Learning From Hegel What Philosophy is All About:
for the Metaphysics of Reason; against the Priority of Meaning[1]

FINAL DRAFT, please cite the published version in Verifiche - Rivista di Scienze Umane 2012.

When it comes to Hegel, disagreement begins with even the most basic questions. Recent debates have focused specifically on so-called “non-metaphysical” interpretations, which raise the most basic question of all: what is Hegel’s philosophy about?

But it can be difficult to state what of philosophical substance is really of issue between opponents in these Hegel debates—at least in a manner that could be agreed to by all sides. As Redding puts it, “it is still not clear which issues dividing them are substantive and which are ultimately verbal.”[1] Certainly the opposing sides make characteristic claims. “Metaphysical” interpreters say that Hegel’s ambitious goals—for example, to give an account of “the absolute”—clearly express metaphysical ambitions.[2] “Non-metaphysical” interpreters argue that they have a trump card in a superior ability to make sense of Hegel’s aim of engaging philosophically with Kant’s critique of metaphysics, rather than merely assuming the viability of a project which Kant has specifically argues is hopeless.[3] But insofar as this disagreement is about whether Hegel pursues “metaphysics,” differences may simply spring from different uses of that term. True, “non-metaphysical” interpreters more specifically compare Hegel’s view with anti-realism, internal realism, or other rejections of “metaphysical realism” or “realism.”[4] But philosophical disagreements about realism are a paradigm case of the lack of agreement between opponents on the issues at stake: anti-realists tend to see a dispute fundamentally in semantics concerning reference and the truth predicate, realists a dispute fundamentally in metaphysics or ontology concerning what exists.[5] So if the Hegel debates concern whether Hegel is realist or anti-realist, then this is again cause to worry whether the sides are talking past one another. Further, some who once called their view “non-metaphysical” now argue that this label produces misunderstanding of the approach, and prefer the label “post-Kantian,”[6] emphasizing Hegel’s engagement with Kant. But my view requires me to contest the implication. For I argue that my metaphysical approach is a better way to understand Hegel’s philosophical engagement with Kant. And so the debate seems drawn ever farther toward terminological questions.

But I will argue here that there is a surprising kind of philosophical substance to the debates about Hegel. For they expose a crucial philosophical question whose import is otherwise easy to miss: the question of metaphilosophy. The question is, what is philosophy itself most fundamentally about? Or, what issues are most fundamental in philosophy?

Thinking in terms of metaphilosophy, in this sense, we will be able characterize the debate in substantively philosophical terms that all parties should accept. What distinguishes “non-metaphysical” interpretations is this: their Hegel takes as fundamental the project of considering our cognition in order to construct an account in epistemology of the possibility of knowledge and/or an account in semantics of the possibility of meaningful thought (section 1). This does not mean a renunciation of all involvement in metaphysics in any and all senses. It means rather that the only legitimate metaphysics will be one that is shaped so as to avoid conflict with, and to play a role in or rest upon the more fundamental project in epistemology and/or semantics. If the emphasis is specifically on the fundamentality of accounting for the possibility of knowledge, then we could mark this priority by referring to an “epistemology-first” metaphilosophy. If the emphasis is rather on accounting for the possibility of meaningful thought, cognition or experience, then we can call it a “semantics-first” metaphilosophy. Insofar as both take as prior consideration of our own cognition and its capacities, we could speak more generally of a “cognition-first” metaphilosophy. Insofar as the so-called “non-metaphysical” readings of Hegel ultimately will rest on emphasis on meaning, I will generally call them “semantics-first” interpretations of Hegel.  

But I reject all forms of epistemology- or semantics-first metaphilosophy. And I think that Hegel rejects them too. Articulating a substantial and successful alternative, however, requires care with the term “metaphysics.” Some people may begin by thinking that philosophy is fundamentally concerned with problems about meaningful thought about or knowledge of an independent world. They will then tend to take “metaphysics” to assume that some things are absolutely independent of our perspective, and that we can potentially know them; they will tend to take “metaphysics” to be inquiry into the absolutely perspective-independent. But we can think of metaphysics differently, and must if we are to understand Hegel (section 2). We need not think of what I call the “metaphysics of perspective-independence.” We should think instead in terms of what I call the “metaphysics of reason.” Metaphysical inquiry, in this sense, is not founded on any special notion of, or assumptions about, perspective-independence. It is founded instead on the basic notion of one thing being any kind of reason for another, or because of another. For example, a monist metaphysics (in this sense) would hold that the whole of everything is the reason for the existence and natures of the parts; an atomist metaphysics would hold the reverse. Perhaps inquiry into reasons, in this sense, is also what is meant when Aristotle’s Metaphysics specifies its concern with “primary causes and principles.” But my aim here is not historical comparison, but rather explanation of this kind of metaphysics and its lasting philosophical appeal in independently accessible terms.

The Hegel debates, then, concern what issues are fundamental. Some say that Hegel’s fundamental project is to account for the possibility of knowledge and (especially) meaningful thought. My alternative is to read Hegel’s project as fundamentally within the metaphysics of reason.

But my opponents will still claim to best make sense of Hegel’s philosophical engagement with Kant’s critique of metaphysics. Matching this trump card requires distinguishing two stands of Kant’s critique. One strand is an epistemology-first critique: philosophy must begin by considering our cognition and constructing an account of the conditions of the possibility of knowledge, which will then show us to be unfit for the projects pursued by pre-Kantian metaphysicians. But Kant’s critique would be weaker if it merely attacked metaphysics from the foreign territory of epistemology. It is stronger insofar as there is another strand: Kant also attacks metaphysics from within, or argues that it contradicts itself from within. That is the argument of the “Transcendental Dialectic,” which makes clear that Kant’s most immediate interest and target is the metaphysics of reason, rather than the metaphysics of perspective-independence (section 3).

I can then match my opponents’ trump card. They see Hegel’s basic project as carrying yet further Kant’s attempt to account for the conditions of the possibility of knowledge and meaningful cognition in the “Transcendental Analytic” from the first Critique, putting this to a different and more radically anti-skeptical end. But I too read Hegel as carrying Kantian considerations yet further. I read Hegel’s basic project as carrying further Kant’s consideration of reason and metaphysics in the “Transcendental Dialectic,” putting this to a different and more metaphysically constructive end (section 4).

Once the metaphysics of reason approach to Hegel is clear, and the trump card matched, the advantages are easy to see. For it is then easy to see that Hegel himself prominently denies that philosophy must first or most fundamentally consider cognition and construct of accounts of our cognitive capacities (section 5). Further, this is not bad news for attempts to relate Hegel to contemporary analytic philosophy; my approach does just as well establishing such relations. And, finally, Hegel’s metaphysics of reason can help us to better understand the philosophical issues themselves. Of course, there is no space here to definitively defend a comprehensive interpretation of Hegel, let alone a comprehensive metaphilosophy. But we can use these terms to better appreciate the underlying terrain on which wars in philosophy are still fought today. In that sense, we can learn something important from Hegel with respect to the question of what philosophy itself is all about.

1. Metaphilosophical Diagnosis of the Hegel Debates

I begin by giving the so-called “non-metaphysical” story about Hegel’s engagement with Kant a sympathetic formulation, in order to see how it depends on “semantics-first” metaphilosophy. So imagine someone—call her SF for “semantics-first”—telling the basic story in three steps:

SF: First, pre-Kantian philosophers have discussed things they thought we could know about.  It doesn’t matter what things—call them “X”. But Kant argues that they have all taken some X to be independent of our knowledge or cognition, something which can in turn explain the possibility of our knowledge and cognition, and something which sets the standard to which our knowledge and cognition is responsible. In other words, they take the standard to be a way of grasping of X in its independence of our point of view—something like a God’s eye view. This package of pre-critical views is called “metaphysical realism” or just “realism.”[7]

I should interject in my own voice that SF here uses the term “realism” to refer to a view about the explanation of the possibility of knowledge and its standard; it is not at base a claim about the existence of a world independent of us. I will continue to use it in SF’s way, since I have no special need of any use of the term. She continues:

 SF: Second, Kant rejects “realism” in this sense, taking it to guarantee skepticism, because we could never know how things would look from a God’s eye view. Kant’s alternative is reflection on our own cognition, aiming to demonstrate that we fix from within a distinction between the subjective and the objective—now in a sense of “objective” that is internal to our cognition, and opposed to a pre-critical sense of absolute independence from us. So Kant’s philosophy parallels 20^th century rejections of “metaphysical realism” in favor of a kind of “internal realism.”

I should interject again that I don’t seek to defend this reading of Kant, but only to tell this one kind of story about Kant and Hegel, which continues:

SF: Third, Hegel argues that Kant should have gone farther. Kant should have taken his argument in epistemology, concerning the possibility of knowledge, and extended it more consistently also to semantics, concerning the possibility of meaningful thought. More specifically, Kant claims that we cannot have knowledge of things as they are in themselves. But just as Kant admits that we cannot explain the possibility of knowledge of such things supposed to be absolutely perspective-independent, Hegel argues that we cannot even explain the possibility of meaningful thought about them. To assume otherwise would be to retain a semantic version of the same “realism” rejected in epistemology. To complete the parallel with the above, the conclusion is that we can meaningfully refer to an objective world only insofar as we fix from within our cognition a distinction between the subjective and the objective—again in a sense of “objective” that is internal rather than the pre-critical sense of absolute perspective-independence. So we must give up as meaningless claims about absolutely independent things in themselves, even Kant’s claims to be ignorant of them.

That is the basic “non-metaphysical” story. Redding has a nice encapsulation of the idea that the Hegelian worry about Kant is that he should apply his epistemological points more completely to semantics:

Kant’s combination of conceivability but unknowability seems to take away with the one hand a quasi-divine epistemic take on the world – the so-called ‘God’s-eye view’ – only to return something like a semantic version of it with the other… (2007, 222)

McDowell’s version of “Hegel's Idealism as Radicalization of Kant” takes its departure (2001, 527) from Pippin’s work. And the crucial elements of recent versions generally do stem from Pippin’s Hegel’s Idealism. The story there is as follows:

RP: Hegel’s basic project focuses on meaning or intelligibility in the sense of considering the conditions of the possibility of “any intelligible experience of an object.” Hegel’s view is built around acceptance, specifically, of Kant’s claim about “the spontaneity and reflexivity of any intelligible experience of an object” (1989, 12). This key claim is a revolt against “realism,” in that the spontaneity of our cognition explains the possibility of intelligible experience, rather than the objects themselves. Realism generally supports skepticism, or “realist skeptical doubts” (e.g. p. 107). Even Kant’s own “‘thing in itself’ skepticism” (p. 6) retains and rests on a residual realism which should, more consistently, be eliminated. Hegel rejects all realism, undercutting realist skepticism, in an attempt to reach the conclusion that our own knowledge is second to no other intelligible standard, and so is “absolute” (p. 94).

It is easy to see why some who tell this story are rethinking the “non-metaphysical” label. For the point is not that Hegel limits himself to only modest knowledge about our own cognition as opposed to knowledge of reality. Rather, in Hegel’s hands, Kantian reflection on the necessary conditions of meaning and knowledge yields knowledge of reality—knowledge that is not in any intelligible respect limited or less than absolute.[8] So the idea is that Hegel is an “anti-realist,” but of a radically anti-skeptical variety. And if “metaphysics” just means inquiry into what there is, then the point is that Hegel’s philosophy comes to metaphysical conclusions in that sense.

But, by the same token, this story has oriented its understanding of Hegel around a debate whose philosophical substance can be elusive. If all sides could agree about Hegel’s conclusions are “metaphysical,” depending on how that term is used, then there is a threat that the debate is merely verbal. Further, attempting to locate Hegel relative to disagreements about anti-realism, brings us to disagreements facing the same threat: Realists will argue that anti-realism simply must be a form of skepticism, social constructivism, or some other denial of objectivity. Anti-skeptical anti-realists will answer that this is a misunderstanding, because their view is anti-skeptical; they will counter-charge that realists conceive of objectivity in a manner that is incoherent. Realists will respond that they do not hold the incoherent conception attributed to them. Anti-realists will charge that realists assume an objectionable account of the truth predicate in semantics. Realists will respond that the debate is not about semantics but about what exists. Etc. Those of us not involved in the fray will naturally wonder whether there can be a substantive philosophical debate if the sides so entirely fail to agree about what is being debated.

Fortunately, I need not try to clarify the opposing positions on the object level of debates about anti-realism. For my view is that it is a mistake to read Hegel as if his basic project were to take any stand in anything like a debate about anti-realism. For my purposes, the most important feature of the story above about Kant and Hegel is the dependence on a view at the meta-level, or a metaphilosophy. The dependence is clear if we examine the need to argue against an opponent, rather than just tell a story. It is clearest if we consider an especially modest sort of pre-critical metaphysician—call him PCM. He might say:

PCM: Look, no philosophy can explain everything. Everyone must take something for granted. I begin with our knowledge of ordinary things, like rocks and trees and their parts, and then I use this to address the metaphysical issues of interest to me—for example, the question of whether these objects have indivisible atomic parts. SF seems to think that I am thereby leaving unanswered some other question about the possibility of knowledge of such things, or the possibility of meaningfully thinking of them. But, if so, then these are questions that I have no interest in answering. Nor, since we clearly can meaningfully think about trees, and most will agree that we can know about them, do I have any need of an account of the conditions of the possibility of such meaning and knowledge.

Now the “non-metaphysical” story about Kant and Hegel begins with the idea that wherever we find such metaphysics, without a prior Kantian reflection on cognition, we are essentially dealing with objectionably “realist” account of the possibility of knowledge. But what could the argument be for this claim against modest PCM, who takes no interest in and does without any account of this topic at all? As far as I can see, there can be an argument here, but only on the basis of strong metaphilosophical claim that the consideration of our cognition and the problem of accounting for the possibility of knowledge is fundamental to philosophy in a way that makes it inescapable. SF, for example, might defend herself as follows:

SF: Claims about any rocks or trees or any X presume the possibility of knowledge of X. So such claims always raise philosophical problems about the explanation of the possibility of this knowledge. If PCM thinks he can ignore those issues, he is mistaken. For he is implicitly committed to some explanation of knowledge of X. Further, such explanations can either be realist or anti-realist. But PCM’s account includes no analysis of our own cognition, only discussions of the objects of knowledge, X. So he cannot account for the possibility of knowledge in an anti-realist manner, in terms of the features of our cognition, such as its spontaneity. So he is committed, know it or not, to the realist view that the objects of knowledge, X, explain the possibility of our having knowledge of them. But an explanans must be distinct from an explanandum, or there is no real explanation. So PCM implicitly takes X to be something independent of our cognition. And that means that PCM implicitly takes our knowledge to be explained by a standard that is independent of our perspective—which is the “realist” view that will inevitably support skepticism, etc.     

This is essentially an argument of the “you are either with us or against us” sort: no one can be neutral, because the issue is too important, and there are only two kinds of response. And SF will eventually turn this general line of argument against Kant as well. We could just as well imagine a modest Kantian, who claims ignorance of things in themselves. SF will have to argue that any such claim is committed to an account in semantics of the possibility of meaningful thought about things in themselves. And that any such semantics must be either realist or anti-realist. And that realism must be rejected in semantics too, undercutting “realist skeptical doubts” including “‘thing in itself’ skepticism.” So the arguments in what was once called “non-metaphysical” readings of Hegel must depend on the metaphilosophical commitment to the fundamentality or inescapability of issues about the possibility of knowledge and, especially, meaning. And we can call this a “priority of semantics” or “semantics-first” approach to Hegel.

Granted, there are any number of ways of developing a “semantics-first” approach that could be more balanced in other respects. Consider the quietist, Q, who argues for this view:

Q: Constructive philosophy must be avoided altogether, because it will inevitably involve commitment to either a “realist” or “anti-realist” account of the possibility of knowledge and meaning, and neither is acceptable.

True, this view gives priority to neither realism nor anti-realism. But the view is still driven by a priority claim at the meta-level: issues about knowledge and meaning are supposedly so fundamental that entanglement in one or another position on just those issues becomes inescapable for any constructive philosophy.

Or consider the defender of a seemingly balanced semantics, BS, who will say this:

BS: Hegel argues that, in order to account in semantics for the possibility of meaningful thought, we must recognize certain ontological commitments about what there is. So Hegel takes such ontology to be just as legitimate, for this reason, as semantics. 

But even if this results in a balance between semantics and ontology at the object-level, what is distinctive about the philosophy here is the way it is shaped by a priority-claim at the meta-level: ontology is legitimate only because and insofar as a kind of ontology can avoid conflict with an account in semantics of the possibility of meaning, and only because and insofar ontology shaped in this way can contribute to semantics. I will argue that Hegel’s project is quite the opposite: he takes a kind metaphysics as fundamental, discussing meaning and knowledge where these do not conflict with but are necessary to advance the fundamentally metaphysical project; for example, discussion of knowledge and meaning will be necessary in a metaphysical account of what we ourselves are and how we fit into the rest of reality—not in the sense of an account of the conditions of the possibility of knowledge or meaning, but rather an account of the ways in which knowledge and meaning make us what we are.

2. Distinguishing the Metaphysics of Reason

Different understandings of “metaphysics” will agree that it is an inquiry into something that is supposed to be privileged or prior, disagreeing on the relevant senses of priority or privilege. The most minimal understanding would take metaphysics to be inquiry into what exists—or what is privileged merely insofar as it exists rather than not. Some might prefer to call that “ontology,” but it makes no difference here. More important is that those who look at philosophy through the epistemological lens of problems concerning the possibility of knowledge of a world independent of our perspective will tend to take “metaphysics” to be inquiry into whatever is privileged in the sense of being perspective-independent. If that is right, then to read as metaphysics any project of accounting for “the absolute” would be to read it as inquiry into something supposed to be absolutely perspective-independent. But it is crucial that this is only one way of understanding metaphysics; it is what I call “the metaphysics of perspective-independence.” We can distinguish “the metaphysics of reason,” which would be inquiry into whatever is privileged insofar as it is a reason for something else. Thinking in this way, a metaphysical account of “the absolute” would be a very different undertaking: inquiry into reasons that are most basic, fundamental or complete—regardless of whether those reasons are independent of us or not.  

To see what I mean by “reasons,” consider again this example of a recognizably metaphysical debate: Some monists will argue that the one whole of all reality is the reason why there are many parts and/or why the parts are as they are or do what they do. Some atomists will take the opposite view: atomic parts are the reason why there is a whole of everything, and why that the whole is as it is and does what it does.[9] Perhaps others would develop and defend a sense in which the whole and parts can be the reasons for one another. But, in any case, note that the sense of “reason” here is not epistemic: the question is not whether we conclude that there are parts for the reason that we know there to be a whole, or vice versa. Further, the question here has nothing to do with whether the whole or the parts is more independent of our perspective. It is open to the monist, for example, to say that they are precisely equal in perspective-independence; the point is that the former is prior in a different sense: the whole is the reason why there are parts, and why they are as they are. So this inquiry need not presuppose any special conception of perspective-independence. This is a dispute within the metaphysics of reason.

Further, this sort of debate can be of interest to almost anyone. Anti-realists who prefer a coherence theory of truth can take an interest in the question just as much as those who prefer a correspondence theory of truth, those who think there is nothing interesting to be said about the truth predicate, and those who just don’t have a position concerning these options. The metaphysics of reason in itself is not built on any of these specific notions of truth, but by what it seeks the truth about: about what is a reason for what.[10]

Similarly, imagine you hold the radically anti-skeptical form of anti-realism that some people see in Hegel, namely: we can have knowledge of the “objective” world in an internal sense; and we cannot even meaningfully or intelligibly engage with any competing accounts of objectivity as absolute or external perspective-independence. If you hold that view, then it will make sense to ask you questions about the objective world you recognize, including the question: what is the reason for what in this objective world? And since your view is radically anti-skeptical—you have dismissed as unintelligible what you call “realist skeptical doubts” including Kant’s claim to ignorance of things in themselves—you will not say that we cannot in principle know the answer to such questions about what is a reason for what in the objective world. So even the sort of anti-realism some see in Hegel would not preclude engagement with the metaphysics of reason.

Perhaps some will think that “metaphysics” must always involve positing a supposedly higher realm, or ground beyond that of finite persons. Perhaps that assumption is explained by a tendency to think of the metaphysics of perspective-independence, and thus to assume that any metaphysics must privilege something independent of us. But being free of such an assumption is an advantage of thinking instead in terms of reasons. Consider this sort of question: is there anything beyond us finite persons that is a reason why we exist and are as we are? It would be odd if our understanding of what sorts of questions, issues or inquiries are “metaphysical” would then classify one answer as metaphysical (the affirmative answer, insofar as it posits a beyond) and another answer as not metaphysical (the negative answer, insofar as it does not). I would take this as a recognizably metaphysical question—a question about what sorts of reasons there are. And then we can more naturally say that different answers equally state positions within metaphysics. “Yes” is the position that there is such a beyond, which I would call “the metaphysics of the beyond”; “no” is the equally metaphysical position that there is no such thing.

Returning to Hegel, then, when I say that I prefer a metaphysical interpretation, I mean that Hegel most fundamentally pursues inquiry into what is a reason for what. I do not mean that Hegel’s basic goal is to address in any particular manner the question of how to account for the possibility of knowledge or meaning—whether realist, anti-realist, or whatever. I mean that addressing such debates is not Hegel’s basic aim, and so not the way to understand his philosophy. Rather, his basic aim is to construct a metaphysics of reason.

To see Hegel’s engagement with the metaphysics of reason, consider the initial example the laws of nature. The natural sciences seek knowledge of what is and what is not a law of nature. But there is a philosophical question here too: what is it to be a law of nature? On a “humean” approach, a law will just be a regularity or a generalization stating a regularity. (It is a matter of debate whether Hume himself is a “humean,” in this sense.[11]) Anti-humeans hold that a law is something else, something which governs events, or something responsible for which regularities hold. Recent humeans have emphasized that theirs is a “non-governing” conception of laws (Beebee 2000). For if laws are just regularities, then they more summarize than govern events.

Hegel clearly holds that the laws of nature are the reason for the events that fall under them, or that laws govern. Consider the Lectures on the History of Philosophy, explaining how an idea of Anaxagoras’ provides the basis for further progress in philosophy:

Anaxagoras was the first to enunciate the doctrine that understanding generally, or reason, governs the world… The movement of the solar system is governed by unalterable laws; these laws are its reason. But neither the sun nor the planets which revolve around it are conscious of them. It is man who abstracts the laws from empirical reality and acquires knowledge of them. Any idea of this kind, that there is reason in nature or that it is governed by unalterable universal laws, does not strike us as in any way strange… (LPH 12:23/34)

And Hegel notes how philosophy builds on the basic idea about reason or governing. Plato builds on it, even while rejecting the way in which Anaxagoras sought reasons merely in “Atmosphere, Ether, Water, and the like.” Indeed, Hegel here treats exceptionless laws of nature governing the motion of matter as a form of reason in the world—but he will also argue that this is the least complete or absolute form of reason. Still, for now I focus on the sense in which laws are at least one form: the laws are the “reason” for the rotation of the planets; and this is just to say that their movements are “governed” by laws. The point is not epistemic—it is not that knowledge of laws is our “reason” for believing or concluding that the planets rotate. Rather, the laws are a form of “governing” or “reason in nature.” The same commitment is clear in Hegel’s rejection of empiricism at the beginning of the Encyclopedia, which formulates and rejects the approach I call “humean,” according to which “universal notions, principles, and laws” (§38) signify nothing over and above “alterations that follow one after the other, and … objects that lie side by side” (§39).

Those who favor humeanism about laws might try to argue that the very idea of a law of nature being a “reason,” or “governing,” is so unclear as to be meaningless. But thinking in terms of Hegel’s interest in reasons of all forms reveals the objection as powerless. For both sides in this debate are equally addressing an issue within the metaphysics of reason: both sides have a position on what is a reason for what. The humeans accept one form of reason in the world—they hold that what actually happens is the reason why there are laws (if there are, because laws are just patterns or regularities in what happens). Some recent humeans say specifically that it is “in virtue of” what happens that there are laws;[12] or, there is a form of “ontological” “grounding” of laws by what happens, or a “reduction” of laws to what happens.[13] These are all proposed ways of understanding the sense in which what happens is supposed to be the reason for the laws. An anti-humean, like Hegel, simply holds the reverse: in cases where there are laws, like the rotation of the planets, the laws are the reason for those happenings. So both sides are talking about reason, but they see different sorts of reason-relation moving in opposite directions. Since both sides are talking about different forms of the same general notion, humeans cannot object that their opponents are too obscure.

Granted, some today might wish to reject both sides in this debate, finding anti-humean talk of “governing” and humean talk of “ontological grounding” both too obscure—perhaps because neither is analyzable in terms of ordinary causality. But this objection too is weak, because it assumes that causality itself is unproblematic, or something not in need of more basic philosophical comprehension. It is however obvious that there is philosophical debate about what causality is. And the debate is similar to the debate about laws. There is, for example, a humean conception of causality: for x to cause y is just for X’s and Y’s to be constantly conjoined throughout space and time. And there are anti-humeans about causality. And this is again a dispute about reason: humeans hold that what happens is the reason why there is causality; anti-humeans hold that, where there are causes, these are the reasons why things happen as they do.

What we are discovering, as we step through these debates, is the fundamentality of the question of what is a reason for what. We need the basic and general notion of one thing being a reason for another in order to engage any of these debates. And so we should accept that notion as basic, and proceed to consider what specific forms of reason there really are, which directions they run in different cases, and how they relate to one another.

Further, the philosophical debates in question cannot be directly resolved by the natural sciences. No matter which specific laws natural science might uncover, doing so will not itself answer the underlying question about what it is to be a law, or the form and direction of reason in terms of which lawhood is best understood. And part of what is at issue here is how different natural sciences relate to one another: are they all seeking reasons in a similar sense, or in divergent senses? If divergent, then how do these different reasons relate? So part of the aim of answering these questions within the metaphysics of reasons is to rationally and systematically comprehend the natural sciences themselves, the specific sorts of reasons they seek, and their relations. So although the sciences seek forms of reason in the world, the metaphysics of reason is distinguished from them in that it seeks to understand those forms of reason and how they relate.

Those who prefer a semantics-first metaphilosophy might see all such issues about laws, monism, causes, etc. as a series of technical issues in isolated or minor areas of philosophy—as “merely” issues in the philosophy of science, for example, peripheral to the core concerns of philosophy with perspective-independence, knowledge, and meaning. But my point here is just that there is an alternative way to look at it. When we think in terms of the metaphysics of reason, we see rather a surprising thread linking issues throughout metaphysics. Take the example of the humean views above: they all stem from a wonderfully clear and incredibly comprehensive humean metaphysics—all there is to reality, says the humean, is a series of disconnected events arranged in space and time; this arrangement is the reason for everything there is—for laws, causality, necessity, etc.—it can all be explained in these same terms. As David Lewis formulates the form of humeanism so important in recent analytic metaphysics, it is “the doctrine that all there is to the world is a vast mosaic of local matters of particular fact, just one little thing and then another” (1986, ix ).

I think that Hegel’s basic aim is to defend a comparably comprehensive position within the metaphysics of reason—albeit one that is about as far away as you can get in content and simplicity as compared to humeanism. The humean view finds the fundamental at the bottom: the reason for everything real is found at the bottom, with “one little thing and then another.” Hegel’s view will find the fundamental at the top, although in a surprising sense. This will mean that the example of Hegel’s anti-humean position on the lowest-level laws of nature is only an introductory example. Hegel’s view is not that the unalterable laws of nature are absolutely fundamental to everything. The rotation of the solar system as an example of something governed by laws of nature. But this does not mean that Hegel sees everything as so governed. Hegel will argue that the behavior of living beings, for example, is teleological, and not governed by exceptionless laws.[14] And Hegel will argue that the laws of nature are only a very limited, incomplete, or minimal form of reason for the events they govern; teleological reasons for the behavior of living beings will turn out to be a more complete form of reason—closer to “the absolute.” And it is not life but rather Geist or spirit that will turn out to manifest the most complete or absolute form of reason of all. So the laws of nature will turn out to be relatively less important in Hegel’s philosophy, even if a good place to begin.

This is to say that Hegel will consider many different accounts of what is a reason for what, in many different senses of a “reason.” This will include those metaphysicians who think about reason as “Grund” (ground), in a sense that Hegel thinks inadequate. Thus Hegel will complain about the way reason is generally understood by advocates of the “Satz des Grundes” or principle of sufficient reason (WL 446/6:82). But what Hegel thinks most important here are those aspects of Leibniz’s use of the principle which direct us rather toward a more complete form of reason, which he calls here “teleological ground” (teleologische Grund). Elsewhere Hegel uses “Vernunft” (reason) to speak of reasons in the world, as when he says of the movement of the solar system that the “laws are its reason,” or the “Gesetze sind die Vernunft desselben” (LPH 12:23/34). And we will see that the term Vernunft establishes an important connection to Kant. But these are in any case different ways of thinking about reasons. The forms of independence most important to Hegel, like what he calls “Selbstständigkeit” (self-subsistence), do not concern at base independence of our point of view—they rather raise issues about whether something is an independent reason for itself; Hegel will argue that this is a standard by which teleological phenomena will be of greater interest in a metaphysics of reason than phenomena governed by exceptionless laws.

3. Kant’s “Transcendental Dialectic” Critique of the Metaphysics of Reason

I turn now to Kant’s critique of metaphysics in the “Transcendental Dialectic” of the first Critique, so that we can later understand Hegel’s response. The thread running throughout this section is Kant’s account of “the faculty of reason” (Vernunft). The basic idea here is that truth is too profligate to serve as any kind of guiding goal for theoretical or rational inquiry. For there are innumerable truths about any number of things, many of them trivial or of no special theoretical interest—such as the exact number of cookies in each box of cookies, the exact distance in miles between the box and the Golden gate bridge, and which of these numbers is higher. The “faculty of the understanding” is capable of coming to true judgments about what is the case in the empirical world; but the understanding nonetheless requires guidance from some more discriminating norm or goal, and it is the “faculty of reason” that provides the guidance. Because reason’s role is to guide, Kant must characterize reason in terms of an “aim” or “end” (Zweck) or even an “interest” (Interesse). And one way of expressing the goal of reason is to say that we are interested, insofar as we are rational, in underlying grounds or conditions of all kinds, or in following a “regress from the conditioned to its condition” (A523/B521).

But Kant exposes difficulties by arguing that this last idea about conditions is not yet an adequate characterization of a guiding aim of the faculty of reason. In a way, conditions threaten to be still too profligate to guide. Imagine for example that we know of some X that is conditioned; if so, the faculty of reason will leave us unsatisfied and interested in the underlying condition. But what if the underlying condition, Y, is also something merely conditioned? Then the same dissatisfaction of reason will persist. So there are many conditions that do not in themselves satisfy reason’s interest. Theoretical inquiry is concerned in conditions only insofar as we assume that, in finding conditions, we are making some progress toward knowledge of an underlying or unifying completion in the series of conditions, or a complete reason—progress toward what Kant calls “the unconditioned.” And so Kant argues that, if we find the condition for something, reason requires that “the condition of its condition thereby has to be sought”; thus:

…we see very well that the proper principle of reason in general (in its logical use) is to find the unconditioned for conditioned cognitions of the understanding, with which its unity will be completed. (A307/B364)

Another way of putting this point is to say that reason seeks complete unity, or a unified underlying complete explanation. And this is the goal or norm by which reason provides the guidance needed by the understanding: “the law of reason to seek unity is necessary, since without it we would have no reason, and without that, no coherent use of the understanding” (A651/B679).

It is worth noting an example crucial to the “Transcendental Analytic,” insofar as it is the topic of the “Second Antinomy”: the understanding might achieve knowledge of an object extended in space; reason will then take an interest in why it fills this volume of space, or the underlying “conditions” in the sense of the parts. Note that this is not the question of the epistemic reason how we know that or whether the object is extended. The question about underlying conditions (Bedingungen) here is a question about the reason—in the sense which the metaphysics of reasons is interested in reasons—why the object fills that space.[15] The idea of the “unconditioned,” then, is not any epistemic idea, such as the idea of something knowable with absolute certainty. Nor is it fundamentally the idea of something independent of our perspective, or the idea of something knowable only from a God’s-eye view. It is the idea of a complete reason why—for example, the complete reason why something extended fills that region of space.  

But Kant is aiming to further argue that we cannot ever attain knowledge of anything unconditioned in this last sense. In other words, we seek the objects of theoretical inquiry under the description of the unconditioned, or the complete reason why—not under the an epistemic description like the absolutely perspective-independent, or the knowable with absolute certainty, or the knowable only from a God’s-eye-perspective, etc. But it turns out, Kant will argue, that what we seek under the former description is in fact unknowable for us. Thus Kant begins the A-preface to the Critique with the resulting tension between reason’s interest and the limits of our knowledge:

Human reason has the peculiar fate in one species of its cognitions that it is burdened with questions which it cannot dismiss, since they are given to it as problems by the nature of reason itself, but which it also cannot answer, since they transcend every capacity of human reason…

Our attempts to answer these questions cannot be conclusive, and so we fall into endless controversies. And this is metaphysics: “[t]he battlefield of these endless controversies is called metaphysics” (Avii-viii). Note that “metaphysics” in this sense has nothing fundamentally to do with absolute perspective-independence, nor with any idea of a God’s-eye view. “Metaphysics,” in this sense is concerned fundamentally with the complete reasons for things, or the unconditioned. It is only on grounds of a further argument that our form of cognition cannot achieve knowledge of anything unconditioned that the derivative result will follow: metaphysics amounts to an interest in something (the unconditioned) that is unknowable from our point of view.

The resulting problem will be this: We seek insofar as we are rational or reasonable to reach conclusions concerning the unconditioned. But, on the face of it, our choices with respect to the existence of unconditioned grounds are that they exist or that they do not. And Kant argues that holding either view is unacceptable.

Consider first the affirmation. Kant argues that we are naturally tempted by it. That is, we are naturally tempted to mistake the maxims or rules guiding reason for objective principles: “In our reason … there lie fundamental rules and maxims for its use, which look entirely like objective principles” (A297/B353). So given reason’s demand that we seek the unconditioned, we are tempted to affirm this further principle: there must always be unconditioned grounds for everything conditioned. This is essentially the “principle of sufficient reason” (PSR) of the early modern rationalists. The idea is that everything not a sufficient ground or reason of itself—everything conditioned—must have an external sufficient or complete ground or reason, or an unconditioned ground. “Rationalism,” in the sense I will use it, is not fundamentally a position in epistemology. Rather, rationalism is any view in metaphysics that endorses a PSR and argues on that basis for the existence of “God” in this sense of a sufficient reason for everything.[16] But note that Spinoza is a paradigmatic rationalist—he employs the PSR in arguing for his “God” at Ethics IP11D. But of course his God is not something separate from the rest of reality, but the single substance that everything real is “in.” This is irrelevant to his status as a paradigmatic rationalist (in my sense here): he argues from a PSR to the existence of an unconditioned ground of everything, or “God” in that sense.

Kant’s response to rationalism is subtle. On the one hand, the interest of our faculty of reason leaves us naturally tempted by rationalism. For example, consider the rationalist arguments for God, in the rationalist sense of “that the concept of which contains within itself the ‘Because’ to every ‘Why?’ … that which is in all ways sufficient as a condition.” Kant holds that reason itself makes this rationalist argument tempting: this is even “the natural course taken by every human reason” (A584-5/B612-3). On the other hand, the Transcendental Dialectic argues that we must learn to avoid asserting theoretical knowledge of such rationalist conclusions.

The problem with rationalism is supposed to be this: it must either contradict itself, or else come to depend on an untenable combination of epistemological claims. The threatened self-contradictions are developed in the first two Antinomies. The contradictions arise from a specifically rationalist line of argument, whose major premise is that there must be completion or sufficiency in a series of conditions: “The entire antinomy of pure reason rests on this dialectical argument: If the conditioned is given, then the whole series of all conditions for it is also given; now objects of the senses are given as conditioned; consequently, etc.”[17] The problem arises specifically with respect to a regress of conditions in time (the first Antinomy) or space (the second). It is best, for the purpose of understanding Hegel’s response, to briefly sketch the second Antinomy.

First, imagine accepting what I will call Assumption A: beginning with an object extended in space, there is an infinite regress to smaller parts. But a rationalist must insist that there is a sufficient reason why there can be any composition here at all. And the only way to find such a reason within the regress would be to hold that there are smallest, simplest parts which explain why there is anything here out of which things could be composed.[18] So a rationalist looking for completeness within the regress must reject the Assumption A and instead endorse smallest spatial parts.

Second, we can try Assumption not-A: there are such smallest parts in space. If they occupy some region of space, then the rationalist must insist that there is a sufficient reason why they occupy that region. Within the regress in space, the only thing that could provide such a reason would be smaller parts which, conjoined, occupy that region. So for any simple or indivisible part in space, the rationalist must hold that there are underlying parts—which is a contradiction.[19] And so the rationalist must reject the Assumption not-A, and hold instead that the regress to smaller parts extends infinitely down.

But it is important that there is a popular way for the rationalist to escape contradiction. He can say that there is another option—unlike both smallest parts and infinite descent into composition—specifically insofar as the sufficient reasons for things might be comprehensible and knowable only by a divine intellect. What God would be able to comprehend is how there could be an infinite regress of conditions, and then also, outside of the infinite regress, an underlying sufficient condition for all of them. For God need not follow a regress one step at a time; Leibniz holds that God could grasp even an infinite series all at once: “there is always, underneath, a reason … even if it is perfectly understood only by God, who alone goes through an infinite series in one act of the mind” (Ariew and Garber, 303). Applying this escape strategy specifically to the regress in space, we get this view: insofar as complete reasons might be such as to be comprehended only by God, there could be an infinite regress of smaller parts in space, which also has a ground or reason from outside itself in the form of non-extended monads.[20]

But Kant will respond that the rationalist escape route requires both affirming and denying oneself the possibility of knowledge of the same thing. It is not that Kant thinks we are entirely unable to comprehend the idea of an intellect able to grasp immediately even an infinity of everything—this is a central feature of what Kant calls intellectual intuition, which “would grasp and present the object immediately and all at once” (8:389). But if the rationalist speaks of sufficient reasons knowable only by such an intellect, superior in kind to our own, then Kant argues that the right conclusion is that we cannot know anything about them—not even whether there exists any such thing. Of course, there is a difference between knowing that something exists and knowing more about it; but Kant argues that if the latter requires divine knowledge, then there is no principled reason why the former should not as well. For example, Kant complains that those who defend Leibnizian monads “would have us be able to cognize things, thus intuit them, even without senses, consequently they would have it that we have a faculty of cognition entirely distinct from the human not merely in degree but even in intuition and kind,” possessed by “not by humans but beings that we cannot even say are possible, let alone how they are constituted” (A277-8/B333-4).

Note that this criticism need not rely on commitments from Kant’s specific claims about the limits of our cognition, its discursivity, etc. There is rather here an independent line of argument, which will offer support for Kant’s specific claims about our epistemic limitations. The point about rationalists is that they box themselves into an untenable epistemic position: they can escape from the contradictions of the Antinomies only by claiming that reality is such as to be knowable only by a divine mind; but this involves claiming to know something while also claiming it is unknowable.

Now to this point I have commented only on Kant’s critique of rationalist metaphysics. But the “Transcendental Dialectic” is not just developing a worry about rationalist metaphysics; it is developing a worry about metaphysics more generally. It presents a problem that is supposed to affect everyone, not just the rationalist. The root of the problem is that we need the faculty of reason, and its guidance by means of ideas of the unconditioned, if we are to pursue any theoretical inquiry at all, and even if there is to be any “coherent use of the understanding” (A651/B679). We cannot rationally hold that there are no unconditioned grounds while also trying to discover some—which is to say, while still engaging in any theoretical inquiry at all. So the denial leads to the skeptical renunciation of the project of reason or theoretical inquiry, “the euthanasia of pure reason,” or “skeptical hopelessness” (A407/B433-4).

A quick way to get a glimpse of the point is to consider a famous scene in Molière: the character Argan is asked why opium interacts with us by putting us to sleep. And Argan responds, famously, that opium has a dormitive virtue or power. Of course, we now know better. We know what opium is made of, and why it does what it does. But consider the farthest point to which we have advanced in the regress of powers or dispositions. Say we find, at the limit of our current knowledge, particles X and Y, where X’s attract Y’s. We might still ask why do X’s attract Y’s? A contemporary Argan would say: on account of their attractive power. But is it rational to take this answer for any sort of conclusion? Kant would deny it. And he would explain his answer in this way: reason demands that we assume, at least for the sake of inquiry, that there is something more complete to be said in answer to the why-question. And reason demands that we inquire into what that more complete explanation might be. True, one could conceivably deny that there is anything more satisfying. But then one has no grounds left for saying that reason favors further inquiry over Argan’s self-satisfaction or utter lack of intellectual curiosity. And that is skeptical hopelessness.

While I think that there is more to say in defense of this last point, this will have to wait for another occasion. For our purposes, the important point is that Kant’s “Dialectic” aims to present a seemingly insoluble problem for everyone. We can affirm or deny rationalism, but this

leads reason into the temptation either to surrender itself to a skeptical hopelessness or else to assume the attitude of a dogmatic stubbornness… Either alternative is the death of a healthy philosophy though the former might also be called the euthanasia of pure reason. (A407/B433-4)

Now Kant raises this problem in order to argue that there is one acceptable alternative, and only one—something he thinks is new and radical. We must conclude that our own knowledge is fundamentally limited or restricted. More specifically, we must conclude that there are specific, principled limits of our knowledge—limits that will preclude knowledge of whether or not there is anything unconditioned, so that we can continue to be guided by ideas of the unconditioned without threat of the rationalist conclusion that we can know any such thing (thus precluding further theoretical inquiry) and without threat of the denial that there is any such thing (thus rendering further inquiry without reason of guidance). What compelling account of our knowledge would leave it limited in principle, in just this way? Kant’s answer is that our understanding is merely “discursive”, or dependent for content on intuitions from a receptive faculty of sensibility, and so also dependent on the a priori forms of our sensibility: space and time. And “in sensibility, i.e. in space and time, every condition to which we can attain in the exposition of given appearances is in turn conditioned” (A508/B536). This account of our limitations gains support in this way: it provides a principled reason for denying knowledge of anything unconditioned, including knowledge of whether or not there is anything unconditioned, thus explaining why the threat of the Antinomy does not justify complete skeptical hopelessness.  

Precisely this limitation of knowledge is what will allow Kant to hold that we can and must always assume, for the sake of inquiry, that there are unconditioned grounds—and then seek them in inquiry.[21] So reason demands that we pursue the unconditioned as an end, a goal; and reason provides perfectly legitimate “regulative” or guiding principles; but we must learn not to mistake this for knowledge of anything unconditioned. We can thereby at least make progress “asymptotically” (A663/B691) as Kant says at one point, toward the goal of reason that guides scientific inquiry. And thus Kant claims to avoid the “skeptical hopelessness,” or the “euthanasia of pure reason,” but without slipping back into the “dogmatism” of rationalist assertions.

This line of argument for the limitation of our knowledge is clear to see in Kant’s own gloss of the argument of the Transcendental Dialectic. Kant says:

That which necessarily drives us to go beyond the boundaries of experience and all appearances is the unconditioned, which reason necessarily and with every right demands… (Bxx)

True, we cannot know that there really is anything unconditioned. But precisely because we still must conceive the unconditioned as a goal, we must conceive of it as unknowable, as “present in things … as things in themselves.” Otherwise, “the unconditioned cannot be thought at all without contradiction” (Bxx). So the threat of contradiction concerning the unconditioned forces us to distinguish the objects of our knowledge from things as they are in themselves, and conclude that our knowledge is merely limited or restricted.

It follows that there is a sense in which metaphysics is impossible for us. Recall that reason gives rise to “endless controversies” on “the battlefield … called metaphysics” (Aviii). If reason is responsible, then these are controversies about the unconditioned. The Dialectic arguments supports the conclusion that such metaphysics is impossible for us: we cannot legitimately assert knowledge of any conclusion about this topic. But this is not to say we can or should forget these controversies: Kant says just here that reason “cannot dismiss” (Avii) those metaphysical questions. He later says that “we will always return to metaphysics as to a beloved from whom we have been estranged” (A850/B878). True, some philosophers think that they have become indifferent to such metaphysics, but Kant sees them as unknowingly entangled in metaphysics: they “always unavoidably fall back into metaphysical assertions, which they yet professed so much to despise” (Ax). So the point is not that metaphysics stems from an optional or misguided interest in knowing things from a God’s eye point of view. It stems from a rational, legitimate, an ineliminable interest in the unconditioned or complete reasons; and we must keep in mind that inescapable interest, precisely in order to guard against mistakenly thinking that we can attain theoretical knowledge that would satisfy it.

4. Hegel’s Non-Rationalist, Post-Kantian Metaphysics of Reason

When we turn back to Hegel, we now have a choice. We need not read Hegel as if his basic goal were to defend one or another approach to explaining the possibility of knowledge and or meaning. So we need not take the basic orienting question to be whether Hegel is more of a realist or anti-realist about that matter. We can instead read Hegel as interested fundamentally in the issues raised by Kant’s Transcendental Dialectic critique of the metaphysics of reason. So the basic orienting question would be how Hegel’s view relates to the options Kant distinguishes there: the rationalist affirmation of the existence of unconditioned grounds, the denial of the same which Kant labels skeptical hopelessness, and Kant’s limitation of our knowledge. Does Hegel go for one of these options, or develop some other response to this same set of issues?

This orienting question makes it easy to see that traditional metaphysical readings, although not always explicit about it, tend to portray Hegel as holding a version of precisely the metaphysical rationalism targeted by Kant. For they portray Hegel as arguing on the basis of an assumption that everything must be completely explicable, so that there must be some complete reason for everything; such readings often take Hegel as accepting Spinoza’s claim that the whole of everything is a ground of everything, and then modifying Spinoza’s account of that whole. But that would be to say that Hegel merely assumes a principle that Kant has argued against, rather than engaging philosophically with Kant’s critique of metaphysics.   

My aim here is to show how the a metaphysics of reason approach makes possible an alternative reading—one which Hegel is not arguing for a return to specific metaphysical view specifically attacked by Kant. The idea is that Hegel follows Kant’s rejections of various basic options: Hegel agrees that problems concerning complete or absolute reasons are inescapable, so philosophers cannot claim to be indifferent to them; Hegel agrees in rejecting rationalism, or the claim that complete or absolute reasons would be something unconditioned, and that there must really be something unconditioned; Hegel also agrees that rejection of the existence of any sort of complete or absolute reasons would be an unacceptable form of “skeptical hopelesness.” But all this agreement does not mean that Hegel accepts Kant’s limitation of our knowledge and denial of the possibility of our reaching conclusions in metaphysics. Rather, Hegel argues that Kant sees only these options because he views the matter in part from the perspective of the aim—so important to earlier section of the first Critique—of defending the faculty of the understanding. So Hegel will argue that, if we begin with Kant’s own account of reason, and re-work philosophy from this perspective, then we can build a new kind of metaphysics—one that finds a new way to avoid both rationalism and yet also skeptical hopelessness. In particular, Hegel will argue that complete or absolute reasons should not be understood as the sort of unconditioned grounds which rationalists think are real, and which Kant thinks are legitimately of interest to reason. Once we better understand completeness or absoluteness of reasons, we will be able to legitimately assert knowledge of them, without asserting rationalism.  

To see the break between Hegel and rationalism, consider Hegel’s account of law-governed natural phenomena, as for example in the Logic discussion of natural kinds of things that are linked by exceptionless laws—which Hegel calls “Chemism,” emphasizing that he is not just talking about chemistry but lawfully interacting kinds of any sort.[22] Here Hegel argues that the nature of such a lawfully interacting thing is such that it cannot be comprehended except in relation to others. As Hegel says, “a chemical object is not comprehensible from itself alone, and the being of one is the being of the other” (WL 6:430/728). What is it to be some law-governed X? It is to react in a specific manners with Y’s, etc. And, further, to be Y will be to interact in characteristic ways with Z’s and X’s. Etc. So the “being” of things of such kinds will depend on a whole interconnected network of kinds and laws within which that are a part, or the network within which they are nodes. For law-governed stuff, the “determinateness” of anything in particular is just one “moment” of a larger “whole” or “Begriff” of the whole: it “is the concrete moment of the individual concept (Begriff) of the whole, which concept is the universal essence, the real kind (Gattung) of the particular object” (WL 6:430/728).

The philosophical pressure towards this kind of metaphysical holism has often been noted in more recent metaphysics, even while holism is often resisted. For example, take Russell:

There are many possible ways of turning some things hitherto regarded as “real” into mere laws concerning the other things. Obviously there must be a limit to this process, or else all the things in the world will merely be each other's washing… (1927, 325)

And Chalmers aims to resist the same holism: “…this would lead,” he says, “to a strangely insubstantial view of the physical world” (1996, 153).

In Chalmers and Russell, the aim of finding more substance in things pushes in the general direction of the view that physical reality is, in itself, mental or somehow akin to the mental. Chalmers reads Russell as attracted to the view that “the intrinsic properties of the physical are themselves a variety of phenomenal property.” And Chalmers considers “protophenomenal properties” (1996, 154). Kant reads Leibnizians as arguing in a similar manner: if there must be non-relational inner features of things, then “what can I think of as inner accidents except for those which my inner sense offers me? - namely that which is either itself thinking or which is analogous to one” (A265-6/B321-2). A more monistic form of rationalism would seek to dispel the seeming lack of substance by finding an unconditioned ground in the whole of all such interconnected kinds—we could also take the whole to be a mind or something similar.

But none of these is Hegel’s view. Hegel’s view is that, when we look to grasp a law-governed thing, it “gets lost” in relations, lost in a regress of dependence; it “becomes something else than it is empirically, confuses cognition” (Phän 3:190/149). Hegel’s view is that lawfully governed things really are “strangely insubstantial” (as Chalmers puts it).  Another way to put the point is to say that, when we look to lawful things, we find only the slightest of reasons for what they do at all. True, the laws are the reason why things do what they do. But the laws are as they are on account of the relational natures of such things. Which is to say that the natures of things are as they are on account of the laws, and the laws on account of the natures, and so on. The regress into relationality results in surprisingly insubstantial reasons.

Hegel’s anti-rationalist view is clearest where he discusses the “weakness” or “powerlessness” (Ohnmacht) (e.g. PN §248) or a degree of “unreason” (Unvernunft) (e.g. PN §250) in nature. Lowest-level law-governed things are weak insofar as they lack a complete or sufficient reason for why they do what they do. They falsify the PSR that is definitive of rationalism. And the point is not that we tend to mistakenly take lawful nature to be weak or unreasonable because we tend to overlook some hidden inner side, like hidden mental grounds in physical particles or in the whole of everything. Rather, lawful nature really is this weakness, or really this degree of unreason. For example, the lack of complete reason is a kind of contingency in the law-governed kinds or forms:

In the sphere of nature contingency and determination from without has its right, and this contingency is at its greatest in the realm of concrete individual forms, which however, as products of nature, are concrete only in an immediate manner … This is the powerlessness of nature. (PN §250)

Hegel will later argue that in some cases, there is a remedy for this lack of substance and reason. In particular, some of the sum total of all the lawfully interacting stuff finds itself part of different living beings—lions, tigers, trees, and you and I. When we look to these higher-level beings, we find some of the substance that had gone missing in lower-level lawful nature. Take a tiger, for example. Why does the tiger have these sharp claws and the corresponding capacities to catch and kill rabbits? Here we do not “get lost” in a regress of dependence. The answer is not that rabbits have a disposition to be caught, and so on. For the tiger has these parts and capacities on account of something about the tiger itself: an account of the intrinsic end or goal of self-preservation. As Hegel says, “the living thing is articulated purposefully; all its members serve only as means to the one end of self-preservation.”[23] This intrinsic end or Zweck is supposed to allow the nature of an organism to be manifest in the determinate way that it relates to the environment, yet without its nature merely dissolving into relations with others. An organism is “the real end or Zweck itself … it preserves itself in the relation to an other” (Phän 156). This is supposed to make life a more complete form of reason, specifically as compared to lawfully governed lower-level things. Or, alternatively, a living being is more substantial—precisely in the sense that lawfully interacting things are strangely insubstantial.

What is really surprising here is that a living being will be more substantial, in the above sense, than even the lower-level law-governed stuff of which it is composed. This can seem surprising when judged from the perspective of what Kant calls the understanding, from which we expect that if X supports Y, then X had better be more solid than Y in order to hold it up, as it were. But Hegel is arguing that matters are different once judged simply from the perspective of the interest in reasons, or the reasons why things do what they do. A tiger, for example, could not exist without the existence of the underlying stuff of which it is composed. But there is also a sense in which the natures of the underlying stuff are a matter of indifference when it comes to what it does. The tiger’s claw, for example, could have had the same capacities and yet be realized in a variety of different underlying stuff.

Note, then, how Hegel’s conception of a complete or absolute reason comes apart from Kant’s conception of the unconditioned: an organism is, in a sense, conditioned by or dependent on the stuff of which it is composed. But insofar as this is a conditioning by something indifferent, this is no limitation of its status as a reason for its own behavior. We can mark this distinction with terms drawn from Hegel: dependence on an “indifferent base”[24] makes no difference to the completeness of a form of reason. So to manifest a complete or absolute form of reason need not mean depending on nothing—it need not mean being unconditioned in this sense. And, since a complete form of reason might depend in that way on something else, clearly a complete form of reason need not be something on which everything else depends—it need not be some form of the rationalists’ God. To be a complete or absolute form of reason is rather to be something that does what it does on the basis of what we might call a complete and internal “principle of activity.”[25]

Now organic life is supposed to be close to an absolute reason, in this sense, but not the most complete form. Although there is no space to even sketch the argument here, it is worth noting how Hegel puts the point in terms drawn from Kant: “idea,” in Kant, refers to a conception of something of interest to the faculty of reason; so Kant’s view is that the “idea” is always a conception of something unconditioned. Hegel’s view is that life (even if not unconditioned) is a form of “the idea,” is more satisfying to reason, although only a partial or incomplete form: “the idea is firstly life” (WL 6:468/760). But Hegel will argue that an absolutely complete intrinsic principle of activity is realized only in the case of our own kind, or what Hegel calls Geist or spirit. This is why Hegel will call Geist “the absolute”: “the absolute is spirit—this is the supreme definition of the absolute” (PG §384). Comparing life and spirit, then, the former retains a smaller degree of the unreason of nature, and only the latter is a complete or absolute form of reason, insofar as it is internally self-determining, or free in this sense:

The highest level to which nature attains is life; but this, as only a natural mode of the idea, is at the mercy of the unreason of externality … whereas in every expression of Geist there is contained the moment of free, universal self-relation. (PN §248An)

We can know the absolute, in this sense—we can have absolute knowledge—in that we can know something with an absolutely intrinsic principle of activity, something that realizes a complete form of reason: we can know and comprehend spirit. This need not mean that there is a complete reason for everything. Nor need it mean that we can know everything—for there are many things without a sufficient reason, which are not absolute. For example, if basic physical particles lack any complete or absolute reason for what they do, then knowledge of the absolute need not be any kind of complete knowledge explaining the location and movement of every physical particle. Knowledge of the absolute is knowledge of something with an absolutely internal principle of activity—knowledge of spirit.

The resulting metaphysics holds that reality has a hierarchical organization of levels.[26] Everything is composed of lower-level stuff, in which is found the most incomplete form of reason. Some of that stuff makes up living beings, whose behavior has more complete reason, or is more completely explicable. And some living beings have capacities of knowing (especially important will be the capacity to know themselves) and meaning, and behave in correspondingly distinct ways, because they are of the kind Geist, and it is here that we are supposed to find a complete form of reason.

Why doesn’t Hegel’s view, insofar as it is neither rationalist nor Kant’s own denial of knowledge, lead to the skeptical hopelessness Kant fears? First of all, Hegel’s view is that rational inquiry does not and need not seek knowledge of the absolutely unconditioned. For it need not have any interest in merely indifferent grounds. Rather, rational inquiry seeks knowledge of the internal principles of things. Such inquiry can be satisfied to increasingly greater degrees on the higher levels of reality. And even within lawful nature inquiry can make a sort of progress insofar as we discover at least a distant, imperfect echo of the sort of independent principle of activity found on the higher levels. For example, the whole network of kinds and laws within lawful nature is a distant but imperfect echo of the sorts of biological wholes we find on higher levels. It is comparable to, but “not yet for itself that totality of self-determination” (WL 6:434/731-2). So we can have some success in explaining nature insofar as we can find more distant approximations of the idea throughout nature, even if these approximations are in truth other than the idea. In Hegel’s terms, nature is “the idea in the form of otherness” (§247).

We can put the final combination of claims in this way: (i) Life approximates, but only Geist or spirit completely realizes, a complete independent principle of activity, or a complete manifestation of “freedom” and “the idea.” (ii) Metaphysically speaking, nature does not depend on Geist, but Geist does depend on the nature in which it is embodied—it “presupposes” nature. However, this is just dependence on an indifferent basis, in the sense discussed above, so it is no limitation of the status of Geist as a complete form of reason, or as a manifestation of “the idea.” And (iii) speaking now epistemologically, there is a kind of dependence of nature on Geist. In particular, the explicability of nature depends on Geist: we seek to explain nature insofar as we find there a distant trace of the kind of self-determining system or complete reason that is fully realized only in the case of spirit—so spirit finds, in this sense, traces of itself in nature, we find the world to be our world, in this sense. I will insert numbers for these points to note how Hegel wraps them all together:

(i) As Geist is free, its manifestation is to (iii) set forth Nature as its world; but because it is reflection, it, in thus setting forth its world, at the same time (ii) presupposes the world as a nature independently existing. (iii) In the intellectual sphere to reveal is thus to create a world as its being – a being in which the mind procures the affirmation and (i) truth of its freedom. (§384)

In this way metaphysics can deal with the problems Kant uncovers in the Dialectic. True, the Second Antinomy shows that rationalists cannot accept there being only compositionality all the way down; rationalists similarly could not accept the idea that lawful nature is only relationality all the way down or around. But Hegel shows how to construct a metaphysics that breaks with rationalism on this point, accepting an anti-rationalist view, while avoiding the skeptically hopeless denial of the existence of complete or absolute reason. The key is to argue that Kant and the rationalists misunderstand what a complete form of reason would be: to be a complete form of reason need not involve being a ground of everything, nor being unconditioned; a complete form of reason would be rather something with an intrinsic principle of activity of its own. Thus Hegel can argue that there is an absolute form of reason—Geist—without arguing that this is the sort of metaphysical ground of everything envisaged by the rationalists. And Hegel can argue that we can know this absolute, without having to somehow know everything or any supposedly complete explanation of everything.

5. For the Metaphysics of Reason and Against Semantics-first Metaphilosophy

I have argued that there is an alternative to “non-metaphysical,” or “semantics-first” interpretations of Hegel, which differ on matters philosophical rather than merely verbal. Semantics-first interpreters sometimes argue that they have a trump card in that they can best make sense of Hegel’s aim of engaging philosophically with Kant’s critique of metaphysics; but my alternative can match this card, once sufficient attention is paid to the “Transcendental Dialectic.” Once it is clear that there is a viable alternative, we can easily see some advantages of this metaphysics of reason approach.

To sneak up on the more complex issues concerning Kant, it is helpful to start instead with Hegel’s response to empiricist critiques of metaphysics. Empiricists will attack by arguing that metaphysicians illegitimately take for granted the possibility of knowledge in metaphysics. The attack is based on an epistemology-first metaphilosophy: it argues that other projects in philosophy are subject to doubts about the possibility of accounting for knowledge on such domains. But Hegel has a powerful response: empiricists are in no position to complain that others merely take for granted knowledge on a favored domain, because precisely in such attacks empiricists do the same: “the truth of the empirical, the truth of feeling and intuition is taken as basic” (§39An). The skeptical worry that we should doubt any kind of knowledge that is supposedly simply basic might be worrying for metaphysics, but only if it is equally worrying for empiricism as well. If we are to take such epistemological worries as fundamental, then the only principled or philosophical position, Hegel argues here, would be the more total skepticism of the ancient skeptics.

Further, Hegel argues that empiricists do not provide any alternative to metaphysics. Just as Kant takes “indifferentists” to “always unavoidably fall back into metaphysical assertions, which they yet professed so much to despise” (Ax), so too Hegel takes empiricists to covertly advance their own metaphysics (§38An). And we have seen one respect in which this is true: they hold a metaphysics of reason which takes disconnected events as the basic reason for everything else.

Note that Hegel’s rejoinder to empiricism anticipates Sellars’ “Empiricism and the Philosophy of Mind” attack on “the myth of the given”—as Sellars notes (1956, 253). But this is no reason to doubt that Hegel pursues metaphysics. Hegel does criticize empiricism for taking “the empirical … as basic” (§39An) or immediate or given. But is point in this section is specifically to defend a position in the metaphysics of reason against empiricist attack: to defend “universal notions, principles, and laws” (§38) which do not, contra empiricists, signify only “alterations that follow one after the other, and … objects that lie side by side” (§39).

Having looked at an epistemology-first criticism of metaphysics that is empiricist, we can now turn to a parallel sort of Kantian critique, considered by Hegel at the beginning of the Encyclopedia:

One of the main points of view in the Critical Philosophy is the following: before we embark upon the cognition of God, or of the essence of things, etc., we should first investigate our faculty of cognition itself, to see whether it is capable of achieving this. (§10A)

Hegel will dismiss this epistemology-first critique, with good reason: If our cognition is suspect when advancing a metaphysics, then it would be equally suspect when reflecting on itself. So reflection on our own cognition cannot be shown to be necessary or even superior to metaphysics on ground of such skeptical worries in epistemology. In Hegel’s terms, the epistemology-first Kantian objection to metaphysics turns on the desire “to have cognition before we have any,” and this is “as absurd as the wise resolve of Scholasticus to learn to swim before he ventured into the water” (§10An). The point is clear: philosophy takes an interest in metaphysical topics, such as “God … the essence of things, etc.”; the epistemology-first Kantian worry provides no reason to refrain from such metaphysics; the only thing to do is to jump in the water and try to learn to swim—which is to say, to start doing such metaphysics, or to start thinking about “God … the essence of things, etc.”, and to try to learn to think about such things well, rather thinking about them unknowingly or uncritically.

Hegel does also think that engagement with metaphysical questions about “God … the essence of things, etc.” prior to Kant tended to be objectionably naïve or uncritical. But clearly Hegel’s worry about being uncritical is not a worry about the lack of adequate prior epistemological reflection on the fitness of our cognition. Rather, Hegel worries that prior metaphysics had not sufficiently noted and responded to the dialectical contradictions which the “Transcendental Dialectic” reveals within metaphysics. Hegel dismisses the epistemology-first Kantian worry about metaphysics; but he seeks to build a non-naïve metaphysics by beginning with the Kantian consideration of metaphysics and reason in the “Dialectic.”[27]

Non-metaphysical or semantics-first interpreters of Hegel might seek to retreat to the view that there is a fault-line in Hegel’s work, dividing a good Hegel who commits to semantics-first philosophy from a bad Hegel who violates his commitment. But I see no evidence for such a fault-line. Imagine taking your pick of any contemporary view, or any historical view; call it view X. We could then proceed to distinguish those claims of Hegel’s which would fit well together with X in an overall philosophical project, and those that would not. That we can make such a distinction is not adequate evidence for Hegel’s philosophy being in tension with itself, or riven by an internal fault-line. For it is no evidence that Hegel is at all concerned with X. And, if it were evidence, then it would be evidence that there are such fault-lines nearly everywhere in every philosopher’s work. In such cases, without further evidence, one should rather conclude that the appearance of a fault-line is an artifact of one’s choosing to view Hegel through the lens of one’s own commitments, such as a commitment to the fundamentality epistemology, semantics, or issues concerning anti-realism and the like.

True, one might specifically seek to understand Hegel in a manner that will bring out connections with contemporary analytic philosophy. And perhaps at some point near the middle of the 20^th Century the dominant analytic orthodoxy would have taken problems about meaning or in semantics as fundamental, and rejected metaphysics by arguing that metaphysical claims are meaningless. But there was analytic philosophy before that point, and the trend for many decades since has been away from that semantics-first critique and back towards metaphysics. As Zimmerman puts it:

There was a period when many analytic philosophers—perhaps even the majority—believed that the problems of metaphysics were … demonstrably meaningless… But it was a relatively short phase… (2004, xiv)

And it is important to note something positive about this trend: it has reopened the possibility of analytic philosophy learning from connections with the history of philosophy. As Zimmerman says, the problems addressed by contemporary analytic metaphysicians

….are not significantly different from those that faced the philosophers of earlier eras; and they defend positions readily identifiable as variously Platonist, Aristotelian, Thomistic, rationalist, Humean, and so on. (xxi)

So a semantics-first reading of Hegel could bring his philosophy into contact with one strand of analytic philosophy. But the same reading threatens to obscure connections between Hegel and the more recent metaphysical trend, which specifically opens analytic philosophy to learning from positions in the history of philosophy—like Hegel’s. For example, just my short sketch here brings Hegel into connection with five lively debates in analytic philosophy, about: (i) the metaphysics of the laws of nature; (ii) the comprehensive program of the humean position in analytic metaphysics, in David Lewis and followers; (iii) the proposal that “ontological grounding” is the subject of metaphysics, a kind of revival of an Aristotelian approach (Schaffer 2009)—I argued above that Hegel’s approach starts with a comparable but broader or more general notion of one thing being the reason or why of another; (iv) the sort of monism which holds that the whole is the “ground” for the parts, or “priority monism” (Schaffer 2010, esp. the comparisons to Hegel at pp. 42 and 67); (v) the issues concerning dispositions, their grounds, and the pressure toward a metaphysical holism noted in Russell (1927) and Chalmers (1996). So we should not prefer semantics-first readings of Hegel on grounds of a desire to bring Hegel into contact with contemporary analytic philosophy; we can do at least as well by reading Hegel in terms of the metaphysics of reason.

I now turn to the philosophical issue of metaphilosophy, briefly setting aside the interpretive issues about Hegel. I myself am convinced by Hegel’s rejoinders that there is no force to epistemology- and semantics-first arguments that philosophy must begin by setting aside metaphysics and focusing on the possibility of knowledge and meaning. As Hegel argues, doubts about the possibility of meaning or knowledge in metaphysics would apply as well to alternative projects in epistemology or semantics. So I conclude that, if we are not to give up philosophy in favor of total skepticism, we should proceed as Hegel recommends: by jumping into the water and trying to learn to swim—or beginning with the metaphysics of reason, and trying to learn to do it well.

Further, it seems to me that our thinking in terms of the metaphysics of reason can help us to better find philosophical significance throughout the history of philosophy, and so to better learn from history. Hegel finds positive philosophical significance to build on even in Anaxagoras, instead of dismissing his ideas as simply meaningless. We too can in this way find philosophical significance not only in Hegel, but in Spinoza, Aristotle, etc. Semantics-first arguments in philosophy, by contrast, tend to seek to deny or close off philosophical significance. Some analytic philosophers once argued, for example, that all of metaphysics is simply meaningless. I think it best to steer as far away from that model of philosophy as I can, because I think it better to learn from the history of philosophy, and that this requires accepting its philosophical significance. So for these reasons I favor the metaphysics of reason metaphilosophy, according to which the basic questions of philosophy concern what is a reason for what.

Note that this metaphysics of reason metaphilosophy does not itself require denying the meaningfulness of questions in epistemology and semantics about how to account for the possibility of meaning and knowledge. I have not tried to show that there are no philosophically significant issues here. Perhaps there is a way of making the issues here precise as issues about reasons, so that the question is whether an objective world is itself the reason why we can know it, or rather whether something about our cognition is the reason for the distinction between subjective and objective. My commitment to the metaphysics of reasons makes me open to such concerns. I see no a priori reason to doubt that they are even meaningful or philosophically significant. I do however, for reasons noted above, find current debates between realist and anti-realist approaches to such issues to display a frustrating lack of agreement about what the real issues are. So I am pleased that my metaphilosophy gives me no reason to think that such debate is in any way fundamental or prior. I see no grounds why other debates about the metaphysics of reason need be held hostage by lack of progress in debates about anti-realism and the like.

In sum, coming back now to Hegel, recent debates about his philosophy tend to take place against the background of the view that problems concerning the possibility of meaning and knowledge are fundamental throughout philosophy. But I have argued that there is an important and promising alternative approach: we might instead take as fundamental problems concerning what is a reason for what, as for example in several familiar and recognizably metaphysical debates. This metaphysics of reason alternative can do justice both to Hegel’s interest in such metaphysical topics from history of philosophy, and yet also to the sense in which Hegel engages philosophically with and accords great importance to Kant’s critique of metaphysics. Once we see this alternative interpretive approach, it is easier to notice that there are some prominent places in which Hegel himself argues against the idea that there is any fundamentality or priority to problems concerning the fitness of our cognition for knowledge or meaning. Finally, consideration of Hegel can lead us in this way to the real and lasting philosophical advantages of a “metaphysics of reason” metaphilosophy. True, we can decisively resolve in this way neither the broadest questions about Hegel nor, certainly, questions about what all of philosophy is most fundamentally about. But thinking about Hegel can help us to come to see that there is a neglected and surprising contender when it comes to metaphilosophy. In this way, we can better appreciate the underlying terrain on which conflicts within philosophy are still fought today. And so I hope to have shown that we can learn something important from Hegel when considering the question of what philosophy is all about.[i]