Naturalism and the Space of Reasons in Mind and World

Abstract This paper aims to show that many criticisms of McDowell’s naturalism of second nature are based on what I call ‘the orthodox interpretation’ of McDowell’s naturalism. The orthodox interpretation is, however, a misinterpretation, which results from the fact that the phrase ‘the space of reasons’ is used equivocally by McDowell in Mind and World. Failing to distinguish two senses of ‘the space of reasons’, I argue that the orthodox interpretation renders McDowell’s naturalism inconsistent with McDowell’s Hegelian thesis that the conceptual is unbounded. My interpretation saves McDowell from being inconsistent. However, the upshot of my interpretation is that what is really at work in McDowell’s diagnosis of the dualism between nature and reason is the Hegelian thesis, not the naturalism of second nature.

space of reasons is reducible to the realm of law. The former, however, renders mysterious the relation between the realm of law and the space of reasons, whereas the latter is welcomed by naturalists.
But McDowell denies that they are the only possibilities: the space of reasons, he holds, is sui generis but no less natural. McDowell's alternative is incompatible with the disenchanted conception of nature. To make room for McDowell's alternative, the disenchanted conception of nature must be in some way rejected. However, McDowell also wants to preserve the achievement of science. Therefore, McDowell wants a naturalism that makes room for both science and the space of reasons. For that, McDowell thinks that our conception of nature must be reconsidered. This re-construed conception of nature is the naturalism of second nature, by virtue of which 'we can keep nature as it were partially enchanted, but without lapsing into pre-scientific superstition' (McDowell, 1996, p. 85). Nature, therefore, is partially within the realm of law and partially within the space of reasons.
The notion of second nature, McDowell claims, is derived from Aristotle's idea that a normal human being is a rational animal. It belongs to our animal nature that we are born with the potentiality to be rational agents. The conditions necessary to actualize it are nothing supernatural. That is why rationality is the second nature of human beings. Thus, the notion of second nature 'gives human reason enough of a foothold in the realm of law to satisfy any proper respect for modern natural science' (McDowell, 1996, p. 84). Our first or animal nature belongs in the realm of law, and our second or rational nature belongs in the space of reasons. McDowell's naturalism of second nature thus consists of two claims: (N1) The realm of law and the space of reasons are distinct from each other: neither one belongs in the other; (N2) The realm of law and the space of reasons are both natural: the nature belonging in the realm of law is disenchanted and the nature belonging in the space of reasons is enchanted.
This construal of McDowell's naturalism, I believe, is the orthodox interpretation in the literature. 1 Moreover, the orthodox interpretation has textual evidence. For example, when McDowell tries to bring out the notion of second nature, he says that even though 'the structure of the space of reasons is alien to the layout of the nature conceived as the realm of law', following Aristotle we can obtain 'the notion of having one's eyes opened to reasons at large by acquiring a second nature ' (1996, p. 84). It appears that McDowell does accept (N1), and meanwhile uses the notion of the second nature to bring out that the space of reasons is no less natural than the realm of law. Despite the textual evidence, I think that the orthodox interpretation of McDowell's naturalism is problematic. I think that many criticisms of McDowell's naturalism in the literature result from the orthodox interpretation. In order to bring out that this interpretation is orthodox and mistaken, I want to discuss two common criticisms on McDowell's naturalism in the literature: (C1) McDowell's naturalism does not make intelligible how the second nature of human beings is possible in a world of first nature; 2 (C2) McDowell's naturalism is another dualism between the realm of law and the space of reasons in disguise. 3 I pick out these two criticisms because they both aim to show that McDowell's naturalism does not dissolve the anxiety caused by the dualism between the realm of law and the space of reasons. (C1) can be taken as claiming that if McDowell's naturalism cannot explain how the second nature of human beings is possible in a world of first nature, it fails to dissolve the anxiety caused by the dualism. And if it fails to dissolve that anxiety, then critics may explain the failure in terms of (C2): the failure demonstrates that McDowell's naturalism is still the dualism between the realm of law and the space of reasons.
It seems that, in Mind and World, McDowell does try to make intelligible how human beings, born as mere animals, can be initiated into their second nature. The initiation into the second nature of human beings is done through learning a human language (1996, p. 125). However, critics of (C1) could ask how human beings, born as mere animals, can learn a language. Consider the acquisition of a first language by a human infant. If the newborn infant's nature is completely within the realm of law, then it is difficult to see how he could respond to his mother's talking to him and then grasps her meaning, since a mere animal only responds to the auditory impression, not to the meaning. But if we take the infant as endowed with linguistic capacities, then it only shows that we are 'born at home in the space of reasons' (McDowell, 1996, p. 125), which contradicts McDowell's view. 4 As discussed above, McDowell tries to bring out the link between first nature and second nature by the 'foothold' remark that second nature is not floating free of potentialities in first nature. But that remark invites questions like (C1) about the link between these two natures. Moreover, one could advance criticisms like (C2) that if McDowell's naturalism is not another dualism in disguise, it must offer an adequate account of the link between first and second nature. For if it fails, it may indicate that in McDowell's naturalism there is no unity that helps bridge the gulf between the realm of law and the space of reasons and that without that unity McDowell's naturalism is still dualistic.
However, McDowell later expresses that he regrets the 'foothold' remark (2000, p. 99). Instead of providing a new account of the relation between first nature and second nature, McDowell thinks that the notion of second nature alone is enough for his purposes: That [foothold] remark seems to promise more, in the way of a continuity between the naturalness of human responsiveness to reasons and the naturalness of phenomena subsumable under natural law, than my purposes require. I think that the only unity there needs to be in the idea of the natural, as it applies, on the one hand, to the intelligibility of physical and merely biological phenomena (themselves needing to be differentiated for some purposes, as I have indicated), and, on the other, to the intelligibility of rational activity, is captured by a contrast with the idea of the supernaturalthe spooky or the occult … That should be enough to reassure us that, for all the sui generis character of responsiveness to reasons, there is nothing spooky about it, and that is all I need from the idea of second nature. (McDowell, 2000, p. 99) Hence McDowell thinks that he needn't answer (C1). His rationale is metaphilosophical. McDowell's metaphilosophical stance is quietism. To dissolve the dualism between the realm of law and the space of reason, instead of searching for a positive account of the link between first nature and second nature, McDowell thinks that it is enough to have the notion of second nature as a reminder that 'there is nothing but a scientistic prejudice in the view that a naturalism of natural science has that default status ' (2006, p. 237). Besides, his reply to (C2) is that the only unity behind first nature and second nature he needs to bring out is being not supernatural. However, I doubt that critics of (C2) would be satisfied with this reply. 5 My concern here is not whether McDowell can meet those criticisms, but rather that those criticismsas well as McDowell's responses to themare based on the misinterpretation of McDowell's naturalism. The criticisms all address the worry that McDowell's naturalism fares no better than the dualism between nature and reason, or even worse, it is just another form of that dualism. Essentially, the worry arises because (N1) asserts that both the realm of law and the space of reasons are distinct from each other. However, the dualism between nature and reason accepts (N1) as well. Therefore, if McDowell does not offer a constructive account that explains why (N2) that the space of reasons is as natural as the realm of law, it seems that the only respect in which McDowell's naturalism differs from the dualism between the realm of law and the space of reasons is that it insists on calling the space of reasons 'second nature' or 're-enchanted nature'. This is, I believe, why McDowell's naturalism, construed as (N1) and (N2), draws criticisms like (C1) and (C2).
Interestingly, it seems, McDowell's responses, presented above, suggest that he himself accepts (N1) and (N2) as the correct interpretation of his naturalism. His quarrels with his critics are all metaphilosophical. His critics demand McDowell to advance some constructive theory to explain how initiating from first nature into second nature is possible. But McDowell resists their demand and emphasizes that his naturalism is diagnostic. Beside the metaphilosophical reminder, McDowell nowhere points out that his critics mistake (N1) or (N2) as a correct characterization of his naturalism.
Despite McDowell's implicit endorsement of (N1) and (N2), I still think that the orthodox interpretation is wrong. The critics of (C1) and (C2) focus on (N2). They argue that there must be a substantial theory that explains (N2), usually in terms of the link between the realm of law and the space of reasons. However, I think that we can leave (N2) as an unsophisticated reminder as McDowell recommends. But (N1) ought to be removed from McDowell's naturalism. Indeed, (N1) should be replaced by (N3): (N3) The realm of law is part of the space of reasons.
The correct interpretation of McDowell's naturalism thus consists of (N2) and (N3). My interpretation can dissolve the worries of (C1) and (C2). Since the realm of law now is only part of the space of reasons, it is obvious that the distinction between first nature and second nature is not another dualism between nature and reason. For the unity behind first and second nature that (C2) demands is being within the space of reasons. As to (C1), it is true that (N3) combined with (N2) is still not a full-blown theory that explains how human infants born as mere animals can be initiated into second nature. However, as I analyze above, the reason why the critics of (C1) expect McDowell's naturalism to be more substantial is that his naturalism shares (N1) with the dualism between nature and reason. Therefore, they argue that (N2) must be a substantial theory about how the second nature of human beings is possible in a world of the realm of law; otherwise, McDowell cannot claim that his naturalism can dissolve the anxiety caused by the dualism between nature and reason. My interpretation, replacing (N1) with (N3), thus roots out (C1) and (C2), since it now shares nothing in common with the dualism between nature and reason. And (N2) can accordingly be as insubstantial as McDowell claims. What makes McDowell's naturalism able to cure the dualism between nature and reason is (N3), not (N2). (N2) is indeed just a reminder that it is a prejudice to think that nature is exhausted by the realm of law and that the space of reasons is natural as well.
My aim is not to argue that my interpretation of McDowell's naturalismcombining (N2) and (N3)is better than other sorts of naturalism, but that it is the correct interpretation of McDowell's naturalism in Mind and World. If so, then criticisms like (C1) and (C2) simply miss their target. However, an immediate problem for my claim is that McDowell, as shown above, appears to endorse (N1). I will present other textual evidence showing that McDowell actually endorse (N3). Given the incompatibility between (N1) and (N3), McDowell seems inconsistent. I am unsure whether McDowell is aware of this inconsistency. But I will explain, in section III, that the apparent inconsistency results from the fact that the phrase 'the space of reasons' is used equivocally. To bring out the correct interpretation of McDowell's naturalism, I will expose two uses of 'the space of reasons' in Mind and World. More important, my interpretation is correct, because it fits McDowell's general diagnosis in Mind and World better. In order to show that, in section II I will argue that the orthodox interpretation is inconsistent with McDowell's other diagnosis of the dualism between nature and reason, that is, the diagnosis of the oscillation between the myth of the Given and coherentism: the space of reasons is unbounded. To make sense of McDowell's diagnosis in Mind and World as a whole, the orthodox interpretation must be replaced by my interpretation.

II. The Unfitness of the Orthodox Interpretation
In Mind and World, McDowell claims that an obvious predicament caused by the dualism between the realm of law and the space of reasons is that philosophers have trouble offering a satisfactory account of the answerability of our thinking to the world, without which our thinking is impossible (1996, pp. xi-xiii). The trouble arises because the answerability of our thinking to the world presumes a Sellarsian thesis: (S) The Sellarsian Thesis: Our thinking, belonging in the space of reasons, is answerable only to objects that belong also in the space of reasons (McDowell, 1996, pp. xvi, 7-8).
(S) is the thought that objects to which our thinking is answerable constitute normative constraints, external to our thinking and according to which our thinking is warranted or correct. The normativity requirement means that the answerability relation of our thinking and the world is rational rather than merely causal. Rational relations must consist of conceptual relations (McDowell, 1996, pp. 3-9). Therefore, objects to which our thinking is answerable must belong in the space of reasons or the conceptual.
Since our thinking must be answerable to the world, given (S) this leads to an astonishing thesis in a spirit of Hegel's idealism: (H) The Hegelian Thesis: The world belongs in the space of reasons (McDowell, 1996, Lecture II).
In McDowell's words, (H) is the thesis that 'the conceptual is unbounded; there is nothing outside it' (1996, p. 44). (H) is astonishing, because (H) appears to be in direct contradiction to natural science. For natural science puts the world within the realm of law. Given the dualism between the realm of law and the space of reasons, (H) seems anti-scientific. However, if we reject (H), the only accounts of the answerability of our thinking to the world left to us are the myth of the Given and coherentism, both of which are unsatisfactory. For the myth of the Given recognizes that there must be an external constraint on our thinking to which our thinking is answerable. However, it places the Given in the realm of law, and given (S) the Given cannot bear any rational relation to our thinking. The Given, therefore, is a myth (McDowell, 1996, pp. 7-9). The denial of the Given leads us to coherentism, which acknowledges that only items in the space of reasons can provide the normative constraint. But since the world is in the realm of law, coherentism acknowledges only items in our thinking as the constraint on thinking. However, the cost of coherentism is that there is no longer any external normative constraint on our thinking, so that it fails to explain how our thinking is answerable to the world (McDowell, 1996, pp. 13-18). Therefore, if we want to secure the answerability of our thinking to the world, according to McDowell, we have to accept (H).
My aim here is not to defend (H). My aim is to indicate that (S) and (H) are in conflict with (N1). To expose the conflict, it is helpful to consider the status of the realm of law. According to (H), there is no longer the dualism between the realm of law and the space of reasons, because nothing is outside the space of reasons, so that the space of reasons must include the realm of law. This shows that a corollary of (H) is (N3): the realm of law is part of the space of reasons and cannot be distinct from the space of reasons. (N1) thus contradicts (H) and should be replaced by (N3).
The orthodox interpretation of McDowell's naturalism, therefore, is inconsistent with McDowell's diagnosis of the answerability of our thinking to the world. The inconsistency is from (N1). To be sure, the inconsistency between (H) and (N1) does not automatically imply that (N1) should be rejected. One may get round this inconsistency by rejecting (H). However, saving (N1) instead of (H) would lead us again to the myth of the Given. For the world that belongs in the realm of law is that to which our natural science is answerable. If the realm of law is outside the space of reasons, then our scientific thinking, given (S), cannot be answerable to the realm of law. The realm of law thus becomes the Given. To save the possibility of science, therefore, following (S), the realm of law must be included within the space of reasons. This means that (N1) should be rejected. 6 Recall that (N1) causes criticisms, such as (C1) and (C2). Both (C1) and (C2) address the worry that McDowell's naturalism is the dualism between the realm of law and the space of reasons in disguise. Now we also see that saving (N1) would lead us to the myth of the Given just as the dualism between the realm of law and the space of reasons does. To dissolve the anxiety caused by that dualism, we need (H). But (N1) is inconsistent with (H); the orthodox interpretation cannot employ (H) to remove the worry of (C1) and (C2). I conclude that (C2) is correct that the orthodox interpretation construes McDowell's naturalism as another version of the dualism between the realm of law and the space of reasons. Therefore, the orthodox interpretation cannot be the correct interpretation of McDowell's naturalism. Not only does it draw criticisms like (C1) and (C2), but also it is in conflict with McDowell's another diagnosis that the space of reasons is unbounded. My interpretation of McDowell's naturalism coheres with McDowell's general picture better, since (N3) is the corollary of (H). Moreover, my interpretation of McDowell's naturalism can avoid (C1) and (C2). The response to (C2) is straightforward. The distinction between first nature and second nature is not another dualism, since both of them belong in the space of reasons. As to (C1), although there is still no positive account of how human beings can be initiated from first nature into second nature, the worry, at least, is relieved. Since first nature is also within the space of reasons, it is less difficult to see how human beings, born as mere animals, can acquire second nature.

III. Evidence for the Orthodox Interpretation?
But if (N1) is inconsistent with (H), why (N1) becomes the orthodox interpretation of McDowell's naturalism? Indeed, as indicated, there seems to be textual evidence for (N1). However, I think that the textual evidence is problematic. The textual evidence seems to support (N1) only because it is ambiguous. For example: For Kant, the ordinary empirical world, which includes nature as the realm of law, is not external to the conceptual. In view of the connection between the conceptual and the kind of intelligibility that belongs to meaning, I have suggested that defending that Kantian thought requires a partial re-enchantment of nature. But it does not require us to rehabilitate the idea that there is meaning in the fall of a sparrow or the movement of the planets, as there is meaning in a text. It is a good teaching of modernity that the realm of law is as such devoid of meaning; its constituent elements are not linked to one another by the relations that constitute the space of reasons. But if our thinking about the natural stops at an appreciation of that point, we cannot properly comprehend the capacity of experience to take in even the meaningless occurrences that constitute the realm of law. We cannot satisfactorily splice spontaneity and receptivity together in our conception of experience, and that means we cannot exploit the Kantian thought that the realm of law, not just the realm of meaningful doings, is not external to the conceptual. (McDowell, 1996, p. 97) This passage is puzzling. 7 On the one hand, McDowell says that the realm of law, together with the realm of the meaningful, belongs in the conceptual. On the other hand, he says that the realm of law is devoid of meaning and thus is distinct from the space of reasons. McDowell here seems to hold that the realm of law is within the conceptual but outside the space of reasons. According to this reading, the conceptual is divided into the realm of law that is devoid of meaning, i.e. the disenchanted conception of nature and the space of reason that is full of meaning, i.e. the re-enchanted conception of nature. This seems to confirm (N1). However, it pays the price: the space of reasons now is only part of the conceptual, rather than equal to the conceptual. But this again is in conflict with (S). (S) requires that the world must belong in the space of reasons, which is the reason why the world belongs in the conceptual. Thus, in (H), the space of reasons is the conceptual. Therefore, in (N1) and (H), the phrase 'the space of reasons' is equivocal: In (H), the space of reasons is the conceptual; In (N1), the space of reasons is the realm of the meaningful, which is part of the conceptual.
For the sake of clarity, I capitalize 'the Space of Reasons' to stand for the one in (H), and leave 'the space of reasons' to stand for the one in (N1). I suggest that the textual evidence appearing to support (N1) should be read as saying that the realm of law is external to the space of reasons, but not to the Space of Reasons ' (2006, p. 238). But (N1*) alone cannot be immune from the dualistic criticism like (C2). It is (H*) or (N3*) that answers (C2) by indicating the underlying unity behind the realm of law and the space of reasons, that is, being in the Space of Reasons.
But why does McDowell use the notion of the space of reasons equivocally? It seems puzzling that he has just proposed the thesis that the Space of Reasons is unbounded in the Lecture II of Mind and World, but unhesitatingly excludes the realm of law from the space of reasons in the Lecture IV. Is McDowell unaware of the fact that (H) is incompatible with (N1)? If McDowell knows that he uses the terms equivocally, why doesn't he spell it out clearly? It is especially manifest that McDowell never answers (C1) or (C2) by the thesis (H*) and (N3*). If McDowell had the equivocation in mind, he could easily respond to (C2) that (N1*) is not a dualistic distinction because of (H*) or (N3*). The worry of (C1) and (C2) that McDowell's naturalism is still dualistic would disappear soon. But he never did that. As I have analyzed, without (H*) or (N3*), McDowell's responses cannot be persuasive to his critics, for (N1) and (N2)the orthodox interpretationcan also be accepted by the dualism between the realm of law and the space of reasons.
I cannot explain conclusively why McDowell seems to forget (H) or (N3*) when he proposes the naturalism of second nature. But it may be interesting to speculate why McDowell equivocates in his use of the phrase 'the space of reasons'. In Lecture IV of Mind and World, McDowell shifts from the Hegelian thesis (H) to the naturalism of second nature. At the beginning of Lecture IV, there is an interesting passage: We had better not aspire to put the lost enchantment back into the merely natural world. According to the picture I have been recommending, our sensibility yields states and occurrences with conceptual content. That enables us to see an experiencing subject as open to facts. The conceptual sphere does not exclude the world we experience. To put it another way: what we experience is not external to the realm of the kind of intelligibility that is proper to meaning. (See Lecture II.) But in so far as what we experience includes merely natural facts, this can look like a call to regress into a pre-scientific superstition, a crazily nostalgic attempt to re-enchant the natural world. (McDowell, 1996, p. 72) This passage is interesting, because McDowell here admits that since scientific facts are what we can experience, they belong in the conceptual sphere -'the realm of the kind of intelligibility that is proper to meaning'so that the natural world seems re-enchanted. On the other hand, McDowell says that he does not want to re-enchant the merely natural world, because it appears to be a pre-scientific superstition. Therefore, in the end of the lecture, McDowell states his naturalism as follows: We tend to be forgetful of the very idea of second nature. I am suggesting that if we can recapture that idea, we can keep nature as it were partially enchanted, but without lapsing into pre-scientific superstition. (McDowell, 1996, p. 85) Thus, it is clear that avoiding lapsing into pre-scientific superstition is, at least, one of the reasons that McDowell wants to leave the merely natural world disenchanted. However, McDowell would be caught in a dilemma: either he calls for a pre-scientific superstition or he commits himself to the myth of the Given. On the one hand, to save the merely natural world from being a myth of the Given, the merely natural world must be placed in the conceptual sphere. Given that the conceptual sphere is the realm of the meaningful, it immediately implies that the merely natural world is enchanted. It seems that McDowell does call for a pre-scientific superstition. On the other hand, to avoid re-enchanting the merely natural world, the merely natural world is left outside the realm of the meaningful, i.e. the conceptual sphere. Given (S), our thinking cannot be answerable to the merely natural world. It seems that McDowell thus commits himself to the myth of the Given.
There is a way out of the dilemma. The crucial point is the idea that the conceptual sphere is 'the realm of the kind of intelligibility that is proper to meaning'. If the conceptual sphere is the realm of the meaningful, then the picture McDowell recommends does imply that the whole nature is re-enchanted, for to enchant the world is to put the world into the realm of the meaningful. Thus, McDowell is caught in the dilemma. However, to say that the conceptual sphere is the realm of the kind of intelligibility that is proper to meaning may be to say that the conceptual sphere is where meaning is possible. What belongs in the realm of the meaningful must belong in the conceptual sphere, but what belongs in the conceptual sphere may not belong in the realm of the meaningful. Therefore, nature can belong in the conceptual sphere without being in the realm of the meaningful, but a re-enchantment must occur within the conceptual sphere. This concurs with my above distinction: the Space of Reasons is the conceptual sphere that includes the whole world, and the space of reasons is the realm of the meaningful that includes only the enchanted nature. Therefore, the dilemma is resolved. We are open to the merely natural world because the merely natural world is within the Space of Reasons, and meanwhile the merely natural world remains disenchanted because it is outside the space of reasons.
This confirms my claim that in Mind and World McDowell uses the phrase 'the space of reasons' equivocally. It can be taken as the conceptual sphere or as the realm of the meaningful. Failing to distinguish them, we must conclude that McDowell is inconsistent in claiming that the realm of law must belong in the conceptual sphere but remains disenchanted. However, this way to save McDowell from being inconsistent costs something. The idea of second nature, belonging in the space of reasons rather than the Space of Reasons, cannot really dissolve the anxiety caused by the dualism between the realm of law and the space of reasons. That anxiety is supposed to be dissolved by the Hegelian thesis that the Space of Reasons is unbounded. The notion of second nature, as McDowell later clarifies, is indeed just a reminder that 'excluding these phenomena [belonging to the space of reasons] from the scope of natural-scientific intelligibility does not imply that they are not natural phenomena' (2006, p. 236). In another place, McDowell emphasizes that all he needs from the notion of second nature is 'for all the sui generis character of responsiveness to reasons, there is nothing spooky about it ' (2000, p. 99). As the critics argue, however, if the realm of law and the space of reasons are distinct, it is hardly satisfactory to see how labelling the space of reasons as 'second nature' helps to show that the space of reasons is natural. Instead, since the realm of law and the space of reasons are both within the Space of Reasons, and if the reason that the realm of law is not spooky is the fact that it is in the Space of Reasons (for it is no longer the myth of the Given), the space of reasons is as natural as the realm of law. In other words, what is really at work in McDowell's diagnosis that the space of reasons is no less natural than the realm of law is not the notion of second nature, but the thesis that the conceptual is unbounded.
If McDowell's real diagnosis of how the realm of the meaningful is as natural as the realm of law is the thesis that both of them belong in the Space of Reasons, McDowell should not worry that his call for re-enchanting nature looks like 'a call to regress into a pre-scientific superstition'. For disenchantment should be regarded as a manifestation of the layout of the Space of Reasons. Disenchantment, indeed, is rationalization. Unfortunately, by claiming that 'if we can recapture that idea [of second nature], we can keep nature as it were partially enchanted, but without lapsing into pre-scientific superstition', McDowell confuses the readers to identify 'the space of reasons' with 'the enchanted nature' and thus to exclude the realm of law again from the Space of Reasons.

IV. Conclusion
I have shown that the orthodox interpretation of McDowell's naturalism cannot be correct, because it is unfit to McDowell's Hegelian thesis (H) that the space of reasons is unbounded. Without (H), the orthodox interpretation, as (C2) correctly indicates, is just another version of the dualism between the realm of law and the space of reasons. The reason why the orthodox interpretation assumes its status in the literature is that McDowell uses the phrase 'the space of reasons' equivocally, which causes misinterpretation of his naturalism. To interpret McDowell correctly, it is crucial to distinguish the Space of Reasons as the conceptual sphere and the space of reasons as the realm of the meaningful. Then the correct interpretation of McDowell's naturalism consists of the claim (N2*)that the realm of law and the space of reasons are both naturaland the claim (N3*)the realm of law and the space of reasons both belong in the Space of Reasons. The upshot of my interpretation is that the naturalism of second nature, (N2), plays no significant role in McDowell's diagnosis. What actually dissolves the anxiety caused by the dualism between nature and reason is (N3*), which is the corollary of the Hegelian thesis (H*).
I want to emphasize that the reason why my interpretation of McDowell's naturalism is correct is not that my interpretation is more philosophically viable than the orthodox interpretation, but that it is supported by textual evidence and saves McDowell's naturalism from being inconsistent with his Hegelian thesis. My interpretation might face its own theoretical difficulties. For example, one may worry that the dualism criticism (C2) is just pushed one step back: it is possible that the Space of Reasons might contain a dualism between the realm of law and the realm of the meaningful. In my reply to (C2), I claim that McDowell can, and should, respond that it is not a dualism, because the unity behind them isnot 'being natural or 'not being supernatural'but rather 'being within the realm of the conceptual'. Whether the unity is enough to unify the realm of law and the realm of the meaningful depends on whether the idea of 'being within the realm of the conceptual' is as empty as the idea of 'being natural' or 'not being supernatural'. Answering this question, however, is beyond the scope of this paper. Nonetheless, my response to the dualism criticism is still an improvement over McDowell's own. For my interpretation directs the critics to the real issue concerning McDowell's naturalism. What is needed from McDowell is not an account of the notion of nature, but the notion of the conceptual. Without a satisfactory account of what is the conceptual, the Sellarsian thesis (S), on which McDowell's diagnosis is relied, is threatened. Therefore, the whole picture McDowell presents in Mind and World might be called into question.

Notes
This paper is derived from my thesis. Many thanks to Michael Rohlf for his supervision and Michael Gorman for his comments. I am particularly grateful to Daniel Whiting and two reviewers of this journal for their helpful comments. Early versions of this paper were presented at the NCCU Graduate Conference, 2008 and the Annual Meeting of Taiwan Philosophical Association, 2010. Thanks to the participants, especially Chung-I Lin, Norman Y. Teng, Tony H. Y. Cheng and Wesley Chai for comments and discussion. 1 Thornton, 2004, p. 226. See also notes 2 and 3. 2 Bernstein, 2002, pp. 19-24;Gaskin, 2006, pp. 29-36;Ginsborg, 2005, pp. 81-91;Larmore, 2002;Wright, 2002, pp. 150-56.