up as a neo-Kantian malgré lui [in spite of himself] — though that is not at all how he thought of or presented himself. It is, however, how Rorty’s hero Wilfrid Sellars regarded Carnap. (Perhaps the revenant neo-Kantian spirit of Heinrich Rickert, passed on through his student Bruno Bauch, Frege’s friend and colleague and Carnap’s Doktorvater [academic mentor, literally “doctor father”], was just too strong to be wholly exorcised by the empiricist rites and rituals practiced by the Vienna Circle.)” (pp. 6-7).

According to Brandom, Rorty criticizes mid-20th century logical empiricism in terms very similar to Brandom’s, except that Brandom is much less inclined to blame Kant.

“But the roots of those foundationalist commitments can be traced back even further, to Descartes. For he assimilated the images delivered by the senses and the thoughts arising in intellect together under the umbrella concept of pensées precisely in virtue of what he saw as their shared epistemic transparency and incorrigibility” (p. 7).

Descartes makes the extravagant assumption that not only that there is one unified subject of all thought, feeling, and sensation in a human, but that it has perfect transparency to itself, and therefore at a certain level cannot be mistaken about itself.

“In rejecting both sensory givenness and meaning- or concept-analytic inferential connections, Rorty relies on the arguments of two of Carnap’s most important and insightful admirers and critics: Sellars in ‘Empiricism and the Philosophy of Mind’ and Quine in ‘Two Dogmas of Empiricism, respectively. (These are in any case surely two of the most important philosophical essays of the 1950s.) Tellingly, and with characteristic insight, Rorty finds a common root in their apparently quite different critiques. Sellars and Quine, he sees, both offer ultimately pragmatist arguments, which find the theoretical postulation of such privileged representations to be unable to explain cardinal features of the practices of applying empirical concepts” (ibid, emphasis in original).

These same two essays are recurringly cited by Brandom. I did not know that Rorty preceded him in that.

When Brandom here mentions “meaning- or concept-analytic inferential connections”, he is referring back to what he earlier called “immediate grasp of the meanings of our terms or the content of our concepts”. (There is a very different sense in which material inference can be seen as simultaneously grounded in, and grounding, concrete meaning and the understanding of meaning, as opposed to formal operations. But in that case, the concrete meaning is something arrived at, not something given or immediately grasped. Meaning is always a question, as Socrates might remind us.)

“Rorty then widens the focus of his own critique by deepening the diagnosis that animates it. The original source of foundationalism in epistemology, he claims, is representationalism in semantics. Thinking of the mind in terms of representation was Descartes’s invention” (ibid, emphasis in original).

Representationalism — the explanation of meaning in terms of reference to (nonlinguistic) objects — still dominates mainstream discussions of semantics. It is associated with a view of truth as a correspondence of claims to a reality that is assumed to be independently accessible. But this is not the only way meaning and truth can be explained. I think that, first and foremost, meaning needs to be explained by relating it to other meaning. Variants of this include dictionaries, the coherence theory of truth, and Brandom’s inferentialism.

The Latin term repraesentatio in fact played a large role in scholastic discourse. The status of sensible and/or intelligible “species” or representations was hotly debated. Scholastic discourse also included quite vigorous and sophisticated debates about nuances of representational semantics, under the rubric of the theory of “supposition”. This refers not to speculation, but rather to something close to the modern notion of reference. Scholastic philosophers even tended to treat questions of knowledge and truth in terms of what we might call questions of referential semantics.

Descartes did not invent thinking about the mind in terms of representation, though he certainly practiced it. Arguably, this goes back even past scholastic theories of species and supposition, to Stoic theories of phantasia. The Stoics also had a somewhat foundationalist outlook. They were the original dogmatic realists in Kant’s sense. But Descartes drew especially vivid conclusions from his claim of the incorrigibility of appearances.

Brandom wants to redeem a positive valuation of Kant from Rorty’s hostility, and he even suggests that Rorty is making a Kantian move without realizing it.

“It is perhaps ironic that in digging down beneath epistemological issues to unearth the semantic presuppositions that shape and enable them, Rorty is following Kant’s example. For Kant’s argument, culminating in the ‘Refutation of Idealism’, was that once we understand how to respond to the threat of semantic skepticism about the intelligibility of the relation between representings and what they represent, there would be no residual issue concerning epistemological skepticism about whether any such relations actually obtained: whether things were ever as we represent them to be” (p. 8, emphasis in original).

Kant was understood by scientifically oriented neo-Kantians as effectively putting epistemology, or the justification of knowledge, before all else. A refreshing aspect of Brandom’s reading of Kant is that he shifts the axis toward questions of meaning and value.

Brandom strongly supports Rorty’s critique of epistemological foundationalism, but thinks that Rorty throws out the baby with the bath water when he claims that all talk about representation is implicitly authoritarian. Momentarily playing devil’s advocate, he reconstructs Rorty’s “global anti-representationalist” argument as follows.

“The starting point is the Cartesian idea that if we are to understand ourselves as knowing the world by representing it (so that error is to be understood as misrepresentation), there must be some kind of thing that we can know nonrepresentationally — namely, our representings themselves. On pain of an infinite regress, knowledge of representeds mediated by representings of them must involve immediate (that is, nonrepresentational) knowledge of at least some representings. Our nonrepresentational relation to these representings will be epistemically privileged, in the sense of being immune to error. For error is construed exclusively as misrepresentation. (This is the representationalist semantic analogue of the justificatory epistemological regress on the side of premises.)”

“Next is the thought that when we ask about our knowledge of the relation between representings and representeds, another potential regress looms if we are obliged to think of this knowledge also in representational terms, that is, as mediated by representings of it. (This is the representationalist semantic analogue of the justificatory epistemological regress on the side of inferences.) On this dimension, too, appeal to nonrepresentational access to representational relations seems necessary…. Rorty saw that according to such a picture, the epistemological choice between foundationalism and skepticism is already built deeply into the structure of the semantic representational model” (p. 9).

Brandom recounts that in 1996 discussions with Rorty on the occasion of Rorty’s anti-authoritarianism lecture in Gerona, “We all agreed that if one found oneself obliged to choose between epistemological skepticism and epistemological foundationalism, then somewhere well upstream something had gone badly wrong conceptually” (p. 11).

I wholeheartedly concur with that conclusion. Skepticism (claiming that there is no knowledge) and foundationalism (claiming to systematically ground knowledge in certainty) are both equally implausible, extreme positions.

“For Rorty, a principal virtue of the sort of pragmatism he endorsed that it had no need and no use for the traditional concepts of experience and representation in talking about how vocabularies help us cope with the vicissitudes of life. Indeed, from a pragmatist point of view, the very distinction between epistemology and semantics becomes unnecessary — a lesson he took himself to have learned from ‘Two Dogmas’…. He sums up his anti-representationalist pragmatism in the pithy slogan: ‘language is for coping, not copying‘” (p. 12).

The kind of experience at issue here is not that which is acquired over a period of time, but the immediate experience that is supposed to be a privileged source of knowledge in empiricism. Brandom recalls John Dewey’s unsuccessful attempt to get the public to change the prevailing notion of “experience” as something immediate.

“Dewey worked tirelessly to give ‘experience’ the processual, interactive, broadly ecological sense of Hegelian ‘Erfahrung,’ rather than the atomic, episodic, self-intimating, epistemically transparent Cartesian sense of ‘Erlebnis’. (Dewey’s is the sense in which, as he says, it is perfectly in order for a job advertisement to specify ‘No experience necessary’. It is not intended to be read in the Cartesian sense, which would invite applications from zombies.) But Dewey signally failed to get the philosophical and generally educated public to shake off the Cartesian associations of the term” (ibid).

Brandom endorses Rorty’s sharp critique of experience talk.

“I was entirely of [Rorty’s] mind as far as the concept of experience is concerned. Outside of explicitly Hegelian contexts, where it figures in his conception of recollective rationality, it is not one of my words…. I agree that the associations and correlated inferential temptations entrained with the term ‘experience’ go too deep, easily to be jettisoned, or even for us to success in habituating ourselves completely to resist. The light of day neither drives out the shadows nor stays the night. We are on the whole better off training ourselves to do without this notion” (pp. 13-14).

But Brandom does not accept Rorty’s “global anti-representationalism”.

“But by contrast to the concept of experience, it seemed to me then, and seems to me still, that things are otherwise with the concept of representation. There are many things that one might mean by ‘anti-representationalism’. When I use the term ‘representationalism’, I mean a particular order of semantic explanation. It starts with a notion of representational content (reference, extension, or truth conditions) and understands proprieties of inference in terms of such already representationally contentful contents. Those contents must accordingly be assumed to be, or made to be, theoretically and explanatorily intelligible antecedently to and independently of the role of representations in inference. ’Representationalism’ in this sense contrasts with inferentialist orders of semantic explanation, which begin with a notion of content understood in terms of its role in reasoning, and proceed from there to explain the representational dimension of discursive content. I recommend and pursue inferentialist rather than representationalist semantic explanations,” (p. 14, emphasis in original).

“But not giving representation a fundamental explanatory role in semantics does not disqualify it from playing any role whatsoever…. [T]here is a big difference between rejecting global representationalism, in the sense of denying that the best semantics for all kinds of expressions assigns them a fundamentally representational role, and being a global anti-representationalist, by insisting that no expressions should be understood semantically to play representational roles” (ibid, emphasis in original).

Here Brandom makes a very Hegelian kind of point. All-or-nothing, black-and-white distinctions like the forced choice between skepticism and foundationalism can only be defended by a kind of sophistry.

“It seemed to me in Gerona, and still does today, that a suitable pragmatist explanatory strategy, beginning with social practices of using expressions to give and ask for reasons, could unobjectionably both underwrite theoretical attributions of representational content to some locutions and also underwrite the viability and utility of the common-sense distinction between what we are saying or thinking and what we are talking or thinking about” (p. 15, emphasis in original).

Brandom takes the reasonable position that we don’t have to ban all talk about representation in order not to be representationalist. What he wants to get away from is the notion of privileged representation that is supposed to be beyond question.

Essence and Explanation

Hegel’s Logic comprises what Robert Pippin calls three separate “logics” — a logic of being, a logic of essence, and a logic of the concept. The first of these, the logic of being, was characterized by Pippin as an out-and-out failure that Hegel deliberately embarks on in order to make an indirect point. Broadly speaking, that failure consists in attempting to explain things or make them intelligible solely by means of simple assertions. The logic of being also shows the impossibility of grounding philosophical explanation in a simple immediacy of sense-certainty or intuition, or in any notion of pure Being or being qua being. It seems to me that what these results have in common is the impossibility of explaining any definiteness or determinacy in terms of what is indeterminate.

So far, there is no indication that the logic of essence will ever be regarded by Hegel or Pippin as a failure like the logic of being. It will be further enriched by the logic of the concept, and we have yet to see the detail of this. But now we have at least reached the beginning of a true beginning, after having completed extensive due diligence toward claims of an easier, more direct kind of beginning that did not pan out. At the same time, the subject matter has changed from mere isolated assertions to what Kant in the Critique of Judgment called reflective judgments.

I have characterized the indirect positive outcome of the logic of being in terms of the primacy of relation and relatedness over discrete “things”. Pippin says that the logic of being also showed the impossibility of a completely presuppositionless beginning. Hegel’s reworking of Kantian reflective judgment now takes the primacy of relatedness as a starting point.

The logic of essence will thus effectively take the constitutive priority of intelligible relations over their respective “things” as its starting point. Relations will constitute things, at least to a greater degree than vice versa. This is what the Preface to the Phenomenology calls the perspective of “otherness”, and what Hegel also, in a special polymorphic sense that has been very badly misunderstood, calls “negativity”.

Rather than futilely trying to explain something determinate from something completely indeterminate, we have now turned to examining the conditions of the constitution of any possible determinacy. Additional normative considerations will be made explicit in the logic of the concept.

Essence is a Latin term that is read backwards into Greek philosophy, due mainly to its use as a translation of Aristotle’s “what it was to be” a thing. As treated by mainstream scholasticism, however, it had a meaning closer to that of Platonic form (see Platonic Truth). Platonic form is eternal, whereas form for Aristotle and Hegel has an irreducible dependency on manifestation and development in time. But Plato in his dialogues treats “essence” or what a thing eternally is as a matter of dialectical discovery subject to a kind of perpetual renewal, whereas the scholastics generally (and Leibniz) held it to be already finally established by God in the act of creation.

I think of human character as a sort of privileged example of Aristotle’s “what it was to be” some particular one. Pippin has given this an excellent development (see Toward Essence; Hegel on Willing). What makes human character a “privileged” example for me is that it makes many nuances visible that are not so applicable to “what it was to be” that chair, for instance. The nuances of interest here concern relations between essence and appearance, which form the main subject matter of the logic of essence.

Here we also have an instance of the Aristotelian and Hegelian point that we gain the most insight from considering the richest examples of anything, rather than the simplest ones.

The moderns learned from Descartes to privilege simple cases, and to aim to systematically reduce complex cases to simple cases. That is an admirable procedure in mathematics, with many applications. But in life more generally, there is no good reason for assuming that richer cases can be explained with no more resources than it takes to explain simpler ones. In mathematics, if we have a proof that some specific class of rich cases can be reduced to some set of simple cases without remainder, then we can make that sort of “reductionist” claim for that particular class of cases. Outside of mathematics, it seems to me that reductionist claims usually turn out to be mere assertions.

What Hegel calls the “problem of indifference” — how are we to judge which particular appearances show aspects of the “essence” or deeper truth of people or things and which do not — is brought to the fore here.

“We can be said to know the ‘what it was to be’ of a thing, neither by direct intellectual intuition (its being-at-work is a process, a way of being, not graspable punctually as itself some object) nor by just observing, say, the life of a living thing or the uses of an artifact” (Pippin, Hegel’s Realm of Shadows, p. 227).

As Pippin puts it, “if essence is to explain anything, it must be the ground of what immediately ‘shines’ or appears. Those seemings must be its own, and they are made sense of by reference to their essence” (ibid).

“In some sense, and it is the task of the logic of essence to explain in what sense, the thing’s actuality is both not its mere seemings, and yet nothing other than those seemings, rightly understood” (p. 228).

This is another very Aristotelian point.

“Determinate specification of something essential in an appearance requires essential predication or specification of some sort — some predicates, not others. But we know which predicates are essential only by already knowing what essence is. This is a problem that assumes different forms but is basically the same, whether posed in the language of classical essentialism and manifestations, or selecting from a large set of ‘grounding’ causal factors the genuinely explanatory one or ones” (Hegel’s Realm of Shadows, pp. 226-227).

Hegel develops the terms “ground” and “grounding” for discussing the generalization from essence to any sort of explanation.

Pippin notes that “Plato, Kant, Locke, Spinoza, and others can all be cited in various ways as expressive of the reflective logic of the appearances of essence, the manifestation of something substantial that is nevertheless not manifest as it is in itself” (p. 231).

He quotes Hegel: “On the one hand, the ground is ground as the immediately reflected content determination of the determinate being [Dasein] which it grounds; on the other, it is that which is posited. It is that on the basis of which that determinate being [Dasein] is supposed to be understood; but, conversely, it is inferred from the latter and is understood from it. The main business of this reflection thus consists in gleaning the ground from a determinate being [Dasein], that is, in converting the immediate determinate being [Dasein] into the form of reflected being; consequently the ground, instead of being self-subsisting in and for itself, is rather that which is positive and derived” (pp. 227-228, Pippin’s emphasis).

Once again, I would note a convergence with Aristotle. Aristotle says that in order to possibly know how things are in and for themselves, we should and do start with how things are “for us”, not with how they supposedly are, full stop. Hegel will eventually amplify this into what he calls the “subjective” (though anything but merely subjective) logic of the concept.

Aristotle and Hegel both want to say that the basis of knowledge and explanation is a partial overlap between how things are initially for us and how they really are. This notion of a partial overlap between essence and appearance is a sort of Aristotelian mean that eliminates the roots of the twin evils of “all is illusion” skepticism, and of foundationalism, or the claim of a certain starting point for knowledge.

“[Hegel] is in effect saying that a putative logic of being is, has shown itself to be, mere seeming, Schein [literally, “shine”]. As [Michael] Theunissen points out, this means that Hegel is actually invoking the notion of Schein in three different senses. There is the unacknowledged Schein that a logic of being has turned out to be. There is the Schein of the mere appearance that the skeptic and idealists claim are all we are able to know. And there is the result of the analysis, that this purported limitation of knowledge to mere Schein is itself Schein, unable to account for itself; what seemed to be mere Schein turns out to be the Schein of essence or Erscheinung [Hegel’s technical term for appearance that is more than just mere appearance]” (p. 229, emphasis in original).

That all appearance is only mere appearance must be itself only a mere appearance, if there is to be any knowledge or meaningful explanation at all.

“In other words, the illusion of any possible absolute presuppositionlessness is what has been demonstrated by showing that Sein [being] must be understood as Wesen [essence], just in order to be understood as Sein. ([Hegel says] ‘Being is as such only the becoming of essence’…)…. Wesen will show itself (and itself as the truth of Sein) as always already conceptually mediated determinacy” (p. 230, emphasis in original).

The brute “things” of mere assertion depend on the richer, subtler “things” considered by reflective judgment for any truth they may have. This is an archetypal Hegelian move.

Pippin points out that the logic of essence gives a new sense to Hegel’s very nonstandard notion of negation. Whereas before, “negation” served to express the dependency of meaningful relational distinction on what else something rules out in order to express what it is, now “Essence’s seemings are its own…, even though no seeming or set of appearances express in their immediacy what that essence really is” (ibid).

He quotes Hegel, “In the becoming of being, it is being which lies at the foundation of determinateness, and determinateness is reference to an other. Reflective movement is by contrast the other as negation in itself, a negation which has being only as self-referring” (p. 231, Pippin’s emphasis).

Rather than addressing an external other, in reflective judgment Hegelian “negation” is now turned on itself — seeking further clarification first and foremost through questioning itself and its own formulations.

Next in this series: Hegel on Reflection

Form and Things

I use the word “thing” in a very general sense for anything at all — real, ideal, or imaginary; abstract or concrete; including properties, actions, processes, and adverbial characteristics.

Kant controversially wanted to assume that things of all sorts have definite ways that they objectively are “in themselves”, i.e., completely independent of our experience and knowledge of them. But for him, there is inevitably a gap between our knowledge and reality. Every attempt to ignore or overleap that gap he called dogmatism. This was his way of practicing what I have called epistemic modesty, or the ethical virtue of avoiding unfounded knowledge claims. When we don’t know, we go ahead and act based on the best beliefs we have, while in principle remaining open to the possibility that our belief may require revision.

Hegel and others have worried that there is something wrong with Kant’s way of expressing the situation — with this gap between knowledge and reality that is inevitable and even virtuous for Kant. Some of Kant’s remarks make it sound as if reality as it really is and our experience could be two entirely nonoverlapping realms. If this were true, Kant’s position could be seen as leading to skepticism, or the conclusion that genuine knowledge of reality is impossible.

Kant himself would have resisted this conclusion with all his might. He does believe we have genuine knowledge; he just wants us to be very careful about what we claim to know. For Kant, genuine knowledge does not require access to things in themselves; rather, it keeps within the bounds of possible experience. It minimally designates an objectivity toward experience, consisting in the absence of dogmatism and an unceasing effort toward unity of apperception.

Hegel agrees with Kant in opposing dogmatism and emphasizing the effort toward unity of apperception. His strongest opposition to Kant’s talk of things in themselves assumes that “in themselves” means “in isolation”, as it would under the Wolffian view (rejected by both Kant and Hegel) that all knowledge is analytical. Hegel emphasizes that unities of apperception are not just individual but also shared. At the same time, he revives the Aristotelian idea that thoughts should be distinguished as forms or meanings shareable in principle with any rational being, and that as such, they are whatever they are independent of subjective presentation. This makes it quite reasonable for Aristotle and Hegel to claim that the form or meaning that is properly being thought is the very same as the form or meaning that is being thought about.

An analogous identity certainly does not apply to experience or consciousness. To assert that would be dogmatism in Kant’s sense. The thought that Aristotle and Hegel identify with form or meaning is not a kind of consciousness. The most fundamental characteristic of consciousness for Hegel is the separation of subject and object, whereas in thought proper there is no such separation, only a succession of forms. In Hegel, the gap between consciousness and its objects takes the place of the gap between knowledge and reality in Kant, and similarly commends to us a practice of epistemic modesty in life.

Another dimension of epistemic modesty in Aristotle and Hegel has to do with the non-univocal character of form in Aristotle, and with Hegel’s repeated warnings about the “falsity” of all fixed representations. Form is said in several ways in Aristotle, e.g., sensible form, mathematical form, linguistic meaning, and the life or soul of a living being. Of these the first two are univocal, but the last two are not.

From the point of view of form, we take a deflationary view of “things”. Things — like the thing in itself in Kant — are in a strict sense indexes delimiting our ignorance rather than univocal “objects” of knowledge.

The fact that our ignorance is delimited means it is not total. We do have knowledge, but insofar as either proper knowledge or Hegelian spirit has “objects”, those objects lack univocal identity.

Descartes Revisited

Descartes is among my least favorite of those conventionally termed great philosophers. My treatment to date has been mainly limited to a few dismissive remarks. Here I’d like to add a few “historiographical” points of demarcation.

Insofar as there is general consensus among scholars, Descartes (1598-1650) first and foremost has claim to fame as a very influential promoter of something recognizably close to modern scientific method. He is often credited with the invention of analytic geometry, based on an early recognition of the systematic isomorphism between geometry and algebra. Galileo (1564-1642) had already taken up an approach to natural science based on mathematical analysis, which Descartes enthusiastically adopted. Descartes particularly promoted a methodology based on clear and distinct ideas, which he held to give certain knowledge. He advocated an orderly progression from the simple to the complex.

On a broad social level, Descartes is remembered for promoting the independence of scientific investigation, particularly from the doctrinal concerns of the Catholic Church. But he was also a religious thinker. While confessing in a private letter that he did not literally believe various details of received scripture, he was very engaged with proofs of the existence of God.

Numerous scholars have pointed out that outside the domains of mathematics and natural science, Descartes in many ways remained close to the Latin scholasticism of which he has been commonly regarded as the slayer. In Descartes and the Modern Mind (1952), for example, Albert Balz argued at length that the thought of Thomas Aquinas was an essential precursor to Descartes. I note that Augustine had already emphasized the importance of the interpretive role of reason, even in matters of faith. On Aquinas’ account, God gives us not only revelation, but also the natural light of reason. In Summa Contra Gentiles, Aquinas addressed questions of philosophy and theology entirely from the point of view of that “natural light”, while aiming to show that the natural light of reason independently leads to many of same truths he attributed to revelation. Descartes makes great use of a similar concept of the natural light of reason.

Both Descartes and Aquinas thought the natural light of reason, properly understood, gives us truths, full stop. I think it gives us invaluable criteria for judgment and interpretation, while always in principle leaving room for discussion about what conclusions should be drawn. I also think the “natural light” itself comes to us in degrees, and is never a simple or unproblematic possession.

A different strand of Latin scholastic thought tended to claim that all human knowledge originates in sensible images, while attributing such a view to Aristotle. (I think this is overly strong, and that Aristotle only meant to defend the pragmatic value of sensation against Platonic skepticism about all deliverances of sense.) Descartes famously argued that sensible images are not the only source of knowledge, and I think that is true, as far as it goes.

Here is where Descartes’ theses about clear and distinct ideas come into play. A methodological discipline based on examining whether our ideas are clear and distinct is an important source of human knowledge. Again, this much I can agree with, but I think clarity and distinctness are relative criteria and not absolutes. As relative criteria, they have been implicitly employed by most if not all serious thinkers. I take such evaluations to have been a major implicit concern of Platonic and Aristotelian dialectic, which in part aims to discern meanings that are more clear and distinct.

Descartes effectively claimed that clarity and distinctness are absolute, decidable properties of ideas. One of the broadly scholastic views he sharply criticized was that our best knowledge of sensible things is only “probable”. Descartes claimed that the results of his methodological analyses were certain, in the same way that mathematical conclusions follow with certainty from their premises. This goes well beyond the claim that there is practical value in such methodology.

Building on arguments of Augustine and Avicenna, he also famously gave great importance to the claim that immediate contents of the mind give evidence of unconditional certainty of the abstract existence of something. The very possibility that I could be deceived implies the abstract existence of an abstract something that could be deceived. Further, if something in any way appears to me to be such and such — even if I am wrong about all the details — independent of all the details, the barest fact of the appearance implies that some appearance generically “exists”.

The “I” that is in this way proven to exist and the appearance that is proven to exist are both extreme abstractions. Even Descartes did not claim that either of these existences by itself gives us any further knowledge. From this basis alone, I could still be entirely mistaken about the kind of being that I am, and about every detail of what appears to me to be the case. In spite of the famous Cartesian doubt, Descartes actually wanted to makes strong claims of certainty and to refute skepticism. Many readers have concluded, though, that he opened the door for a new, more global form of skepticism, because what he clearly establishes as certain seems so utterly lacking in content.

I would hasten to add that unreasonable, excessive skepticism about human knowledge is best refuted by successful achievements of goals in real-life situations. Only a hypocrite could claim to live in the world with no well-founded beliefs whatsoever. The ancient Skeptics were only “skeptical” about theoretical accounts of things, not about practical concerns of everyday life.

By setting the bar too high and aiming at absolute certainty, Descartes actually opened the door for more radically subjectivist views that no one in the ancient world would have taken seriously (and not just because ancient people were naive). At the same time, he was very impatient with “dialectic”, and tended to foreshorten discussions of meaning and interpretation, in favor of claims that certain contents are unequivocally clear and distinct. Thus the ultimate result of his thought oscillates unstably between extremes of “Cartesian” skepticism and dogmatism.

Another point on which Descartes has been very influential is his strong representationalism. For Descartes, strictly speaking we never have practical knowledge of things in the world, only knowledge about contents of our mental representations, insofar as they are clear and distinct. In particular, we only know bodies through our mental representations of them. Rather than consisting in an interpretive stance of a situated being in the world, the Cartesian cogito achieves its purely subjective certainty in a way that is supposed to be peculiarly “outside” the real world altogether.

Unlike the representationalism of Locke, which is grounded in a kind of empirical psychology, that of Descartes is closely bound up with an ontological mind/world dualism more radical than anything Plotinus, Augustine, or Avicenna ever contemplated. For Plotinus, Augustine, and Avicenna, the soul was a very special kind of “something” existing in the real world, even if for Plotinus and Augustine it was not a “subject” in the sense of something underlying something else. For Locke — the other great early modern promoter of representationalism — our mental worlds are ultimately contained within the natural world. For Descartes, there is the world and there is the soul, and never the twain shall meet. The soul has its own mental world where it seems to relate directly only to God, and human knowledge occurs only in that mental world.

It is due to this unprecedentedly radical mind/world dualism of Descartes, I think, that virtually no one — even among his admirers — wants to uphold his metaphysics. This is an extreme example of what Hegel called “alienation”.

Hegel on Skepticism

The next shape of self-consciousness after “Stoicism” in Hegel’s Phenomenology is “Skepticism”. H. S. Harris in his commentary thinks some of Hegel’s remarks apply specifically to Carneades, perhaps the best known “Academic” Skeptic, who shocked the Romans by arguing for opposite theses on alternating days, as an exercise on Platonic dialectic. Carneades also wrote a work arguing in detail against the great early Stoic Chryssipus. Although I like to stress the less textually obvious role of Aristotelian dialectic in Hegel’s work, Hegel’s explicit remarks emphasize a kind of Platonic dialectic with Skeptical inflections (see Three Logical Moments).

For Hegel, neither pure Understanding — which excels in clarity, utility, and systematic development but tends toward dogmatism — nor a skeptically inclined Dialectic, whose movement undoes everything that is apparently solid — is adequate to characterize what he wants to call Thought. Thought ought to involve a sort of Aristotelian mean that combines the insights of both.

Hegel writes, “Skepticism is the realization of that of which Stoicism is merely the notion, and is the actual experience of what freedom of thought is…. [I]ndependent existence or permanent determinateness has, in contrast to that reflexion, dropped as a matter of fact out of the infinitude of thought” (Baillie trans., p. 246). “Skeptical self-consciousness thus discovers, in the flux and alternation of all that would stand secure in its presence, its own freedom, as given by and received from its own self…. [This] consciousness itself is thoroughgoing dialectical restlessness, this melée of presentations derived from sense and thought, whose differences collapse into oneness, and whose identity is similarly again resolved and dissolved…. This consciousness, however, as a matter of fact, instead of being a self-same consciousness, is here neither more nor less than an entirely fortuitous embroglio, the giddy whirl of a perpetually self-creating disorder” (pp. 248-249).

Harris comments, “[T]he Stoics had to be taught by the Sceptics that no Vorstellung [representation] (not even that of the great cosmic cycle) could comprehend Erscheinung [appearance]” (Hegel’s Ladder I, p. 393). “[Skepticism] knows that ‘language is truer’… than the simple assumption that truth is the name of an extralinguistic Sache [thing or content]” (ibid).

“The Sceptic ideal is to be untroubled in the face of the sensory flux. Sceptical reason tells us not to worry about what we cannot help” (ibid). But “Far from behaving like one who is undisturbed, [Carneades] enjoys being an active disturber; and on the practical side his life has to be controlled by the felt motive actually present at a given moment” (p. 394). “Achieving ‘suspension of judgment’, by setting whatever contingent arguments one can discern in the whirl against those that someone else offers, is a cheat. The Sceptic lives in the world, and allows himself to be guided by senses which he says we ought not to trust” (ibid).

“Every effort the Stoic makes to realize his freedom is tantamount to a serf’s fantasy that he really owns the land” (p. 395). “The Sceptic is a laughing sage because he has the Stoic to laugh at. We laugh at both of them” (ibid).

“[The Skeptic] has not recognized that the [self] he identifies with is only a formal ideal by which the concretely actual self can measure itself. Nobody is that ideal self. No finite consciousness can be that self (by definition). There is no Lord walking the earth: not the one that the serf fears; not the Stoic who thinks he is free; and not the Sceptic who knows what thinking is, and what it is not” (p. 396).

Next in this series: The Unhappy Consciousness

Foundations?

Foundationalism is the mistaken notion that some certain knowledge comes to us ready-made, and does not depend on anything else. One common sort involves what Wilfrid Sellars called the Myth of the Given.

Certainty comes from proof. A mathematical construction is certain. Nothing in philosophy or ordinary life is like that. There are many things we have no good reason to doubt, but without proof, that still does not make them certain.

In life, high confidence is all we need. Extreme skepticism is refuted by experience. It is not possible to live a life without practical confidence in many things.

Truth, however, is a result, not a starting point. It must be earned. There are no self-certifying truths, and truth cannot be an unexplained explainer.

In philosophy, we have dialectical criticism or analysis that can be applied from any starting point, then iteratively improved, and a certain nonpropositional faith in reason to get us going. All we need is the ability to question, an awareness of what we do not know, and a little faith. We can always move forward. It is the ability to move forward that is key. (See also Interpretation; Brandom on Truth; The Autonomy of Reason.)

Stoicism, Skepticism

Brandom makes interesting connections between Hegel’s rather idiosyncratic discussion of Stoicism and Skepticism in the Phenomenology and the preceding discussion of Mastery. Stoicism and Skepticism for Hegel each in a different way reflect aspects of Mastery’s attitude that wants to claim total independence.

Hegel’s criticism of Stoicism in this context is rather different from my previously expressed issues with its foundationalism, claims of a completed system, and what the ancients called its dogmatism. My remarks probably apply more to the system of Zeno and Chrysippus, whereas Hegel’s apply more to the narrower ethical concerns of someone like Epictetus.

Zeno and Chrysippus are known only from references in other authors; none of their original works survive. Surviving references to early Stoic teaching often tend to be somewhat anonymous and generic. The details of the system are quite fascinating and worthy of study in their own right (see the collected fragments in Long and Sedley, The Hellenistic Philosophers; also Sambursky, Stoic Physics; Mates, Stoic Logic; and Nussbaum, Therapies of Desire).

Ancient Skepticism is also quite worthy of study in its own right. In addition to fragments, a number of works by the late author Sextus Empiricus survive. Ancient skeptics were mainly skeptical about theoretical developments. (The more extreme skepticism many modern authors have worried about in the third person seems to be a post-Cartesian development.)

Brandom says Hegel’s Stoics and Skeptics both refuse the experience of error that is crucial to the elicitation of conceptual content. On his reading, Hegel’s Stoic in, say, refusing to recognize physical pain, is both just being stubborn and refusing to address what turn out to be incompatible commitments, effectively denying the reality of the object in order to maintain the independence of consciousness at all cost. The Skeptic is just refusing to make any commitments at all, which is another attempt to maintain the independence of consciousness at all cost. Hegel’s point is that this attitude of wanting to maintain the total independence of consciousness from anything other is unsustainable.