1 In volume 2 of The Works of John Locke, 12th edit ion, 9 volumes. London, 1824.

2 The main contemporary proponent of the mode interpretation is Uzgalis, William ‘Relative Identity and Locke's Principle of Individuation,’ History of Philosophy Quarterly 7 (1990) 283–97.Google Scholar Lowe, E.J. Locke (New York: Routledge 2005),Google Scholar also suggests a mode interpretation. Vere Chappell, ‘Locke on the Ontology of Matter, Living Things and Persons,’ Philosophical Studies 60 (1990) 19-32 holds that consciousness is a mode but — from my point of view surprisingly — denies that persons are modes.

3 That is, Book 2, Chapter 12, section 4 of Locke, John An Essay Concerning Human Understanding, Nidditch, Peter H. ed. (Oxford: Clarendon Press 1975).Google Scholar

4 Abstract ideas and their objects may fall outside this taxonomy, but we can safely leave them aside for current purposes. Whatever else persons are, they are not abstract.

5 Martha Bolton points out that Locke never claims that his taxonomy of complex ideas is exhaustive. And — as Leibniz's disease example shows — certain ideas do not fit the taxonomy well. (A disease is obviously not a relation or a substance, and it cannot be a mode because it has a real essence amenable to empirical investigation.) Nevertheless, I believe that the trivision is intended to be exhaustive. Book II of the Essay is meant to show that all ideas derive from sensation or reflection, and the way Locke tries to establish this is by showing that ideas of substances, modes and relations are composed of simple ideas and that simple ideas derive from sensation or reflection. Moreover, even if Locke would allow the possibility of a fourth sort of complex ideas not discussed in the Essay, it is not plausible that an idea discussed in as much detail as the idea of a person would fall into that fourth category.

6 Carson, Emily ‘Locke on Simple and Mixed Modes,’ Locke Studies 5 (2005) 19–38Google Scholar argues that the distinction between simple and mixed modes is important, though problematic, for Locke: ‘geometrical modes are tied to the simple idea of space in a way that undermines their claim to ideality, and thus undermines Locke's account of our knowledge of them’ (32). On her view, although Locke claims universal truth, reality, and adequacy for ideas of both simple and mixed modes, the reasons should be different in the two cases. Mixed mode ideas are adequate because they are their own archetypes; simple mode ideas are adequate for roughly the same reasons as simple ideas. However, Carson also grants that ‘Locke himself often attributes to the objects of mathematics features that are only appropriate to mixed modes’ (24). One example of this is the comparison between mathematical and moral ideas at 4.4.6-9, where moral knowledge can be as certain as mathematical knowledge because both their objects are ideas. (Cf. 2.31.3, 4.12.8.) Since I am interested in Locke's views on mathematics only in so far as they help us understand his views on morality, I focus on precisely those passages. However, I agree that Locke ought not to assimilate mathematical and moral knowledge in the way he does.

7 This example brings up a host of questions concerning Locke's views on natural kinds. However, these questions and the scholarly debate they have occasioned can safely be bracketed out here. My example simply requires Locke to allow that we sometimes decide that our substance ideas include features they should not include. And although he is far more concerned with a second sort of case, where our substance ideas should include more features than they do — and although he may have trouble describing how the first sort of case works in his theoretical framework — I find it hard to believe Locke could deny that cases of the first sort ever occur.

8 Here I am bracketing out various complications that arise when we intend our mode ideas to correspond to those of other people.

9 A third potential source of unity may be worth noting as well. Perhaps substances are unified because all their qualities flow from their internal constitution. This, unlike the real essence suggestion, provides unity to individuals and not just to the idea of substances of a certain kind in general.

10 As Carson, ‘Locke on Simple and Mixed Modes,’ points out, we actually need to consider the nature of space as well. This complicates the situation somewhat. Carson suggests that simple modes are constrained by the nature of space and hence not entirely arbitrary in the way that mixed mode ideas are arbitrary. If so, the epistemology of simple and mixed modes is entirely different. However, I am inclined to think that Locke does not fully realize the extent to which he is appealing to the nature of space here. After all, the 3.3.18 passage just quoted says explicitly that all the properties of a triangle flow from its real essence.

11 It's clear why Locke assumes that all rational or thinking beings are ipso facto conscious: recall the famous claim of 1.2.5 that it is ‘near a contradiction, to say that there are truths imprinted on the soul, which it perceives or understands not.’ However, personhood requires more than momentary consciousness: it requires the ability to consider oneself as a being persisting through time. I am unsure whether Locke would accept the possibility of conscious beings that lacked this ability. If he does, then we should say that a person is a certain kind of conscious being. This possibility will make no difference for the following discussion.

12 In this section I use the word ‘property’ in its inclusive, 21st century sense.

13 I borrow this suggestion from Cleve, James van Problems from Kant (New York: Oxford University Press 1999), 212,Google Scholar who characterizes Spinoza's modes in this way.

14 I owe this objection to an anonymous referee for this journal.

15 Winkler, Kenneth ‘Locke on Personal Identity,’ in Chappell, ed., Locke (Oxford: Oxford University Press).Google Scholar

16 Kaufman, Dan ‘Locke on Individuation and the Corpuscular Basis of Kinds,’ Philosophy and Phenomenological Research 75 (2007) 499–534.Google Scholar Kaufman's objection is actually directed at the mode interpretation of organisms. But one might also direct this objection against the mode interpretation of persons.

17 In conversation.

18 The oak and the mass of matter constituting it, Winkler thinks, would be two bodily substances in the same place at the same time. Thus one could also reply that ‘body’ is not the relevant kind: organism and mass of matter are the two relevant kinds.

19 I owe this clarification to personal correspondence with Winkler.

20 It is unsurprising that we do so, as we are often confused about what ideas the terms ‘man’ and ‘person’ stand for. One goal of 2.27 is to distinguish the ideas of man and person, after all, and it was not until the 2nd edition of the Essay that Locke wrote about identity and hence came to understand the relation between the ideas of persons and organisms.

21 I think my reading is far more natural than Winkler’s. But I also think that such judgments of naturalness tend to be conditioned by pre-existing commitments. Perhaps, then, the passage in itself is genuinely ambiguous. But recall where we are in the dialectic: Winkler advances this passage as an objection to the mode interpretation, so if the passage is actually ambiguous, this is enough to defeat his objection.

22 I owe this point to an anonymous referee.

23 Since Kaufman directs this objection against the mode interpretation of organisms, let me note that I do think the arbitrariness of modes counts against construing organisms as modes.

24 However, consider 2.23.28, where Locke seems to say that the transfer of motion from one body to another, though inconceivable, is actual.

25 On this reading, the present moment of consciousness is not all that counts in constituting the self. When I appropriate the little girl into myself, I appropriate all of her past — everything she has appropriated — whether or not I am now conscious of it. Thomas Reid objected in his 1785 Essays on the Intellectual Powers of Man that Locke's account leads to contradictory results. If a general remembers being a young officer capturing a flag, the young officer remembers being a boy stealing fruit, and the general does not remember stealing the fruit, then the general both is and is not the boy. If Winkler and I are right, this just misunderstands how appropriation works. When the general appropriates the young officer into himself, he thereby appropriates the boy as well.

26 The objection, first proposed by Joseph Butler in his 1736 Analogy of Religion, is that Locke's account is circular: what makes a the same person as b is that a remembers b, but a's memory is only veridical if she is the same person as b.

27 Locke thinks that body-switching, at least, has happened in the past and will happen again: Christ was resurrected and we will be resurrected. However, it is interesting to note that he does not invoke these examples in 2.27, preferring to stick with examples that he would consider fanciful.

28 There is surprisingly little discussion of Locke's thought experiments. An exception is David Soles and Bradfield, Katherine ‘Some Remarks on Locke's Use of Thought Experiments,’ Locke Studies 1 (2001) 31–62.Google Scholar Soles and Bradfield argue that the thought experiments of 2.27 ‘are designed to convince … readers that they do, in fact, employ identity of consciousness as their criterion of individuation of persons and doing so makes sense only if they already implicitly accept Locke's analysis of what it is to be a person’ (53). I agree with this whole-heartedly. My goal is to determine why Locke thinks we can legitimately assume that our implicit criterion of personal identity is correct.

Soles and Bradfield do not tie their discussion into issues about real essence. However, it's worth noting that they distinguish two kinds of Lockean thought experiments: those intended to help with conceptual analysis, like the ones in 2.27, and those that rely on analogies between unobservables and observed objects.

29 The one exception I am aware of is the man with microscopical eyes at 2.23.12.

30 Again, Carson shows that this account of the reality of mathematical knowledge is not really consistent with Locke's claim that mathematical objects are simple rather than mixed modes. Given his explanation of the genesis of simple modes, he ought to allow that the reality of mathematical modes depends on their agreement with space as well.

31 This does not imply the unacceptable result that our ideas of organisms and souls must be mode ideas too. For Locke does not suggest that distinguishing our idea of a person from those other ideas yields interesting results about souls or organisms, only about persons.

32 The reader may worry that a parallel argument implies the false conclusion that ideas of organisms are mode ideas: don't sections 4-8 use a similar procedure to arrive at the conclusion that ‘the same animal … is the same continued life communicated to different particles of matter, as they happen successively to be united to that organized living body’ (2.27.8)? I reply that the procedure of 2.27.4-8 is importantly different in that it proceeds by considering cases we have actually experienced and hence constitutes probability. (Recall that Locke goes to great lengths to defend the credentials of the author of the story of Prince Maurice and the parrot, in order to show that we can accept the story on the basis of testimony.)

33 I think that Locke should not be committed to even the possibility of knowledge of the real essences of substances, but nothing in my argument hangs on that. For even if you think that knowledge of substantial real essences is possible and that a demonstrative science of morality is merely possible and not actual, the text still makes it clear that the reason for the possibility of a demonstrative science of morality is that moral ideas are modes.

34 The distinction between trifling and informative propositions is, I think, closely related to the distinction drawn in section II between what is implied by an idea and what is exhibited in the idea. The proposition ‘man is an animal’ (like other trifling propositions such as ‘a fetiche is a fetiche’ (4.8.3) is trifling because anyone possessing the idea of a man ipso facto sees that a man is an animal. (Recall that Locke holds that the idea of a man is the idea of a living human organism.) In contrast, propositions like ‘the internal angles of a triangle add up to 180 degrees’ is instructive because adding up to 180 degrees is not a feature exhibited in the idea of a triangle.

35 Locke allows one exception to his claim that in our current state, demonstrative sciences must begin from ideas of modes. We have demonstrative knowledge of God's existence and — as we shall see in a moment — the idea of God is central to the demonstrative science of morality. And surely God is a substance and not a mode! However, God differs from created substances in many ways, and one widely recognized difference helps us see why the idea of God should be able to enter into a demonstrative science when no other substance idea currently can. At 4.10.12 Locke tells us that God's ‘omniscience, power, and providence will be established, and all his other attributes follow’ from ‘the necessary existence of an eternal mind’ — something he has just demonstrated. (I do not claim to understand how one derives providence from necessary existence.) Thus, although Locke does not put it in these words, we have at least partial knowledge of the real essence of God. This explains why the idea of God can enter into demonstrative science while our ideas of created substances cannot.

36 Ruth Mattern, ‘Moral Science and the Concept of Persons in Locke,’ in Chappell, ed., Locke points out the significance of this passage (273).

37 This requires a bit of qualification. As noted earlier, all rational beings are ipso facto conscious for Locke. His argument for this claim does not establish that all rational, conscious beings can consider themselves as themselves in the appropriate way; perhaps he thinks there could be conscious, rational beings who thus do not amount to persons. However, we clearly are not such beings, nor do we conceive of ourselves as such.

38 Locke, John Two Treatises of Government, Laslett, Peter ed. (New York: Cambridge University Press 1988)CrossRefGoogle Scholar

39 For an extended argument that a substantial portion of Locke's political theory is intended to be a demonstrative science, see Grant, Ruth John Locke's Liberalism (Chicago: University of Chicago Press 1987),CrossRefGoogle Scholar ch. 1.

40 See, for example, Ayers, Michael Locke (New York: Routledge 1993), vol. 2, 206.Google Scholar Kaufman argues that the tension is irresolvable.

41 That is, the tension between Locke's anti-essentialism and his account of personal identity dissolves. A tension between his anti-essentialism and his account of the identity of living organisms remains.

42 I do not claim that Locke succeeds in doing this. His account of personal identity makes the individual self the sole legislator of its constitution: whether a certain consciousness is part of me depends on whether or not I have appropriated it. I am not sure that our actual practices of ascribing responsibility fit in with this, or whether they presuppose some further criterion of correct appropriation. Considering Locke's account of why we punish the sober man for actions performed while drunk is relevant here. However, I think Locke would allow some back-andforth between the idea of a person and the practice of ascribing responsibility: a constraint on the idea is that it grounds the practice, but once we make the idea clear and distinct we could very well see reason for amending the practice somewhat.

43 A similar point can be made concerning the demonstrative science of morality. Since many consistent sets of mode ideas are possible, many demonstrative sciences — and hence many different sets of moral rules — are possible. However, this does not yield relativism because one such set is privileged: the set of rules that best conduces to the preservation of mankind or, to put it differently, the set that is handed down by God in the form of natural law.

44 Versions of this paper were presented at the July 2007 Atlantic Canada Seminar in Early Modern Philosophy and the 4th Biennial Margaret Wilson Conference, at Cornell in August 2008. I would like to thank both audiences for helpful comments. Martha Bolton, Stewart Duncan, Paul Lodge, Walter Ott, and Kenneth Winkler read previous versions of this paper. The end result has benefited greatly from all their suggestions. In addition, I would like to thank two anonymous reviewers for this journal, both of whom gave helpful and detailed comments that went well beyond the call of duty.