Ethics 3p5: Resolving a Contradiction in Spinoza’s Metaphysics John R. T. Grey Abstract: The following triad of claims from Spinoza's Ethics is prima facie inconsistent. First, Spinoza writes, “Whatever is, is in God... ” (Ip15). Second, “ Things...cannot be in the same subject insofar as one can destroy the other ” (IIIp5). Third, for any singular thing, “there is another more powerful by which the first can be destroyed ” (Iva1). In this paper, I discuss a number of different attempts at resolving this apparent inconsistency, focusing on different ways of interpreting the second claim. The proper interpretation of IIIp5, I claim, is that two modes or qualities cannot inhere in the same subject insofar as one can destroy the other. However, I then argue that this interpretation of IIIp5 forces us to take very seriously Spinoza’s notion that the world has different aspects or perspectives. From the perspective of eternity, nothing destroys anything; it is only from particular finite perspectives that destruction ever occurs.

The following triad of claims from Spinoza’s Ethics is prima facie inconsistent. First, Spinoza writes, “Whatever is, is in God...” (Ip15). 1 Second, “Things...cannot be in the same subject insofar as one can destroy the other” (IIIp5). Third, for any singular thing, “there is another more powerful by which the first can be destroyed” (Iva1). Paraphrasing and simplifying, we have: (1) all things are in the same unique subject; and (2) no things that can destroy one another can be in the same subject; yet (3) something can destroy something else. These three claims cannot all be true. In what follows, I will argue that the real culprit behind this apparent contradiction is the second claim, which is based on a faulty interpretation of IIIp5. In §1, I show why Spinoza cannot afford to simply jettison any of the three propositions just cited—despite appearances, each plays an important role in his overarching philosophical work. Nonetheless, there are a number of different ways one might try to reinterpret IIIp5 in order to avoid the contradiction. In §2, I consider several such interpretations, showing how each either fails to resolve to contradiction or fails to make sense of Spinoza’s demonstration of the proposition. Of particular interest are the solutions proposed by Jonathan Bennett and Don Garrett. Bennett (1984) discusses this apparent contradiction, but I will argue that his proposed solution fails on a number of different grounds. Garrett (2002) attempts a resolution as well, and his approach is more promising. Garrett’s

1

approach, which I discuss in §3, takes IIIp5 to be the claim that if two qualities are in the same subject, one cannot destroy the other. This approach is still subject to an important flaw, however. If we adopt a strict reading of Spinoza’s monism according to which all particular objects are qualities of substance, the contradiction reappears: (1’) all things are qualities inhering in the same unique subject; (2’) no qualities that can destroy one another can inhere in the same subject; and (3) something can destroy something else. Having considered the most viable extant readings of IIIp5, in §4 I present the final piece of interpretive work that is needed to save Spinoza. The proposition at issue contains Spinoza’s ubiquitous qualifying phrase, “insofar as [quatenus] ”, which we might initially take to mean something like “to the degree that”. If we take “insofar as ” in that sense, it does us no good. However, Spinoza uses this phrase in two distinct ways, sometimes to signal that something is subject to variation by degrees, and sometimes to signal that something is subject to variation depending upon the way in which we conceive of it. If we take the “insofar as ” of IIIp5 to have this latter sense—as I will argue we should—then Spinoza can consistently reject the claim that particular things are in substance insofar as they can destroy one another. That is, he can admit that there is a way of conceiving of particular things according to which they can destroy one another, and a way of conceiving of particular things according to which they are qualities of substance, but there need be no single way of conceiving of particular things according to which they can destroy one another and inhere in substance. From the point of view of eternity, nothing destroys anything; it is only from particular finite points of view that destruction ever occurs. This, I will argue, is how we ought to resolve the apparent contradiction.

1 Before jumping to any conclusions, it is important to survey each of the three cited passages in full. Spinoza’s meaning often hinges on the placement of a single word, and it may be that my elisions did not do justice to his claims. The first passage is proposition 15:

2

Whatever is, is in God [Quicquid est, in Deo est], and nothing can be or be conceived without God.

Spinoza’s claim here is nothing less than a declaration of the absolute metaphysical dependence of all things on the same, unique substance, God. This is, I think, an uncontroversial reading of the proposition, although there has been some debate about how best to understand Spinoza’s notion of metaphysical dependence.2 Notice that the proposition is even stronger than we need in order to derive the contradiction; all we require is the first conjunct, “Whatever is, is in God.” The second passage is a bit more tricky. IIIp5 is part of Spinoza’s notoriously difficult argument for his doctrine of conatus or striving. Spinoza writes, Things [Res] are of a contrary nature, that is, cannot be in the same subject [subjecto], insofar as one can destroy the other.

‘Subject’ here (following standard early modern usage) seems to have the sense of the subject of a proposition or the subject of predication.3 Presumably God is such a subject, given that it has attributes and modes (EId3-5).4 This suffices to derive the contradiction, since it allows us to infer that things cannot be in substance insofar as one can destroy the other. Whereas the first two passages have been the subject of voluminous commentary, the sole axiom of Ethics IV is not often the focus of philosophical debate. The axiom states, There is no singular thing [res singularis] in nature than which there is not another more powerful and stronger. Whatever one is given, there is another more powerful by which the first can be destroyed.

This claim is easy to understand, though less easy to motivate. The idea is that for any finite, determinate mode, there is another such mode capable of destroying it.5 Set aside concerns about the motivation for this thesis. Whatever else follows from it, surely this does: something can destroy something else. And this, in conjunction with the first two claims, is all that is required to generate the contradiction. If something can destroy something else, then according to IIIp5, those two things cannot be in the same subject. Yet according to Ip15, they must be. So it seems that this is indeed a problem for Spinoza, and not merely a superficial gloss on his claims.6 The straightforward approach to resolving the contradiction is simply to reject one of the three claims. Some of the claims seem more important than others, and so one might hope to reject one of them while preserving the spirit of the Ethics. I think this cannot be done. Each of the three 3

claims turns out to be important for a different part of Spinoza’s system; none are idle wheels in the machinery of the Ethics. Consider the first claim. This claim is among the most fundamental parts of the metaphysical picture presented in the Ethics, as it derives from Spinoza’s substance monism. Indeed, the vision of all things being “in God ” that Spinoza expresses at Ip15 is what he uses to justify his claim that all particular things are merely “affections of God’s attributes” (Ip25c). And this metaphysical picture is the basis for most of the rest of the Ethics—a bit of Spinoza’s account of psychology does not rely upon the picture, but most everything else does. The first claim is therefore central to his philosophical program. Consider instead the second claim. It is cited only a few other places in the Ethics, and though these propositions are in turn used later in the text, it seems that they can each be demonstrated without appeal to the second claim.7 However, Spinoza also appeals to it in the demonstration of IIIp6, the lynchpin of his conatus doctrine. This claim, in turn, is the basis for much of Spinoza’s ethical theory. For example, consider some of the most characteristic propositions of book IV: The more each one strives, and is able, to seek his own advantage....the more he is endowed with virtue (IVp20); Knowledge of God is the mind’s greatest good (IVp28); The good which everyone who seeks virtue wants for himself, he also desires for other men (IVp37).

Spinoza’s demonstrations for each of these ethical claims rely upon the conatus, either directly or indirectly. The second claim plays an important role in generating all that is characteristically ethical about Spinoza’s Ethics.8 So it remains only to consider the last claim of the inconsistent triad, and the most straightforward and innocuous-looking of the bunch: something can destroy something else. It goes without saying that this claim has the strongest intuitive pull of the three. Beyond intuitions, though, there are reasons that can be marshaled in its defense as well. I have (historically, sad to say) caused a glass to shatter. The glass existed before I shattered it, and it ceased to exist after I

4

shattered it; I destroyed that glass. Hence something has destroyed something else. Hence something can destroy something else. Even setting such examples aside—as indeed Spinoza might, given that they involve so many ideas of the imagination—it turns out that at least eight propositions in Ethics IV depend upon its axiom, several of which are vital to the practical therapeutic program Spinoza wants the Ethics to include.9 The key use of the axiom is to derive the claim, “An affect cannot be restrained or taken away except by an affect opposite to, and stronger than, the affect to be restrained” (IVp7).10 This idea is central to Spinoza’s program for overcoming the bondage of the passions, and so IVa1 is (like the other two claims) an important part of Spinoza’s system. Now the bitter choice available to the Spinozist can be surveyed in full. The Ethics appears to contain a contradiction: All things inhere in a unique subject; and no two things that can destroy one another can be in the same subject; yet something can destroy something else. Moreover, each of the three is deeply entwined in Spinoza’s overall philosophical picture. More might be said about each claim’s relative significance, but at that point, we would seem merely to be haggling over the contradiction’s philosophical price.

2 Then again, perhaps the above interpretation of Spinoza’s claims was unfair, somewhere. I assumed that Spinoza’s terms meant the same thing in each of their occurrences, but perhaps the meaning of Spinoza’s terms has shifted in the course of the three claims. In particular, there are several plausible ways of reading Ethics IIIp5, one of the thornier propositions of the conatus argument. Scholarly discussions of the conatus argument have typically resorted to one of two interpretations of IIIp5. In this section, I will discuss the reading proposed by Bennett (1984), then add two more, none of which ultimately work. After dismissing these readings, I will turn to the reading proposed by Garrett (2002).

The mereological reading

5

Debate about IIIp5 has focused upon that troublesome phrase, “in the same subject [in eodem subjecto] ”. The first and least abstract reading we might consider is that “in the same subject” just means “parts of the same whole”. Call this the mereological reading. On this reading, the relation of being “in the same subject” is supposed to be equivalent to mereological underlap. It takes as its relata concrete things, and is true whenever those things are both parts of the same concrete whole.11 The mereological reading fits well with Spinoza’s demonstration of IIIp6, in which he infers from IIIp5 that each thing “is opposed to everything which can take its existence away”, and from this concludes that each thing “strives to persevere in its being.” Since IIIp6 seems to be about concrete individuals, and it is supposed to derive partly from IIIp5, it seems reasonable also to read IIIp5 as about concrete individuals. And since the most obvious way that two individuals can be “in the same subject” is for them to be parts of the same concrete whole, the mereological interpretation is apparently vindicated by these considerations. One problem with the mereological reading is that the argument Spinoza gives for IIIp5 simply does not support it. The demonstration reads, For if [things that can destroy each other] could agree with one another, or be in the same subject at once, then there could be something in the same subject which could destroy it, which (by IIIp4) is absurd. Therefore, things and so on, q.e.d.

On the mereological reading, this argument is really quite bad. It would have to involve the illicit assumption that there is no way for some of a thing’s parts to be destroyed without also destroying the whole. But given the vast array of counterexamples to that principle, it is hard to imagine Spinoza simply assuming it.12 Furthermore, earlier in the Ethics, Spinoza explicitly argues that as long as an individual’s parts are replaced “by others of the same nature”, the individual will persist through the change (L4 after IIp13s). This seems directly contrary to what we would have to assume to get the mereological reading of IIIp5d to work. So if we interpret IIIp5 to be about parts of concrete things, then Spinoza has provided us with no argument for it, and it is difficult to see what argument could be offered on his behalf. A more serious problem with the mereological reading, however, is that it yields a variant of the contradiction described in the previous section. Spinoza takes the universe to be a single

6

“infinite individual” of which all bodies are parts (L7 after IIp13s; cf. Ep. 64). So given the mereological reading, the contradiction simply reappears for material parthood: all bodies are parts of a unique whole; and no bodies that can destroy one another can be parts of the same whole; yet some bodies destroy one another. We have gained no ground.

The modal deflationist reading A second option, and one that might be adopted in conjunction with the mereological reading, is to claim that the modal qualification in IIIp5 is mere window dressing. The proposition reads, “Things are of a contrary nature, that is, cannot be in the same subject, insofar as one can destroy [potest destruere] the other. ” We might well ask, under what circumstances can one thing destroy another? We could simply read “can destroy” in this context to mean “does destroy,” and then the apparent contradiction just boils down to the truism that if one thing destroys another, they don’t both exist afterward. Call this the modal deflationist reading. The contradiction depended on the tension between IIIp5 and Spinoza’s earlier claim that “ Whatever is, is in God ”. That contradiction is resolved on this reading because whenever one object can destroy another, it does destroy the other, and so the latter object ceases to exist. Since the latter object is no longer in the class of “Whatever is,” the two objects are not “in the same subject,” which is consistent with IIIp5. So, unlike the mereological reading, the modal deflationist reading manages to resolve the contradiction that I outlined in section 1. So far, so good. The problem with this reading becomes apparent, however, when we consider applying it to the third claim of the inconsistent triad, IVa1, “Whatever [thing] is given, there is another more powerful by which the first can be destroyed.” If we strip this claim of its modal qualifiers, it becomes monstrously false: for any object, there is another object that (presently) destroys it. Given that at least some objects exist for more than a single moment—and the discussion of composite individuals after Ethics IIp13s assures us that some do—the axiom in question cannot be maintained. An advocate of the modal deflationist reading might here object to extending the reading

7

beyond the context of IIIp5, or at least beyond the context of the conatus argument. Applying the reading to IVa1 is no fair (he might complain) since it is clear in that context that Spinoza intends his modal language to be taken seriously. But this objection is not a strong one: in the absence of further textual evidence, it amounts to the modal deflationist admitting that his reading is ad hoc.13 It saves Spinoza from contradiction, but that is all it does. If there is another reading that saves Spinoza, has a strong grounding in the text, and can be applied without ad hoc restrictions, it will be strongly preferable to the modal deflationist interpretation.

The restriction reading Yet another option is to restrict the extension of IIIp5 to finite things. The term ‘subject’ is strange, for it is not one Spinoza commonly uses in the Ethics. The term is neither defined explicitly in the text nor used frequently enough for its meaning to be inferred from the immediate context. One might reasonably be concerned that Spinoza would not want to count God as a subject in the requisite sense: Without the claim that God is a subject, the contradiction is resolved ab initio. Read this way, IIIp5 simply says that things of a contrary nature cannot be in the same mode, so there is no tension with Ip15. Call this the restriction reading. While this reading does the trick of resolving the contradiction, there is a strong case to be made that Spinoza intends IIIp5 to be fully general and unrestricted in its scope. A philosophical difficulty with the reading is that the demonstration Spinoza provides for IIIp5 does not make any appeal to features peculiar to modes. If the argument he gives is good—and surely Spinoza thought that it was, even if we disagree with his judgment—then Spinoza has given an argument that IIIp5 applies to substances as well as to modes. We would face a brand new problem, on this reading: identifying how the proposition can be restricted to modes alone while the demonstration can be fully general.14 There is a textual difficulty with this reading as well. Spinoza’s philosophical influences and contemporaries commonly spoke of substances [substantiae] being subjects [subjecta], or being like subjects. This is because substances were supposed to support or ground their modes in

8

a manner analogous to the way the subject of a proposition grounds its predicate. A striking example of this occurs in Descartes’ writings, with which Spinoza was thoroughly familiar. In the “ Second Reply”, Descartes complies with Mersenne’s request to lay out the argument of the Meditations in geometrical fashion. He there defines substance as anything in which there “inheres, as in a subject [ut in subjecto]...something we perceive, that is, some property, or quality, or attribute... ” Descartes (1904,Vol. VII, 161, emph. added).15 Spinoza apparently looked to these definitions when he composed his own geometrical rendition of the Principles; he uses almost precisely the same definition of substance (PPC Id5), so it is clear that this usage of the term was familiar to him. Unless there is textual evidence to the contrary, it is most natural to read Spinoza’s use of the term ‘subject’ as having the meaning it commonly did in that period, denoting any thing in which properties, qualities, or attributes inhere. And Spinoza’s substance is certainly such a thing; even if (with Curley) we reject the claim that particular things inhere in God, God nonetheless has properties. These philosophical and textual considerations suggest that Spinoza intended the term ‘subject’ to apply indiscriminately to substances and modes, and that IIIp5 should not be read as restricted to modes. If this is right (as I think it is), the restriction reading fails.16 Having explored a series of plausible, consistent interpretations on which IIIp5 is about concrete objects rather than qualities, each interpretation has been found wanting. We are left with the original conclusion that the Ethics contains three claims that are prima facie inconsistent. Now, however, having delved below the surface, we have reason to think there is no easy way out of the inconsistency. Fortunately for Spinoza, there is a possibility that has been left out of the picture so far.

3 The readings so far have faced an interpretive Scylla and Charybdis: on the one side, our interpretation of IIIp5 must defeat the contradiction outlined in §1; on the other side, it must be able

9

to make sense of Spinoza’s demonstration. The foregoing readings have each failed to resolve one or the other (or both) of these problems. Now, each of the foregoing readings involved conceiving the “things ” of IIIp5 as concrete objects. An alternative option is to take up an entirely different interpretation of the term ‘thing [res]’, with accompanying reinterpretations of the relations of inherence and destruction. Specifically, let’s suppose that the objects of IIIp5 are qualities, and it should be understood as the claim that no two qualities, one of which can destroy the other, can be in the same subject. Call this the quality reading.17 This reading does the best job of making sense of IIIp5 and Spinoza’s demonstration of it, as well as resolving the contradiction described in the previous section. In order to make my case, I first need to explain a bit further how we should understand IIIp5 on this reading. In particular, I need to explain what it means for one quality to be able to destroy another, and what it means for two qualities to be in the same subject. Something like the quality reading is championed by Don Garrett (2002). On Garrett’s interpretation, the tricky phrase “in the same subject” does not concern the way a thing’s parts are in the whole. Rather, it concerns a special sort of metaphysical priority or inherence, as for instance the way modes are in a substance (and, for Spinoza, substance is in itself). The meaning of ‘in’ here is neither containment nor parthood. Instead, it denotes the relation of inherence that binds an instance of a quality to the subject that bears it. This relation of inherence is not itself very clear, though, so it is worth saying a bit more about how the quality reading of “in the same subject” is supposed to work. There is a clause in IIIp5 that I left in the background while discussing previous readings of the passage, but that turns out to be quite helpful on the quality reading. Spinoza draws an equivalence between things that “are of contrary nature” and things that “cannot be in the same subject”. While this equivalence is rather mysterious if we take the things in question to be distinct objects, it makes a great deal more sense if we take the things in question to be distinct qualities. It also suggests a simple way of interpreting the relation of being “in the same subject”—namely as the relation that holds between two qualities when a single thing has both of them. So, for example,

10

the qualities being human and having two arms would bear this relation to each other, because there is something that is human and has two arms. This explains the equivalence Spinoza draws in IIIp5, for it is a straightforward result of the law of noncontradiction that the same subject cannot bear contrary qualities: for any individual x and quality F, it is not the case that both Fx and ¬Fx. (No surprise that Spinoza should think it unnecessary to argue for this!) We can get even clearer about this relation. Garrett argues convincingly that Spinoza is committed to the following inference: if one thing, x, is in another thing, y, then (1) x is conceived through y, and (2) x is caused by y.18 This inference is fairly easy to make sense of if we construe x to be a quality and y to be a subject that bears it. The idea, I think, is that in order to conceive of a quality, one must be able to conceive of a thing that possesses that quality—in order to conceive of redness, one must be able to conceive of a particular red patch, for instance. Furthermore, in such a case, Spinoza has it that the instance of that quality is caused by the thing that possesses it.19 The best way that I can think of to make sense of this inference is in terms of the grounds of the causal power of a quality. An instance of the quality of being in motion might (illicitly) be thought of as affording a body with certain causal powers, such as the power to communicate some degree of motion to adjacent bodies. It would be a mistake, however, to take the quality to be what grounds that power. It is rather the body in which motion inheres that is the ground of that power, and it is in virtue of the fact that the body has the power to communicate some degree of motion (among other powers) that we say it has the quality of being in motion. In this sense, a subject is the cause of its qualities. So on the quality reading, the meaning of IIIp5 is that if two qualities, x and y, are both conceived through or caused by the same subject, x cannot destroy y. Now, in the overwhelming majority of cases, what we call parts are neither conceived through nor caused by their whole. Usually it is the other way around; for instance, we recognize something as a human being in virtue of the fact that he or she has certain parts bearing certain relations to one another—we do not generally recognize something as a human arm in virtue of the fact that it is part of a human being. Hence two parts of the same whole may well not be “in the same subject,” for Spinoza, and this is

11

how the quality reading evades the contradiction raised by the mereological reading. It is perfectly consistent with IIIp5, on the inherence interpretation, for one part of a thing to destroy another part of it. In such cases, Garrett notes, all Spinoza is forced to conclude is that one of the parts was not in (conceived through/caused by) the whole; therefore such cases are also consistent with Spinoza’s claim that there is an infinite individual, Nature. However, Garrett’s work on this passage leaves unclear what it might mean for one quality to be able to destroy another. Fortunately this is a fairly easy blank to fill in. We can understand one quality being able to destroy another in terms of the instantiation of the one precluding the instantiation of the other. In other words, quality F can destroy quality G if and only if whenever something bears F, it cannot bear G. So the qualities of being human and of being a plant destroy one another in this sense, for whenever something is human, it is not a plant, and vice versa. For clarity, I will refer to this notion of destruction as quality destruction; it is to be contrasted with the distinct notion of destruction applicable to concrete things, which I will call object destruction. The quality reading therefore gives us a way to skirt the contradiction of §1, for it allows us to read Spinoza as using two different notions of destruction. On this reading, the predicate “x can destroy y” is ambiguous, but context clarifies which sense is at issue: in the context of IIIp5 it denotes quality destruction, while in the context of IVa1 it denotes object destruction. Spinoza may consistently deny that substance possesses any mutually exclusive qualities and at the same time affirm that for every singular thing, there is something else that can destroy it. The contradiction generated by Ip15, IIIp5, and IVa1 vanishes. Of course, as with other readings that avoid the contradiction, we might be worried that the interpretation involved is ad hoc and ungrounded. However, the quality reading turns out to have a firm grounding in the text. There are three main pieces of evidence in its favor. One. Something that most interpreters of Spinoza gloss over is the peculiar terminology Spinoza uses at IIIp5.20 He is considering things that are “in the same subject [subjecto] ” a term he uses in only two different passages in the Ethics (by my count), and which does not appear in the rest of the propositions involved in the conatus doctrine. As I argued in discussing the

12

restriction reading, it is most natural to read ‘subject’ here as picking out any thing in which properties or qualities inhere. The quality reading therefore provides a similarly natural way of reading the phrase “in the same subject,” namely as denoting the relation of co-inherence or coinstantiation that two qualities bear when they inhere in the same subject. The interpretation I have provided for the phrase “in the same subject” ought, I think, to be the default interpretation; in that regard, it is not ad hoc. Two. I said that the quality reading can avoid the contradiction of §1 by appealing to an ambiguity between two senses of destruction, quality destruction and object destruction. I claimed that Spinoza uses the former sense of destruction in IIIp5 and the latter in IVa1. I think this claim can be motivated by noting that in IIIp5, Spinoza is discussing the destruction of things that are in a subject—qualities, as I have argued, that inhere in either a substance or a mode—while at IVa1 he is discussing the destruction of singular things. For Spinoza, ‘singular thing’ is a technical term meaning a thing “that is finite and has a determinate existence” (IId7). Singular things act, and are acted upon; they are finite agents. Yet the only place besides IIIp5 that Spinoza uses the phrase “in the same subject,” he is not speaking of agents, but of their actions: If two contrary actions are aroused in the same subject, a change will have to occur, either in both of them, or in only one, until they cease to be contrary. (Va1; emph. added)

It seems to me that Spinoza’s use of the term ‘subject’ is in both cases meant to be a cue that the ‘in’ denotes inherence of qualities (or, what is closely related on Spinoza’s view, actions). If that is right, then the sense of destruction at issue in IIIp5 simply cannot be the same as that which is at issue in IVa1. The proposed shift in meaning between the two claims is not ad hoc. Three. Perhaps the strongest evidence for the quality reading is the fact that its interpretations of “in the same subject” and “can destroy” render sound Spinoza’s argument for IIIp5. A very straightforward proof of IIIp5 by contradiction runs as follows. 1.

Some property F can destroy some property G. (Hypothesis)

2.

There is some individual, x, such that Fx and Gx. (Hypothesis)

3.

If Fx, then ~Gx. (1, Def. of quality destruction)

13

4.

Fx. (2)

5.

~Gx. (3, 4)

6. Gx. (2) Since (5) and (6) contradict one another, we must reject one or more of the hypotheses. (1) is true, so (2) is false—there can be no individuals that bear any two properties such that one can destroy the other. More importantly, there is no mystery as to why such an individual would be impossible: its impossibility is due to the logical inconsistency that would be involved in its nature. And this seems to be precisely what Spinoza is driving at in his argument for IIIp5: For if [things that can destroy one another] could agree with one another, or be in the same subject at once, then there could be something in the same subject which could destroy it, which (by IIIp4) is absurd. (IIIp5dem)

I argued earlier that this argument simply makes no sense if we consider the “things ” that are “in the same subject” to be concrete parts of the same whole. Spinoza explicitly acknowledges that such parts can be replaced, and so their destruction need not lead to the destruction of the whole. By contrast, this argument makes perfect sense if we take it to be about mutually exclusive qualities; Spinoza seems to have had in mind something very like the proof by contradiction just provided.21 In short, the quality reading seems to be correct. At first blush, it appears to avoid the contradiction outlined in §1, and this speaks in its favor. The fact that it can make sense of Spinoza’s argument for IIIp5 is also very strong evidence, for that argument appears quite strange on most other readings. Unfortunately, the quality reading blocks only the contradiction between Ip15-IIIp5-IVa1. It does not block a distinct but very similar contradiction that can be generated when we consider Spinoza’s monism. Supposing that the quality reading of IIIp5 is right, it says that two qualities cannot destroy one another to the extent that they are in (conceived through, caused by) the same subject. Then if there is a single subject in which all qualities inhere, it follows that no quality can destroy any other. I argued in the previous section that Spinoza is committed to the antecedent— ” Whatever is, is in God ”—so he seems also to be committed to the consequent. If we interpret his substance-mode terminology most naturally but also most strictly, Spinoza takes particular objects to be qualities of a unique underlying substance.22 I will call an interpretation of Spinoza’s monism 14

strict if, on that interpretation, particular things are qualities or properties predicated of substance. Variations of the strict interpretation are favored by the majority of commentators, including Bennett (1984), Garrett (2002), Nadler (2008), Della Rocca (2008), and Melamed (2009).23 On any such view, the quality reading of IIIp5 presents a serious problem. The problem is this: if all particular things are qualities of substance, then the distinction between quality destruction and object destruction collapses, for every instance of object destruction will be an instance of quality destruction. Once the distinction between the two kinds of destruction collapses, the quality reading of IIIp5 gives rise to a contradiction very similar to what was considered in §1. Specifically, it again follows that nothing can destroy anything else, contrary to both IVa1 and common sense. To resolve this final difficulty, we need to look more closely at the way in which Spinoza thinks particular things relate to substance.

4 The distinction between quality destruction and object destruction was premised upon a distinction between qualities and objects. However, many interpreters have thought that Spinoza does away with this distinction: “Particular things are nothing but affections of God’s attributes, or modes by which God’s attributes are expressed in a certain and determinate way” (Ip25c). While it is no easy task to establish what Spinoza means when he says that particular things are “modes ” or “affections ”, it seems that he takes concrete objects to be qualities, or perhaps instances of qualities. This is the sort of reading that Bennett endorses: As applied to extension, the story goes like this: The one extended substance is the one infinite and eternal spatial item, namely, Space. What we call the existence of bodies in Space is really a qualitative variegation of Space itself. If at a given moment there are bodies of such-and-such sizes and shapes surrounded by empty space, the underlying fact that makes that so is that Space is qualitatively various in certain ways. (Bennett 1994, 16)

On this reading, Spinoza’s substance monism is thoroughgoing: strictly speaking, there are no finite concrete objects; all that exists is the infinite substance and its qualities.

15

The problem I outlined at the end of §3 arises when we combine this reading of Spinoza’s monism with the quality reading of IIIp5. Since I think the quality reading of IIIp5 is right, and I also think something close to Bennett’s reading of Spinoza’s monism is right, I want now to try to respond to that problem. Phrased as a question: If all particular things are qualities of substance, and no two qualities of substance can destroy one another, how can particular things destroy one another?24 The question admits of a number of responses, but only one that fits Spinoza’s philosophy. The answer, I think, is that Spinoza rejects the claim that particular things are in substance insofar as they can destroy one another. To see why, let’s look more closely at the wording of IIIp5: “Things...cannot be in the same subject insofar as [quatenus] one can destroy the other.” I have not made much of Spinoza’s use of the qualification ‘quatenus’, but it is crucial for resolving this final contradiction to which IIIp5 seems to lead. What does Spinoza’s use of ‘quatenus’ add to the claim of IIIp5? Usually that qualification means something like “to the degree that” (as it clearly does in IVp16, 17, and 20, for example). Then IIIp5 would say that two qualities cannot be in the same subject to the degree that one can destroy the other. This reading does not get Spinoza out of trouble, though, for it has the consequence of rendering destruction ungrounded. There are two cases: either x is not in substance to the extent that it can destroy y, or y cannot be in substance to the extent that it can be destroyed by x. Neither option leads to good results. Recall that, for Spinoza, all causal power inheres in substance—he thinks that the fact that all particular things are modes of God implies that “whatever exists expresses...the power of God ” (Ip36dem). So to say that x is not in substance insofar as it can destroy y is just to say that it can destroy y, but not in virtue of any causal power. Alternatively, to say that y is not in substance to the extent that it can be destroyed by x is just to say that the part of y that grounds the destruction relation between x and y—that part in virtue of which it is the case that x destroys y—is not in substance. But if it is not in substance, we cannot appeal to it to ground the destruction relation. In either case, the destruction relation is rendered groundless. Admittedly, the attempt to rescue Spinoza by appeal to degrees of existence could be approached in any number

16

of ways, and there are certainly possibilities I have not considered here. The point is that simply appealing to degrees of existence does not alone provide us with a straightforward solution to the problems facing the quality reading of IIIp5. Occasionally, though, Spinoza uses the qualifier quatenus in a different sense, usually when he is talking about things being in God. When an object has a certain feature insofar as it is in or related to God, Spinoza does not seem to mean that it has that feature to a certain degree. Rather, he seems to mean that there is a certain way of conceiving the object (such as the infamous quadam specie aeternitatis) according to which it has that feature. If this is right, we can take IIIp5 as signifying something quite different than a consideration about degrees of contrariety. Instead, it would say that two qualities cannot be in the same subject according to any way of conceiving of them on which those qualities can destroy one another. The key passages that use this sense of quatenus are found surprisingly early in the Ethics. Of particular importance is the following claim: The modes of each attribute have God for their cause only insofar as [quatenus] he is considered under the attribute of which they are modes, and not insofar as he is considered under any other attribute. (IIp6)

So, according to this proposition, a mode x of a given attribute A has God for its cause only insofar as God is considered under A. It seems obvious to me that this claim is not supposed to mean that God is a partial cause of x, or the cause of x to some limited degree. After all, “God is the efficient cause of all things which can fall under an infinite intellect” (Ip16c1). Rather, it means that x is wholly caused by God, but only on some particular way of conceiving of God’s nature—namely, under the attribute A. Here, then, we have a different model for understanding the qualification “insofar as ”: It is a phrase used to indicate that something has a given feature only on a particular way of conceiving of it. It is worth mentioning some other passages to solidify my case that Spinoza sometimes uses quatenus in the sense I have attributed to him. Elsewhere in Book II, Spinoza asserts: ...so long as singular things do not exist, except insofar as [quatenus] they are comprehended in God ’s attributes, their objective being, or ideas, do not exist except insofar as God’s infinite idea exists. And when singular things are said to exist, not only insofar as they are comprehended in God ’s attributes, but insofar as they are

17

said to have duration, their ideas also involve the existence through which they are said to have duration. (IIp8c)

This passage is part of Spinoza’s attempt to handle the apparent asymmetry between thought and extension that occurs when we have ideas of nonexistent things. His approach is to distinguish between (a) something existing and (b) something existing insofar as it is comprehended through something else. The ideas of nonexistent modes only exist in the latter sense. Again, given the context of this passage, the philosophical work it is intended to do simply cannot be accomplished if “insofar as ” is understood to denote “to the degree that”. The two different notions of existence, (a) and (b), do not stand on either side of a scale of degrees of existence. Rather, they are entirely distinct ways of conceiving of existence.25 Spinoza is saying that even if a mode does not exist in the ordinary, concrete sense, there is nonetheless a way of conceiving of the mode according to which it exists—namely, as part of “the order of the whole of Nature” (IIp7s). Again, something has a given feature insofar as it is is conceived in a certain way, but not as a matter of degree.26 Appealing to this alternative sense of quatenus or “insofar as ” allows us to resolve the final contradiction. The problem, recall, was that on a strict interpretation of Spinoza’s monism, the quality reading of IIIp5 seems to imply that nothing can destroy anything else: All particular things are (instances of) qualities of substance, and no two qualities can be in the same subject “insofar as one can destroy the other. ” The solution suggested by the foregoing discussion of the phrase “insofar as ” is this: Any way of conceiving of two modes according to which one can destroy the other is different than the way of conceiving of those modes according to which they are in substance. Insofar as any two modes are in substance, those modes cannot destroy one another, and insofar as two modes can destroy one another, they are not in substance. When Spinoza uses quatenus in this sense, it is a cue that the claim in question is relative to different metaphysical perspectives, as it were. Although there is only one substance, that substance contains every possible perspective of itself: as of a finite mode, as of an infinite mode or an attribute, or as of the one infinite and eternal thing, God. This is how we can (and should) make sense of the “insofar as ” of IIIp5: Insofar as a quality inheres in God, it is to be conceived not from the perspective of finite modes, but from the perspective of eternity. And from this perspective,

18

nevertheless the quality is not destroyed. A tree’s leaves in autumn take on a beautiful dark red color; as winter comes on, the leaves fall from the tree and decay and are destroyed. The leaves are modes (qualities) of extension; their coloration and various states, qualities of said modes. Spinoza seems pressed to admit that when the leaves decay, certain qualities of extension have been destroyed. This because the leaf-qualities cease to be instantiated in virtue of the fact that extended substance now instantiates other qualities, qualities which preclude the instantiation of leaf-qualities. It is this implication that I think the modifier ïnsofar as ” is intended to resist. Instead, Spinoza need only affirm that from the perspective of the finite, those leaf-qualities have been destroyed. But from that perspective, the leafqualities are not in substance, and so there is no contradiction with IIIp5. When we shift to conceiving the world from the perspective of eternity, reason assures us that the leaf-qualities are still in substance, and that they are eternal and indestructible (or so Spinoza claims). There is no contradiction with IIIp5. Again, this does not mean that there is a certain degree to which the modes can destroy one another, and they are not in substance to that degree. It means that there are at least two ways of conceiving of the modes, one according to which they can destroy one another, and another according to which they cannot. The strange implication that nothing can destroy anything else is thus relativized to a certain way of conceiving the world. Nothing can destroy anything else, according to the way of conceiving the world on which all things are in substance. So we have resolved the last difficulty that the quality reading of IIIp5 seemed to leave for Spinoza. Even on a strict interpretation of Spinoza’s monism, according to which all particular things are qualities of substance, the quality reading of IIIp5 does not lead to a contradiction. Given such an interpretation of Spinoza’s monism, to say that no two qualities can be in the same subject insofar as one can destroy the other is perfectly consistent with, e.g., the claim of IVa1 that each singular thing can be destroyed by some other singular thing. As before, we might worry that this solution is ad hoc, and that aside from the fact that it renders IIIp5 consistent with Spinoza’s monism, there is no reason to adopt the proposed reading.

19

Yet I think that there is textual evidence in its favor as well. We can appeal once more to IIp8 to see the two levels of conception at work in Spinoza’s thinking about destruction. When one singular thing is destroyed by another, it is destroyed only according to our conception of that thing “through which it is said to have duration,” but not insofar as that thing is conceived through an attribute of substance. Similarly, Spinoza explicitly affirms in Ethics V that there are several different ways of conceiving of things, and that objects will have different features according to these different ways of conceiving of them: We conceive things as actual in two ways: either insofar as [quatenus…eatenus] we conceive them to exist in relation to a certain time and place, or insofar as we conceive them to be contained in God and to follow from the necessity of the divine nature. But the things we conceive in this second way as true, or real, we conceive under a species of eternity, and their ideas involve the eternal and infinite essence of God. (Vp29s)

This passage affirms even more than what is suggested by the corollary to IIp8 quoted above. Not only is there a way of conceiving of things as having a kind of eternal formal existence, when we conceive of their essences as contained within the attributes of God. There is also a way of conceiving of things as having eternal actual existence, when we conceive of the actual things as inhering in God. The quoted passage (combined with numerous similar passages in the latter half of Ethics V) is strong evidence for my reading of IIIp5. There is a way of conceiving of actual things as eternal, and that way of conceiving of them presents them as inhering in God. Then when we consider IIIp5, it is no stretch to read it as implying that if we conceive of two modes as being in God, then according to that way of conceiving of them they cannot destroy one another. The contrapositive follows by logic: If we conceive of two modes as being able to destroy one another, then according to that way of conceiving of those modes, they are not in God. And this is the solution I proposed above.

Conclusion I began this paper by noting that Spinoza appears to be committed to three contradictory claims: (1)

20

all things are in the same subject; (2) no things that can destroy one another can be in the same subject; (3) something can destroy something else. I then articulated a promising reading of IIIp5, the quality reading, according to which IIIp5 states that no two qualities that can destroy one another can be in the same subject. This reading avoids the problems that beset the others, though, as I showed at the end of §3, if we assume a strict interpretation of Spinoza’s monism, the quality reading faces a different version of the same formal contradiction. To handle that contradiction, in §4 I outlined a new interpretation of Spinoza’s ubiquitous phrase “insofar as ”. The reading of IIIp5 on which I finally settled was this: Two qualities cannot be in the same subject according to any way of conceiving of them on which those qualities can destroy one another. This reading avoids the contradiction because, as Spinoza suggests in a number of passages, the way of conceiving of two modes according to which one can destroy the other is different than the way of conceiving of those modes according to which they are in substance. Finally, I adduced several passages from the Ethics that motivate this reading of IIIp5. Ethics IIIp5 has often been given lip service or even overlooked entirely in interpretations of Spinoza’s work. In this paper, I have laid out the main difficulties in the interpretation of that proposition—namely, avoiding the contradiction that appears to be generated by Ip15-IIIp5-IVa1 while simultaneously explaining the demonstration of IIIp5. I have also provided an interpretation that resolves all these difficulties: it avoids the contradiction and explains the demonstration. While there may be other interpretations that accomplish this, I provided ample textual motivation for the modified quality reading developed in §4. A surprising result of this interpretation is that it invokes part of Spinoza’s doctrine of the eternity of the mind in order to explain IIIp5, part of the conatus doctrine. If I am right, both doctrines require admitting that one and the same object can have different but objective features according to different ways of conceiving of it—the different levels of conception at work in Spinoza’s argument for the mind’s eternity, on the one hand, and on the other hand, the different levels of conception needed to render IIIp5 consistent. In defending the reading of “insofar as ” that we need for IIIp5, we came to identify these two hierarchies of ways of conceiving. The way of

21

conceiving of two modes according to which they are in God is, on this reading, the same as the way of conceiving them “under a species of eternity” that is so important in Spinoza’s argument for the eternity of the mind. And the way of conceiving of two modes according to which one can destroy another is, on this reading, the same as the way of conceiving them as existing at “a certain time and place”. Two of Spinoza’s doctrines that appear at first only distantly related are thereby shown to depend upon the same premise. A final point about my approach bears mention. It may be surprising that in all of this, I have not made much of the role IIIp5 plays in Spinoza’s conatus argument, IIIp6dem. I have avoided this approach because it seems to me that we should first develop an understanding of the meaning of IIIp5, and then apply that understanding to determine the role of IIIp5 in the conatus argument. This paper is intended to contribute to the former project in the hopes that it will clarify the latter.

Appendix A formal proof of the contradiction can be developed easily if we assume that there is no equivocation among the terms in each of the three claims. Let’s stipulate that our quantifiers will range over only over things both concrete and actual, and ‘ ≤ ‘ means “is in”. Then we have (1)

∃x ∀y (y ≤ x).

I think that this should be an uncontroversial formal representation of Spinoza’s first claim. The idea is that there is one thing, substance, and each body (or physical individual, or extended mode) is in substance. The work of formalizing the second claim is more tricky, but not impossible. Let x DES y stand for “x can destroy y. ” Set aside for now the issue of analyzing the meaning of this difficult notion. Notice that we have buried a modal operator in our simple binary symbolization. The contradiction can be derived without analysis of the modal term ‘can’, which term presents serious interpretational difficulties for readers of Spinoza. The views expressed in Garrett (1991) and

22

Curley & Walski (1999) capture fairly well the variety of available interpretations that I am eliding. I also assume (contrary to the restriction reading discussed in §2) that God is understood as one of the ‘subjects’ quantified over in IIIp5. Then we can formalize the second claim as: 
 (2)

∀x, y, z (x ≤ z ∧y ≤ z → ¬x DES y).

Finally, we need only a very weak implication of the third claim. By ‘singular things,’ Spinoza tells us he means, “things that are finite and have a determinate existence” (IId7). We need not let the meaning of the term slow us down: singular things are things, and hence it certainly follows that: 
 (3)

∃x ∀y (x DES y).

The contradiction is derived as follows. Instantiate with x = i, y = j, giving us 
 (4)

i DES j.

Recall that by (1), we also can infer that there is something, s, such that 
 (5)

i ≤ s ∧ j ≤ s.

Finally, we can instantiate (2) with x = i, y = j and z = s, giving us 
 (6)

(i ≤ s ∧ j ≤ s) → ¬i DES j,

and hence 
 (7)

¬i DES j.

But (4) and (7) are contradictory. Since there were no assumptions beyond the three Spinozist claims formalized as (1), (2), and (3), it is pretty clearly that triad of claims that is responsible for the inconsistency. One concern is that, as I mentioned above, I have buried a modal operator in the relation DES. This does not turn out to have an impact on the derivation of the contradiction. To see this, we can unpack the relation into its modal components. Let D(a, b) mean “a destroys b ” (instead of “...can destroy... ”) Then (2) and (3) become: 


23

(2’)

∀x, y, z (x ≤ z ∧y ≤ z → ¬◊D(x, y)),

(3’)

∃x, y ◊D(x, y).

But (1), (2’) and (3’) imply ◊D(x, y) ∧ ¬◊D(x, y), a contradiction. Another translation of the second claim does seem at first to resolve the contradiction. Instead of (2’), translate the second claim as: 
 (2’’)

∀x, y, z □ (x ≤ z ∧ y ≤ z → ¬D(x, y))

(Read: “Things cannot be in the same subject insofar as one destroys the other. ”) This appears to block the contradiction. Suppose that at the actual world, x and y are in God, and it is possible that x destroys y, but x does not actually destroy y. Then if there is another world where x destroys y, but x and y are not both in God, all three of the claims (1), (2’’), and (3’) will be true. However, if we also refine (1) to reflect our shift to a language including the modal operators, the contradiction reappears. The first claim becomes: 
 (1’)

∃x □ ∀y (y ≤ x).

Then the contradiction is back, for there will be no worlds such that D(x, y) but x and y are not in the same subject. Since (1’) seems to be what Spinoza intended by Ip15, going modal does not get us out of these logical difficulties.27

References Bennett, J. (1984). A Study of Spinoza’s Ethics. Indianapolis: Hackett Publishing. Bennett, J. (1994). “Eight questions about Spinoza.” In Y. Yovel (Ed.) Spinoza on Knowledge and the Human Mind, (pp. 11-26). Leiden: Brill. Carriero, J. P. (1995). “On the relationship between mode and substance in Spinoza’s metaphysics. ” Journal of the History of Philosophy 33, 245-273. Curley, E. (1969). Spinoza’s Metaphysics: An Essay in Interpretation. Cambridge: Harvard

24

University Press. Curley, E. (1988). Behind the Geometrical Method: A Reading of Spinoza’s Ethics. Princeton: Princeton University Press. Curley, E., & Walski, G. (1999). “Spinoza’s necessitarianism reconsidered.” In R. Gennaro, & C. Huenemann (Eds.) New Essays on the Rationalists, (pp. 241-263). New York: Oxford University Press. Della Rocca, M. (2008). Spinoza. New York: Routledge. Descartes, R. (1904). Ouvres de Descartes. Edited by C. Adam and P. Tannery. Paris: Leopold Cerf. Garrett, D. (1991). “Spinoza’s necessitarianism.” In Y. Yovel (Ed.) God and Nature in Spinoza’s Metaphysics, (pp. 191-218). Leiden: Brill. Garrett, D. (2002). “Spinoza’s conatus argument. ” In O. Koistinen, & J. Biro (Eds.) Spinoza: Metaphysical Themes, (pp. 127-159). New York: Oxford University Press. Lebuffe, M. (2010). From Bondage to Freedom: Spinoza on Human Excellence. New York: Oxford University Press. Lin, M. (2004). “Spinoza’s metaphysics of desire.” Archiv für Geschichte der Philosophie 86, 2155. Melamed, Y. Y. (2009). “Spinoza’s metaphysics of substance: The substance-mode relation as a relation of inherence and predication.” Philosophy and Phenomenological Research 78, 17-82. Nadler, S. (2008). “‘Whatever is, is in God’: Substance and things in Spinoza’s metaphysics.” In C. Huenemann (Ed.) Interpreting Spinoza: Critical Essays, (pp. 53-70). Cambridge: Cambridge University Press. Newlands, S. (2010). “Another kind of Spinozistic monism.” Nous 44, 469-502. Spinoza, B. (1882). Spinoza Opera. Edited by C. Gebhardt. Heidelberg: Carl Winters Universitätsverlag. Spinoza, B. (1988). The Collected Works of Spinoza. Translated and edited by E. Curley.

25

Princeton: Princeton University Press. Spinoza, B. (1995). The Letters. Translated and edited by S. Shirley. Indianapolis: Hackett Publishing. Viljanen, V. (2008). “On the derivation and meaning of Spinoza’s Conatus doctrine. ” In D. Garber, & S. Nadler (Eds.) Oxford Studies in Early Modern Philosophy, Vol. 4, (pp. 89112). New York: Oxford University Press. Wilson, M. D. (1999). “Spinoza’s causal axiom (Ethics I Axiom 4). ” In Ideas and Mechanism: Essays on Early Modern Philosophy, (pp. 141-165). Princeton: Princeton University Press. Wolfson, H. A. (1983). The Philosophy of Spinoza: Unfolding the Latent Processes of his Reasoning. Cambridge: Harvard University Press.

Footnotes: 1

Quotations from the Ethics (E) and Descartes’ Principles of Philosophy (PPC) use Curley’s translation in Spinoza (1988), unless otherwise noted. Quotations from Spinoza’s letters (Ep.) are taken from Shirley’s translation in Spinoza (1995). Quotations in Latin are from the Gebhardt edition, Spinoza (1925). 2

See, for instance, Curley (1969, 3-43), Della Rocca (2008, 61-9), Nadler (2008), and Newlands (2010). See also Bennett (1984, 240-2) and Garrett (2002) for discussion of Spinozistic inherence as it pertains to the conatus doctrine. 3

Carriero (1995) provides a useful discussion of the Aristotelian and Cartesian usage of the term ‘subject’. 4

Whether this assumption is reasonable is explored below in §2, “The restriction reading”.

5

Spinoza presents the claim as an axiom, and so it is difficult at first to see what could be his philosophical motivation for endorsing it. Of course, this fits well with his general approach: Spinoza wants readers to swallow the axiom first, only later to see the line of propositions to which it is hooked. Nonetheless, we can provide a few of the reasons for the axiom that he might have had in mind. According to Spinoza’s definition, a finite mode is one “that can be limited by another of the same nature,” a claim he illustrates with the example that “a body is called finite because we always conceive another that is greater” (Id2). While the Ethics does not specify this exactly, it seems plausible to suppose that a finite mode is conceptually limited in every direction, as it were— including duration. In order for a finite mode to be conceptually limited in duration, we need to be able to conceive adequately of the existence of another that can destroy it. But if we can adequately conceive of the existence of a potential destroyer, then (on Spinoza’s view) it in fact exists. And this would give us much of the content of IVa1. However, here is a problem about Ethics IVa1 that might be worth some further consideration. Spinoza seems to be asserting, among other things, that a destroyer is always “more

26

powerful and stronger [potentior et fortior] ” than the thing destroyed. Why should this be the case? If a human is killed by a bullet, in what sense does this imply that the bullet was more powerful and stronger than the human? (It may be that the human in question is the one who crafted the bullet in the first place!) However we decide to treat this problem, it will have broader ramifications for our understanding of Spinoza’s conception of destruction. 6

See the appendix for a formal presentation of the contradiction.

7

Aside from the demonstration of IIIp6, IIIp5 is only cited at IIIp10, IVp7, and IVp30. IIIp10 and IVp30 can both be derived using IIIp4 instead of IIIp5; and IVp7 can be derived from a much weaker version of IIIp5, essentially the law of the excluded middle. 8

It is worth mentioning that not everyone agrees about the role of IIIp5 in the conatus argument, even though Spinoza sees fit to reference it in the demonstration. Spinoza’s demonstration of the conatus doctrine (IIIp6dem) runs as follows: For singular things are modes by which God ’s attributes are expressed in a certain and determinate way (by Ip25c), that is (by Ip34), things that express, in a certain and determinate way, God’s power, by which God is and acts. And no thing has anything in itself by which it can be destroyed, or which takes its existence away (by IIIp4). On the contrary, it is opposed to everything which can take its existence away (by IIIp5). Therefore, as far as it can, and it lies in itself, it strives to persevere in its being, q.e.d.

Now, there are a number of ways to read this argument. On one reading, the argument relies exclusively upon the fact that modes are merely expressions of God’s power. The idea is that since God’s essence is existence, it follows directly that every expression of God’s power strives to persevere in existing—singular things just are finite manifestations of God’s infinite power to exist. On this reading, the reference to IIIp5 is not really necessary to the conatus argument, and hence we could jettison the second claim without worry. Viljanen (2008, 91-94) notes that Alexandre Matheron and Henry Allison both favor this reading. Lin (2004) claims that the pairings of Ip25c/Ip34 and IIIp4/IIIp5 each constitute independent arguments for the conatus. His view would therefore also allow the Spinozist to reject IIIp5 while maintaining the conatus doctrine. Garrett (2002, 145) also argues that IIIp4/5 are not required for the conatus argument, but he proposes that they “contribute to the argument by demonstrating more fully how the conatus doctrine is true.” Yet these readings run against the grain of the text, for Spinoza clearly thinks the cited propositions each add something to the argument. It seems that the derivation of the conatus doctrine requires, at the very least, some further constraints on the way the power of finite things is directed, and these constraints are what Spinoza intended IIIp4 and 5 jointly to provide. I will therefore assume that each of the propositions that Spinoza cites in the conatus argument are ones that are necessary for generating his intended conclusion. 9

E.g., IVp3, IVp7, IVp14, IVp15, IVp16, IVp43, IVp44, possibly IVp59 (I am unconvinced that the demonstration actually requires the use of IVp43, and hence it may not depend upon IVa1), and IVp69. 10

It may at first be unclear how the axiom that every singular thing can be overpowered or destroyed by some other singular thing is related to the restraint of the affects. Spinoza’s view (set out in IVp7d) is that every affect of the mind corresponds to some bodily affection, and every bodily affection is itself a singular thing. Then since every singular thing can be overpowered or destroyed by some other singular thing, the bodily affection can likewise be overpowered or destroyed. And this event manifests in the mind as the restraint of the affect corresponding to that bodily affection. The axiom is therefore needed to provide the justification for Spinoza’s inference that an affect can be overcome only by another that is stronger. I say that this claim is key to the Ethics because it features prominently throughout the remainder of book IV. One of Spinoza’s main aims in book IV, remember, is to establish the

27

difference “between a man who is led only by an affect, or by an opinion, and one who is led by reason ” (IVp66s), and to show us how we can become more like the latter sort of person, the “free man. ” It is to this end that Spinoza prescribes nobility and tenacity, as affects that help us master our passions. For Spinoza acknowledges that our ability to be guided by reason is frequently overcome, explaining that, “A desire arising from a true knowledge of good and evil can be extinguished or restrained by many other desires which arise from the affects by which we are tormented” (IVp15). This is why we need a non-intellectual method for mastering the passions: without it, Spinoza would be forced to give up the Stoic arm of his philosophy. Lebuffe (2010, 9098) notes that Spinoza provides two ways to overcome the passions, only one of which depends upon IVp7. The method which relies upon IVp7 focuses on the general cultivation of affects that overcome one’s passions—countering sadness with happy memories, hatred with fear, and so on. The alternative (and, Lebuffe argues, preferable) method involves forming adequate ideas of particular passions, on the grounds that once a passion is truly understood, it loses a lot of its bite. Although, on Spinoza’s view, this latter method is better and more reliable, the former method is also necessary for escaping the bondage of the passions. Not everyone is capable of forming adequate ideas of the passions, but most everyone can marshal some of their affects to restrain others. This fact is what underwrites Spinoza’s claim, in the Tractatus Theologico-Politicus, that obedience without understanding can still lead to salvation. Furthermore, even very wise people can occasionally be overcome by passions, since their power of acting is always only finite. The method of reordering one’s affects gives the wise person some recourse in this situation. 11

This reading is proposed by Bennett:

On [a mereological] reading [IIIp5] says something to the effect that if x and y are both ‘individuals’...and y can destroy x, then they cannot both be parts of the same individual unless it is much bigger than either of them. The stipulation about relative size is needed: the universe is supposed to be an ‘individual’ in Spinoza’s sense, and p5 must not imply that if x can destroy y then they cannot coexist in the same universe. (Bennett 1984, 242).

The “stipulation about relative size” is a monstrous appendage to IIIp5, which contains no hint of such a restriction (but cf. note 13). To assuage any worries about my use of the term ‘concrete things’, I should note that I take ideas to be concrete things in Spinoza’s ontology—they are actually existing particulars with causal power, and I do not know what other desiderata we might reasonably ask of concrete things. I therefore do not intend to suggest that the mereological reading excludes ideas from the scope of IIIp5. 12

Spinoza does appear to invoke the principle that if some part of a thing is destroyed, so is the whole. As an anonymous commenter observed, he seems to appeal to it at Ip15s in his argument that “corporeal substance, insofar as it is a substance, cannot be divided.” In spite of appearances, however, Spinoza does not in fact an appeal to that principle— which is good, for he would not be entitled to it. By way of argument for the quoted conclusion, Spinoza asks two rhetorical questions: “For if corporeal substance could be so divided that its parts were really distinct, why, then, could one part not be annihilated, the rest remaining connected with one another as before? And why must they all be so fitted together that there is no vacuum?” Now, Spinoza’s reasons for thinking that a vacuum is impossible are conceptual ones. Suppose that a vacuum is a region of extension neither is nor contains corporeal substance. The region nonetheless has certain properties, such as a certain height, width, and depth. But the only kind of thing that can bear such properties is corporeal substance, so (contrary to hypothesis) a vacuum must have substance after all (PPC IIp3). This is what Spinoza means when he calls the vacuum absurd because “Nothing has no properties” (Ep. 13; cf. Ep. 6). When he infers from this that corporeal substance cannot be divided, he does not need to make use of the dubious principle that if some part of a thing is destroyed, the thing itself is destroyed. Instead, he needs (and does) only point to the fact that the divisibility of corporeal substance entails the possibility of a vacuum, which is absurd.

28

13

The advocate of the modal deflationist reading might appeal to the fact that in the context of IIIp5, we are considering things (x and y, say) that are metaphysically proximate, in that they are both in the same subject. When x and y are metaphysically proximate, as in IIIp5, any potential destruction relations between them are activated. Hence modal deflationism applies in the context of IIIp5, but need not apply in the context of IVa1. While this avenue might be worth developing (especially in light of the long and storied debate about Spinoza’s necessitarianism), the contradiction will still arise on the very popular inherence reading of Spinoza, discussed in § 3 below. For if we read “in the same subject” as referring to inherence in the same subject, and if we grant that God is such a subject, it will follow from Ip15 that all things are metaphysically proximate in the sense just defined. My thanks to an anonymous commenter for raising this interesting objection. 14

It is important to note that although IIIp5 is used in proving IIIp6, which does appear to be restricted to modes, it does not thereby follow that the former proposition is intended to apply only to modes. IIIp6 appears to be restricted to modes because its demonstration begins with a claim about “singular things [res singulares] ”, which are finite modes. It is true that this entails that the term ‘subject’ must encompass modes. However, this does not tell against a fully general reading of IIIp5, on which both modes and substances may be subjects. 15

Cf. Descartes’ Comments on a Certain Broadsheet, in which Descartes readily adopts Regius’ use of the term ‘subject’ as synonymous with ‘substance’. Also compare Carriero (1995, 246-53). Although I think Carriero is correct, I am not here opposing the interpretation developed by Curley (1969) & (1988). Curley argues that for Spinoza, unlike for Descartes, modes do not inhere in substance as predicates inhere in subjects. What I am claiming presently is only that when reading Spinoza, the term ‘subject’ should be understood to refer to substances (as well as to modes). 16

My thanks to an anonymous commenter for raising the possibility of the restriction reading of IIIp5. The fact that IIIp5 is used in the demonstration of IIIp6 suggests a connection between the ‘subject’ of IIIp5 and the ‘singular things’ of IIIp6dem, which is important to recognize. 17

It should be emphasized that I use the term “quality” only as a way of distinguishing what I have in mind from Spinoza’s conception of modes. Although I think that these two notions run together for Spinoza, not everyone agrees. Curley (1969), Curley (1988) and Wolfson (1983) both offer interpretations on which Spinoza’s modes are not understood as qualities or properties of substance. In a world where everyone agreed about Spinoza’s conception of modes, I would happily call this interpretation the mode reading of IIIp5. 18

The equivalence of ‘x is in y’ with ‘x is conceived through y’ is made clear in Ip4d. Spinoza there writes: “Whatever is, is either in itself or in another (by Ia1), that is (by Id3 and Id5), outside the intellect there is nothing except substances and their affections.” By aligning the predicate in itself with Id3 (about that which is conceived through itself) and in another with Id5 (about that which is conceived through another), Spinoza is equating in with conceived through. On this notion of inherence, see Garrett (2002, 135) and also Della Rocca (2008, 65-9). 19

Garrett (2002, 136-7) calls this the “Conception Implies Causation” doctrine, and argues that Spinoza’s claim at IIId1 commits him to it. (Garrett also alludes to Ia4 as evidence of Spinoza’s acceptance of this doctrine. However, as Wilson (1999) has argued, that axiom tells us only that causal relations imply conceptual ones, not the other way around.) 20

Cf. the relevant interpretations in Bennett (1984, 240-43), Curley (1988, 108-9).

29

21

One might be dubious that if something possesses two mutually exclusive qualities, then the ensuing contradiction “destroys ” it. But there is good textual evidence that Spinoza was open to this way of thinking about contradictory qualities. At Ip11dem, he explains that: For each thing there must be assigned a cause, or reason, both for its existence and for its nonexistence. ... But this reason, or cause, must either be contained in the nature of the thing, or be outside it. For example, the very nature of a square circle indicates the reason why it does not exist, namely, because it involves a contradiction.

By a fairly trivial extension of Spinoza’s terminology, we can say that something is destroyed if and only if it is caused not to exist. Then (given Spinoza’s version of the principle of sufficient reason) for everything that is destroyed, there is a cause of its destruction. In the case of a square circle, the cause of its destruction is simply that it bears two contradictory qualities, being a square and being a circle. 22

But see Curley (1969) for an alternative interpretation.

23

See also Carriero (1995). Carriero usefully distinguishes between taking particular things to inhere in substance and taking them to be predicated of substance, and argues that Spinoza is only committed to the former. It might seem that he thereby evades the problems caused by IIIp5. Yet I think that the problem with IIIp5 will reappear for his interpretation of Spinoza’s monism as well: IIIp5 speaks only of two things being “in the same subject,” and this ‘in’ will have to be read consistently with the ‘in’ at work earlier in the Ethics, at least in the absence of compelling textual evidence that Spinoza meant something else by it. If, elsewhere in the Ethics, Spinoza uses ‘in’ to denote the predication of particular things to substance, then IIIp5’s ‘in’ denotes predication. If, to the contrary, Spinoza elsewhere uses ‘in’ to denote the inherence of particular things in substance, then IIIp5’s ‘in’ denotes inherence. When we tailor the quality reading to suit the chosen interpretation of ‘in’, the problem will reappear. 24

The question has the roughly the same force even if we think, with Bennett, that Spinoza takes particular things to be instances of qualities of the one substance: If all particular things are instances of qualities of substance, and no two instances of qualities of substance can destroy one another, how are we to account for the destruction of particular things? 25

I am glossing over some subtle distinctions in Spinoza’s terminology—Spinoza appears to use ‘consider [considero]’ and ‘comprehend [comprehendo]’ interchangeably with ‘conceive [concipio]’, though it is difficult to be certain on this score. 26

There are a number of other passages that are best read using this sense of ‘quatenus’: IIp11c, IIp32, IIp34dem, IIp36dem, IIp49, IVp8, and IVp18 are all good candidates. It is debateable which sense of ‘quatenus’ Spinoza intends when discussing adequate and inadequate ideas, but my proposed reading fits well with the distinction between God (conceived as infinite) and God (conceived as the cause of the human mind) that Spinoza makes in IIp9dem and elsewhere. Most other passages where Spinoza uses ‘quatenus’ do seem to have the normal sense of “to the degree that”. 27

This point is owed to —, who rightly suggested that my reticence to explore the modal logic was unfair to Spinoza.

30