Kant and Hegel on Teleology and Life: a New Pespective from Debates about Free Will

Forthcoming 2012 in The Freedom of Life, edited by Thomas Khurana, a third volume in the Freiheit und Gesetz series. Please cite the final version from there; this is a late draft:

Kant’s treatment of teleology and life in the Critique of the Power of Judgment is complicated and difficult to interpret; Hegel’s response adds considerable complexity. One of Kant’s central claims is that we can never know whether anything natural is truly explicable in teleological terms. Hegel responds by arguing that Kant’s own analysis should have led him to deny that skepticism, and to affirm teleological explanation of natural living beings. But many of the views attributed to Kant by interpreters fail to cohere with or support his skepticism. And without an understanding of Kant’s argument for the skeptical part of his view, there is little hope of understanding Hegel’s response.

My aim here is to propose a new way of understanding the underlying philosophical issues in this debate, allowing a new appreciation of the basic structure of the arguments on both sides. My new way is unusual: I use for an interpretive lens some formal features of familiar debates about freedom of the will. These debates, I argue, allow a better understanding of the underlying structure of a great many philosophical issues. But I do not aim to interpret what Kant or Hegel has to say about freedom of the will. The idea is rather to use the interpretive lens of free will debates to better understand the philosophical issues at stake in their disagreement concerning teleology and life. This will clarify the precise philosophical burden that must be met by Kant’s argument in defense of his skepticism, and why his case has considerable philosophical force. But it will also explain why Kant’s argument itself inevitably provides the opening for Hegel’s reply, and sets a standard that Hegel will meet in a surprising way. Finally, this approach will show that the arguments in both Kant and Hegel retain significant philosophical force and importance, regardless of their ignorance of the intervening years of such great progress in the biological sciences. So by looking to Kant and Hegel we can better understand the structure of underlying philosophical terrain of the issues concerning teleology and life—terrain we are still fighting over today.

1. The Usefulness of Debates about Freedom of the Will

In debates about freedom of the will, the libertarian position is comprised of two claims. The first is incompatibilism: freedom of the will is not compatible with determinism. The second is that we have free will. Duns Scotus is often cited as an historical example of a libertarian. But what is the alternative to libertarianism? It can seem that compatibilism is the alternative, as for example in Hume.

But care is needed to guard against a mistake here. Imagine a compatibilist were to decisively prove that determinism is true. This would, to be sure, refute the libertarian view. And so we might be tempted to think that it would support compatibilism. But it would neither directly tell against incompatibilism, nor directly support compatibilism. For it would in itself leave entirely open the possibility of retaining incompatibilism and drawing the conclusion that we are not free—or the “hard determinist” conclusion. If we neglect that alternative, then we can easily misunderstand what sorts of arguments really support or undermine which positions.

To guard against the mistake, we need only recognize that this debate engages at once two orthogonal issues, or issues along two dimensions. The issue at stake along the first dimension is broader than just the question of whether free will is compatible with determinism. The fundamental issue is: what is free will? Or: what would have to be the case for there to be free will? The broadest issue fundamentally at stake along the second dimension of debate is this: is there any such thing as free will? Or, in terms of our more narrow concern: do we have free will? With respect to the question of whether we have free will, the most obvious possible answers are “yes” and “no.” I will call these “optimism” and “pessimism” (although the latter need not be pessimistic in any other sense than denying free will—some may argue that this is a happy and welcome conclusion). With respect to the first question—what is free will?—there are more and less inflationary or deflationary views. Compare incompatibilism and compatibilism, for example. Incompatibilists hold that, for my will to be free I must be able to will without being caused or determined by anything prior. Compatibilists hold that free will includes no such demanding requirement. So compatibilism is a deflationary view in the sense of holding that less is required for free will. Incompatibilism is inflationary by comparison:

This basic scheme makes room for additional complexity that might be needed. It may become important to distinguish the yet more inflationary agent-causal incompatibilism from the less inflationary event-causal incompatibilism. Perhaps semi-compatibilism is akin to a position between optimism and pessimism. And there is logical space for pessimistic deflationism, although in practice I know of no such views.

This bi-dimensional structure underlies many other debates as well. But is less easily recognized in other cases. So in other analogous debates, alternatives are easier to neglect, and the corresponding mistakes are easier to make. Consider for example debates about causality. A deflationist will hold that causality is just constant conjunction. An inflationist will hold that causality is a form of necessary connection not reducible to constant conjunction. It can seem that there is an easy way to argue for the constant conjunction view, on grounds of an argument that we cannot have knowledge of necessary connections. But this is, at best, too hasty. For those epistemological worries, even if entirely justified, also leave open the possibility of combining a necessary connection view of what causality is with a denial of the possibility of our knowing causes.

What is going on here, again, is that the debate addresses two orthogonal issues. The first issue again concerns the what it is question: what is causality? And here again there are inflationary and deflationary answers. But the other issue most prominent in this debate does not concern existence, as in the question of whether there is any such thing as free will; it concerns rather the epistemological question: can we have knowledge of causality? Here optimists say “yes,” and pessimists “no.”

Some may wish to give a skeptical argument that philosophy should never take what is X? issues seriously; if there were space, I would argue that philosophy cannot avoid doing so, and that such skeptical arguments themselves presuppose answers to other what is X? issues, such as what is meaning? But here it is more important to consider what general lessons we can draw about such debates, if we are to engage them. First of all, the availability of both existential and epistemological questions about optimism and pessimism suggests the need to proceed carefully. Where it is more prominent to debate about the existence question—as in debates about free will—we should keep in mind not only the possibility of arguing for an inflationary pessimism that would deny existence, but also distinguish the possibility of an inflationary pessimism that would deny knowledge. For example, we should not neglect the possibility of arguing for incompatibilism about free will coupled with a denial of the possibility of our having knowledge of whether or not we have free will—a position I will return to below.

An simpler general lesson is that potential errors stem from neglecting alternatives made possible by the bi-dimensional structure—often the alternative of pessimistic forms of inflationism. For example, some might wish to argue that inflationism has an epistemological disadvantage. But that is not exactly correct. An inflationary account of X can always be combined with a pessimistic view, as epistemologically modest as you like, that we cannot have knowledge of X’s. Similarly, some might wish to argue that deflationism has the advantage of ontological simplicity, or to object that inflationism is metaphysically extravagant. Again, that is not really right. For an inflationism about X can always be combined with an ontology that is as simple as you like by pessimistically denying that there are any X’s.

When assessing the impact of scientific progress, it can be tempting to instead mistakenly neglect deflationary possibilities. For example, contemporary psychologists sometimes claim to have shown that we do not have free will. But they tend to merely assume inflationary views about what free will would have to be—for example, that it would have to involve a cause that is uncaused, immediately conscious, immaterial, etc. The more inflationary the demands, the easier to prove that we do not meet them. But regardless of the extent of the evidence that we do not meet those demands, the conclusion is unsupported insofar as there is no philosophical argument for inflationism.[1]

So it is no surprise that, in debates with this structure, the arguments of historical figures can remain of importance. Compared to us, for example, Hume is incredibly ignorant about the brain. And yet his deflationary account of liberty and moral responsibility remains of philosophical importance. Indeed, we must remember the possibility of this kind of deflationary position if we are to avoid the last mistake, concerning the relevance of all our newer scientific knowledge. Similarly, Hume is comparatively ignorant about physics. And yet his arguments about causality remain of importance. In general, given any inflationary account of X in terms of some requirements A, B, and C, scientific results might well show that there is nothing that meets those requirements. But this result in itself always leaves open the philosophical question of whether the right conclusion is that there is no X—or rather whether ABC is an overly inflationary and so mistaken way of understanding what X would be. And there is no obvious or widely agreed way in which scientific results might in themselves have entirely resolved or rendered obsolete the question of what would have to be the case for there to be free will. Nor for there to be causality. Nor, as we shall see, for there to be natural teleology.

Finally, it is important to note what burdens an inflationist must and must not generally carry. We have seen that inflationism in itself is not subject to worries about epistemological immodesty, metaphysical extravagance, or conflict with science. Still, an inflationist cannot just assert inflationism; she must argue. For example, an incompatibilist cannot just say that free will is incompatible with determinism. Perhaps she will argue that (a) any account of free will must support an account of moral responsibility, and then that (b) moral responsibility requires incompatibilism. And, in general, the inflationist’s burden is to show that (a) there is some widely agreed feature of X, which (b) requires an inflationary account.

2. Teleology’s Explanatory Implications as the Basis of Kant’s Inflationism

Kant’s discussion of teleology and life in the Critique of the Power of Judgment focuses on the concept of a Naturzweck. I have previously approached this material in terms more internal to the KU, aiming to consider all of the evidence there.[2] My focus here is on using the interpretive lens above to better understand the underlying philosophical issues and the structure of Kant’s argument.

Interpreters often look to what Kant says about the concept of a Naturzweck for his views about living beings. And they often seek to understand the concept of a Naturzweck in terms of the features of living beings discussed by Kant: mutually compensating parts, taking nutrition, reproduction (KU 5:371f.). The underlying idea of such approaches is that “Naturzweck” is Kant’s “expression for biological organisms.”[3] And this approach can seem to charitably attribute to Kant a kind of deflationism: a Naturzweck/organism is not something metaphysically extravagant; to be a Naturzweck just is to have features that we can clearly observe, like reproduction and nutrition.

But any first step or assumption along these lines will in fact doom us to confusion. For Kant denies that we can know whether there are any Naturzwecke. The concept of a Naturzweck is “problematic”—when employing it “one does not know whether one is judging about something or nothing” (KU 5:397). And: “the concept of things as natural ends [Naturzwecke], places reason as a cause into a relation with such things, as the ground of their possibility, in a way which we cannot know through any experience” (EE 20:234). Or we can put the point in terms of explanation: a Naturzweck would be something natural explicable in teleological terms, and Kant denies that we can ever have the knowledge required to explain anything natural in teleological terms. But this pessimism or skepticism is obviously confused if Naturzweck is just Kant’s expression for biological organisms, or for beings that take nutrition and reproduce, or something along these deflationary lines. For we obviously can know that there are biological organisms. And we can know that organisms reproduce and draw nutrition from the environment. And we know this, plainly, from experience.

To correct the original misstep, we need only recall the structure of free will debates. It should be obvious that we cannot understand the issue at stake between compatibilists and incompatibilists if we mistake it for an issue concerning our actual capacities. The issue is rather orthogonal to the question of whether we have free will at all. And Kant’s discussion of teleology and life similarly engages two orthogonal issues. The first is: what is a Naturzweck (natural end or purpose)? Or, roughly, what would it take for there to be natural teleology, or something natural and genuinely explicable in teleological terms? The second, orthogonal issue is epistemic: Can we know that there is any natural teleology? Or, more narrowly, can we know whether living beings are Naturzwecke?

As we have just seen, Kant’s answer to the question of knowledge of Naturzwecke is “no.” But his overall view is subtle: We cannot know whether there are any Naturzwecke; but actual living beings do have certain features—nutrition, reproduction, etc.—which inevitably lead us to think of them as Naturzwecke—so teleological judgment of them is necessary for us; and such teleological judgment plays an irreplaceable role in guiding our scientific inquiry into the non-teleological explanations of things, even if we must always avoid claiming theoretical knowledge that they are Naturzweck and so avoid the claim to explain anything natural in teleological terms.

Kant, then, is not a deflationist but an inflationary pessimist about the teleological concept of a Naturzweck. He is arguing that the concept of a natural teleological system includes requirements demanding enough that we can never know whether anything meets them. Again, there is no threat of metaphysical extravagance here. Kant’s inflationism is epistemologically modest insofar as he denies the possibility of theoretical knowledge of Naturzwecke. His view is that we must not assert or deny in any theoretical context, whether scientific or philosophical, including metaphysics, that there are any Naturzwecke.

But while Kant’s inflationism is not vulnerable to those worries, he still must argue for it. The inflationist’s burden, we have seen, is to argue (a) that there is some feature of X to which all must agree, and (b) that this feature requires inflationism. Kant’s strategy is to argue (a) that all must agree that the notion of natural teleology carries explanatory implications, and (b) that those explanatory implications require an inflationary account precluding our having knowledge here.

Kant’s initial analysis consists in two requirements on the relation of part to whole within a complex system. The first is a requirement for anything to be a teleological system or Zweck, which would include artifacts produced by external design. The second narrows the analysis to Naturzwecke, or systems that are teleological by nature rather than external design.

What is crucial about the first requirement is the explanatory demand it imposes. The idea here is that, for a teleological system, it is not enough that the parts might benefit other parts or the whole. Kant considers an example: there are features of the arctic which benefit human survival there, including sea creatures that provide nourishment. The benefit is certainly real, and we can know that it exists. But it does not justify the conclusion that such fish, say, have the teleological purpose of nourishing humans. We can look at this point in terms of the connection between teleology and normativity: “[a] teleological judgment compares the concept of a product of nature as it is with one of what it ought to be” (EE 20:240). So to judge the arctic ecosystem in teleological terms would be to judge that, should the fish flourish but become impossible for humans to catch, or less nutritious for humans, then the fish would be failing to fulfill their purpose, and the arctic ecosystem would be malfunctioning. But clearly the fact that humans benefit does not alone justify such teleological conclusions. For, as Kant says, “one does not see why human beings have to live there at all” (KU 5:369). So any teleological concept of a complex system must include the demand that such benefit is not merely incidental: parts of this beneficial form must exist here because of the role they play within the whole, or for that reason, or on that account, so that they are explained thereby. Kant puts this in terms of dependence: “for a thing as a natural end [Naturzweck] it is requisite, first, that its parts (as far as their existence and their form are concerned) are possible only through their relation to the whole” (KU 5:373).

(A1) If you analyze some concept of a kind of complex system, Y, without including the requirement that the existence and form of the parts depend on a role played within the whole, then there can be cases of evidence that a system is Y without evidence that it is a teleological system.

(A2) Thus any analysis of a genuinely teleological concept of a system must include the explanatory requirement that the existence and form of the parts depends on the role they play within the whole.

To complete the analysis of a Naturzweck, Kant seeks a second requirement to rule out cases of teleological systems that are created by external designers, leaving only teleological systems by nature—a requirement that will distinguish such “inner purposiveness” from the “external purposiveness” of artifacts. The idea here is that the parts of artifacts are means to an end only insofar as an external designer imposed some overall structure or organization; a Naturzweck, by contrast, would have to “self-organizing” (KU 5:374). Stated in terms of part-whole relations, the second requirement must demand that the structure or organization of the whole is determined not by something external to the system but rather internally—and so by the parts of the system itself. But for a part to contribute to the determination of the structure would be to contribute toward determining what other kinds of parts are present and in what arrangement. So each part must form the others. Or, for a Naturzweck, it is required “second, that its parts be combined into a whole by being reciprocally the cause and effect of their form” (KU 5:373).

3. Completing the Argument for Inflationary Pessimism

How does this analysis support a yet more inflationary account? The key argument connects the concept of any Zweck (whether a Naturzweck or not) with a prior determining representation of a concept of the whole system. But this argument is complicated by Kant’s epistemic modesty, holding that our cognition is restricted or limited to knowledge of a spatio-temporal empirical world. So insofar as this particular argument turns on appeal to temporal order, its results are limited to things we can know about. That said, I would understand the key argument in this way:

(B1) A part of a certain form can play a role within the whole only once the part exists in that form within that whole.

(B2) The role played by parts in the whole can precede the existence of parts with those forms only insofar as there is a preceding representation of the whole and the roles to be played by the parts.

(B3) The role of parts with certain forms can be responsible for the existence and form of those parts only insofar as the whole is determined by a representation of the whole.

In Kant’s terms, in the order of “real causes,” an end or purpose (Zweck) cannot precede and thereby influence its own causes, so it can do so only as “ideal,” or as first represented (KU 5:372).[4] And so Kant infers directly from the explanatory requirement to the need for determination by a representation of a concept or idea of the whole:

[F]or a thing as a natural end it is requisite, first, that its parts (as far as their existence and their form are concerned) are possible only through their relation to the whole. For the thing itself is an end, and is thus comprehended under a concept or an idea that must determine a priori everything that is to be contained in it. (KU 5:373)

Kant then proceeds immediately to note how this requirement is satisfied in the case of artifacts, and to consider whether it might also be satisfied in a case of natural teleology.

Again, we must keep Kant’s epistemic modesty in mind. Strictly speaking, the conclusion is this: in the spatio-temporal empirical world, the first requirement can be met, and so there can be a teleological system, only where the system is the product of a prior representation of the whole. Or: anything knowable by us can be a teleological system only if it is the product of a prior representation.

It is a mistake to worry, with MacFarland, that Kant’s point here is scientifically outdated.[5] Kant is not claiming that actual organisms originate in representations. And he is certainly not trying to explain in terms of design something that we have since learned to explain in terms of natural selection. On the contrary, he is arguing that we can never have the knowledge necessary explain nature in any teleological terms at all.

More specifically, Kant argues as follows: If the purposiveness of a system is to be inner, or if it is to meet the second requirement, the parts themselves would have to determine the structure of the whole, and – in a spatio-temporal, empirical world – they would have to do so in a manner guided by a representation of the whole. Now, the parts of the real complex systems of which we have empirical knowledge, such as living beings, are matter. And matter cannot represent concepts or intend to act in accordance: “no intention in the strict sense of the term can be attributed to any lifeless matter” (KU 5:383). The point is not that the first requirement cannot be met by a material system: it can be met by material artifacts, which are the products of external design. The point is that Kant’s two requirements, applied to an exhaustively material system, would be incompatible. For any reason in favor of thinking that the structure of an exclusively material whole is due to its own parts would also be (because the parts are matter) reason to deny that the whole is determined by a representation, and so to deny that it is a teleological system at all. So:

[I]f we consider a material whole, as far as its form is concerned, as a product of the parts and of their forces and their capacity to combine by themselves […] we represent a mechanical kind of generation. But from this there arises no concept of a whole as an end [Zweck]. (KU 5:408)

We are, then, not just dealing with a theory that is comparable to incompatibilism about free will in that both are forms of inflationism; Kant’s account of natural teleology is a form of incompatibilism, applied to this topic—at least when it comes to anything of which we can have knowledge.

And yet it is crucial that Kant is arguing that there is at least a bare or minimal logical possibility of a Naturzweck. He is, in a sense, holding out the merely logical possibility of a kind of compatibilism that must remain unknowable and incomprehensible for us. For there is an open possibility that we cannot rule out on merely logical grounds—that is to say, an open merely logical possibility. There could be something that is a Naturzweck in virtue of a non-spatio-temporal “supersensible real ground of nature” or a “thing in itself (which is not an appearance) as substratum” (KU 5:409). We can have neither knowledge of any of such substratum of nature, nor knowledge of whether there are any Naturzwecke. We cannot even have any positive comprehension of how there could be such a Naturzweck. We can only conceive, by contrast with our own intellect, of an intellect that would be superior in kind—an “intuitive understanding” (KU §77)—which might have knowledge of beings that are Naturzwecke in virtue of a supersensible real ground of nature.

I can see two ways to try to think our way towards this logical possibility, despite its remaining incomprehensible for us. One way would be to note that supersensible stuff underlying matter might be (unlike matter itself) capable of representing the concept of a whole and organizing itself in accordance. The resulting organized being could satisfy Kant’s formulations of the notion of a purpose or Zweck in general, even where he formulates the notion in terms of the need for a determining or originating concept (e.g. KU 5:220, 5:373). But such a being would not be the product of merely external design—it would be the product of the self-organization of its own underlying parts or stuff (which would be capable of representing concepts and organizing themselves in accordance). If a “mechanistic” whole is just one in which the structure of the whole is due to “the parts and of their forces and their capacity to combine by themselves,” then this sort of being would also manifest a (for us unknowable) compatibility of teleology and mechanism.

The second way is to note that, once we are conceiving of a non-spatio-temporal, supersensible substrate, we have already stepped beyond the reach of the argument that a Naturzweck requires an originating representation of the concept of the whole. For the argument turns on a representation being the only way that the whole can precede itself in time. So the need for the determining representation follows only for anything that could be the object of our non-divine understanding, or anything that could be known or comprehended by us. A fundamentally non-spatio-temporal being knowable only by an intellect superior in kind might—for all we can know or comprehend—organize itself from within without need of representations of concepts at all. So even if we take “mechanism” to require not only something explicable in terms of its parts themselves, but also to rule out teleological explanations in terms of representations of ends—still we cannot rule out on logical grounds a Naturzweck that would be also “mechanistic” in virtue of its grounds in an underlying non-spatio-temporal substrate of nature. As Kant later says:

[A]nother (higher) understanding than the human one might be able to find the ground of the possibility of such products of nature even in the mechanism of nature, i.e., in a causal connection for which an understanding does not have to be exclusively assumed as a cause. (KU 5:406)[6]

Fortunately, we need not wade any further into Kant’s complicated discussions of the intuitive understanding. What is crucial for our purposes is that the merely logical but incomprehensible possibility cannot in any case apply to any exhaustively material system, nor to any system of which we can in principle have exhaustive knowledge. Similarly, if a “mechanistic” system is one whose structure is explained by material “parts and of their forces and their capacity to combine by themselves,” then there is no possibility whatsoever of a compatibility of natural teleology and mechanism.

While there is a slight sense in which the open logical possibility mitigates Kant’s incompatibilism, it in no way mitigates Kant’s inflationism. The concept of a Naturzweck remains extremely demanding: it requires either an origin in a concept of the whole represented by the parts, or an origin in an unknowable and incomprehensible (for us) supersensible non-spatio-temporal ground of nature. This is a demanding requirement—demanding enough that nothing we could ever know could ever meet it.

Of course, yet again, the point is not to argue that living beings originate in either representations or anything supersensible. On the contrary, the point is that we cannot know any of this to be the case. Precisely something unknowable yet logically possible is what is needed for Kant to argue that the concept of a Naturzweck is useful for reasons unrelated to knowability.

To begin with, preserving the logical possibility of a Naturzweck leaves room for Kant to claim that living beings are such that we will think of them or judge them as self-organizing Naturzwecke, even though we cannot have knowledge of this. We will think of them as Zwecke insofar as they are so organized that we cannot see how the benefits of parts to whole could be merely incidental, for “nature, considered as mere mechanism, could have structured itself differently in a thousand ways without hitting on precisely the unity in terms of a principle of purposes” (KU 5:360). Further, we will think of them as self-organizing, or in terms of the inner purposiveness of a Naturzweck, insofar as they have mutually compensating parts, incorporate matter in order to grow, and they generate new living beings by reproduction (KU 5:371f.). In Kant’s terms, our experience “exhibits” but nonetheless cannot “prove” the existence of real Naturzwecke (EE 20:234).

Second, Kant can then argue that thinking of living beings in such teleological terms provides us with an indispensable heuristic aid in scientific inquiry seeking non-teleological explanations—so Kant will argue that we cannot make scientific progress toward understanding them without teleology as a guide.[7] Teleological judgment serves as “a heuristic principle for researching the particular laws of nature, even granted that we would want to make no use of it for explaining nature itself” (KU 5:410). In fact Kant will argue that we require guidance by thinking of nature as a whole in these terms, if we are to seek knowledge of the laws of nature. But it is crucial that the point is not to establish the existence, likely existence, or even merely possible existence (in any but the merely logical sense) of Naturzwecke. The point is to establish that the concept of natural teleological system is so demanding that there is some hope of our using it for guidance without confusing this with the assertion of knowledge.

This Kantian position is not rendered obsolete by progress in biology. It is rather a kind of neglected alternative in today’s debates about teleology and life. Some today argue that the explanatory implications of teleology require external design, and that this justifies the deflationary conclusion that the notion of function of use in biology is weaker, and not teleological—it is just the notion of a part of a system susceptible to a kind of functional analysis.[8] The resulting notion can apply pretty much anywhere. The arctic ecosystem can be functionally analyzed in such a way that fish will be correctly said to have the “function” (in this non-teleological sense) of feeding humans. The other side in the contemporary debate is occupied by those who argue that the explanatory implications of teleology can be satisfied by natural selection, without need of design, concluding that the parts of living beings do have functions in a teleological sense.[9] These new teleologists can argue that biology requires use of a more restrictive notion of function than the idea of functional analysis applicable anywhere. The anti-teleologists can reply that natural selection, properly understood, cannot really satisfy the explanatory implications of a teleological notion of function.[10] Both lines of argument are worth taking seriously, but once we understand Kant we can see that neither is sufficient to support either contemporary approach. For there is a neglected alternative. What is neglected is the possibility that biology needs a more restrictive and so teleological notion of function, but only as a guide for inquiry and never as the object of knowledge or scientific conclusion. A contemporary Kantian, then, can accept the just-noted arguments made by both sides against the other, and can claim this as a unique strength of a Kantian view. So Kant’s view is not obsolete. On the contrary, understanding Kant allows us to better understand that the underlying philosophical terrain is more complex than is apparent if we judge only by current debates.

Finally, we can compare a kind of debate about free will. Imagine on one side there are agent-causal libertarians arguing that compatibilism unacceptably deflates the notion of free will. And imagine on the other side there are compatibilists arguing that everything we know about the world makes plain that there is not and cannot be any evidence in favor of our being agent causes. Both sides would be neglecting an alternative that could accept both of their arguments—namely, this view: To have free will would be to be an uncaused agent-cause. But we cannot know whether we really are agent-causes. So we cannot know whether we have free will. Still, we cannot avoid taking up a practical standpoint, or the standpoint from which we deliberate about what we have reason to do. And taking up that standpoint involves thinking of ourselves as free, or as uncaused agent-causes. So we must think of ourselves as free. Further, such thinking plays an irreplaceable and beneficial role in moral life. But we still must avoid claiming theoretical knowledge of whether or not we have free will, or asserting this in the context of any theoretical philosophy or science.

Note that this is not a deflationist account on which having free will is just taking up a practical standpoint, or deliberating. A deflationist reading makes the view confused, since we obviously can know that we do deliberate. Further, note that the inescapability and benefit of thinking of ourselves as free, on this view, has nothing to do with any assertion of theoretical knowledge that we are free, likely free, or even possibly free in anything but a merely logical sense.

I think that this last view about free will—this form of inflationary pessimism—not only illuminates by comparison Kant’s position on teleology and life, but would also be a promising approach to Kant’s position on free will itself in much of the critical period. Although I cannot develop or defend that approach further here, I will just note this striking parallel: Kant’s aim is to argue that Naturzwecke “can be conceived without contradiction but cannot be comprehended” (KU 5:371). The KU Introduction, in explaining the importance of the topic of the power of judgment and its connection to Kant’s previous work, claims that “freedom” involves a special sort of “cause,” “the possibility of which cannot of course be understood, although the objection that there is an alleged contradiction in it can be adequately refuted” (KU 5:195).

4. Hegel’s Strategy

Hegel’s Logic response to Kant concerning life and teleology is by no means entirely critical. Hegel finds Kant’s analysis and the whole idea of “internal purposiveness” to be of great importance—for philosophy generally and not just for philosophical issues concerning life. In Hegel’s terms, “[w]ith this concept of internal purposiveness, Kant has resuscitated the Idea in general and especially the Idea of life.” (EL §204A)[11] But Hegel aims to draw on Kant’s analysis to argue against Kant’s own skepticism: Hegel argues that living beings do manifest true “internal purposiveness,” or are Naturzwecke, that their structure and development is explicable in teleological terms, and that we can have objective knowledge of this natural teleology. So Kant should not, Hegel says, have been satisfied with natural teleology as “mere maxims of a subjective cognition”; “the end relation” is on the contrary “the absolute truth that judges objectively and determines external objectivity absolutely” (WL 6:444/739).

I have elsewhere approached the “Life” section in Hegel’s Logic from a perspective more internal to Hegel’s project.[12] My aim here is to focus on how the interpretive lens of free will debates clarifies the standard to be met by the argument, and how Hegel meets the standard. We can in particular understand the challenge facing Hegel by comparing this sort of argument for an inflationary pessimism about free will:

(C1) Free will must be something such that our having it would account for our being morally responsible for the things we do.

The argument itself will structure the terrain on which optimists might counter-attack. A first possibility is an entirely inflationary optimism: One could accept (C1-3) and then try to give an argument in metaphysics that we are agent-causes. It will be challenging, of course, to square uncaused agent causes with what we know about the laws of nature (regardless of whether determinism is true).[13]

A second possibility is an entirely deflationary response, but it is hard to see the appeal. One could admit that there is for the given reason no moral responsibility, but then argue for a purely deflationary account on which free will faces no such problem because it need not have anything to do with whether anyone is really responsible for what they do. That is, one could admit that (C1) would force us to accept everything else all the way to (C4), and then one could reject (C1).

A third possibility is the most obvious opening: accept the inflationist’s own standard for what counts as free will, in (C1), but try to block the move to any further inflationism (C2-3) by giving a compatibilist account of moral responsibility.

I think the point about an obvious opening for rejoinder will generalize. Compare Kant’s argument about Naturzwecke. He argues that teleology carries an explanatory requirement that must be recognized by all sides, and then argues that this demands an inflationary account requiring (in anything knowable for us) originating representations. Interpretations of Hegel’s response tend to gravitate toward the most extreme possibilities.

One extreme possibility would be to defend an entirely inflationary form of optimism. One could grant all of Kant’s argument concerning what would be a required for a Naturzweck, but then simply argue (contra Kant) that we can know that living beings satisfy all of his inflationary demands. The resulting view would be this: There is indeed a dualism separating ordinary intellect from a higher, divine form of understanding; but (contra Kant) we can know that there is such higher understanding, because we can break through the barrier and ascend to this higher perspective. And, once ascended, we gain a access to a supersensible ground of matter itself, and can see that it is self-organizing in a way that is either animated by something like a soul representing concepts, or else in some non-spatio-temporal manner utterly incomprehensible for anything short of this superior-in-kind higher intellect. So we can know in this way that there are Naturzwecke. Perhaps by ascending to a divine immediate grasp of all reality at once, we can even know reality itself to be a Naturzweck. I will argue that this is not Hegel’s approach.

A second possibility, on the other extreme, would be an entirely deflationary response to Kant. Sometimes this takes the form of a complaint that Kant confuses teleology with efficient causality.[14] The idea would be this: first, grant that Kant’s explanatory requirement, that parts of a certain form must be present because of their roles in the whole, would force us to the rest of Kant’s inflationism and pessimism; second, reject Kant’s explanatory requirement and so escape his conclusions. The result would be an analysis of teleology on which it carries no implications about why parts of a certain form are present or exist—no implication about what is really responsible for the origin and shaping of a complex system. And so we would have two obviously compatible ways of talking about one and the same bit of nature: we could talk about whatever really brings about, determines, produces, or is responsible for parts of this form in this arrangement; and we could also talk in teleological terms carrying no implications about any of that. This would be as unproblematic as saying that one and the same thing can be both square and heavy. To understand teleology (on this view of it) would be to see that there can in no case be any difficulty concerning incompatibility.

But such entirely deflationary proposals are no more compelling than the idea of ceding incompatibilism about moral responsibility and then trying to argue for the strangely deflationary claim that free will is safe because it doesn’t have anything to do with whether one is responsible for what one does. And such proposals are no match for Kant’s argument for the explanatory requirement. Consider again the system in which arctic fish benefit humans. The deflationary proposal holds that it does not matter, for teleology, whether fish of this form came to be present there on because of this role, relative to humans, in the whole system. But then the deflationary proposal will allow that, from the fact of benefit, it would follow that this is the teleological purpose of the fish, so that fish becoming smarter and impossible for us to catch would be a malfunction in the fish. But that teleological conclusion is clearly not justified by the fact of benefit. And any way I see of tightening the deflationary proposal would allow altering the example to generate the same false implication. So the purely deflationary proposal does not capture genuine teleology. A teleological system requires at least some kind of real dependence (which in no way need be narrowly a causal relation between events) of the existence and form of the parts on the whole.

Fortunately, it is clear that Hegel’s approach is not purely deflationary in this sense. Hegel says that teleology and mechanism cannot be shown to be mutually “indifferent” simply by noting that they differ:

[I]f mechanism and purposiveness stand opposed to one another, they cannot for that very reason be taken as indifferent concepts, each of which is correct on its own account, possessing as much validity as the other. (WL 6:437/734)

Nor does an “equally validity” of both follow simply because “we have them both” (WL 6:437/734). Hegel, then, accepts that the idea of natural teleology presents a challenge or a problem not so easily defused. And he accepts the burden of providing an argument that the challenge can be met. And he will argue that life is a kind of special case in which the challenge can be met.

But once we look through the lens of free will debates, we easily see that talk of pure inflationism or pure deflationism is distracting us from a neglected alternative that is obvious and inviting: a kind of mixed response. One could (a) accept a bit of inflationism in the form of Kant’s explanatory requirement for any teleological system and then (b) add a bit of deflationism in the form of trying to argue against the inference from here to any further inflationary requirement for a determining representation on the part of the underlying matter or anything else.

And we can easily see that Hegel must be pursuing such a mixed strategy. For we have just seen that Hegel’s response is not pure deflationism; Hegel will clearly (a) accept a bit of inflationism in the form of Kant’s explanatory requirement. But Hegel will clearly add (b) a bit of deflationism, in that he rejects any further inflationary claim that a natural teleological system requires an originating or determining representation explaining how the parts of a system originally came to be present. This is clear because Hegel so emphatically denies that the possibility of natural teleology depends on any prior determining representation, or on anything like an intelligence capable of representation. For example, when Hegel begins the “Teleology” section in the EL he says: “In dealing with the purpose [Zweck], we must not think at once (or merely) of the form in which it occurs in consciousness as a determination that is present in representation” (EL §204A). Hegel proceeds immediately to the importance of Kant’s analysis of the Naturzweck; his aim is to take this and then use it in some way that will block the connection to determining representations present in some prior consciousness.

5. Hegel’s Argument

Hegel argues by constructing a concept of life out of three requirements. The point is not to attempt an a priori argument about what life must be like. Rather, the point is to argue that anything meeting these three requirements—which are clearly met by actual living beings—can be known for that reason to also satisfy Kant’s analysis of a Naturzweck, without need of any originating representation of the whole on the part of the underlying matter or anything else. The three part analysis of life, structures the “Life” section in the self-standing and the Encyclopedia Logic, and also discussions of biology in the Philosophy of Nature. The “Life” section in the Logic consists of: “A. The Living Individual”, “B. The Life Process”, and “C. Kind (Gattung).”

The first requirement mirrors Kant’s analysis in terms of reciprocal relations between part and whole: “all the members are reciprocally momentary means as well as momentary ends” (EL §216, transl. modified). Or: “the organs are the means of life, and these very means, the organs themselves, are also the element in which life realizes and maintains itself […] [T]his is self-preservation” (VPR 17:508/336). Second, the concept of life also demands that a complex system itself requires some kind of assimilation from the outside environment by which “it maintains itself, develops itself and objectifies itself” (EL §219). Third, Hegel’s concept of life also demands that individuals must reproduce within a species. So anything satisfying Hegel’s concept must also pursue self-preservation in an additional sense: it must aim to reproduce itself—it “produces itself as another individual of the same species” (PP 4:32/142). It produces another with the same structure: in reproduction, “there emerges a determination of the entire structure” (EN §355Z). In Hegel’s terms, the third requirement demands the “process of the Gattung” (kind or species) or the Gattungsprozess.[15]

In combination the three requirements demand something that is (a) organized to preserve itself through the activities of (b) assimilation and (c) reproduction. Hegel seeks to prove that this yields teleology—a “system of activities which is actualized into a system of organs through which those activities proceed” so that “in this way the living thing is articulated purposefully; all its members serve only as means to the one end of self-preservation” (VPA 13:193/1:145).

One way to bring out the philosophical point of all this is to attend to the influence of Aristotle. I set aside the question of the accuracy of Hegel’s reading of Aristotle. But Hegel sees in Aristotle the claim that all life aims at self-preservation. For example, a seed “directed solely to self-preservation” is Hegel’s example of Aristotle’s “concept of the end as immanent” (EN §245Z). And it is easy to see what Hegel is thinking of in Aristotle: the claim that all life, including plants, shares the “nutritive soul” in the sense of having the end of self-preservation, not only in the sense of the nutrition by which the individual survives but also reproduction by which the species survives.[16] But Hegel sees in Aristotle also the denial that this has anything to do with a claim that plants or their parts or a seed must represent that end and organize in accordance with its representation. It is again easy to see what Hegel is thinking of in Aristotle: for example, “[i]t is absurd to suppose that purpose is not present” simply because there is no “agent deliberating.”[17] (Perhaps Aristotle and/or Hegel refer “souls” in order to argue further that being directed toward self-preservation also must involve some kind of capacity to detect ways to reach it, and perhaps that would mean something like a proto-representational capacity; but this would not cede Kant’s claim, which concerns the need in a knowable teleological system that it originally came about as a result of a prior representation of that system.[18]) Finally, Hegel sees in Aristotle the claim that in reproduction an organism “produces itself as another individual of the same species” (PP 4:32/142).[19] Hegel writes his terms for this point into his construal of a passage (Physics 198a) from Aristotle:

That which is produced is as such in the ground, that is, it is an end [Zweck], kind [Gattung] in itself, it is by the same token prior, before it becomes actual, as potentiality. Man generates men; what the product is, is also the producer. (VGP 19:176)

To follow the implications of these points for Kant’s argument, we need to think of the connection between species (Gattung) and individual, type and token. A token elm tree, for example, “produces itself” in that what it produces is the same in type. From the perspective of the later elm, it has been produced by itself “prior, before it becomes actual.” A part of the elm, a leaf (token), has been produced by prior generations which share that very same part (type) “prior, before it becomes actual.” Now our new leaf (token) certainly benefits the whole elm (token)—it assimilates from the environment. The question raised by Kant’s analysis is whether we can know that this benefit is why a leaf of this form is present at all. And we can. The new leaf (token) is only possible insofar as the leaf (type) benefits the whole elm (type). For if this (type of) leaf did not, then a system (of this type) could not survive and reproduce, and the new leaf (token) would never have been produced at all. With life generally, a part (token) is possible only insofar as that part (type) plays its beneficial role in relation to the whole (type). And each organism meets Kant’s demand that, in a teleological system the “parts (as far as their existence and their form are concerned) are possible only through their relation to the whole” (KU 5:373). It is not just that the parts benefit the whole; the parts are present specifically on account of the way in which they are a benefit to the whole. So we can say that the leaf is a means to the end, purpose or Zweck of assimilating from the environment, and in turn the end of self-preservation—it is there for that reason.

Or we can put the point in terms of Kant’s own connection between teleology and normativity. Failure cannot just mean that something no longer benefits another—for this would suggest that fish suddenly eluding capture by humans would be normative failures. So Kant’s view is that this possibility of normative malfunction in anything we could know about could only be relative to a determining representation of a concept. Hegel’s response makes the possibility of normative failure in life relative to the species or kind. Thus Hegel sometimes equates “Gattung” with an “ought” or Sollen (e.g. WL 6:306/662). In biological cases specifically, the possibility of “defect” is relative to “the rule, the characteristic of the species or class (Gattungs- oder Klassenbestimmtheit)” (EN §368Z in the German, §370Z in the English edition).

To follow the argument in Hegel, we need to follow his use of the term “immediate” in the Logic. The first step of Kant’s analysis governs only part-whole relations or “the process of the living being inside itself” (EL §218). But in such analysis there is no mediation through which we could comprehend the possibility of the origin or genesis of a teleological system. At this point, teleology can only be an immediate presupposition: “the living individual, at first disengaged from the universal concept of life, is a presupposition that is not as yet authenticated by the living individual itself.” But now, given Hegel’s focus on self-preservation including reproduction within the species, “its genesis, which was an act of presupposing, now becomes its production” (WL 6:484/772-3). The significance of the third requirement and the completed analysis is that “the living individual, which was at first presupposed as immediate, is now seen to be mediated and generated” (EL §221). So Hegel cedes that the genesis of a Naturzweck is problematic; what he argues is that reproduction within the species solves that problem, explaining the genesis of a Naturzweck in a way that we can understand and know about.

Strictly speaking, it remains for Hegel to argue in the discussions of life in the Philosophy of Nature that our knowledge of plant and animal biology fits the analysis. But what is needed here will be uncontroversial—after all, there are living beings, and they do assimilate and reproduce. The argument in the Logic shows why, in knowing this, we know such living beings as natural teleological systems, or Naturzwecke.

In the present case, Kantians will certainly be tempted to respond that Hegel’s natural teleology is not genuine teleology at all, because it lacks an origin in any power of matter to represent an organized whole and organize itself in accordance, thus fixing genuinely teleological purposes. But Kantians are in no position to so respond. For they cannot just stipulate that teleology requires something like a prior representation. If they did, then Hegel would be free to respond that he is defending teleology in a sense that accounts for every feature of it aside from the merely stipulated addition about a determining representation. So Kant must instead argue. And Kant’s argument sets the standard for what will count as a genuinely teleological system: the existence and form of the parts must depend on the whole. Kant chooses this standard in order to defend the conclusion that a knowable or comprehensible Naturzweck would require a determining representation of the whole. So if Hegel can meet that same standard without need of a determining representation of the whole, then Kantians are in no position to protest that this is not genuine teleology.

Further, recall the specifics of Kant’s first requirement, defining the notion of a purpose or Zweck in general: the existence and form of the parts must be possible only through their relation to the whole. But Kant cannot mean to require the parts be possible only through their relation to this very token whole. For it is a strength of Kant’s analysis that it explains why artifacts are purposes, with parts that can malfunction (even if they are not natural purposes, or Naturzwecke). The first requirement is met by artifacts in this way: the existence and form of the token parts are possible only through the representation of a concept of the whole—only through the relation between parts of this type to wholes of this type. So Kant must allow identification of token and type in defending his first requirement, or his analysis of purposes or Zwecke. Thus he has no principled grounds to resist Hegel’s meeting both requirements by appeal to a connection between tokens and a different sort of type—the type fixed by reproduction, rather than representation.

Similarly, Kantians cannot protest that Hegel’s rejoinder fails to explain how matter can organize itself. On the face of it, we can at least conceive of a logically possible non-material universe, and inner purposiveness or self-organization there would have nothing to do with matter. So the analysis should, and does, concern only rather relations between parts and whole. If Hegel’s argument succeeds, then it demonstrates how we can know that the analysis is satisfied without needing any knowledge of the underlying matter. If so, then the underlying matter is not relevant to the question of the possibility and knowability of Naturzwecke.

6. Metaphysically Robust Compatibilism about Natural Teleology

Another way in which comparison with free will debates can help us is by highlighting the fact that Hegel’s view in “Life” is not just analogous to compatibilism—it is a form of compatibilism. I noted above Kant’s view is that teleology and mechanism, applied to a material system, would be incompatible. Hegel’s argument is also a case for the compatibility of determination of underlying parts by necessary law and teleological determination of the whole. Thus Hegel praises Aristotle’s philosophy of nature for defending “two determinations: the conception of end and the conception of necessity” (VGP 19:173/2:156).

To be sure, Hegel does not hold that living beings can also be explained in non-teleological terms. The basic reason is that a living being has an intrinsic end or purpose. And it has parts or “members” which are themselves means to the intrinsic end. Neither matter nor chemical substances fit the analysis of life, and neither have intrinsic ends in this sense. So the nature of living beings and their “members” is neither mechanical nor chemical. In Hegel’s terms, “the organism is a manifold, not of parts but of members.” (WL 6:476/766)[20] And “members” in this sense “are what they are only by and in relation to their unity” (EL §216Z)—only insofar as they are means to the end of the whole.[21]

But this is not to deny the applicability of lower-level forms of mechanical and chemical explanation within the spatio-temporal bounds of a living being. The point is rather that, so long as we have no teleological ends or purposes in view, what we explain by this means would not itself be living being as such nor its “members.” So Hegel says of the living being that “the mechanical or chemical relationship does not attach to it.” He adds, however, that “as externality it is indeed capable of such relationships, but to that extent it is not a living being.” Hegel then concludes that, “[w]hen the living thing is taken as a whole consisting of parts, or as anything operated on by mechanical or chemical causes […] it is taken as a dead thing.” But we can also “grasp” it as “living being” in terms of a “purposiveness” that is genuinely “inner” (WL 2:419/766).

The transition from assimilated external elements into blood, for example, cannot be understood in terms of necessitating causes (WL 6:228/562). But Hegel does not deny the possibility of analyzing what is going on within the blood stream in terms of underlying chemical elements; this simply does not explain blood as such: “blood which has been analyzed into these constituents is no longer living blood” (EN §365Z; see also EL §219Z). To be blood is not to have a certain chemical constitution, but to serve in distinctive ways as a means to the end of self-preservation in particular kinds of organism. More broadly, we can explain the behavior of the chemical substances and reactions found along the way of the broader process of assimilation in “inorganic” terms, in which case their interconnection or organization will be “superfluous.” But all of these elements are present, in this particular arrangement, for the sake of an end, so that “still the course of organic being in itself occurs for its own sake” (EN §365Z).

Note that Hegel’s compatibilism here is inconsistent with other interpretive approaches noted above. On the one hand, a purely deflationary reading would portray Hegel as arguing that teleological concepts carry no implication about what is responsible for the existence and form of the parts depending on the whole. This would yield a kind of compatibilism, to be sure. But this purely deflationary compatibilism would render superfluous all of the complex work Hegel does in “Life,” all of the points he draws from Aristotle, and in particular the emphasis on reproduction—all of it would be neither needed for nor even relevant to a defense of compatibilism about natural teleology. But Hegel’s view is that it is specifically reproduction which makes possible compatibilism here. What makes this possible is that an individual “produces itself as another individual of the same species” (PP 4:32/142), or (where Hegel is reading Aristotle) “what the product is, is also the producer” (VGP 19:176), so that we can understand how “its genesis, which was an act of presupposing, now becomes its production” (WL 6:484/772-3). Clearly Hegel’s view is that the broader question of what produces a system is—contra pure deflationism—relevant to teleology and key to his compatibilism.

On the other hand, imagine reading Hegel as arguing that matter has a soul that represents concepts and organizes itself in accordance. This would be to argue for a kind of resistance to or interruption of the necessity of lower-level laws, which would contradict Hegel’s compatibilism.

Note that Hegel gives no defense of teleological explanation of the historical development of a species. Hegel does not require that there must have been a time when elms lacked leaves of a certain form, and then acquired them on account of some purpose or goal. So he cannot be defending the claim that elms came over time to have leaves for the sake of any end. Hegel’s argument, if successful, demonstrates that questions about how and whether a species changes, or how life originated, are irrelevant to the resolution of the philosophical problem, left by Kant, concerning how teleology might explain the structure and behavior of a complex natural system, such as an individual organism. So while Hegel might well have any number of false beliefs about historical development of biological species, this is irrelevant to his philosophical argument against Kant. Note in particular that Kantians cannot require that Hegel provide an account of how, from a state of lifelessness, life emerged. Kant himself thinks that he has no explanation of how life came about, and yet he does not consider this reason to doubt whether we can know whether there are living beings—we can see them with our own eyes. If Hegel provides reasons for also thinking that living beings also satisfy Kant’s analysis of the concept of a Naturzweck without appeal to any required ancient history, then he shows that ancient history is beside the point. Note that there is another parallel here with compatibilism about free will, for that view too must advocate the irrelevance of sufficiently distant history: a compatibilist account of free will must leave open the possibility that what I do is determined by the state of the universe prior to human beings, holding that this possibility is irrelevant to the question of my free will.

As with Kant’s, Hegel’s view is not rendered obsolete but rather neglected in contemporary debate. There are those today who defend teleology by appealing to natural selection. This approach will require many of the elements of Hegel’s. This obviously includes the requirement of reproduction within a species. But it also includes the robust explanatory role for general types of features of a species. Millikan’s definition of normative or proper function, for example, refers to “traits having been causally efficacious”.[22] But if Hegel’s argument works, then natural teleology can be defended by means of these elements without need of any view about the historical development of a species, and thus without need of any appeal to natural selection. Insofar as the addition means that the new teleologists must defend themselves against the worry that natural selection cannot do what they require it to do,[23] a contemporary Hegelian would likely see here a strength of Hegel’s view by comparison.

Further, the argument in “Life” is evidence against the idea that Hegel’s metaphysics is an organic monism, holding that everything real is part of an all-encompassing organism, which is a Naturzweck, so that its nature and purpose explain the existence and form of everything real. For “Life” argues that there can be a true Naturzweck only when there are multiple individuals engaged in a process of reproduction and drawing nutrition from outside themselves.[24]

But it is important that Hegelian compatibilism is not a view that blandly seeks comfort in a lack of surprising or far-reaching metaphysical claims. On the contrary, Hegel’s compatibilism is part of an ambitious project in metaphysics. To begin with, we should not underestimate the degree to which Hegel’s view marries some deflationism (the lack of a requirement for representations) with a significant dose of inflationism. In particular, Kant argues that, without representation, a Naturzweck would require a supersensible non-spatio-temporal ground, knowable and comprehensible only by a higher form of intellect. I have focused mainly on the sense in which Hegel rejects this requirement: we can know and comprehend a Naturzweck simply by knowing about reproduction within a species. But there is also a sense in which Hegel’s response is an explanation of how the requirement can be met. For there is a sense in which Hegel thinks that we take up a superior perspective on life when we come to understand how an organism “produces itself,” or to understand this kind of type-token connection in reproduction. And insofar as we take up that superior perspective, we understand why the order of time does not prevent a sense in which something can contribute to its prior genesis: we understand how “[t]hat which is produced is as such in the ground […] prior, before it becomes actual” (VGP 19:176).

Further, Hegel’s argument in “Life” directs attention toward his metaphysical views about “concepts” or Begriffe. For although we are supposed to be able to know about a form of natural teleology that needs no determining representations, Hegel does not take this to mean that it needs no determining concepts. Rather, Hegel is aiming to argue that there is a kind of “objective concept,” or Begriff, that can be of explanatory relevance without needing to be represented.[25] In biological cases “the concept” is the kind or species [Gattung]. It makes sense to use the term “concept” here insofar as the Gattung is something general or universal—insofar as there are multiple instances of one and the same kind. But “the concept” in this sense is in no way dependent on its being represented. Nor is it dependent on its somehow “containing” representations of necessary and sufficient conditions of its application. The concept or kind distinguishes itself in the ways that its tokens struggle to survive: “the animal establishes and preserves itself as an independent existence, that is, distinguishes itself from others” (EN §368A in the German, §370A in the English edition). And such individuals bind themselves together into the kind by relations of reproduction, so that the “product” of this process is “the realized species [Gattung], which has posited itself identical with the concept [der Begriff]” (WL, 6:486/774). Further, insofar as biological normativity is relative to the species, this is one case of Hegel’s general position that where something ought to be X this is determined by its concept or Begriff. I take Hegel’s claim for the importance of such an objective concept or Begriff to be crucial throughout his metaphysics.[26]

But to see the extent of Hegel’s metaphysical ambitions, we also need to consider the worry that Hegel’s compatibilism would allow underlying non-teleological explanation to be ultimate or fundamental, thus preempting and rendering superfluous teleological explanation. Hegel will argue, by contrast, that law-governed lower levels of reality are of less explanatory relevance than higher teleological levels, preventing any preemption. One contemporary way of working towards Hegel’s distinctive position here would begin with a famous scene in Molière. The character Argan is asked why opium puts us to sleep, and he responds that opium has a dormitive virtue or power. Fortunately, we can nowadays do more to explain here. Imagine for the moment that we could go so far as to explain the phenomena in terms of the arrangement of the underlying electrons and protons in opium. It is natural to think that some such underlying story has all of the explanatory relevance, leaving Argan’s laughable story with none. But now consider drawing the conclusion that any appeal to a supposed natural kind of thing and its supposed dispositions or powers must always be preempted, and can never be itself of any explanatory relevance. One problem is that even our new underlying explanation in terms of protons and electrons cannot be any different on this score: we explain in terms of powers of attraction between unlike charges, or the like. And there is no hope of science discovering any explanations different in kind on this score. For science can only learn about whatever there is to things in virtue of which they affect one another and ultimately our observations. As Blackburn puts it, “any conceivable improvement in science will give us only a better pattern of dispositions and powers.”[27] So if we hold that powers and dispositions of general kinds of things are always explanatorily irrelevant, then either we must always remain ignorant of the real explanations of things, or else nothing ever happens for any reason at all.

Having noted the general problem from a contemporary perspective, we can now approach Hegel’s view: We are tempted to go astray by thinking that scientific talk of low-level nature is akin to talk about hidden physical objects, which can end a regress of why-questions by explaining everything by appeal to something like the seemingly self-evident manner that a billiard ball’s impact affects another. We are tempted, Hegel says, to give “a physical meaning of independent forces” (EN §270A) to the results of the sciences. But Hegel argues that the truth is quite otherwise. Consider Hegel’s account of “chemism,” by which he means not chemistry specifically but any lower-level lawfully interacting substances (WL 6:429/727). What science tells us about such kinds of things is how they interact with others—their dispositions and powers. To take this seriously is to hold that what such a thing is will be merely dependent on what it interacts with and how—it “is not comprehensible from itself alone, and the being of one is the being of the other” (WL 6:430/728). When we seek to understand such a lawful physical thing, and why it does what it does, it “gets lost” in relations, or a regress of dependence; it “becomes something else than it is empirically, confuses cognition” (PhG 3:190/149). So we expect to find a sort of ultimate independence or substance in lawful nature, and we find only dependence and a kind of lack of substantiality.

Part of the point of Hegel’s account of life is that living beings are higher-level phenomena that have exactly the sort of independence or substance that goes missing on the lower law-governed levels of nature. For example, why does an elm have its leaves and so power to take in energy? Here we do not “get lost” in a regress of dependence. The answer is not that sunlight has disposition to be caught, and so on. For the elm has leaves and its capacities because of something about the elm itself: because of its intrinsic end or goal of self-preservation. 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 getting lost in relations with others. An organism, by contrast, is “the real End [Zweck] itself […] it preserves itself in the relation to an other” (PhG 3:198/156). So we can explain what happens in terms of the elm’s power to take in energy and thereby to preserve itself. And such teleological explanations of living beings have a kind of greater explanatory relevance than lower-level lawful explanations.

But 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 and on which it, in a sense, depends. The downward dependence does not in any way lesson the greater explanatory force of the higher-level teleological phenomena. For what is depended on is, in a sense, a matter of indifference. A tiger, for example, is dependent in that it 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 stuffs are a matter of indifference. The tiger’s claws could be made out of other chemical substances. And if they were, then those other substances would be present in virtue of the way they contribute to the end of self-preservation. So there is here an independent or immediate explanatory force that is not undercut by its dependence on or mediation by something underlying—life has a kind of mediated immediacy, in Hegelian terms. And there should be no question of preemption by something more fundamental below.

In sum, Hegel’s position on teleology and life is not as inflationary as Kant’s. Nor does it involve a claim that matter can somehow represent concepts and organize itself in accordance. But this is not to say that Hegel’s view blandly seeks comfort in a lack of surprising or far-reaching metaphysical claims. On the contrary, Hegel’s compatibilism is part of an ambitious project of arguing for the importance of objective concepts, and for a metaphysical priority of higher-level teleology over lower-level necessity.

Conclusion

Before coming to an end, it is worth casting a brief glance again back at debates about free will. With respect to the philosophical issues, Hegel’s distinctive metaphysics suggests some possibilities worth pursuing. Probably the most prominent argument today against our having free will or being morally responsible turns on the claim that, given what we know about the laws of nature, none of us can be an “ultimate source” of what we do.[28] It would be worth investigating the force of the Hegel-inspired reply that our knowledge of the laws of nature cannot possibly suggest any competing “ultimate source” for our behavior—for everything within lawful nature, Hegel argues, is merely dependent. Perhaps this reply would support an argument that our own lack of ultimate sourcehood does not count against our being responsible for what we do unless there was reason to expect some other, competing ultimate source.

With respect to Hegel’s own account of free will, this is a difficult and complex topic, to say the least. But here too debate tends to cluster toward extreme possibilities. On the one hand, some worry that Hegel treats the topic of free will in terms of his general metaphysics, in which the defense of natural teleology plays a large role. The worry is that this will suggest that having free will is just accepting one’s place and role within and subordinated to the ends of a larger organism that is all of reality. On the other hand, the way to escape such a reading can seem to be a purely deflationary interpretation, on which Hegel attempts to reject narrowly causal theories of freedom, and this is supposed to render moot all traditional metaphysical problems about freedom, leaving a kind of free will supposed to be independent of any implications about what really produces or brings about or is responsible for my behavior. But our results here suggest that there is a neglected and more attractive possibility. Hegel advocates a metaphysics that is compatibilist, defending the distinctness and explanatory priority of higher levels of reality over lower—even while recognizing a sense in which higher levels depend on the lower levels. So we should expect Hegel to have a similar metaphysical approach to our freedom, not a purely deflationary one. True, we should expect Hegel to recognize a kind of dependence of our freedom on life—to hold, for example, that the purposive action of embodied beings is possible only insofar as there are natural teleological systems, or life. But we should also expect him to argue that this sort of dependence does not collapse freedom into life or anything natural. On the contrary, we should expect Hegel to argue that free beings have a sort of responsibility for what they do that contrasts with anything on any lower level below. So freedom should depend on nature, but in a way that leaves it also independent—it too should have a form of mediated immediacy. We should not, then, expect Hegel’s account of freedom to be independent of metaphysics. Rather, we should expect it to complete the metaphysics on which the highest levels manifest the greatest independence or freedom.

But free will itself is a topic that will have to wait for another day. The aim of my argument here has been to use familiar features of debates about free will in order to clarify the underlying issues concerning teleology and life at stake between Kant and Hegel. Once we understand these issues correctly, we can see how they demand a certain structure from Kant’s argument and Hegel’s rejoinder. We can then understand both the considerable philosophical force of Kant’s argument, and also how it nonetheless opens itself to a powerful Hegelian rejoinder. The arguments on both sides here remain of considerable philosophical force and interest, despite the incredible subsequent progress in the biological sciences. So understanding the arguments in Kant and Hegel can still help us to better understand the underlying philosophical issues concerning teleology and life.

* For feedback and helpful discussion on this material and other work in this area, I want to thank Fred Beiser, Richard Boyd, Paul Hurley, Thomas Khurana, Daniel Moerner, Dean Moyar, and Peter Thielke. I have used and modified in this work some parts of my essay “The Logic of Life: Hegel's Philosophical Defense of Teleological Explanation of Living Beings,” in: Frederick C. Beiser (ed.), The Cambridge Companion to Hegel and Nineteenth-Century Philosophy, Cambridge: Cambridge University Press 2008, pp. 344–377.

[1] See e.g. Richard Holton, “Review of The Illusion of Conscious Will by Daniel Wegner,” Mind 113:449 (2004), pp. 218–21.

[2] James Kreines, “The Inexplicability of Kant's Naturzweck: Kant on Teleology, Explanation and Biology,” Archiv für Geschichte der Philosophie 87:3 (2005), pp. 270–311.

[3] Clark Zumbach, The Transcendent Science. Kant’s Conception of Biological Methodology, The Hague: Nijhoff 1984, p. 19; see the similar approach in John D. McFarland, Kant’s Concept of Teleology, Edinburgh: University Press 1970, p. 102 and Willem deVries, “The Dialectic of Teleology,” Philosophical Topics 19:2 (1991), p. 53.

[4] On this argument see also McFarland, Kant’s Concept of Teleology, p. 106, Rachel Zuckert, Kant on Beauty and Biology. An Interpretation of the “Critique of Judgment”, Cambridge: Cambridge University Press 2007, ch. 2, and Paul Guyer, “Organisms and the Unity of Science,” in: E. Watkins (ed.), Kant and the Sciences, Oxford: Oxford University Press 2001, p. 265.

[5] See McFarland, Kant’s Concept of Teleology, p. 106.

[6] I am greatly indebted to Thomas Khurana’s comments on the whole, but especially those that prompted the addition of this paragraph.

[7] That there cannot be a Newton for a blade of grass is a point meant to leave open the possibility that organisms really originate in “mere mechanism” (KU 5:400). But limitations of what it is possible for “humans” to “grasp” leave us needing the heuristic guidance of teleology if we are to make progress.

[8] See Robert C. Cummins, “Functional Analysis,” Journal of Philosophy 72:20 (1975), pp. 741–765.

[9] See Ruth G. Millikan, Language, Thought and other Biological Categories, Cambridge: MIT Press 1984; Karen Neander, “The Teleological Notion of ‘Function’,” Australasian Journal of Philosophy 69:4 (1991), pp. 454–468.

[10] See e.g. Robert C. Cummins, “Neo-teleology,” in: Andrew Ariew, Robert Cummins, Mark Perlman (eds.), Functions: New Essays in the Philosophy of Psychology and Biology, Oxford: Oxford University Press 2002 pp. 157–173.

[11] See also EL §55A and WL 4:440-1/737.

[12] See Kreines, “The Logic of Life.”

[13] See e.g. Derek Pereboom, Living Without Free Will, Cambridge: Cambridge University Press 2001.

[14] E.g. deVries, “The Dialectic of Teleology,” p. 56.

[15] E.g. WL 6:486/774; EL §221; EN §367ff; VL 213.

[16] De Anima 415a.

[17] Physics 199b.

[18] I thank Richard Boyd for comments throughout, and especially on this point.

[19] Michael Thompson (see Michael Thompson, “The Representation of Life,” in: eds. Rosalind Hursthouse, Gavin Lawrence, Warren Quinn (eds.), Virtues and Reasons: Philippa Foot and Moral Theory, Oxford: Oxford University Press 1991, pp. 247–296) also stresses the importance of the role of an organism in a species. If I understand him, our claims are similar in this respect but the philosophy different. I am defending Kant’s challenge: on my view, accounts in this neighborhood fail, like the purely deflationary account, if they do not both (i) distinguish the questions about the analysis of natural teleology from questions about living beings, and (ii) resolve Kant’s problem concerning the need in a teleological system for the role of the parts to explain their existence and form. I am defending Hegel insofar as I think he recognizes the inescapable problem; what he argues is that attention to species resolves it. As far as I can see, Thompson doesn’t address that particular problem about natural teleology; what he argues that reference to the species is necessary to answer a different question, about living beings or the form of description appropriate to them.

[20] See also VL pp. 210-211.

[21] See also WL 2:419-20/766; EN §350Z.

[22] Ruth G. Millikan, White Queen Psychology and Other Essays for Alice, Cambridge: MIT Press 1993, p. 41.

[23] See e.g. Cummins, “Neo-Teleology.”

[24] For accounts of monism and teleology in Hegel, see for example Rolf-Peter Horstmann, Die Grenzen der Vernunft. Eine Untersuchung zu Zielen und Motiven des Deutschen Idealismus, Frankfurt a.M.: Anton Hain 1991, pp. 177–82 and Frederick C. Beiser, Hegel, New York: Routledge 2005, chs. 3–4. McTaggart notices the tension between “Life” and organic monism, but takes Hegel to be trying to support organic monism but doing a very poor job of it given the tension (John McTaggart, A Commentary on Hegel's Logic, Cambridge: Cambridge University Press 1910, pp. 275–6).

[25] On “objective concept,” see, for example, WL 6:271/597.

[26] On this theme in Hegel’s broader metaphysics, see my “Hegel: Metaphysics Without Pre-Critical Monism,” Bulletin of the Hegel Society of Great Britain 57/58 (2008), pp. 48–70.

[27] Simon B. Blackburn, “Filling in Space,” Analysis 50:2 (1990), pp. 62–65, here p. 63.

[28] E.g. Pereboom, Living Without Free Will.