1. McTaggart’s insight

In this section, I will show that Hegel rejects the Modification View by considering McTaggart’s version of it. McTaggart’s view is useful because he shows that the Modification View is inconsistent with Hegel’s own conception of the method. McTaggart thinks the right conclusion to draw is that Hegel was wrong to reject the Modification View. In later sections, I will try to show that Hegel was right to think about the method as he did.

On McTaggart’s view (and I agree with him), the subject matter of the Logic is the nature of thought. More specifically: it is an account of the various pure or a priori categories that constitute the nature of thought (cf. McTaggart Reference McTaggart and McTaggart1896, sec. 1). He thinks that the nature of thought involves what he calls a “super-organic unity,” such that one cannot correctly grasp any aspect of it without grasping all of it.Footnote ⁶ Each of our accounts prior to the last one only gets at some aspects of the nature of thought, and so they are incorrect. The dialectical method of the Logic consists (in part) in showing how each partial account is wrong because it excludes other aspects of thought. So, the method is that by means of which we progress from erroneous accounts to more accurate ones. This means, McTaggart argues, that the method does not characterize our account of the nature of thought itself, but only our coming to grasp it: as he puts it, “the dialectic is subjective” (sec. 122).

It might seem that McTaggart’s claim that the dialectic is subjective rests on his particular way of thinking of the subject matter of the Logic as having “super-organic unity.” Many interpreters have thought that McTaggart’s appeal to this unity entails a strong form of ontological monism which is not only precritical but also unHegelian (cf. Kreines Reference Kreines2015, 262ff.). I am sympathetic with this criticism. But I think this part of McTaggart’s view is inessential to his argument for the conclusion that the method is subjective, and I think focusing on his objectionable monism has led people to misunderstand what actually leads McTaggart to this conclusion.Footnote ⁷ I will show that the claim that the dialectic is subjective follows just from commitment to the Modification View, the view that the dialectical method essentially involves making and overcoming errors.

According to the Modification View, the results arrived at by the method do not always accurately characterize the subject matter of the Logic. Instead, they are our attempts to characterize that subject matter. As our attempts lead us to make errors, the method can be nothing more than a tool we use to arrive at the truth about the subject matter. One can think of this tool as similar to Wittgenstein’s ladder, such that once we have used it to arrive at an account of thought, we no longer need it and so throw it away. That is, once we have our account of thought, we no longer need to employ the method and can just advert to this account directly. Or one can think of this tool as always important, so that we can only ever grasp the nature of thought by using the method. In this latter case, the method would be like a telescope, essential any time we want to observe the heavens closely. Whichever metaphor is more appropriate, what matters is that the tool and the account arrived at by using it are different from one another: an account of the tool cannot be an account of what we know by using the tool, except by a mere accident (e.g., if we use a telescope to look at another telescope). This is what McTaggart means when he claims that the dialectical method is subjective, and this follows not from his commitment to a monistic interpretation of Hegel but just from his commitment to a version of the Modification View. So, the Modification View entails that the method of the Logic is subjective, or that it is merely an account of how we come to know the subject matter of the Logic.

Hegel denied the subjectivity of the dialectical method, as McTaggart recognized: “The conclusion at which we are thus compelled to arrive must be admitted, I think, to have no warrant in Hegel… . [T]he procession of the categories, with its advance through oppositions and reconciliations, he apparently regarded as presenting absolute truth—as fully expressing the deepest nature of pure thought” (1896, sec. 120). That is, McTaggart sees that Hegel denied that the method was subjective.

McTaggart does not mention particular passages in which Hegel expresses views inconsistent with the subjectivity of the dialectical method. Perhaps McTaggart has in mind passages like this one from the introduction to the Logic: the method “is nothing different [Untershiedenes] from its subject matter and content;—because it is the content in itself [in sich], the dialectic that the content has in itself [an ihm selbst], that moves it along” (WdL 5.50/21.38). The content or subject matter is the method, nothing different from it. That means that an account of the method is not an account of anything other than the subject matter. Hence, the method is the subject matter’s own nature. And this is not accidentally true, as when we use a tool to work on that very tool: Hegel’s point in this passage is that the method is identical with its subject matter as such, not that it is identical with its subject matter in this one instance. Moreover, Hegel claims that the course of the method “is the course of the matter [Sache] itself” (WdL 5.50/21.38). That is, the activity of “employing” the method is not our activity as against the activity of the subject matter, as it would have to be if the method were a tool; rather, it is the activity of the subject matter itself.

Hegel claims that the method is not our method, in the sense that we apply it to the subject matter. It is rather the method of the subject matter itself. Nothing that I have said so far sheds light on this difficult idea; illuminating it will be the task of section 4. My point at present is just to note, with McTaggart, that this claim is inconsistent with thinking of the method as subjective, and so inconsistent with any form (monistic or otherwise) of the Modification View.

This gives us an exegetical reason to pause before accepting the Modification View. But it would be helpful to consider a possible philosophical argument against the Modification View as well. And it would be even more helpful if the argument considered was given with an eye towards making sense of the exegetical evidence given in this section. I propose to consider just such an argument in the following three sections.

2. Kreines’s version of the Modification View

In this and the next section, I want to show that the Modification View presupposes that the dialectical method is a form of theoretical cognition. I mean “theoretical cognition” in the Kantian sense, as cognition which is not the cause of the existence of what it knows, such that what one cognizes must be able to be given to one. Hegel discusses this form of cognition in “The Idea of the True” in the Logic, noting that in this form of cognition “the content still has the determination of something given.”Footnote ⁸ Being able to be given something, in this context, just means that the existence of what is cognized theoretically is independent of the act of cognizing it such that one must have some way of receiving it if one is to cognize it. In section 4 I will offer a reason to think that no form of theoretical cognition in this sense can account for the subject matter of the Logic (the nature of thought).

In Kant, theoretical cognition is naturally contrasted with practical cognition; the former is not the cause of the existence of what it knows, the latter is (cf. Kant Bx). Hegel also has this contrast in view, as he distinguishes between what he calls the Idea of the True and the Idea of the Good. But my claim that no form of theoretical cognition can account for the subject matter of the Logic does not entail for Hegel (as it might for Kant) that the account of the subject matter of the Logic is a form of practical cognition. For Hegel recognizes, in addition to theoretical and practical cognition, another form of cognition which he tends to call “speculative.”Footnote ⁹ The nature of speculative cognition is quite contentious in Hegel scholarship; my claim here is that it is not theoretical cognition in the sense just specified.Footnote ¹⁰ But it is worth noting that this does not mean that it is practical cognition.

Further, it is also worth noting that there is in general no problem with a form of thought that involves givenness in the sense at issue in theoretical cognition. To be given something in the sense at issue in theoretical cognition is not to be Given something in some mythical, unacceptable sense. Being given something in the relevant sense is no impediment to being known. And theoretical cognition is an objectively valid form of cognition; it is that form that is employed in our knowledge of the world through perception, for instance.Footnote ¹¹ So, the problem is not with the form of theoretical cognition just by itself. It’s with this form being used to know this object, thought. The point I will try to make in this section and the next is that the Modification View involves treating thought as knowable in the same way, or through the same form, as what we know through perception. I will then show why Hegel thinks that cannot work in section 4.

The argument that the Modification View presupposes that the dialectical method is a form of theoretical cognition can be put abstractly in two steps. First, if the dialectical method can result in errors, then my (dialectical) account of the subject matter of the Logic cannot be grounded in itself, since the form of my cognition is compatible with my erring or getting my subject matter wrong. Second, since I must be able to determine whether my account of the subject matter is correct, and since (from the first point) my account cannot justify itself, I must have some access to the subject matter which is independent of my account of it. Hence, it must be existentially independent of my account, and so it must be able to be given to me. So, the Modification View presupposes that the method of the Logic is a form of theoretical cognition. In what follows in this and the next section, I will work out this argument in more detail against specific versions of the Modification View that seem like they can resist this conclusion.

To start with, consider James Kreines’s version of the Modification View: at each nonfinal step, we advance a proposal or hypothesis which turns out to be false. Here is how Kreines puts it:

So, on Kreines’s view, we put forward some proposal about the absolute (the subject matter of the Logic), or form a hypothesis about it, only to discover that this hypothesis leads to a contradiction that requires or at least suggests making a particular revision in our account of the absolute.

The method of hypothesis testing is familiar from both everyday life and natural science. A hypothesis is such that one advances it as possibly true, without yet having sufficient grounds to guarantee its truth. For hypothesis testing to work, we must have access to the subject matter which is independent of our hypothesis about it. Our access to the subject matter may well be highly mediated by various kinds of tools or by a background theory. Regardless, we can only disconfirm or confirm a hypothesis because our access to the subject matter is not exhausted by the hypothesis.

The point that we must have hypothesis-independent access to the subject matter must apply to forming hypotheses about the absolute too. Were our grasp of the absolute exhausted by our hypothesis about it, then we could have no access to anything about the absolute which could possibly disconfirm or confirm the hypothesis, and the very idea of advancing a hypothesis about it would lack sense.

So, we must have access to our subject matter which is independent of our hypothesis about it. That is just to say that we must be able to be given the subject matter in question: what we have access to is what it is regardless of our account of it, and our receptive access to it can then serve as a basis for checking our accounts.

To this, Kreines might object that his hypothesis testing does not involve needing to be able to be given anything. For we can simply say of some account: generalize this to see if it can hold of everything. And it might seem like this generalization involves no appeal to the objects as they can be given to us. That is, we only need to think through the hypothesis itself to realize that it is incoherent. We see some evidence that he might respond in this way in his account of the “pure mechanism hypothesis,” according to which I explain everything by appealing to its parts (Reference Kreines2015, 39). But if I explain everything by appealing to its parts, Kreines notes, then there is no explaining anything because, in the final analysis, there turns out to be no explanatorily significant parts to which I can appeal. So, the hypothesis that I can explain everything by appealing to its parts “undercuts” itself (40). This argument does not appear to rest on an appeal to anything that can be given.

But in fact the argument does rest on an appeal to something that can be given. To see whether mechanical explanation can explain everything, we must have some prior understanding of everything that is independent of our understanding of mechanical explanation. For otherwise our understanding of mechanical explanation would exhaust our conception of what it was to be an explicable thing, and in rejecting the hypothesis of “pure mechanism” we would be rejecting the very idea of explanation. So, in formulating the hypothesis, there must at least be some content to our understanding of everything which is independent of our idea of mechanical explanation.

We could imagine different conceptions of everything: it constitutes a delimited whole, it is consistent with a benevolent and omnipotent God, it is as simple as possible in its governing laws. Or, in Kreines’s proposal, among all things there must be something in the world that is completely explicable (Reference Kreines2015, 129, 233–34). Kreines is explicit that the need for something completely explicable is generated by a constraint on theoretical inquiry (by the fact that “we must take this not to be pointless” as he puts it in the block quote above). What matters for our purposes is not which conception of everything one operates with; nor is it a question of whether the conception of everything is adequately justified or rather dogmatically presupposed. Let’s stipulate that Kreines’s conception of everything is right, and that something must be explicable. Let’s even stipulate, for the sake of argument, that this conception has been adequately justified. Still, this view of how the method works turns it into a form of theoretical cognition.

On Kreines’s account, our hypotheses are thus of the form: can this form of explanation capture the fact that something is completely explicable? But this obviously entails that the intelligibility of our procedure rests on a feature of the subject matter that it has whatever our account of it. Even if we thought this feature was fully justified, we would still be treating it as a data point, something that our account must get right. And so, relative to our hypotheses, we would be taking this feature for granted, as something given. So, the hypothesis-testing conception of the Modification View presupposes that the method of the Logic is a form of theoretical cognition.

At this point, it would be helpful to contrast my approach towards Hegel’s method with that of Stephen Houlgate’s. As I did in section 1, Houlgate insists that the method is not ours as against the subject matter’s. He explains and defends this claim by pointing to Hegel’s demand that the Logic make no presuppositions about its subject matter (cf. Houglate Reference Houlgate2006, 48; and chap. 2, more generally). If, as Houlgate’s account suggests, Hegel’s sole or primary concern were with eschewing all presuppositions, it might seem like the Logic could proceed as follows: at the outset, we presuppose no argumentative standard to which we hold our account of thought but, making no such presuppositions, we are able to arrive at some such standard. We can then use that standard to determine the adequacy of our successive accounts. Arguing in this way, it seems like we could arrive at a Modification View.Footnote ¹³ Specifically, we would check whether the accounts of thought that emerge from successive applications of the standard met the standard and reject or modify those that did not.Footnote ¹⁴

A version of the Modification View which sought to justify some argumentative standard without starting from any presuppositions in the manner just outlined would still qualify as a form of theoretical cognition. For our assessment of our account of thought would consist in comparing our account with the standard that we have justified and determining whether our account met that standard would be our method for assessing whether the account suitably matched up with the world. Of course, it would have to be our account of thought which we tested, as the standard is precisely justified in virtue of us knowing it to reveal whether our account is adequate—it would make no sense to hold the subject matter itself to such a standard, deeming it to be modified or not according to whether it met the standard. The nature of the subject matter, on this view, would be existentially independent of our account of it; it must, then, somehow be available to us, and not simply through or in our account of it. Thus, the application of the standard would presuppose that we could be given the form of objectively valid thought.

My argument, unlike Houlgate’s, does not rest on any appeal to being without presuppositions. Similarly, the justifiability or validity of whatever standard we seek to hold our account to is also irrelevant to my argument. Theoretical cognition is, after all, a perfectly valid form of cognition; its standards are justified. Indeed, I believe Hegel thinks the Logic justifies the standards of theoretical cognition without starting from any presuppositions. But that does not mean that those standards can be used everywhere, for any subject matter. There might well be a subject matter to which those standards do not apply because that subject matter is one that we do not relate to as something that is what it is whether we know it to be so or not. In section 4, I will argue that the standards of theoretical cognition—however valid, however well justified—cannot apply to the subject matter of the Logic, and that that is what is at stake in Hegel’s claim that the method of the Logic is not ours as against the subject matter’s.

3. The Conceivability View as theoretical cognition

At the beginning of section 2, I argued that the Modification View entails that the dialectical method is a form of theoretical cognition: for, according to it, our accounts can be wrong, and so cannot be grounded in themselves, and so our accounts must relate to something that can be given to us. Couldn’t one object to the first step, however? That is, might one not adopt a different version of the Modification View according to which all opposing accounts are simply incoherent, without resting on an appeal to anything given? In that case, my account would be grounded in itself since its negation would be incoherent and we could know that without appealing to anything beyond the account itself.

We might call this view the Conceivability View, a specific variant of the Modification View. The idea is that I arrive at the conclusion that the subject matter of the Logic, the nature of thought, is P by showing that the claim that it is not-P is contradictory. This view has some right to be called the dominant conception of Hegel’s method: the idea that the dialectical method is supposed to show that “all roads lead to Hegel,” or that Hegel’s justification for his position consists in showing that every other possible position is contradictory. In this section, I want to show that, despite initial appearances, this view also implicitly presupposes that the method is a form of theoretical cognition.

For the Conceivability View to be a Modification View, I have to make errors in my account of the nature of thought. The method then consists in revising my account until I finally arrive at one that is not contradictory. We are, of course, quite familiar with one kind of contradiction from everyday life: it occurs when I affirm, by inattention or because the subject matter is complicated or for some other reason, both P and not-P without realizing it. A condition on the possibility of this kind of contradiction is that I have in view some subject matter about which I am making claims, such that my access to the subject matter is not exhausted by my claims about it. Of my claims P and not-P, one is true and one is false. Which one is true will not be determined just by my account of the subject matter; it must be determined by how things stand with the subject matter. If my access to it were exhausted by my claims about it, then of course I could not land in a contradiction (at least not of this garden variety sort), as the very thought P (or not-P) would suffice for me to know the truth of that thought.

On one version of the Conceivability View, Hegel makes the mistake of thinking we can do without an appeal to the subject matter by somehow thinking that every position but one is contradictory. But, by the nature of garden-variety contradiction, that makes no sense at all: for of course either P or not-P could be true. It is typical to avoid attributing this absurd error to Hegel by adding some starting account of the subject matter, such that either P or not-P contradicts it. We then regard that starting point as giving us access to the subject matter that is independent of our account of it, some feature we are seeking to capture in advancing the claim that P or the claim that not-P. And that is just to admit that the view is a form of theoretical cognition, albeit one in which the law of noncontradiction is being pressed into doing more work than is typical.

This, for instance, is Rolf-Peter Horstmann’s reasoning in his account of the method of the Phenomenology. On his view, the Phenomenology is meant to justify a monistic thesis in ontology. He considers the possibility that Hegel’s method in the Phenomenology consists in showing that all nonmonistic theories (“pluralist” theories) are “contradictory or inconsistent.”Footnote ¹⁵ But, Horstmann notes, this procedure will not work. So, instead, Hegel follows what Horstmann calls a “negative transcendentalistic method” that

That is, Hegel’s method consists in starting with an assumption that we can make sense of some subject knowing some object (as it really is). We then rule out different accounts of what a subject is and what an object is because they do not capture that relation between them. That starting assumption turns the subject matter into something given, in just the same way as Kreines’s claim that the world must be such that we can find explanatory satisfaction in our theoretical inquiries into it. Even if the characterization is right, and completely justified, by treating it as a standard that our accounts must meet, our accounts are theoretical. The subject matter, for Horstmann, is characterized as some subject knowing some object. We then treat the subject matter (so characterized) as what it is in a way that is existentially independent of our accounts of it; and so we treat the subject matter as given to us.

It might seem like Horstmann’s view, or one like it, might be saved if we attend to the fact that the contradiction emerges between the content of the subject’s claim about (or conception of) its object and the subject’s activity in claiming (or conceiving). The object itself need play no role in the contradiction, and so it might seem like we do not have to regard the object as given. We might think of this as a variant on the Conceivability View on which the contradiction in question is no longer a garden-variety contradiction but rather what is often called a “pragmatic contradiction.”Footnote ¹⁷

But this suggestion does not help. On this version of the Modification View, unlike the version we find in Kreines or that is more explicitly operative in Horstmann’s account, we do not generate an argumentative standard or characterization of the subject matter against which we then check our accounts. Rather, we check our accounts of thought by relating them to the acts of putting forward those very accounts. The pragmatic contradiction lies between what is explicitly claimed by our account and what is implied about thought by the fact that our act of putting forward that account is an act of thought. It follows that we treat our subject matter as given to us in regarding our act, which is an instance or manifestation of our subject matter, as given to us.

To spell this out a little bit more: I think something, and then I reflect on that. In reflecting, I regard my prior act of thinking (together, perhaps, with its content) as something given, a “data point” which my theory—a theory of acts and contents of that kind—needs to accommodate. In relation to the act of reflection, the initial act of thought is whatever it is whether I reflect on it or not; it is existentially independent of my act of reflection. The act is thus given to me, and I revise my account to capture its nature. So, this version of the Conceivability View rests on being given the subject matter.

The idea that on this version of the Conceivability View what we are given is itself an act of thought is strange. But the strangeness really has its source in the view itself, in the alienation implicit in it: in the act of reflection (as understood by this view), I do not relate to the act of thought as something which I affirm; rather, I relate to it as I would to something I am given in perception.

We could envision an alternative way of thinking of the act of reflection, on which the act reflected on was no different from the reflection on that act. This would be an act which is conscious of itself as what it is.

Acts of theoretical cognition, on at least some views, can be self-conscious in this way as well: in judging that some igneous rocks are volcanic, I am not given the judgment and in judging I am conscious of it as as my judgment. But I am, of course, given the rocks that I judge about. To arrive at a way of thinking of the unity of putting forward the account and its reflection on itself that is nontheoretical, we have to think that the self-consciousness of the act is sufficient, all by itself, to determine its own validity. That is, we must envision an act of thinking in which the validity of what is thought would be settled just by and in the act of thinking itself: no need to appeal to any perception of rocks or anything else to justify the thought. In such a case, the subject matter would not be given, and the form of cognition would not be theoretical. Precisely for that reason, any progress that might be made in this kind of account could not result from making and correcting errors: the thoughts would validate themselves.Footnote ¹⁸

In the introduction, I noted that the Modification View can seem compulsory because it can seem that to make progress from one account to another requires some kind of revision or modification of our account. With this idea of a self-validating thought, we can begin to see what an alternative to the Modification View might look like.

Consider the transition from mechanism to chemism, the stage that follows mechanism in The Science of Logic. Suppose that this transition occurs in the manner just suggested: that the validation of mechanism consists in nothing other than the thought of it. On my view, this validation just is the progression from mechanism to chemism. So, chemism must be contained in the very thought of mechanism. Or, as Hegel puts it, the method “finds the further determination of its initial universal entirely in it” and so “is in this respect analytic” (WdL 6.556–57/12.242).

That the method is analytic in this way raises a problem: how can it involve genuine progress? Through analysis we might clarify our concepts, but it is unclear how we might arrive at anything genuinely new or anything “richer and more concrete” (WdL 6.569/12.250).

The idea that, within the Logic, our accounts reflect on themselves provides some help. In our account of mechanism, we think it in such a way as to determine its validity solely in our thought of it. That is, our account of mechanism does not merely consist in listing the various aspects of mechanism. It also involves holding together or synthesizing those various aspects in such a manner as to determine the validity of mechanism. Or, as Hegel puts it, the method is not only analytic; it is “just as much synthetic, in that its subject matter, determined immediately as simple universal, shows itself as an other through the determinacy that it has in its immediacy and universality themselves” (WdL 6.557/12.242). Hegel thus thinks that through the synthesis of the various aspects of mechanism, one determines the validity of that notion. Moreover, he also thinks that this synthesis yields a new—richer and more concrete—notion (chemism).

At this point, one would quite reasonably be dubious about this whole procedure. Wouldn’t holding together the various aspects of mechanism just get you … mechanism? How could it yield an advance? And how could that advance determine the validity of mechanism? Indeed, how could thinking a concept ever determine the validity of that concept? In this essay, I will not be able to answer these questions. I will, in the next section, give a brief example of how this kind of progression goes. And I will also motivate the need to find some alternative to the Modification View, even though space will prevent me from fully clarifying the alternative that I think one finds in Hegel’s descriptions of the method.

4. Hegel’s argument against the Modification View

In sections 2–3, I argued that the Modification View is committed to an account of the method of the Logic on which it is theoretical, in that (according to it) the method treats the subject matter as something given in its accounts. This is not, by itself, an objection, but it is a consequence many proponents of the Modification View would contest since it suggests that their view saddles Hegel with a kind of precritical rationalism. But that suggestion is not by itself an argument against the view. In this section, I will give just such an argument.

The starting point for my argument is a description of the subject matter of the Logic. At the outset of both versions of his Logic, Hegel claims that its subject matter is, in some sense, the nature of thought. He further clarifies that it is, in particular, the nature of thought insofar as it comprehends what is the case: “thought or more determinately comprehending thought [begreifende Denken]” (WdL 5.35/21.27). That is, the Logic is about the nature of thought insofar as it is able to arrive at knowledge of the world. Hegel thinks that thought that can arrive at knowledge comes in various fundamental, necessary forms. The Logic is an account of those forms. Or, as I will put it, the Logic is an account of the forms of objectively valid thought.

Moreover, since the Logic is about the nature of thought insofar as it is able to arrive at knowledge, it is also true that it is an account of the world. Hegel puts this point in terms of the ancients’ idea that “nous, thought, is the principle of the world” (WdL 5.44/21.34; cf. EL sec. 24A). His point is that the account of the “necessary forms” of thought, insofar as thought can be knowledge, is an account of the necessary forms of the world (WdL 5.44/21.34).

With that initial characterization in mind, I turn to Hegel’s argument for the claim that the method of the Logic is nonaccidentally about itself. The basic idea of the argument, which I will work out in what follows, is that an account of the objective validity of a form of thought must employ that form of thought and so cannot be given it. This idea is expressed in the passage considered in section 1, in support of McTaggart’s contention that Hegel denies that the method is subjective. There we saw Hegel note that the method “is nothing different from its subject matter and content;—because it is the content in itself, the dialectic that the content has in itself, that moves it along. It is clear that no presentations can count as scientific which do not follow the course of this method and are not in accordance with its simple rhythm, for it is the course of the matter itself [der Sache selbst]” (WdL 5.50/21.38). In this passage, Hegel claims that the method by which we advance in the Logic is nothing other than the very topic of the Logic. The topic of the Logic, I have claimed, is the forms of objectively valid thought. So, the forms of objectively valid thought are their own method. That suggests that one accounts for a form of objectively valid thought by employing that very form of thought.

This claim—that one accounts for the forms of objectively valid thought precisely by employing them—can be usefully contrasted with both an empiricism and a rationalism about those forms. According to an empiricist account of the forms of objectively valid thought, we might claim that what I take to be a form of thought actually comes from my perception. But thought must have some form that it does not receive from what we perceive. For otherwise there would be no way in which what we perceive could be cognized at all. For instance, thought must bring with it whatever form is required for thinking multiplicity. Without that form, no diversity of perceptions would be able to make an impression on our thought, such that we could think that there are many things: whatever was given would not be differentiated by thought, unless thought already had the form required for drawing some distinctions.

According to a rationalist account of the forms of objectively valid thought, they are a priori and innate. By innate, I mean that they are in our mind waiting to be discovered, and their presence in our mind is not dependent on our cognition of them or of their objective validity. To stick with the example from the previous paragraph, on this line of thought we discover the possibility of multiplicity as something innately built into thought. We must, moreover, discover not only that this structure is innate to the mind, but also that it is objectively valid, or that the world contains the possibility of multiplicity. To know that the form is innate is one thing; to know that it is objectively valid is another. For the rationalist account to work, we need to know both. The problem with the rationalist view lies precisely in this difference between knowing that a form of thought is innate in the mind and knowing that it is objectively valid.

To determine whether the innate forms of thought are objectively valid, I must appeal to some standard that establishes their objective validity. Perhaps, for instance, I have as my standard for objective validity that the form of thought can generalize (can hold of everything) or that it is coherent, to echo two of the proposals considered in previous sections. Whatever the standard is (and however well justified it is), the crucial point is that the forms of thought satisfy (or fail to satisfy) that standard regardless of whether we know that they do.

The problem with this proposal lies in our inability to be given the form of thought such that we could then determine whether it satisfies the standard. The point Hegel makes when he insists that the method of the Logic is its subject matter is that the forms of thought we are investigating are inseparable from our investigation of them, in a way that objects of everyday consciousness are not.Footnote ¹⁹ I suggest that his point can be spelled out like this: To represent a form of thought as objectively valid is to represent not just it but also its relation to the world. But this relation cannot be thought from outside of the very form of thought in question. For I must employ the form of thought in question to determine whether the world is as that form of thought represents it to be. For instance, the only way in which I can determine whether the world has distinctions runs through my thought that the world has distinctions. To ask: “Does the world indeed have distinctions?” involves the thought that the world has distinctions, and so employs the very form of thought the objective validity of which is being asked about.

This inseparability of the forms of thought and our investigation of their validity can be contrasted with the separability of objects of everyday cognition and our investigation of them. I might, for instance, wonder what that object in front of me is. Suppose it is an apple, though I do not yet know it to be. I am able to investigate it, since it is given to me in perception. After my investigation, I conclude that it is an apple. My investigation need not start by deploying the concept apple; indeed, to perceive the apple, I need not possess the concept apple at all. I can think about the apple without going through my thought that the object is an apple. By contrast, there is no way of investigating whether some form of thought is objectively valid which does not go through the form of thought under investigation.

The key point is that we have cognitive access to the objects of everyday consciousness that does not run through our empirical concepts of them. We can perceive them.Footnote ²⁰ But we cannot perceive a form of thought. Indeed, this is part of what was shown in responding to the empiricist treatment of the forms of thought: for our perception to give us cognitive access to objects in the world, we must be able to take up what we perceive into thought. Thought must have certain forms for this to occur, forms which inform or structure our perception (at least insofar as we take up what we perceive into thought) and so cannot be derived from it.

So far, I have shown that the forms of thought must be employed in our investigation of their objective validity, in a way that differentiates them from the objects of everyday consciousness. It is a short step to show that the forms of thought cannot be given. To be given something, I must have access to it in a way which is independent of it and so can inform my theory of it: I must have some kind of perception of it. I cannot have a perception of the forms of thought, for I can have no cognitive access to them that is independent of my “theory” of them, or my understanding of what they are. So, I cannot be given them. The method for cognizing them in their objective validity cannot be theoretical.Footnote ²¹ To reemphasize: it does not matter whether the standard that I would hold the forms of thought to is justified; it does not matter if that standard is itself valid; it does not even matter whether we have arrived at that standard without making any presuppositions. What matters is that we cannot sufficiently separate our accounts from the forms of thought we are investigating and so cannot create the logical space in which we could apply those standards to our account.

I argued in sections 2–3 that the Modification View presupposes that our subject matter can be given to us. It thus involves the idea that our representations of that subject matter may or may not agree with that subject matter, such that we can check those representations against the subject matter. This is to ignore the way in which the forms of thought are, always already, present in the activity of our thinking anything. It is thus incompatible with giving an account of the forms of thought that reveals them to be objectively valid.

As previously noted, space does not permit me to try to work out in any detail what a more satisfying explanation of the method of the Logic looks like. I have noted, in the prior section, that it must involve nothing more than an analysis of whatever form of thought is being put forward and that synthesizing the aspects of that form of thought (as arrived at in the analysis) both determines the validity of it and yields a new form of thought. Rather than try to say more about the method in general, I will instead provide a brief example of how the progression might go.

The example I will consider is the one I have already mentioned a few times: the transition from mechanism to chemism.Footnote ²² Hegel explains mechanism very generally in terms of a whole composed of objects that “exist [bestehen] in self-standing indifference as objects outside of one another” (WdL 6.408/12.132). That is, the whole is a “composite” or “aggregate” of objects (EL sec. 195; cf. also WdL 6.410/12.133). The nature of the mechanical whole is such that the objects that make it up interact with one another, for instance by transferring motion or heat (two of his examples) from one to another. The resulting whole is defined in terms of the initial make-up of the objects (e.g., their position, direction, and velocity), and then a set of rules or laws that govern their interaction.Footnote ²³ For chemism, the objects are essentially related to one another such that the resulting whole cannot be decomposed into the initial make-up of the objects plus laws that govern their interaction: the tendency of one chemical object to bond with (or resist bonding with) another defines it, and so there is no way of specifying the initial make-up of the chemical object without already referring to the laws that govern its interaction with other objects. Rather, their initial nature is already defined in terms of their interaction, or the law that determines their interaction is (as Hegel puts it) an “immanent law of the objects themselves” (WdL 6.409/12.132).

There are two aspects of mechanism to highlight: first, the initial make-up of the objects that compose the resulting whole and, second, the rules or laws that determine, relative to the initial make-up of the objects, what the whole looks like. The progress to chemism consists in synthesizing these two aspects.

To synthesize these two aspects is to think the connection between the objects and the laws that govern their interaction. First, the laws that govern the interaction must be present in the objects in their initial make-up; otherwise, the resulting behavior of the objects would not be in accordance with those laws. So, we have to think of the objects in such a fashion that the laws of interaction characterize them. However, second, this relation between the objects and the laws that characterize them is not intelligible through mechanism. For mechanism is defined by the claim that the objects as they are in their initial state are intelligible independently of their interaction. Mechanism does nothing to make intelligible how a law describing the interaction between objects can characterize an object in its initial state.Footnote ²⁴ But we can only hold together the two aspects of mechanism if we can grasp a law of interaction that characterizes an object in its initial state.Footnote ²⁵

To grasp an object that is initially characterized by laws that govern its interaction with other objects is to go beyond mechanism. So, to grasp mechanism, and in particular to grasp a mechanical whole, requires advancing beyond mechanism, to a different kind of whole: a whole that is not intelligible in terms of the initial make-up of its components plus the laws that govern their interaction, because there is no way to specify the initial make-up without already referring to the connection between the objects as expressed by the laws that govern their interaction. It is in virtue of the intelligibility of that kind of whole that we can think of a mechanical whole as a way in which things can be, or as an objectively valid form of thought. Arriving at this conclusion involves (indeed, Hegel thinks, consists in nothing but) thinking through the form of thought present in the first stage, mechanism. And it legitimates that form of thought, as mechanism is intelligible in light of chemism (which, it will turn out, is intelligible only in light of other forms of thought).

The most pressing objection to this kind of account is how one can arrive at something genuinely new from an analysis of a concept. In this case, it might seem like Hegel must have smuggled in a demand for a kind of unity of object and law that is neither present in nor required by mechanism. Once that’s accomplished, of course it is no surprise that we need something richer and more concrete than mechanism—but that is simply because Hegel has presupposed what he set out to show. One might raise this objection somewhat more sympathetically as follows: What power must Hegel be attributing to thought, such that the thought of one concept can not only validate it but also yield a new concept? And with what right does he attribute that power to thought?

The only way to answer these questions, and so defend the kind of account exemplified above, is to work out in some detail how Hegel arrives at and justifies his account of the power of thought, and also to work through more examples in much greater detail to illustrate this power of thought. I cannot hope to do that here; I only hope to have indicated the direction one must go to avoid the Modification View.

The contrast between this account and the Modification View can be made more obvious by once more considering Kreines’s explanation of the transition from mechanism to chemism. Kreines argues that we can limit the objective validity of mechanism, and make room for a nonmechanical and yet objectively valid form of thought, by considering the hypothesis that mechanism characterizes everything. This hypothesis is incoherent or self-undermining, because there is nothing that we can appeal to that would actually compose any whole: it would be aggregates all the way down, and that makes no sense.Footnote ²⁶ This hypothesis-testing might initially seem very similar to my proposal: that is, it might seem like it consists in nothing more than thinking through what is involved in mechanical explanation. But, in fact, it does not consist in that. For, on this proposal, if all I did was employ the form of thought or analyze it, then I would encounter nothing but mechanically explicable objects. Nothing would force me to realize that I need to move beyond mechanism. I am only brought to that realization by stepping back from the mechanical form of thought and considering the hypothesis that it characterizes everything. That is, I must check and see whether it is fit for the job of explaining everything. That is not a job I have in view when I employ the mechanical form of thought; nor is it a job I have in view when I analyze that form. It is a job I only have in view when I consider whether the mechanical form of thought can be enlisted in a very particular metaphysical program. And that shows that Kreines’s proposal does not consist in simply thinking through the mechanical form of thought.

Indeed, Kreines’s proposal is rather like the rationalist proposal I considered above: I possess a form of thought (either innately or in some other way) and I must determine its validity. I cannot do so simply by thinking through the form of thought; rather, I must consider whether it matches up with the world. But this procedure assumes that we have a more basic grasp of the world in light of which we can check the validity of the form of thought in question. And Hegel’s idea is that what differentiates the forms of thought is precisely that we do not and cannot determine their validity (or relate them to one another) through checking to see how they fit together in our more basic grasp of the world. They are our most basic grasp of the world.

The proposal I have suggested (of simply analyzing a form of thought and then holding together the aspects that result from it) may or may not work out in the end.Footnote ²⁷ But I hope working through the example has made the philosophical point clear: if we are considering those forms of thought which articulate the most basic grasp we have of the world, then we have to find a way of thinking their objective validity that does not consist in checking how they match up with the world.

5. What about the Phenomenology of Spirit?

As I noted in the introduction, Hegel claims that the method of the Logic is the same method employed throughout his works (cf. WdL 5.17/21.8, 5.35/21.27, 5.48–49/21.37–38). It follows that the method of his other works cannot be theoretical either. Though I do not have the space to consider in detail why that should be, the basic reason will be the same as for the Logic: in all of his works, Hegel is articulating the a priori form of whatever his topic is. The validity of that form cannot be explained in an empiricist or a rationalist way. So, it cannot be known theoretically.

Generalizing the results of the last section to Hegel’s other works will immediately raise an objection: perhaps my argument holds good for the Logic, but there is no way it can hold good for the Phenomenology of Spirit. For (so the objection) it is obvious that the Modification View is right of the Phenomenology. Footnote ²⁸ After all, Hegel describes the Phenomenology as “the path of despair,” specifying that natural consciousness “loses its truth on this path” and that it “is the conscious insight into the untruth of knowledge as it appears” (PhG ¶ 78). Such a loss appears to entail, indeed obviously to entail, the Modification View.

Any adequate response to the Modification View must offer an account of the Phenomenology which is both consistent with its character as the path of despair and with rejecting the claim that its method consists in modifying the accounts put forth in the various stages. I suggest that one can do this by focusing on Hegel’s distinction between two ways in which the “path of despair” is understood: there is how the consciousness that undergoes it understands it, and there is how the philosopher who “looks on” understands it (cf. PhG ¶ 86-7). That distinction admits of many different interpretations, but at least prima facie it seems like one feature of the distinction is that consciousness does not understand the method by which the path is constructed. As Hegel puts it, “the emergence of the new subject matter,” the next shape of consciousness, “shows itself to consciousness, without consciousness knowing how it occurs; it proceeds for us, as it were, behind its back” (PhG ¶ 87). He also notes that, to consciousness, the next shape of consciousness is found as though “by accident” (PhG ¶ 87). And he claims that consciousness relates to the next shape as though it perceived it or happened to encounter it receptively: it is a “pure apprehension [Auffassen],” and is not active in determining what the next shape will be (PhG ¶ 87). These passages suggest that it is only the philosopher who understands the method by which the path is constructed, and in particular only the philosopher who can recognize the necessity (nonaccidentality) of the emergence of one shape of consciousness from another; it also suggests that consciousness thinks about the path of despair theoretically. Reading these passages in this way is not uncontroversial; Michael Forster, for instance, denies that this is the import of these passages.Footnote ²⁹ But if we accept this interpretation of the distinction between what consciousness understands and what the philosopher understands, then it is at least open to us to think that the path only seems to involve error to consciousness, while to the philosopher the despair and its removal do not involve any errors.

On the view that I have just sketched, the difference between consciousness’s understanding of the path of despair and the philosopher’s grasp of that path is quite significant: it amounts to a difference in the logical character or form of the transitions, as it were. On many other views, the difference is not nearly so significant. For instance, as Ludwig Siep (Reference Siep and Smyth2014, 67) explains it, the philosopher merely contributes an appreciation of the fact that a subsequent shape of consciousness “contains moments of” the prior conception.Footnote ³⁰ As an example, he mentions that “a dynamic conception of nature or a holistic concept of truth might arise through the unification of features belonging to the old conception (mind-independence, permanence, intersubjective authority, verifiability, etc.) with moments of knowledge (laws of motion, verification conditions)” (67). These moments are moments that consciousness itself grasps, as they are just aspects of its conceptions of knowledge and reality. But if consciousness grasps that which the philosopher relates, then what the philosopher contributes will only be posterior to the shapes themselves—an organizational principle that does not make the shapes intelligible in the first place.

On my view, by contrast, the point of Hegel’s claim that the philosopher is able to explain the emergence of one shape of consciousness from another is precisely that the relation that holds between them—such that the one emerges from the other—is that without which we cannot grasp the aspects of the prior stage. It is not an accident, relative to what the first shape is, that the second shape follows it; that the second shape follows it is precisely internal to the first shape. In the terms I have used, the philosopher effects the synthesis of the aspects that consciousness is unable to hold together.Footnote ³¹

I have not here tried to explain the actual course of the method of the Phenomenology, or the necessity of the despair as it is experienced by consciousness, or the relation between what is available to consciousness and what is not available to consciousness, or what justifies the appeal to the method in the Phenomenology. My point with respect to the Phenomenology has only been negative: the Modification View, though it is internal to the method in that it articulates how consciousness understands the path of despair, is not an adequate account of the method. Indeed, if the interpretation I have suggested is on the right track, Hegel thinks it has no comprehension of the method at all.