[If you’re just joining us, please click on the cover icon on the right side of the page to see the post that gathers all the discussions of Braver Reading Group, or click here]
One of the coolest things about Chapter 3 is how beautifully the previous two chapters are explained as part of Hegel’s story.
Unfortunately (well actually fortunately for the reader; order the book!) there is so much great philosophy in this chapter that I can’t do a book report.
Instead, I made a groovy chart that shows how Braver has retold the Phenomenology of Spirit in his first three chapters, with page numbers to the relevant sections in Braver’s book. I present this chart in lieu of a detailed exposition. “R” refers to a realist take on the row’s thesis, “A” refers to an anti-realist take. Expressions such as “6d” occurring in cell X denote that the argument given by Braver in the page numbers given in row 6, column d are part of why cell X gets an A or R. Also note that if you still need your reading glasses after left-clicking on the picture (it’s linked to the full-sized jpeg), then click HERE for the .doc version.
I hope that this chart is actually more helpful as a reading guide than the book report I’d started working on.
2. A Little Bit of Recapitulation.
Braver presents Hegel’s undermining of Kant as coming from two directions: (a) allowing historical contingency to undermine the transcendental subject’s claim to be the repository for necessary truth [72-76], and (b) subjecting the role of the noumena to in transcendental idealism to “imminent critique” (self-reflexively applying a position to itself) via something like Fichte’s affection argument [76-93]. This leads first to what Meillassoux calls “correlationism” (in this context, the rejection of Kant’s noumena while retaining phenomena in that what exists is in some strong sense dependent upon human minds; being is “correlated with” thought) and then to Absolute Idealism when necessity is now explained in terms of inevitable historical movement itself (as opposed to Kant’s transcendental subject).
In this previous post I clarified what I took the affection argument to be, and specified four ways it crops up in the Kantian tradition, the first version of which was one of the key moments in the development of post-Kantian philosophy (as Braver shows). In Schopenhauer and the Wild Years of Philosophy, Safranski writes of this version,
Fichte, basing himself on Schopenhauer’s Goettingen philosophy teacher Gottlob Ernst Schulze (who had published his crituqe of Kant under the pseudonym of ‘Aenesidemus’) and on Maimon, had discovered a faulty derivation of the ‘thing in itself’ by Kant. He argued like this: “The assumption that the world as it appeared to us was concealing a world as it was in itself, and that this world-in-itself as ‘material’ was ultimately the cause of what, by means of our senses and and reason, we would transform into the phenomenal world – this, as it were, ‘realistic’ assumption was itself only made by means of the causality principle, i.e. by means of our intellect. In other words, the causality principle, which was valid only for the phenomenal world, was being applied to a sphere that lay beyond appearance. Thus the ‘thing in itself’, which was beyond experience and intellect, was derived only with the aid of causation, which, however, applied only to the phenomenal world.
In consequence it was no longer a ‘thing in itself’ but a ‘thing for ourselves’. (Schopenhauer was subsequently to adopt this argument against Kant). (Safranski, 126)
It is of course a fair question whether Kant was being fairly read by these people. In particular I think the reading of the practical philosophy as primary (Mikhail has discussed this in a number of pre-Braver Perverse Egalitarian discussions; Daniel raised it in the overly heated discussion started by my previous post) is really interesting in this context for a few reasons: (1) the resulting philosophical picture is not prima facie implausible, and it is also non-trivial. The resulting Kant still says groundbreaking things about normativity in reaction to Hume’s worries. (2) It is an open question whether the resulting Kant can still be free of the affection argument. For example in After Finitude Quentin Meillassoux argues (circa page 35) that the “weak correlationism” (the view that a reality independent of the human mind is unknowable but thinkable) assumed by Kant in his discussion of the Ontological Proof collapses into “strong correlationism” (the view that a reality independent of the human mind is unthinkable). Likewise, Braver’s presentation of Hegel (discussed below) actually proceeds in this same way, starting with Kant as a weak correlationist and arguing that this collapses into strong correlationism, and ditching the noumenal as a result. [Also note that Fitch’s paradoxical logical derivation from the claim that all propositions are knowable to the conclusion that all propositions are known is highly relevant to these debates (see Salerno’s excellent recent anthology)]. If the reading of Kant where the practical is primary (so that avoiding the full paradoxes of totality expressed in the Dialectic is not a central part of his thought) can be sustained by weak correlationism and hence avoid affection arguments (and I took this to be Daniel’s argument), one would still have to respond to Meillassoux’s argument (and Braver’s Hegel and issues surrounding Fitch’s Paradox, for that matter). (3) Though the resulting Kant may not face an affection problem, you then raise the interesting issue of whether this Kant still has groundbreaking things to say about problem of modality, totality, and externality. But maybe if the practical issues are the important ones we shouldn’t care so much about that (this would be a moral realist Rorty, that is, John McDowell in some modes). And (4) Fichte’s first book on ethics and religion (Essay of a Critique of All Revelation, from Hartung in Koeningsburerg) does something similar with respect to moral claims and metaphysical claims concerning religion. I would be interested to know if the elderly Kant took the early Fichte’s reading as expressing his views, as a friendly amendment, or as a criticism. [I’m sure scholars have written about this. Kant actually got Fichte’s first book book published. When it came out anonymously people actually thought it was Kant’s own work, anonymous to escape censors. It was only a letter by Kant that let the cat out of the bag, making Fichte famous overnight.]
In any case, if the real Kant can be isolated from all forms of the affection argument, it is still the case that the “misreading” was the motor of a vast amount of the important philosophy that followed. So give us license to talk about Kant* (pronounced “Kant star”) for the time being.
3. Hegel’s Absolute Idealism (and Nietzsche’s Stage Six Physics, Heidegger’s Phenomenological Ontology) and Where This is Going.
In one of the footnotes in Sense and Sensibilia J.L. Austin says there is the part where a philosopher says something, and then the part where they take it back. The big example of this is Berkeley telling us that all that exists are minds and perceptions. But then you ask what happens when everybody leaves the room and he tells you that the objects still exist as thoughts in God’s mind. After this taking back, you always wonder if the difference between “matter” and “thoughts in God’s mind” is a difference that makes a difference.
The above chart shows, according to Braver, the way in which Hegel “takes it back” in the move to Absolute Idealism, where necessity is reposited as holding in virtue of historical necessity. As Braver really fascinatingly shows Chapters 4 and 5, this kind of repositi(o)n(in)g of necessity becomes the model for Nietzsche’s stage six physics and the early Heidegger’s temporal idealism from the second part of Being and Time.
What Braver will go on to do with the late Heidegger is apply Hegel’s own anti-Kantian contingency critique ((a) above) to the later “taking back” part of the Hegelian thinkers. Note that, on Braver’s account, though he rejects Absolute Idealism late Heidegger still accepts the Hegelian strong correlationism.
Interestingly, some of the arguments of Graham Priest and Meillassoux can be interpreted in part as applying Hegel’s anti-Kantian imminent critique ((b) above) not just against the phenomena-noumena distinction, but also against the correlationism that the late Heidegger retains. So, given that they accept the imminent critique (all forms of affectation arguments are varieties of this) against the noumenal-phenomenal distinction, the speculative realist has to find a way to defend what Levi Bryant refers to as “neo-pre-critical philosophy.”
Obviously, in this set of blog posts we will not be able to settle this controversy. I’m just bringing this to bear because I’m trying to show how important Braver’s book is to a proper understanding of dialectical space surrounding realism/anti-realism debates.
4. Hegel’s Rejection of Noumena.
This is such a rich discussion that what follows is going to be a little bit of a cartoon.
4.1. Hegel’s Affection Against the Noumena
Here’s the main bit of text.
Since our concepts determine all the meaning that ideas like “existence” or “thing” can have, the idea of a thing wholly independent of a conceptual scheme is at best meaningless, at worse an outright contradiction. Hegel considers the Kantian noumenon, the object beyond all access to experience, not ultimate reality but rather a confusion. Since things are constituted by us (the truth that Kant hit upon), removing this contribution does not yield true reality but just “a pure direction or a blank space” (Hegel, PS 47, Section73; See also Forster 1998, 156n72).
The Thing-in-itself . . . expresses the object when we leave out of sight all that consciousness makes of it. . . . It is easy to see what is left–utter abstraction, total emptiness, only described still as an “other-world.” . . . Nor dos it require much penetration to see that this caput mortuum is still only a product of thought, such as accrues when thought is carried on to abstraction unalloyed. (Hegel, HL 72, Section 44; see also 162, Section 112)
. . . .The thing-in-itself would be an object stripped of all the features that make objects objects, leaving nothing; Kant’s Being (thing-in-itself) turns out to be nothing (wholly indeterminate and meaningless; see Hegel, HL 127, Section 87). Kant says that things that are in principle outside of our experience are “nothing to us” (CI A105; see also A370-71, A383), but he leaves a noumenal remainder which simply exists. Hegel downgrades this “nothing to us” to “nothing at all”: “What is not present for consciousness as something existing in its own right, i.e., what does not appear, is for consciousness nothing at all” (Hegel, PS 151, Section 249) (Braver, 81-82)
Again, if things are constituted by us, we can’t say anything about reality as it is in itself. Kant tried to have this both ways, and according to Hegel he cannot. This is the basic form of the affection argument.
To point this out is not to be a correlationist, because you could react to it by holding that things are not constituted by us (albeit at the cost of not having anything like a Kantian solution to the antinomies).
4.2. Hegel’s Correlationism
Here’s some of the relevant interesting text where we get the next step.
These notions of reality or thing-in-itself have meaning because they play a role in our experience, just not the role that Kant gives them; that role is both unintelligible in itself and doesn’t match our actual employment of the concept of reality in our continuous quest to know it. “Hegel shows that what must be presupposed is not a distinction between the ‘in-itself’ and the ‘for-us’ as two worlds, or even as two ontologically or epistemologically separable worlds. Rather, the in-itselfness of the world is an aspect of experience which belongs to appearance itself” (Flay 1984, 77) (Braver, 83).
Absolute Knowing takes its start from Kant’s “highest principle of all synthetic judgments . . . that the conditions of the possibility of experience in general are likewise conditions of the possibility of the objects of experience” (Kant, C1 A158/B197). What Hegel then does (and all of the thinkers we will examine follow him in this in various ways) is to remove the final qualifying phrase “of experience,” so that the conditions of the possibility of experience are the conditions of objects. Or better, he leaves the phrase in place but eliminates the contrast that gave it ists previous meaning, since the only reality we can speak intelligibly of is the one accessible too us. After all,
it argues an utter want of consistency to say, on the one hand, that the understanding only knows phenomena, and, on the other, assert the absolute character of this knoweldge, by such statements as “Cognition can go not futher”; “Here is the natural and absolute limit of human knowledge.” . . . No one knows, or even feels, that anything is a limit or defect, until he is at the same time above and beyond it. (Hegel, HL 91-92, Section 60)
Hegel argues that if we cannot know noumena, then there is no reason to call them real, and the world we experience simply becomes the world. What appears to us is not mere appearance but reality full stop, since we cannot talk about another realm which would supply the invidious contrast. Without a contrasting term, there should be no hesitation in calling the world we experience the real world, or even the really real (ontos on) world. (Braver, 83-85)
The only thing I want to say about this is that it would be way, way to quick to dismiss it as “verificationist.” Naive verificationism about meaning holds that for each sentence there is a determinant set of verifying and falsifying experiences that somehow give the meaning of that sentence. The above argument does not assume anything like that. Instead (in common with Wittgensteinian arguments as Braver points out here and in the discussion of Heidegger), what is being reduced to absurdity is the claim that we could talk about something that is in-principle unknowable. The “verificationist” assumption is only that if a concept were such that no evidence could ever tell in favor or against assertions logically committed (by presupposition, entailment, or implicature) to instances of that concept existing, then that concept isn’t really a coherent concept (Wittgenstein’s beetle-in-the-box argument). This does not in any way entail that concepts/meanings are clearly individuated in terms of canonical evidence (the kind of verificationism presupposed by logical behaviorism, and that died a much needed death during the unfolding of early analytic philosophy).
So with the beetle in the box argument, weak correlationism (the view that a reality independent of the human mind is unknowable but thinkable) becomes strong correlationism (the view that a reality independent of the human mind is unthinkable). I take this to be part of Braver’s reconstruction of Hegel.
Again, since strong correlationism is subject to affection arguments, it is incompatible with there being a coherent idea of a reality independent of human minds. Strong correlationism gets rid of the noumena, but in a sense keeps the phenomena. For the strong correlationist, all that exists is like Kant’s phenomena in that it cannot be shaved off from thinking.
This is a really, really important point for everything that follows in the dialectic Braver is considering. Everybody considered (Nietzsche in some modes, early and late Heidegger, Foucault, and Derrida) remains a strong correlationist. A realist companion volume to Braver’s would have to at the very least recap all of his first three chapters to show how we got here, then respond to the Hegelian argument by minimally somehow dealing with Kant’s Dialectic (and the full paradoxes of totality they gave rise to such as Cantor’s paradox, Russell’s paradox, etc.) while still holding that a reality independent of the human mind is knowable.
4.3. The Doubling of Meaning
A brief issue that will really come to the fore in the early Heidegger chapter, but (properly) starts to surface here with Hegel. In After Finitude, Meillassoux describes how a correlationist has to describe claims about the universe, the truth-makers of which predate all possible human existence (these truth makers are arche-fossils).
The logical (constitutive, originary) anteriority of givenness over the being of the given therefore enjoins us to subordinated the apparent sense of the ancestral statement to a more profound counter-sense, which is alone capable of delivering its meaning: it is not ancestrality which precedes givenness, but that which is given in the present which retrojects a seemingly ancestral past. To understand the fossil, it is necessary to proceed from the present to the past, following a logical order, rather than from the past to the present, following a chronological order.
Accordingly, any attempt to refute dogmatism forces two decisions upon the philosopher [remember, he is talking about the correlationist here] faced with ancestrality: the doubling of meaning and retrojection. The deeper sense of ancestrality resides in the logical retrojection imposed upon its superficially chronological sense. Try as we might, we do not see any other way to make sense of the arche-fossil while remaining faithful to the injunctions of the correlation. (Meillassoux, 16)
The “doubling of meaning” is the more profound way in which the philosopher “takes it back” in Austin’s sense. Again, think of Berkeley. First you say the table isn’t really real, but then when Butler kicks the rock you say something like, “In one sense (the sense in which the realist means it) the table isn’t real, but in another sense, what it is to be real is to be a thought.” So when someone like Meillassoux tries to talk about objects existing independently of human thinking, you double the meaning, trying to show that in one sense the claim is false or meaningless and in another reconstructed sense it’s true.
In the Hegelian context, the claim construed realistically becomes not false, but meaningless,
Hegel’s idealism is not the metaphysical, Berkeleyan claim that only minds and ideas exist, but rather the negative, anti-realist claim that we have no way of talking about input ab extra. When noumena go, discussion of genuinely external sensory data must follow. As Hegel puts it: “Now, because in this way, the pure essentiality of things, like there difference, belongs to Reason, we can, strictly speaking, no longer talk of things at all, i.e., of something which would be for consciousness merely the negative of itself” (Hegel, PS 143, Section 236) (Braver, 88) [emphasis added]
But of course, loosely speaking, of course we can still talk of things, and the correlationists show how their views buy us this right. Braver’s nice discussion from p. 88-93 eludes easy encapsulation in a quote. Nonetheless,
The reality which we contribute to is reality itself, and the features we give it are its actual features, not mere subjective projections onto phenomena (see Margolis 2003, 49-50, 54-55). Hegel’s version of Kant’s highets principle is “Logic therefore conicides with Metaphysics” (Hegel, HL 36, Section 24).
So when the man on the street loosely talks about the really real or objects existing when there are no people, we can reconstruct the meaning of their assertions so that they are saying something true.
I think that this becomes a clearer issue with respect to Braver’s discussion of the early Heidegger. So I’ll postpone any philosophical thoughts about it until I present that. Let me say that I worry that it generates self-reflexive problems about how to construe the philosophical claims that concern the distinction (between assertions construed as truth valueless because presupposing realism and those same assertions construed as true and not presupposing realism). But it really won’t be possible to discuss this until we get to Heidegger, because the connections between Heidegger and Wittgenstein (who almost certainly in part because of the childhood influence of Shopenhauer, both adpopted Fichte’s affection arguement, and faced this problem head-on in his early work) are so brilliantly clarified by Braver.
5. Brief comment.
I just wanted to say how nice the discussion of bivalence is in this chapter.
I really love the way Braver raises the point about how when you idealize enough facets of meaning to start representing things in standard logic, you lose the ability to represent the way theory change happens historically. Any interesting case of theory change will not be a matter of redistributing truth values over a set of propositions with uniform meaning.
If Vaihinger’s philosophy of as-if has purchase anywhere, I think it does in the things we routinely presuppose as part of communication. To draw inferences from what you say, I presupposed that we both mean the same by our words and that these meanings are non-vague. These are normative imperatives we bring to conversation. There is nothing in each of our heads or the world such that we can produce an algorithm from that to ensure sameness of meaning. Rather we agree to agree on meaning. And then later one of the ways to signal unfruitfulness of conversation is to stop this pretense.
If we push on the way this pretense makes logical inference possible just the least little bit we get the sorites paradox, but as theorists and interlocutors we’re pretty good in practice at not falling down the slippery slopes that constantly surround us. Once we start to talk about big differences in theories of the world we can’t help but to slip.
Braver’s discussion of this kind of phenomena in the context of Davidson and Dummettian worries here and later in the text is really thought provoking.