1

Spinoza’s Metaphysics Revisited

Edwin Curley

Some years ago I published a book called Spinoza’s Metaphysics [hereafter: SMC],¹ which claimed, first, that Spinoza’s monism, his doctrine that there is only one substance, God, of which everything else is a mode, should be understood as asserting that modes depend causally on substance, and not that they inhere in it, or can be predicated of it; second, that Spinoza was a moderate necessitarian, who believed that everything which happens is in some sense necessary, but not that it is necessary in a sense which would have committed him to the anti-Leibnizian claim that the actual world is the only possible world; and finally, that Spinoza’s monism does not make him a pantheist, not, at least, in what appears to be the most common understanding of that term.

Not everyone, alas, has accepted all my conclusions. I suppose this was to be expected, given the difficulty of the subject. But lately I have come to think that some of my critics have seriously misunderstood the reasoning by which I reached those conclusions, and that this was, to some degree, my fault, for not explaining myself as well as I might have. I now think that if I had set out my argument in a way which came closer to the order by which I first arrived at my conclusions, I might have been better understood, and perhaps more successful in persuading some dissenters. Perhaps. One can always hope, though perhaps one should not hope for too much.

The critic I’ll focus on is Yitzhak Melamed, whose own Spinoza’s Metaphysics [hereafter: SMM]² is the most extensive, and potentially most influential, critique of my book I know. Serious scholars, whose ability I generally respect, have taken his work seriously.3 But it’s also, I regret to say, a critique which exemplifies, and may promote, the most common and most serious misunderstandings of my work.⁴ In this chapter I propose to respond to his criticisms, and then to re-present the main arguments I intended to make, in the different order I wish I had adopted in my book, and adding further arguments which I hope will make the case for my interpretation more persuasive.

I.

First, though, the misunderstandings. Melamed takes my “main argument” [SMM, 40] to be this: Spinoza’s modes are (or at least include) particular things; so to read Spinoza’s claim that modes exist in substance as meaning that modes inhere in, or can be predicated of, substance would imply that particular things can be thought of as predicates of substance; but that would be to attribute a category mistake to Spinoza, and so violate the principle of charity, which requires us not to attribute gross errors to great philosophers; we should, therefore, adopt a different interpretation of the mode-substance relationship, which takes it to be one of causal dependence.

I protest. Not only was this not my main argument, it’s not an argument I ever made, or would now endorse, not if we understand the terms of the argument as Melamed seems to. Take the principle of charity. Whether I would accept that will depend, of course, on what we mean by it. In SMM Melamed doesn’t define that principle, but explains it, first, by appeal to Quine’s discussion of interpreting logical connectives in Word and Object, and then by offering two hypothetical examples of historical interpretations which go astray by applying a principle of charity suggested by Quine’s discussion. Quine had imagined an anthropologist encountering a native who spoke only an entirely unfamiliar language, and accepted sentences which appeared to require a translation which would be explicitly self-contradictory. He suggested that rather than simply dismissing the native as holding an absurd view, we would (and should) “impose our logic” on him, understanding his sentences as not violating the law of contradiction:

The maxim of translation underlying all this is that assertions startlingly false on the face of them are likely to turn on hidden differences of language […]. The common sense behind the maxim is that one’s interlocutor’s silliness, beyond a certain point, is less likely than bad translation.⁵

In his text Quine didn’t call this maxim “the principle of charity,” but in a footnote he invited us to compare it with a principle Neil Wilson had proposed for interpreting statements about individuals, which Wilson called “the principle of charity”: that we should “select as the designatum [of a person’s singular terms] that individual which will make the largest possible number of [his] statements true.”6 So charitable interpretation has come to be understood as a principle which requires us to maximize the agreement between ourselves and others whose utterances we’re trying to understand. So, Davidson:

Charity in interpreting the words and thoughts of others is unavoidable in another direction as well: just as we must maximize agreement, or risk not understanding what the alien is talking about, so we must maximize the self-consistency we attribute to him, on pain of not understanding him.⁷

Davidson’s principle is controversial.⁸ But arguably we do often use something like it. Consider Hugh Hewitt’s interview with Donald Trump in the 2016 campaign, when Hewitt offered a charitable reading of Trump’s claim that President Obama was the founder of ISIS.

HH:		I know what you meant. You meant that he created the vacuum, he lost the peace.
DT:		No, I meant he’s the founder of ISIS. I do. He was the most valuable player. I give him the most valuable player award […].
HH:		But he’s not sympathetic to them. He hates them. He’s trying to kill them.
DT:		I don’t care. He was the founder.9

Hewitt’s reaction here seems a natural one: he’s puzzled by Trump’s words and tries to find an interpretation of them which will make them say something a reasonable man might believe. Defiantly, but shrewdly, Trump resists. He knows outrageous falsehoods get more attention than sober truths. (It seems that they can even help win elections.)

Whatever we may think of charity as a general principle in philosophy of language, it’s clearly dangerous in the history of philosophy. Melamed supplements his paraphrase of Quine by imagining two possible historical applications of the principle. In the first, a student of Aristotle proposes to interpret his statements about the natural rightness of slavery so that they’re consistent with our modern notion that slavery is evil, understanding (the Greek words for) “slave” and “master” as meaning “employee” and “employer.” In the second, a student of Descartes proposes to clear him of the charge of having committed a category mistake by revising our understanding of his position on the mind-body relation. On this reading Descartes comes across sounding remarkably like Ryle in The Concept of Mind.

These are hypothetical examples, which it may be hard to imagine any serious historian endorsing. So I’ve been dismayed to be accused of doing the same thing. I quite agree with Melamed that the principle of charity, understood in this way, encourages anachronistic interpretations, and should be rejected [SMM, 42]. That’s why I rejected it long ago.10 Or to be more precise, that’s why I rejected it in the form I’ve so far been discussing.

In SMM Melamed’s account of the principle of charity is a matter of the interpreter’s maximizing the agreement between himself and another person by interpreting the other person’s words so that they say something he can accept. He objects that the principle of charity, so understood, not only encourages anachronism, but more crucially, deprives us of valuable challenges to “our own fundamental conceptions,” with the implication that we ought to be willing to reconsider those views. In the same year in which SMM appeared, Melamed also published an article on charitable interpretation [hereafter: CI], as allegedly practiced by various commentators on Spinoza.¹¹ Though both the book and the article appeared in the same year, CI seems to represent a later, improved version of his argument.¹² Whereas SMM did not explicitly define the principle Melamed wished to criticize, CI does. There Melamed writes:

The logic of charitable interpretations is rather simple. Suppose a Past Philosopher (PP) makes a statement S. We believe that S, read literally, is clearly unacceptable. Since we appreciate PP as a great mind, we cannot believe that he or she could have uttered such foolishness. Thus, instead of ascribing S to PP, we ascribe S’, which is different from, and sometimes even utterly opposed to S. [CI, 260]

This is broader than the implicit definition in SMM. It doesn’t require the charitable interpreter to impose his own views on the past philosopher. All he must do, on this definition, is to offer a reading of the philosopher which has him saying something different from the “clearly unacceptable” statement which initially puzzled us. Presumably this “something different” will be something not clearly unacceptable (or at least, not so clearly unacceptable as the original statement). But this definition does not require the “something different” to be something the interpreter thinks true (or even very plausible).

The broader definition casts a wider net, permitting Melamed to attack those who reject the narrower version of the principle. But the shift to a broader definition doesn’t prevent him from continuing to criticize his opponents on grounds which presuppose the narrower definition—e.g., as being “philosophical narcissists,” who lazily try to find their own image everywhere [CI, 263]. The shift also makes it possible for him to allege actual textual evidence supporting his claim that his opponents are charitable interpreters, a nicety SMM had neglected. So in CI he cites the following, from an article Greg Walski and I co-authored:

We operate on the methodological principle that views which are tremendously implausible should not be attributed to the great, dead philosophers without pretty strong textual evidence.13

This is essentially the principle I once called “the principle of respect for the intelligence of the great, dead philosophers.”¹⁴ It’s obviously not Melamed’s principle of charity, even on the broader interpretation of that principle. It doesn’t prescribe that the historian substitute a more plausible reading for the one he finds “tremendously implausible.” It doesn’t exclude the possibility that even a very great philosopher may have said something “startlingly false.” All it says is: if you think a great philosopher has said something very implausible, think twice before you accuse him of a gross error. Consider the possibility that you may have misunderstood him. Be sure the texts support your charge.¹⁵

My embrace of the principle of respect was a reaction against the positivists’ wholesale dismissal of metaphysics as nonsense. It still seems to me sensible advice, in all our discourse, though given the difficulties we have understanding one another in philosophy, it’s perhaps especially apposite there. If Melamed wants to call this a version of the principle of charity, fine. I accept that version. But then he might need some argument to show that that version has consequences as unacceptable as the versions I’ve rejected.

Melamed says that in SMC I held that the predicative interpretation of the mode-substance relation is untenable because it ascribes a category mistake to Spinoza. I can’t find that I ever spoke of “category mistakes.” But I did write that:

Spinoza’s modes are, prima facie, of the wrong logical type to be related to substance in the same way Descartes’ modes are related to substance, for they are particular things (E1p25c), not qualities. And it’s difficult to know what it would mean to say that particular things inhere in a substance. When qualities are said to inhere in substance, this may be viewed as a way of saying that they are predicated of it. What it would mean to say that one thing is predicated of another is a mystery that needs solving.¹⁶

This may be enough to justify talk of category mistakes. But note: I said only that Spinoza’s modes are prima facie of the wrong logical type to be predicated of substance, that if Spinoza claims that I can be predicated of God, it’s unclear what that would mean. I didn’t say the mystery was insoluble. In the immediately following paragraph, I pointed out that Bayle had offered us one natural way of getting around the difficulty. In his Dictionary article he assumed that when Spinoza says a finite thing is a mode of God, what he means is just that whatever properties we might have attributed to the finite thing should by rights be attributed to God. As Bayle puts it:

If it were true, as Spinoza maintains, that men are modes of God, we would speak falsely when we said “Peter denies this, he wills that, he affirms such-and-such.” For really, according to this system, it’s God who denies, wills, and affirms. Hence, all the denominations which result from the thoughts of all men, belong properly and physically to God’s substance.17

When we think we’re saying something about a finite subject, we’re really saying something about God. (The same thing, in fact.)¹⁸

Melamed’s own interpretation of the mode-substance relation is essentially a modernized version of Bayle’s, couched in more recent jargon. He argues that Spinoza takes particular things to be bundles of tropes (i.e., particular property instances), so that they “bridge, or even undermine” the distinction between things and properties. They’re both things and properties [SMM, 59]. For all his talk about giving a hearing to voices which challenge our most fundamental conceptions, Melamed doesn’t really want to make Spinoza speak nonsense. He offers an interpretation according to which treating particular things as modes doesn’t involve a category mistake [SMM, 49, 54–57]. We can regard Mt. Rushmore as a property of God because in the end Mt. Rushmore is just a bundle of property instances.

In 1969 I thought (as I still do) that the real problem was not that Spinoza made a category mistake, but that if we make the move Bayle does, to make sense of what Spinoza says, we encounter other, more serious, difficulties, which Bayle pointed out quite forcefully:

(1)  Because different finite beings have contradictory properties—Brutus loved Caesar, Cassius didn’t—if God is the real subject of all predications, then if those propositions are true, God would have to have contradictory properties. [IV, 260–261; Gros, 568–571; Charles-Daubert and Moreau, 65–68; Popkin, 308–311]

(2)  Because some finite beings do wicked things—Cain murdered Abel, Tarquin raped Lucrece—if God is the real subject of all predications, then if those propositions are true, God would be guilty of great wickedness. [IV, 261; Gros, 571–572; Charles-Daubert & Moreau, 68–70; Popkin, 311–312]

And finally,

(3)  Because finite beings are constantly changing, if God is the real subject of all predications, then if the propositions describing those changes are true, God would be constantly changing [IV, 260; Gros, 565–568; Charles-Daubert & Moreau, 63–65; Popkin, 307–308]. Whenever a leaf flutters in the breeze, God changes.

Bayle claimed that these consequences were enough to make Spinoza’s system “the most monstrous hypothesis imaginable, the most absurd and most diametrically opposed to the most evident notions of our mind” [IV, 259; Gros, 527; Charles-Daubert and Moreau, 23; Popkin, 296–297]. Melamed thinks the consequences either don’t follow or are consequences Spinoza would accept.

That has some plausibility regarding the second objection. Arguably, Spinoza has views on good and evil which would allow him to accept, without qualm, our attributing to God acts we might think wicked. On the other hand, some of the consequences Bayle mentions look problematic, no matter what Spinoza thought about good and evil. As Bayle pointed out, transitive sentences in the active voice, describing the action of one thing on another, can be transformed into the passive voice, making the original object of the action the subject of an equivalent sentence. So on this reading, whenever one finite being does something to another finite being, God is doing that to himself. Or (perhaps more precisely), it’s God as the first finite being who’s doing it to God as the second finite being. As Bayle put it:

In Spinoza’s system all those who say The Germans killed ten thousand Turks speak badly and falsely, unless they mean God, modified in Germans, killed God, modified in ten thousand Turks. [Remark N, IV, 261; Gros, 572; Charles-Daubert & Moreau, 69; Popkin, 312]

We need not think such killing wicked to think that a rather bizarre act to ascribe to the Almighty. On this reading, when Tarquin rapes Lucrece, it’s really God (modified as Tarquin?) who is raping God (modified as Lucrece?). For now I leave obscene variations on this theme as an exercise for the reader. I won’t say that the strangeness of these consequences shows that Spinoza could not have accepted them. But I do think we ought to wonder how he could have thought it was consistent with his firm rejection of anthropomorphism19 to say that the things we normally predicate of human beings should instead be predicated of God, and that the things we normally think people do to one another, God does both to others and to himself. Bayle reports that some of Spinoza’s followers complained that he misunderstood Spinoza’s teaching. [See: Remarque DD, 4, 268–270; Gros, 529, 603–612; Charles-Daubert and Moreau, 26, 100–110; Popkin, 303, 329–338.] He says he’s never been able to find a Spinozist who could explain to him where he went wrong. But it doesn’t actually seem that hard to see why they might have thought this. The commitment to anthropomorphism is a fundamental problem with Bayle’s interpretation. I don’t think I saw the issue in these terms in 1969. But now that I have seen it that way, the objection seems to me both clear and fatal.

As for the first objection, Melamed thinks Spinoza can avoid the (unwelcome) consequence of accepting violations of the principle of non-contradiction by saying, not that God simpliciter both loved and did not love Caesar, but that God qua Brutus loved him and that God qua Cassius did not.20 This is similar to the move Bayle makes in his “Spinozistic” analysis of the statement that the Germans killed 10,000 Turks.²¹ In correspondence Melamed explained this language by using the following analogy. Suppose a man is a citizen of two countries—in his example, Poland and Romania. Some things might be true of him qua citizen of Poland (say, he has a right to vote in Polish elections) which are not true of him qua citizen of Romania.

Does this help? Take the sentence “Peter qua Polish citizen can vote in Polish elections.” There the qua locution is the equivalent of a causal subordinate clause. We can paraphrase the sentence salva significatione by saying: “Because Peter is a Polish citizen, he can vote in Polish elections.”²² This use of qua would not avoid the attribution of human qualities to God. From “Peter qua Polish citizen can vote in Polish elections” we can infer “Peter can vote in Polish elections.” So if the analogy holds, “God qua Tarquin raped Lucrece” will entail “God raped Lucrece.” And, I suppose, by Bayle’s reasoning, not only did God have-sexual-intercourse-with-her-against-her-will, he also had-sexual-intercourse-with-himself—and against his will, no less!

However, as Melamed reminds us, we must always be open to challenges to our most fundamental conceptions. Who knows what Spinoza might have thought plausible? So let’s waive this difficulty. It’s still unclear how we’re supposed to apply this analysis to the kind of example Bayle presents as an objection to Spinoza. Consider two people, Joshua and Nicholas, one of whom believes the sun moves around the earth, the other of whom doesn’t. Apparently we’re supposed to avoid ascribing contradictory beliefs to God by understanding this on the following model:

God qua Joshua believes the sun moves around the earth; but God qua Nicholas doesn’t.

What does this mean? If the qua clauses function as they do in the case of Peter’s voting rights, we would understand these sentences along the following lines:

Because God is a Joshua, he believes the sun moves around the earth; but because God is a Nicholas, he doesn’t.

But this is hardly grammatical. Where the qua locution involves a proper name, it looks as though the “is” will have to express identity, not class membership: “Because God is Joshua … ”, “Because God is Nicholas …”. I don’t think Melamed would actually want to say that. But if he did, wouldn’t he be conceding that God, who is identical with each of two beings which have contradictory properties, himself has contradictory properties?

However the first two objections turn out, the third is so blatantly unavoidable that Melamed accepts the consequence, and argues, evidence to the contrary notwithstanding, that Spinoza admitted change in God [SMM, 38–40]. This may not be as historically unlikely as I once thought. Melamed notes that some of the Kabbalists and rabbis of the Talmud thought God was changeable. Clearly a changeable God will be easier to reconcile with the anthropomorphism of the Bible than the immutable God of philosophers like Anselm, Descartes, and Leibniz. But Spinoza’s own references to the Kabbalists and rabbis don’t suggest that he had much respect for their opinions:

I’ve read, and for that matter, known personally, certain Kabbalistic triflers. I’ve never been able to be sufficiently amazed by their madness. [TTP ch. ix, §34/C II 217/G III 135-136]

The rabbis are completely crazy. [TTP ch. ix, §28/C II 216/G III 134]

It seems doubtful that the Kabbalistic or rabbinic precedents would have impressed Spinoza.23 Throughout his work he insists strongly on God’s perfection,²⁴ historically an attribute theistic philosophers have thought a compelling reason to hold that God must be immutable.

In SMC I relied simply on E1p20c2—“God, or [sive] all of God’s attributes, are immutable”²⁵—as evidence that Spinoza would not have accepted the conclusion that God is mutable. It hadn’t occurred to me that anyone would find this evidence ambiguous or insufficient. But Melamed takes the sive clause to introduce a qualification, not an equivalence. On his reading, all Spinoza is saying in E1p20c2 is that God’s essence is immutable, not that God himself is immutable. He grants that this interpretation makes that corollary trivial, since Spinoza would have thought all essences are immutable [SMM, 39].

In principle this is possible. I’ve argued elsewhere that Spinoza’s use of sive is much more complicated than is generally appreciated.²⁶ Sometimes it marks a simple equivalence between the terms it links. Sometimes it introduces what I think Spinoza intends to be a more precise way of putting things. Sometimes it has still other meanings. So it could introduce a qualification here. However, E1p20c2 is only one of a number of passages in which Spinoza seems to identify God with the totality of his attributes.27 So I think we should take sive here as indicating an equivalence, and read that corollary as affirming God’s immutability, not merely the immutability of his essence.

Suppose I’m wrong about that. There are still other passages affirming God’s immutability which don’t offer Melamed any obvious way of discounting them. He could dismiss the Metaphysical Thoughts easily enough, arguing that there Spinoza is just restating a standard Cartesian view.²⁸ But this ground is not available for a passage like E5p20s, which says that clear and distinct knowledge

begets a love towards a thing immutable and eternal (see E5p15), which we really fully possess (E2p45), and which therefore cannot be tainted by any of the vices which exist in ordinary love, but can always be greater and greater (by E5p15), and occupy the greatest part of the mind (by E5p16), and affect it extensively. [E5p20s]

Melamed says this passage asserts immutability only “in relation to Natura naturans.”²⁹ In his view the love Spinoza is commending in E5p20s is just a love of God’s essence, not a love of God. This is a pretty dubious reading of that scholium. The citation of E5p15 in the first line is strong evidence that the immutable and eternal thing which is the object of the love Spinoza’s talking about is God. So is the citation of E5p16 in the last line. So, for that matter, is the statement of E5p20 and its demonstration. Context matters … to most of us.

In SMM Melamed says he’s not aware of any late text which contradicts his conclusion [SMM, 39]. I suppose he intends this to excuse him from discussing the frequent and unequivocal assertions of God’s immutability in the Short Treatise.³⁰ He doesn’t say what he means by “a late text,” but he treats the Ethics as late. Since the primary metaphysical portions of the Ethics—Parts 1 and 2—seem to have been essentially in place by 1665, and not to have been significantly revised after that date,³¹ that would make anything dated 1665 or later a late text. So the Theological-Political Treatise will be a late text.³² It contains two passages which seem to have escaped Melamed’s vigilance:

Since […] the laws of nature extend to infinitely many things, and we conceive them under a certain species of eternity, and nature proceeds according to them in a definite and immutable order, to that extent they indicate to us God’s infinity, eternity and immutability. [TTP, ch. vi, §25/C II 158/G III 86, ll. 14–19]

Similarly, in TTP, ch. vi, §68, Spinoza appeals to the fact that nature observes a fixed and immutable order to show that “God has been the same in all ages,” a teaching Spinoza finds in Ecclesiastes. Later I’ll develop an interpretation of Spinoza which makes this connection between the immutability of the order of nature and the immutability of God highly significant.

Even in the Ethics, there are relevant passages Melamed doesn’t discuss: E1p33s2, where Spinoza argues that it follows from God’s perfection that his decrees could not be or have been different from they are, and that they cannot now change; and E2p10s, where he asserts that “substance is, by its nature, infinite, immutable, indivisible, etc.” Finally, in a letter to Tschirnhaus written in the last year of his life we find Spinoza saying:

From the mere fact that I define God to be a being to whose essence existence pertains, I infer many of his properties: that he exists necessarily, that he is unique, immutable, infinite, etc. [Ep. 83/C II 48/G IV 335]

Spinozan texts don’t get much later than that. Melamed’s consideration of the relevant textual evidence is not as thorough as we might have wished.

Melamed claims, correctly, I think, that “the issue of divine immutability seems to stand or fall with that of pantheism” [SMM, 40]. That is, if we conceive God as immutable, we can’t also conceive of him as being identical with everything there is, since some things change. Conversely, if some things change, and God is everything there is, then God is not immutable. So a lot hangs on the question of divine immutability. Every argument in favor of God’s immutability is an argument against reading God as a pantheist.

Since Melamed denies God’s immutability, he’s obliged to deny also that we can equate God with Natura naturans, and to argue instead that God is both Natura naturans and Natura naturata. I had thought E1p29s was pretty conclusive that Spinoza identified God with Natura naturans:

By Natura naturans we must understand what is in itself and is conceived through itself, or [sive] such attributes of substance as express an eternal and infinite essence, i.e. (by E1p14c1 and E1p17c2), God, insofar as [quatenus] he is considered as a free cause.

But by Natura naturata I understand whatever follows from the necessity of God’s nature, or from any of God’s attributes, i.e., all the modes of God’s attributes insofar as they are considered as things which are in God, and can neither be nor be conceived without God. [E1p29s]

Melamed grants that “at first sight” this scholium seems to endorse my identification of God with Natura naturans, but argues that the quatenus clause qualifies that identification, implying that insofar as God is not considered as a free cause, he’s identical, not with Natura naturans, but with Natura naturata.33 Spinoza doesn’t say this, of course, but Melamed suggests that he refrains from doing so because he finds it “uncomfortable” to describe God as finite or compelled.

I suppose he would find that uncomfortable, if he were ever tempted to say it. But since he claims to have demonstrated only a few pages earlier, without any qualification, that God is a free cause [E1p17c2], it seems unlikely that he would have thought he needed to address the possibility that God is considered as not a free cause. We can account for the presence of the quatenus clause in the first paragraph of E1p29s, without a matching quatenus clause in the second, by noting the ambiguity of quatenus. Usually it indicates scope, and implies some limit to the scope. But sometimes it has causal force. When it’s used in that sense there’s no reason to expect a matching clause of the form “insofar as God is considered as not a free cause.”34

In the parallel passage in the Short Treatise there are no quatenus clauses to muddy the waters and suggest (what Melamed requires) that perhaps Spinoza identifies God with Natura naturans plus Natura naturata. There the identification is clearly just with Natura naturans:

Here, before we proceed to anything else, we shall briefly divide the whole of Nature into Natura naturans and Natura naturata. By Natura naturans we understand a being that we conceive clearly and distinctly through itself, without needing anything other than itself (like all the attributes we have so far described), i.e., God […]. We shall divide Natura naturata in two: a universal and a particular. [KV, I, ch. viii/C I 91/G I 47, ll. 20-29]

He then goes on to identify universal Natura naturata with the infinite modes, and particular Natura naturata with the finite modes.

The passage in the Short Treatise seems to me quite important. It’s one of only two passages in which Spinoza explains what he means by the contrast between Natura naturans and Natura naturata; it unequivocally identifies God with Natura naturans, i.e., with the attributes; and it speaks of God as one division of the whole of nature. If we understand pantheism to involve identifying God with the whole of nature, this passage counts strongly against interpreting Spinoza as a pantheist.³⁵ I don’t think Spinoza was a pantheist in that sense, though I grant that there may be other senses in which we can call him a pantheist. More of this later.

Melamed’s Baylean interpretation of the mode-substance relation seems to be a very natural, even obvious, interpretation. I’ve met students of Spinoza who think that the mere fact that Spinoza says modes “exist in” substance shows that he must think they’re predicable of it. But this begs the question. When we think this interpretation through, we see that it clearly entails consequences Spinoza would never have accepted. Later we’ll see that there’s direct textual evidence against it.

II.

Melamed ascribes to me an argument I didn’t make. That’s unfortunate, but these things happen. What’s more serious, from my point of view, is that he ignores what I considered my main argument. So far I’ve followed him in discussing only negative arguments, reasons why, on my view, it’s problematic to think of the mode-substance relation in Spinoza as one a property has to its subject. I haven’t discussed any arguments which might provide a positive reason for thinking the relation is one of causal dependence. The negative arguments would not be sufficient without some positive argument for the alternative interpretation. Contrary to what my order of presentation in SMC evidently suggested, my central argument was a positive one. I discussed the difficulties of the predicative interpretation mainly to prepare the way for what I hoped would be a sympathetic hearing for an interpretation whose grounds were quite different. That doesn’t seem to have worked out very well. Perhaps the restatement of the negative arguments in the first section of this chapter will succeed where the original statement failed. In any case, this section of the chapter will restate my positive argument, not simply repeating what I said earlier, which would be tedious and pointless, but trying to make that line of argument clearer than I did in 1969, and adding certain refinements, based on things I’ve learned since.

My starting point was a question about Spinoza’s necessitarianism, not about the relation between substance and mode. Reading Hampshire, I found him saying that Spinoza held, against Leibniz, that the actual world is not merely one of many possible worlds, but is, in fact, the only possible world.36 Let’s call this “the anti-Leibnizian proposition.” Spinoza himself never used this language to express his position on necessity. Hampshire was trying to convey, in terms now more familiar, a view Spinoza himself expressed in different language:

In nature there is nothing contingent, but all things have been determined from the necessity of the divine nature to exist and produce an effect in a certain way. [E1p29]

And

Things could have been produced by God in no other way, and in no other order, than they have been produced. [E1p33]

These propositions clearly show Spinoza asserting that everything which happens is necessary. So there’s no question but what Spinoza was in some sense a necessitarian, who held that in some sense every truth is a necessary truth. I’ve never denied this, though some people treat the label “necessitarian” as if it could only apply to those who hold the anti-Leibnizian proposition.37 But there’s nothing here about possible worlds, no claim that only the actual world is possible. These propositions just make claims about God’s causality: that everything which happens was determined by God to happen as it does, and that, his nature being what it is, he couldn’t have acted differently. I will claim for my interpretation the virtue of explaining why Spinoza thought those propositions were true. But first let’s look at the interpretation of his necessitarianism to which I’m opposed.

It seems to have been Leibniz himself who first suggested that Spinoza might be committed to the anti-Leibnizian proposition. This emerges in his discussion in the Théodicée of Bayle’s criticism of Spinoza’s necessitarianism. In his Dictionary article on Chrysippus, Bayle wrote:

Today it’s a great embarrassment for the Spinozists that according to their hypothesis it’s been as impossible from all eternity that Spinoza not die in The Hague as it would be for two and two to make six. They know very well that this is a necessary consequence of their teaching, which shocks people, and puts them off, because it involves an absurdity diametrically opposed to common sense. They’re not very happy for us to know that they’re overturning a maxim as universal, as evident, as this: That whatever implies a contradiction is impossible, and whatever does not imply a contradiction is possible. What contradiction would there have been in supposing that Spinoza died in Leiden?³⁸

So Bayle reads Spinoza as holding that it was as impossible for a particular event, like a person’s death, to have happened differently than it did, as it would be for a mathematical truth to be false. In each case the supposition inherently involves a contradiction. Bayle doesn’t use the concept of a possible world here. But when Leibniz discusses this passage in the Théodicée, he introduces it. After quoting the passage I’ve just quoted, and writing that Bayle’s words are “sufficiently to my liking” [assez à mon gré], he comments that Spinoza

seems to have explicitly taught a blind necessity, when he denied both intellect and will to the author of things, imagining that the good and perfection are related only to us, and not to him. It’s true that there’s something obscure in Spinoza’s thought on this subject. For after having denied God intellect, he gives him thought … Nevertheless, insofar as we can understand him, he does not acknowledge goodness in God, strictly speaking, and teaches that everything exists by the necessity of the divine nature, without God’s making any choice. We won’t waste time here refuting an opinion so wrong and even so inexplicable. [Théodicée §173]

Now so far, so long as we make the omission my ellipsis indicates, I don’t think Spinoza could complain about Leibniz’s exposition of his views. It’s certainly true that, though Spinoza attributes thought to God, he denies him intellect, that he doesn’t acknowledge goodness in God, or represent the existence of this world as the consequence of a personal agent’s having made a wise choice. In that sense Spinoza does “teach a blind necessity.” But in the process of making this criticism, Leibniz makes another, which seems to me much more questionable:

Our own opinion is established by the nature of possibles, that is, of things which don’t imply a contradiction. I don’t believe a Spinozist will say that all the stories we can imagine either really exist now, or have existed [at some time], or will exist yet, somewhere in the universe. Nevertheless, we can’t deny that these stories are possible. [Théodicée §173]

This “refutation” of Spinoza contends that there’s a plurality of possible worlds, where the possibility of a world is a matter of its being conceivable without contradiction. The refutation will work only if Spinoza held that the actual world is the only one conceivable without contradiction. No doubt it’s convenient for Leibniz to imply that Spinoza did hold that, as a way of distinguishing himself from a predecessor he’s anxious not to be associated with. His own doctrine that God was required by his goodness to choose the best of all possible worlds brings him uncomfortably close to the full-bore necessitarianism he accuses Spinoza of.39 But it remains to be shown that Spinoza thought that only the actual world is possible.

Sometimes Leibniz displays reservations about this interpretation. In Theodicée §174 he remarks that Bayle spoiled what he had rightly said when he criticized Spinoza for rejecting the maxim that whatever does not imply a contradiction is possible. And in a sentence I omitted from the first quote, he acknowledges that “there are passages where [Spinoza] softens his position on necessity.” He doesn’t say there what those passages are, but in a recently published text he does:

Even Spinoza is forced in the end to recognize that not everything is absolutely necessary. For he says [in E1p33s1] that “Something is impossible either because its definition involves a contradiction or because there is no determinate cause for producing such a thing.”⁴⁰

Now I would contest Leibniz’s claim that Spinoza was “forced in the end” to recognize that not everything is absolutely necessary, or that passages like this represent a “softening” of his position. The passage he quotes is not one which comes from a work later than the Ethics, in which Spinoza is responding to criticism of that work. It comes from the Ethics itself, and simply explains what Spinoza meant, a few propositions earlier, when he said that there is nothing contingent in nature: some truths are absolutely necessary, because their denial involves a contradiction; others are not absolutely necessary, but necessary only because their denial would be inconsistent with the causal order. E1p33s1 is a “softening” of Spinoza’s position on necessity only if we must interpret E1p29 and E1p33 as involving a commitment to the proposition that everything which happens is absolutely necessary. And so far we have no reason to think that.

Leibniz is right to criticize Bayle for saying that Spinoza held every false proposition to be as self-contradictory as the denial of a mathematical truth. E1p33s1 shows that as plainly as we could wish. (So, for that matter, does E2a1.) Spinoza’s necessitarianism does not require him to hold that there’s an inherent contradiction in supposing that Spinoza died in Leiden. The contradiction is between that proposition and the previous history of the world. But by implying that Spinoza thought only one world is possible, Leibniz planted a seed which has borne much fruit.

When I first wrote about Spinoza’s necessitarianism, I didn’t know this discussion. I hadn’t read Bayle’s article on Chrysippus or Leibniz’s discussion of it in the Théodicée, much less the fragment recently published in the Akademie edition. But even without these texts, I could see that E1p33s1 is difficult to reconcile with the idea that only the actual world is really possible. I didn’t say that the anti-Leibnizian thesis was so shocking to common sense that no competent philosopher could have accepted it. (I was not, after all, a Melamedian charitable interpreter.) What I did say was that if we interpreted Spinoza as holding that the actual world is the only one possible, his view would be open to “very strong objections,”41 that the textual evidence for that interpretation was unclear and that before we ascribed such a paradox to Spinoza, we should try to understand how he might have defended it, if he held it.

Here’s the way I tried to deal with these issues. Suppose we imagine a complete and accurate description of the whole world, including all past, present, and future statements about particulars.⁴² Because these statements describe a reality changing over time, they’ll have to be temporally indexed, if the description is to be consistent. A complete and accurate description of the actual world will include such truths as “From 20 January 2001 to 20 January 2009, George W. Bush was President of the United States” and “From 20 January 2009 to 20 January 2017 Barack Obama was President of the U.S.”

Given this apparatus, we can represent the supposed disagreement between Leibniz and Spinoza in the following way: Leibniz believes in a plurality of possible worlds; this means that, in addition to the description of the actual world, there are other world-descriptions which differ from that of the actual world in one or more of the claims they make, and so would not be accurate descriptions of this world, though those world-descriptions are still logically consistent. In the actual world the Supreme Court settled the election of 2000 by deciding in favor of Bush. But we can conceive a world in which they decided for Gore, and he became President in 2001. If determinism is true, then to be consistently describable such a possible world would have to differ from our world in infinitely many other ways. A determinist will hold that Bush’s election followed, in accordance with the laws of nature, from the prior history of the world.43 So the history of any alternative world leading up to Gore’s election would have to have been different from the history which in our world led up to Bush’s. But on the face of it, there must be many consistently describable alternative worlds. This was Leibniz’s dominant view in the Théodicée. When he represents Spinoza as an opponent who can be refuted by pointing this out, he implies that Spinoza would have disagreed. Wrongly, I argue.

To understand Spinoza’s rejection of contingency, I thought the first move was to understand his determinism, which he seemed to formulate most helpfully in the Preface to Ethics Part 3:

Nothing happens in nature which can be attributed to any defect in it, for nature is always the same, and its virtue and power of acting are everywhere one and the same, i.e., the laws and rules of nature, according to which all things happen, and change from one form to another, are always and everywhere the same. So the way of understanding the nature of anything, of whatever kind, must also be the same, viz. through the universal laws and rules of nature. [E3pr]

Here Spinoza clearly commits himself to the idea that particular events must be explained through the laws of nature, a view he arguably inherited from Descartes.⁴⁴ In the Theological-Political Treatise Spinoza connects this idea with the notion of God’s governance:

By God’s governance [directio] I understand the fixed and immutable order of nature, or [sive] the connection of natural things. For […] the universal laws of nature, according to which all things happen and are determined, are nothing but the eternal decrees of God, which always involve eternal truth and necessity. [TTP ch. iii, §§7-8/C II 112/G III 45-46]⁴⁵

Here the laws of nature function not merely as the proper means for understanding why things happen as they do, but also as the way of understanding how God acts in the world, what the connection is between God and the things he causes. This idea seems to me so crucial to the understanding of Spinoza’s metaphysics that I would say: any interpretation of Spinoza which does not accord a central role to the laws of nature in God’s causation must be wrong. The only questions are: How is that role to be spelled out? And what does it entail?

Here’s a second point, no less crucial than the first. The laws of nature are universal propositions, from which, by themselves, no particular conclusion follows.46 So to understand what happens in the world, we must also understand the particular circumstances in which the laws were operating. We need to be quite clear about this, because it has important consequences. First, it’s a fundamental truth of Aristotelian logic that no particular conclusions follow from universal propositions alone. Unless we’re willing to say that Spinoza was woefully ignorant of traditional logic, we must presume he knew you couldn’t deduce any particular event from the laws of nature alone. (And how, we might ask, could he have successfully practiced the science by which he made his living, if he didn’t know this?) But we don’t have to rely on presumptions. Spinoza acknowledges the logical point in the Treatise on the Emendation of the Intellect:

From universal axioms alone the intellect cannot descend to singulars, since axioms extend to infinity, and do not determine the intellect to the contemplation of one singular thing rather than another. [TIE, §93/C I 39/G II 34, ll. 20–23]

Spinoza’s awareness of this point appears also in his discussion of miracles in the Theological-Political Treatise, where he recognizes two kinds of ignorance which can make people think miracles occur: one is ignorance of the relevant laws, “the principles of natural things” [TTP, ch. vi, §§14–15/C II 155/G III 84, l. 4]; the other is ignorance of the particular circumstances obtaining at the time [TTP, ch. vi, §§45–51/C II 162–164/G III 90–91].

Sometimes Melamed suggests that I attributed this theory of explanation to Spinoza because I wanted to credit him with ideas “in vogue” at the time I was writing [CI, 264], and in general, that I advocated my interpretation because it “bestowed upon Spinoza a certain aura of modernity and philosophical respectability” and made his view “more attractive” to contemporary philosophers [SMM, 12, 42]. That’s what charitable interpreters do. Perhaps if he paid more careful attention to the textual evidence for his opponents’ arguments, he wouldn’t attribute their interpretations to discreditable motives.

Reflecting on the precise form Spinoza’s determinism takes is helpful because it explains why he might hold that some descriptions of alternative worlds which seem consistent are in fact inconsistent. Take the apparently possible world in which Spinoza died in Leiden. “Where’s the contradiction in that?” Bayle wants to know. Well, there’s no contradiction in that proposition considered in itself. But if determinism is true, and everything which happens can be deduced from the laws of nature plus the past history of the world, then if we embed the proposition that Spinoza died in Leiden in a rich enough context, which contains, in addition to the laws of nature, the history of the events which actually led to Spinoza’s death, there will be a contradiction. That rich set of propositions will entail that he died in The Hague, and that will contradict any supposition that he died elsewhere. In imagining a counterfactual history you can’t change just one thing. You must change many. Infinitely many.

So far, then, reflecting on Spinoza’s determinism helps us to see why many alternate worlds which might seem possible aren’t really possible. That gets us some distance to seeing how he might have thought that the actual world is the only one possible. We can get still closer by noting that Spinoza thought the laws of nature are themselves absolutely necessary truths, which follow from the nature or definition of the things they apply to.47 At the time I was writing, this seemed very much a minority view. Some philosophers had defended it, on the ground that the laws of nature support counterfactual inference, which seemed difficult to explain otherwise. But I think Hume’s view that the laws of nature are contingent was dominant, and I would guess that it still is.⁴⁸

Setting aside questions about the truth or popularity of Spinoza’s view, it’s clear that he will think any world in which the laws of nature are different from those in this world is not really possible. That’s why he says, in E1p33, that God could have produced the things he produces in no other way, and in no other order, than they have been produced. I claim it as an advantage of my interpretation that it explains why Spinoza thought E1p33 is true. I have no idea how the predicative interpretation explains this.

If we want to understand how it might have seemed to Spinoza that everything was necessary, this should feel like progress. But if we pursue this line of thought further, we will quickly see that it cannot, in the end, justify the anti-Leibnizian proposition. Suppose we contemplate a world whose description differs from that of the actual world, not only in supposing that Spinoza died in Leiden, but also in having a history which leads up to that conclusion. This will involve an infinite regress, so it’s not a process we could hope to complete ourselves, even if we had vastly more knowledge than we do. But this thought experiment does explain how a mind not subject to our limitations could construct a description of a world different from the actual world and yet consistent.

I’ve generally taken this line of reasoning to be grounds for denying that Spinoza thought the actual world is the only possible world. Perhaps I should have limited myself to a more modest conclusion: that insofar as Spinoza bases his necessitarianism on determinism, and understands determinism as he does, he must allow that there are possible worlds which aren’t actual. Whether he realized this is another question. Perhaps he never actually considered the question in those Leibnizian terms. (I know of no textual evidence that he did.) Perhaps if he had considered the Leibnizian question, he would have insisted that the actual world is the only one possible. But I’d need clearer textual evidence than I’ve seen so far before I drew that conclusion.49

These reflections on Spinoza’s necessitarianism seemed to lead to an interpretation of his ontology which solved certain classic problems about his philosophy. In trying to explain his necessitarianism I had imagined a complete description of the world, which would contain, not only a description of his death in The Hague, and not only descriptions of all prior, concurrent and future events, but also statements of the laws of nature by which the earlier events led to the later ones. This prompted me to wonder: if a complete and accurate description of the world must include propositions of these sorts, what must the world be like? Can we use the logical features of this description as a guide to the logical structure of the world?

Suppose we think that if a proposition is true, it must be true in virtue of some feature of the world it describes. The Metaphysical Thoughts suggests that Spinoza thought that: “an idea is called true when it shows us the thing as it is in itself, and false when it shows us the thing otherwise than it really is.”⁵⁰ So I thought Spinoza held that a true idea will represent its object as being what it is, and doing what it does; and a false idea will represent it as having qualities it does not have or doing things it does not do.

This theory of truth looks like it should have an important ontological consequence: if we’re given a complete and accurate description of the world, then the world must have a corresponding structure, embodying the features required to make the propositions in that description true. We need a term for these truthmaking features. From the logical atomists (early Russell, early Wittgenstein) we can borrow the term “facts.”⁵¹ The description of the world will be a guide to the different kind of facts there are and their relations to one another: the logically different kinds of proposition in the description will be matched by correspondingly different kinds of fact; the logical relations between the different kinds of proposition will be matched by causal relations between the different kinds of fact.

In SMC I proposed using a version of this atomist metaphysics to interpret Spinoza, if only as a heuristic device. I didn’t suppose that Spinoza ever explicitly thought in these categories. But I thought it useful to ask: If he had, how would he have articulated his vision of the world? Suppose he’d accepted the idea that the world is the totality of facts, not the totality of things. How might he have developed that idea further? And what seemed most interesting about this thought experiment: Can entertaining this hypothesis give us any insight into his system which we might not otherwise have had? Can it explain why he said some of the things he said? Or how he might have responded to objections he never actually addressed?52

It seemed to me that on one important point where the atomists disagreed—concerning the existence of general facts—Spinoza would have sided with Russell against Wittgenstein. Russell argued that since general propositions can’t be reduced to conjunctions of singular propositions, we must recognize general facts as part of “the furniture of the universe,” distinct from any combination of particular facts. I thought that given Spinoza’s insistence on the importance of the laws of nature, and given the implications of Wittgenstein’s denial of general facts, Spinoza would have therefore sided with Russell against Wittgenstein. Consider the consequences of siding with Wittgenstein. His view, which denies general facts, also denies that particular facts are causally dependent on one another, and that there is a causal nexus which would justify the inference of one state of affairs from another, and that there can be a necessity for one thing to happen because another has happened.⁵³ Clearly Spinoza would not have accepted these propositions. In these respects he is the anti-Wittgenstein.

Implicit in this ontology is a conception of causality different from those found in Aristotle or Hume. On my account Spinoza thinks that if there’s a relation of logical consequence between the propositions describing certain facts, there must be a causal relation between the facts, and conversely. I knew Spinoza had been criticized for confusing the relation of logical consequence with that of causality.⁵⁴ It’s indisputable that he does use language which implies a close connection between these relations. What I thought questionable was that this way of speaking necessarily involves a confusion.

The conception of causality I hypothesized was very like one Kenneth Clatterbaugh has recently attributed to Descartes: a cause is any proposition which occurs as a premise in a scientific explanation.⁵⁵ For Spinoza, I would amend Clatterbaugh’s account of Descartes: causes are things in the world, not in what we say about the world; so I would not say that a cause is a proposition of some sort, but that a cause of a phenomenon is any feature of reality described by a proposition which occurs in a correct scientific explanation of the phenomenon. This implies that the traditional Aristotelian classification of causes will not fit easily into Spinoza’s system.⁵⁶

I don’t claim that this conception of causality is preferable to more familiar conceptions. But I do think it has certain attractions. It recognizes the causal role, not only of the particular events we normally think of as the causes of other particular events—the spark, say, which ignited the gas coming from the burner on the stove—but also of the background conditions we’re apt to ignore because they are typically present and are not happily thought of as events—like the presence of oxygen in the environment.57 And what is most crucial for our purposes: it assigns a causal role to those general features of reality which are always present, which the laws of nature describe. Without such general facts as that connecting sparks and the presence of oxygen with ignition, events like the spark would not necessitate an event like the ignition. If you think a cause should be something which explains an event (or contributes to its explanation) by showing (helping to show) that it was necessary under the circumstances, you may find this way of thinking of causality preferable to the Aristotelian conception.

Another attraction is that because this conception of causality is broad enough to allow for causal relations between general features of reality, it can accommodate the fact that science doesn’t try to explain only particular events, but also lower level regularities, which it seeks to deduce from more general regularities. Cartesian science certainly attempted this. The first two chapters of Descartes’s Dioptrique provide an example, when they try to explain light rays’ bending in the way they do, as they pass from one medium to another, by showing that it follows from facts about the nature of light (that it’s an action, or inclination to move), combined with facts about the laws governing motion. No doubt Descartes’s explanations would look very crude to a modern physicist. But his attempt does show a kind of ambition I think scientists still have.⁵⁸

Descartes held not only that scientists can deduce less general laws of nature from more general laws, thereby explaining the less general laws, but that in an ideal formulation of science, all the laws of nature would find their place in a deductive system. We find this program expressed metaphorically in the preface he wrote to the French edition of his Principles of Philosophy:

The whole of philosophy is like a tree, whose roots are metaphysics, whose trunk is physics, and whose branches are all the other sciences, which reduce to three principal sciences: medicine, mechanics and morals. [CSM I 186/AT IX-2 14]

In the Principles Descartes tries to make good on this program by deducing the most basic laws of his physics, the laws of motion, from the nature of God, and the laws of the less fundamental sciences from the laws of motion. The first deduction is supposed to depend on God’s immutability, which entails that when he created the world, he could not create a world in which the quantity of motion was not preserved.59 From this conservation principle it’s supposed to follow that in any world God created, bodies would have to obey such laws as the principle of inertia and conservation principles governing the transfer of motion between bodies which impact one another. From these laws of motion, which apply to all bodies, Descartes hoped to deduce the laws of all the more special sciences, such as the laws of reflection and refraction in his Dioptrique.

The idea of a unified science has been popular in recent philosophy of science, so Melamed may have ignored this aspect of my interpretation because he thought it just another deplorable attempt to make Spinoza appear respectable to twentieth-century analytic philosophers. (God forbid that Spinoza might have held any view more recent philosophers would find attractive!) So it may be worth insisting on this connection with Cartesian science, and its clear textual support. It was Stuart Hampshire whose work suggested to me the importance this Cartesian program had for understanding Spinoza. Hampshire wrote:

If we are to provide a complete explanation of the existence and activity of anything in the Universe, we must be able to deduce the existence and activity of the thing studied from the essential attributes and modes of the self-creating God or Nature. This so-called pantheistic doctrine can in fact be fairly represented as the metaphysical expression of the ideal or programme of a unified science, that is, of a completed science which would enable every natural change to be shown as a completely determined effect within a single system of causes; everything must be explicable within a single theory.⁶⁰

I didn’t think Hampshire gave us proper evidence for this idea, or completely realized its potential, but I found it wonderfully suggestive. In many respects my interpretation of Spinoza’s metaphysics was a development of this idea.

That Spinoza agreed with Descartes about the logical structure of science, and that he thought this structure had ontological implications, seemed to me clearest in the Treatise on the Intellect, where he contrasts “singular, changeable things” with “the fixed and eternal things.” He begins with the observation that:

To unite and order all our perceptions, we must ask […] whether there is a certain being—and if so, what sort of being it is—which is the cause of all things, so that its objective essence may also be the cause of all our ideas. Then our mind will […] reproduce nature as much as possible. It will have nature’s essence, order and unity objectively. [TIE, §99/C I 41/G II 36, ll. 7-13]

Some things in the continuation of this passage may be unclear, but one thing is clear in what we have so far: Spinoza is envisaging a deduction which would begin with the idea of God (the cause of all things), and end in ideas of particular things. For our mind to reproduce nature as much as possible, Spinoza says, we must:

Deduce all our ideas […] from real beings, proceeding, as far as possible, according to the series of causes, from one real being to another. [TIE, §99/C I 41/G II 36, ll. 15-17]

He’s emphatic that by “the series of causes” he does not mean “the series of singular changeable things,” but only “the series of fixed and eternal things” [TIE, §100].

What are these fixed and eternal things? My suggestion was that they are a series of general facts which are truthmakers for the laws of nature. Spinoza says they have “laws inscribed in them, as in their true codices.” The Nagelate Schriften translates codices as wetboeken, lawbooks. On my reading that’s an apt metaphor. As a lawbook gives authority to the civil laws inscribed in it, so what grounds the laws of nature are the features of the world which make those laws true.

The first member of this series, on my account, would be a set of general facts constituting one of God’s attributes, and expressing its nature. When Spinoza argues in E2 that we have adequate knowledge of God’s essence [E2p47], this is knowledge, not of the definition of God in E1d6, but of one of God’s attributes,61 a knowledge Spinoza thinks our ideas of the particular things instantiating those attributes presuppose. He tells us more about the attribute of extension than about the attribute of thought, so I concentrate on extension. E2p45 tells us that having the idea of a particular extended thing requires us to have the idea of extension. This seems obvious. E2p46 adds that this idea must be adequate. This seems reasonable, if you think we have an immense number of ideas of extended things, and that inadequacy results only from having an unrepresentative selection of ideas belonging to a certain class.

E2p49 tells us that ideas inherently involve an affirmation or negation. That is, they have a propositional structure.⁶² So I took it that having a clear idea of extension requires understanding certain general propositions about extended things which are implicit in that idea. The concepts of general things involve the laws governing the behavior of the particular things which exemplify the general things. From general propositions about all extended things there would follow, first, general propositions about the properties all extended things have, such as motion and rest, and then propositions about the properties particular kinds of extended things have.

Spinoza could not follow Descartes in this program without making significant modifications to it. His ideal science would not begin with a personal God, creating the world by an act of free will, an unfettered power to do whatever at the moment might please him. That would explain nothing. And it would also create problems about God’s immutability, since it requires an unchangeable God to act differently at one time than he had previously.63 Rather Spinoza thought that the most basic assumptions of the science which would explain the phenomena of physical nature—the axioms explicating the nature of extension—could stand on their own, and did not depend in any way on the will of an omnipotent person. Their truth was inherently necessary and evident. But Spinoza did, I think, agree with Descartes that the various sciences dealing with particular kinds of extended things could be organized into a deductive system.⁶⁴ The foundation of this system would not be any assumption about the immutability of God’s will. Its axioms would be the most general statements we can make about bodies. But as in Descartes, the laws of motion would be crucial to deducing the lower level laws.

TIE §§99–101 projects the possibility of a deductive science of bodies, and on my view the physical excursus following E2p13 gives us a sketch of the way Spinoza at one stage thought that science might begin.⁶⁵ There he purports to deduce the principle of inertia and other physical principles from certain obvious sounding propositions about bodies. Did Spinoza really think the deduction could culminate in propositions describing the existence and changes of each particular thing in the universe, without any other information? TIE §101 concludes with a statement which might seem to say this:

Though these fixed and eternal things are singular, nevertheless, because of their presence everywhere, and most extensive power, they will be to us like universals, or genera of the definitions of singular, changeable things, and the proximate causes of all things. [TIE, §100/C I 41/G II 37, ll. 5-9]

We might infer from this that Spinoza held the fixed and eternal things to be, by themselves, logically sufficient conditions for the existence and activities of singular changeable things. But I think that’s the wrong conclusion to draw. For one thing, it would be inconsistent with what we said previously about Spinoza’s understanding of scientific explanation: that he understood the role laws play in such explanation, but also understood that because laws are universal propositions, they couldn’t be sufficient by themselves to explain particular events. Spinoza does think that to understand how a thing derives from God, we must attend first to the series of fixed and eternal things—or, to the laws of nature “inscribed” in them—and not to the series of singular, changeable things. One reason he gives for this is that the series of singular, changeable things is infinite, and so beyond our grasp, implying, I think, that the series of fixed and eternal things is finite and graspable.⁶⁶ But he also says that because there are “infinitely many circumstances in one and the same thing, any one of which might be a cause of the existence or nonexistence” of the thing we’re trying to explain, we should not try to understand their series. I take this to imply that other singular things are among the causes of the singular thing to be explained, but that it would be difficult for us to identify which singular things ought to be included. So the fixed and eternal things are not by themselves sufficient.

I suspect Spinoza thought that unless we know the relevant laws of nature, we can’t know which of the infinitely many circumstances antecedent to a phenomenon is causally relevant to its occurrence. Consider an example he uses in the Theological-Political Treatise: the ancient Israelites did not know the laws of optics which explain parhelia; so they didn’t appreciate that when ice crystals are present in the atmosphere, the sun’s light may last longer than we would have expected. In their ignorance, they attributed the greater duration of the light at Jericho to God’s having made the sun stand still, not to the presence of those ice crystals.67 Whether Spinoza’s right about that or not, he’s clearly thinking of the unusual duration of the light as explicable only by a deduction which includes, among its premises, both laws of nature and statements of antecedent conditions. The existence and activities of singular changeable things are caused both by the fixed and eternal things and by other singular changeable things. Each of these kinds of cause is necessary. Only jointly are they sufficient.⁶⁸

One reason which led me to propose that the relation between modes and substance was causal rather than predicative was that this picture of the world as consisting of facts of these kinds, causally related to one another in these ways, did not allow for modes to be predicated of substance. If this ontology is correct, predication takes place within modes and attributes, not between them. As noted above, Spinoza’s ideas have a propositional structure, which involves predicating a property of a subject [E2p49]. So their corresponding modes in the attribute of extension must have a similarly complex structure.

Difficulties of the kind Bayle had alleged were certainly a factor in my proposing a causal interpretation. Taking the relation to be predicative seemed to lead to consequences Spinoza would not have accepted. If there were no alternative to the predicative interpretation, we might just have to accept that Spinoza’s system is marred by the kind of incoherence Bayle alleged. But I knew there were alternatives. Sometimes Melamed writes as if, when I offered my interpretation, I was setting myself against the universal opinion of previous scholars:

In order to avoid these absurdities, so skillfully pointed out by Bayle, Curley suggests that we should do away with the traditional, literal interpretation of the substance-mode relation in Spinoza as a relation of inherence [in favor of an interpretation which stresses causal dependence]. [SMM, 9]69

The definite article here is Melamed’s, but the emphasis on it is mine. Where does that article come from? In 1969 it didn’t seem to me that there was a consensus among Spinoza scholars about the interpretation of the definitions of substance and mode. I knew that some previous readers had taken the doctrine that modes “exist in” substance as Bayle had, and that this was a very natural reading. But I knew that others hadn’t, notably Wolfson, who rejected the predicative interpretation quite firmly.⁷⁰ I didn’t think Wolfson’s own alternative was that promising [SMC, 28–36]. But I tried to show that in the Cartesian tradition Spinoza was working in there was a more promising alternative and that he knew it. Sometimes Descartes defines substance as the subject in which properties exist (e.g., in the Geometrical Exposition, Def. V [CSM II 114/AT VII 161]); sometimes he defines it as what exists independently of anything else (e.g., in Principles I, 51[CSM I 210/AT VIII-1 24]). Spinoza was aware of this ambiguity, since he reproduced versions of both definitions when he expounded Descartes geometrically.⁷¹ As far as maintaining consistency with Cartesian usage was concerned, he might have adopted either definition.⁷²

The primary reason I favored a causal understanding of the definition was that the textual evidence seemed to show that Spinoza endorsed:

(i)      a theory of scientific explanation which emphasized that such explanations involve deduction from the laws of nature, conjoined with statements of antecedent conditions;

(ii)     a theory of causality which held that the causes of a feature of reality are those other features of reality which would be described by a correct scientific explanation of the feature;

(iii)    a theory of the ideal nature of science according to which its laws could be organized into a deductive system, whose axioms would be the most general statements we can make about the world, and whose theorems would be the less general laws, with laws of motion playing a crucial role in the deduction of the lower level laws;

and finally,

(iv)    a theory of truth which held that propositions are true just in case they correctly describe the feature of reality they aim to describe.

I don’t claim to know that the propositions I ascribe to Spinoza are true. Probably some of them are false. But I do think they’re propositions an intelligent philosopher in the seventeenth century, convinced of the promise of the new philosophy, might easily have believed.

These commitments—all, I think, well supported by considerations of text and context—suggested the following correlation between Spinoza’s language and the language of the interpretation:

(1)  God’s attributes have an essential nature explicated by the axioms of the science explaining the behavior of things possessing that attribute. We can identify the essential nature of extension with the most general facts about extended things.

(2)  The infinite modes of an attribute are the general facts which are truthmakers for the theorems of the science explaining the behavior of things possessing the attribute.

(3)  The finite modes of an attribute are the facts which are truthmakers for the propositions describing the behavior of the particular things possessing that attribute.

Items in the first two categories can be identified with the fixed and eternal things of the Treatise on the Intellect. Items in the third category are the singular, changeable things of the Treatise on the Intellect. Items in the first category are the adequate cause of items in the second category. That is, the infinite modes are deducible from the attributes of which they are modes. Items in the first two categories are partial, but not adequate, causes of items in the third category.73

My method was hypothetical in the following sense. Although I believed I had good textual support for commitments (i)–(iv), what moved me to suggest the identifications in (1)–(3) was the thought that if we understood Spinoza in this way, that would enable us to answer certain questions about, or criticisms of, his philosophy.

Consider, for example, the question about what Spinoza is referring to when he talks about God. This is a point on which Melamed and I disagree. He takes “God” to refer to the whole of nature, i.e., Natura naturans plus Natura naturata. This is a common view among Spinoza scholars, usually supported by appeal to Spinoza’s use of the phrase Deus sive Natura, which occurs twice in E4pr, and twice again in the demonstration of E4p4. I, on the other hand, take “God” in Spinoza to refer to Natura naturans, resting my case mainly on E1p29s and KV, I, ch. viii.⁷⁴

Melamed’s way of explaining the reference of “God” has one attraction. If we understand God to be the whole of nature, everything that exists, there can be nothing “outside of” God which could be his cause. If God has a cause, he must be his own cause. We may stumble if we try to explain how the whole of nature could be its own cause in any positive sense. But at least it’s easy to give the notion of God’s being causa sui a negative interpretation: there is, by definition, nothing else the whole of nature could be caused by.

This advantage, though, brings with it a certain disadvantage. It makes the question of God’s existence a little too easy. Who would deny that the whole of nature exists? We may disagree about what kinds of thing comprise the whole of nature, or whether there are infinitely many such things, or only finitely many. But if “the whole of nature” refers to everything which exists, then to say that it exists is to say that everything which exists exists. As Horatio says, “There needs no ghost […] come from the grave to tell us this.”

If, on the other hand, we take “God” to refer to the general features of the universe which are truthmakers for the most fundamental laws of nature, we are at least taking “God” to refer to something about whose existence there might be some dispute. Is it really possible, in principle, to construct a science of extended nature which would have fundamental laws from which all the other laws needed to understand the behavior of extended things could be deduced? Can we hope that that science will have fundamental laws so evident that they need no explanation? The answer might be “yes,” in which case we would have grounds for saying that there is a first cause of all things which is its own cause. This “something” might differ in many ways from the God of traditional philosophical theology, but its being a first cause would mean it had something in common with God, as philosophers have generally conceived God. On the other hand, the answer might be “no.” In that case we would have to give up this way of defending the idea that there is a God.

This way of thinking about Spinoza also, I thought, offered what seemed a pretty clear way of explaining how God might be the cause of all particular finite things. When a leaf flutters in the breeze, this must be explicable by the laws of extended nature and the conditions obtaining antecedently among other finite extended things. On the hypothesis under consideration here, we could, in principle, connect God with this event through a chain of laws which leads down from the most fundamental laws to the lowest level generalizations needed for this case. By contrast, if we take God to be the whole of nature, it’s unclear how that whole causes what happens among its parts. No doubt we can imagine that the various parts of the whole are connected with one another in such a way that a change in any one part can effect changes in all the others. But to say this is not to say that there is a causal relation between the whole and its parts. The causal relation would be between some of the parts and others.

One of the most important advantages of my interpretation, I’ve always thought, has been that it gives a plausible explanation of the infinite modes. Many commentators have wondered why Spinoza thought God must have some modes which are infinite and eternal.75 They’ve also wondered what these modes might be. There are no such entities in Descartes’s ontology. Why did Spinoza think he must make room for them in his? This has been a puzzle ever since Spinoza began circulating the manuscript of the Ethics among his friends.

I thought my interpretation explained why there must be such things, and what in general they might be. It postulates a deductive science of extended things, from whose axioms there will follow a system of laws capable of explaining whatever happens among particular extended things. Spinoza’s commitment to a version of the correspondence theory requires that there be “truthmakers” for the laws which follow from the axioms. Since those laws are strictly universal generalizations, not limited to a particular time or place, but holding everywhere and at all times, since they are (on this conception of laws) necessary truths, and since the facts they describe are thought of as having causal power, it makes sense to say of these general facts what Spinoza says of the fixed and eternal things in the Treatise on the Emendation of the Intellect: that they are present and powerful everywhere, the proximate causes of all things. It also makes sense to say of them what Spinoza says of universal modes in the Short Treatise: that they “have been from all eternity, and will remain to all eternity, immutable.” Given the importance of laws of motion in the new science, it’s not surprising that when Spinoza wants to give an example of an infinite mode in the attribute of extension, the example he repeatedly gives is motion (or motion and rest).⁷⁶ Nor is it surprising that he should beg off from saying more about the infinite modes than he does by claiming that it belongs more properly to a treatise on natural science than to a treatise in philosophy [KV, I, ch. ix, §1].

So far, then, we have this: my interpretation gives a plausible account of what “God” refers to, of why Spinoza thinks of that being as God, and of how God, so understood, could be causally related to the infinitely many things he is supposed to cause. It also explains why God’s nature should cause the existence of infinite modes, and what those infinite modes might be. These are not small matters. But now I turn to one final—and particularly important—problem I thought my interpretation could solve: How are we to explain, in Spinozistic terms, the existence of finite modes?

Why is this a problem? Well, E1p16 says that from the necessity of the divine nature, there must follow:

Infinitely many things in infinitely many modes (i.e., everything which can fall under an infinite intellect). [E1p16]

It has seemed to many readers of this text that if finite things follow with logical necessity from the nature of a being whose existence is itself logically necessary, they ought to share the necessity of their cause. But they manifestly don’t share that necessity. If they did, they wouldn’t be finite in their existence; they would neither come into being nor pass away.77 But they do come into being and pass away.

On my reading, the explanation for this is that the causation of finite modes is more complicated than the causation of the infinite modes. God—or more precisely, one of his attributes—is a logically sufficient condition for each infinite mode he causes. The laws describing the infinite modes follow from the most basic laws of one of God’s attributes without any other propositions being necessary for their deduction. So the general facts which are truthmakers for those laws will have the corresponding relation of causal dependence. Spinoza says that the infinite modes follow from the absolute nature of one of God’s attributes [E1p21 and E1p23]. I take this to mean: the infinite modes follow from the nature of God’s attributes unconditionally, without it being necessary to make any further assumption.⁷⁸ That’s why the infinite modes are infinite and eternal. They are truthmakers for laws which require no temporal or spatial limitation in their statement.

But no attribute of God is by itself a logically sufficient condition for any of its finite modes. The model of explanation we’ve found in Spinoza calls for us to explain finite happenings by appealing, not only to the laws of nature, but also to the antecedent conditions of the happening. So finite modes follow from the relevant infinite modes only because the appropriate finite modes—particular facts existing among the antecedents of the particular facts to be explained, but not existing always and everywhere—are also a part of the story. In the language of the Treatise on the Emendation of the Intellect, the fixed and eternal things are either causa sui or caused ultimately and adequately by something which is causa sui. But the singular, changeable things are caused by the combination of a finite series of fixed and eternal things with an infinite series of other singular, changeable things.

There is, of course, a price to be paid for this solution. The feature of my interpretation which explains how it is that finite modes can be finite, not infinite and eternal—i.e., the claim that the laws explain particular facts only if we are able to bring in information about the antecedent conditions—is also the feature of my interpretation which entails a rejection of extreme necessitarianism. It’s because the explanation of finite facts requires the assumption of other finite facts—i.e., because laws alone, without further assumptions about antecedent conditions, cannot explain any finite thing—that there are possible worlds which are not actual. So if you think, as many interpreters seem to, that Spinoza cannot be a moderate necessitarian, that he must have thought that the actual world is the only possible world, you will want to reject this solution. But if you do take this step, you should be honest about the price you pay for it.

You may also object—and I’m sure some Spinoza scholars will object—that this solution implies a restriction on the principle of sufficient reason which Spinoza could not accept. Spinoza claims that for each thing that exists, there must be a cause or reason for its existence, and that for each thing which does not exist, there must be a cause or reason why it does not exist [E1p11d2]. Does my interpretation satisfy this principle? It does entail that each particular mode—whether infinite or finite—has an adequate causal explanation. It does not entail that the totality of finite modes under a given attribute has a causal explanation. Rather, it rejects that idea as inconsistent with the fundamental idea that no particular conclusion can follow from universal premises alone. In my view we ought to accept this result as an inevitable consequence of the only form of explanation Spinoza clearly supported. But if anyone can show me how Spinoza could get to a stronger conclusion without giving up any of his fundamental commitments, I would like to see how that works.

III.

In 1969 I argued for my interpretation of the substance-mode relation by the kind of indirect reasoning so far described. That reasoning evidently seemed to me so strong that I did not feel the need to cite the more direct textual evidence which I must have known existed. By 1987 I’d written a paper on Bennett’s predicative interpretation which mentioned TIE §92 as a passage which supported my reading. But even then I didn’t give it the prominence I now think I should have, mentioning it only in a footnote.79 In the TIE Spinoza had written:

If a thing is in itself, or [sive] as is commonly said, is the cause of itself, then it must be understood through its essence alone. But if it is not in itself, but requires a cause to exist, then it must be understood through its proximate cause. [TIE, §92/C I 38-39/G II 34, ll. 10–13]

I take this as an explicit endorsement of my interpretation of the definitions of substance and mode. What Spinoza says here is that for something to exist in itself, it must be its own cause; but if something is not in itself, it must be caused by something else. Spinoza could hardly have made it clearer that he understood the substance-mode relation to be, by definition, one of causal dependence. If Melamed discussed this passage, he would no doubt find some way to discount it. After all, it comes from one of those early works we are supposed to disregard in deciding what the later works mean. But so far as I can discover, this is the only passage in which Spinoza explains the language he would later use in the central definitions of the Ethics. And I know of no later passage where he contradicts what he says here. So I really think it’s important evidence of his meaning.

I come, finally, to an objection Melamed must be very fond of, since he repeats it several times. We can call it the “good old theism” objection:

The claim that Napoleon is a mode of God should, according to Curley, amount to nothing over and above the claim that God is the (efficient) cause of Napoleon. Under this interpretation the claim that all things are modes of God appears to be completely innocent (in fact, too innocent) insofar as it ascribes to Spinoza nothing more than the common theistic view that God is the cause of all things. [SMM, 9]

In a later variation on this theme Melamed writes:

If, as Curley suggests, Spinoza takes modes to be just effects of the substance, then … Spinoza turns out to be much closer to good old theism. For many, this may seem to disappointingly flatten Spinoza’s far more bold and interesting position. The price we pay for making Spinoza like us is that it is no longer clear why we should have an interest in Spinoza (we have plenty of ourselves even without Spinoza, and we have plenty of other theists in the seventeenth century). [SMM, 43]80

Now I would not concede that my interpretation actually makes Spinoza like us. Granted, I make use of certain twentieth-century ideas to formulate the position I attribute to Spinoza. But I don’t know of any philosopher in any century who has put these ideas together in quite the way I claim Spinoza did. I would guess that many of “us”—say, those of us who think that ideally the best way to explain what happens in nature is by appealing to its laws—would find some of it attractive. But even those who find that attractive might doubt whether it’s possible to formulate an ideal science which would explain everything by appeal to laws which follow logically from basic laws so evident that we cannot doubt them, or whether, if such a science should turn out to be possible, its existence would entail the existence of general facts which can serve as truthmakers for its laws. I would guess that the jury is still out on these matters.

But the claim that my interpretation makes Spinoza like us is just rhetoric, which assumes facts about what we believe which are not in evidence. Melamed’s more substantive objection is that my interpretation makes it difficult to explain why intelligent contemporary readers like Leibniz found Spinoza’s monism so unorthodox, a reaction which is supposed to be more intelligible on Melamed’s predicative reading. In framing this objection Melamed makes a number of assumptions I wouldn’t accept.81 But I grant that if my interpretation could not explain why someone like Leibniz might have found Spinoza’s system disturbing and unacceptable, that would count against it. My main response to this objection is to its assumption that my interpretation of the substance-mode relation ascribes to Spinoza nothing more than the common theistic view that God is the cause of all things.

Melamed’s insistence on inserting the term “efficient” in his paraphrase of my view is a clear sign that something has gone wrong here. That Spinoza should think of God as being, in some sense, the cause of all things is necessary, I would say, for his conception of God to count as a conception of God. The house of the monotheistic traditions may have many rooms, but theists do generally insist pretty strongly that God must be the cause of the world. When we ask, though, whether Spinoza is saying something bold, and interesting, and different from most theistic views, it actually makes some difference how we think his God causes things. As I thought I had said clearly enough in SMC, the traditional Aristotelian categories are not much help in trying to understand Spinoza’s conception of causality. For Spinoza, I claim, a cause of a phenomenon is a feature of reality described by a proposition which occurs in its scientific explanation, where a scientific explanation is required to subsume the phenomenon under the laws of nature. On that view, God is the cause of all things in the sense that everything which exists or happens in the world can in principle be explained by a system of scientific laws which explicates the most fundamental properties of being, combined with statements about the prior history of the world. God’s causal actions are necessary because the fundamental laws of nature are necessary truths, which transmit their necessity to all the other laws derivable from them. Contrary to Leibniz, the ultimate cause of all things is not a choice made by an intelligent agent, sub specie boni.

Spinoza’s metaphysics is not, then, “good, old theism” in any form Leibniz would have found acceptable. My interpretation provides firm grounds for his complaint: Spinoza does not locate the ultimate cause of all things in the wisdom and goodness of an all-powerful personal being. It’s one thing to say that Spinoza’s understanding of God represents a possible way of conceiving of God: it’s a way of thinking of God not so remote from common monotheistic ways of thinking that Spinoza’s use of the term “God” is a joke worthy of Lewis Carroll. But there’s nothing in my Spinoza for those who hope to be reassured that whatever happens, happens for the best, because it’s part of the plan of an infinitely wise and good creator, who can see the good which will result from allowing a few judiciously chosen evils, a creator who loves us and wishes us well. Spinoza’s God makes no choices designed to bring about any ends. “He” has no ends. And unlike the God of the Hebrew Bible, he is no lawgiver. My interpretation makes it quite understandable that Leibniz should think Spinoza posed a danger for ordinary theism. In SMC I borrowed a phrase from Pollock to characterize Spinoza’s relation to the tradition when I said that his use of the ontological argument was “a crucial step in … his ‘euthanasia of theology’” [SMC, 41, 82]. How clearly do you have to say something to avoid misconstruction?

IV.

If we take “pantheist” to mean what Melamed means by it, Spinoza is no pantheist. He does not identify God with the whole of nature. But “pantheism” is an ambiguous term. Suppose we start with a broader definition: “pantheism” need not mean only the doctrine that God is everything and everything is God. Any philosophical theory which holds that God is immanent in the universe may also count as pantheistic.82 This is more promising. Spinoza does say that “God is the immanent, not the transitive [transiens], cause of all things” [E1p18]. E1p18d suggests that an immanent cause is the cause of an effect which is “not outside” the cause, whereas a transitive cause is one which causes something “outside” itself. Since E1p15 has assured us that whatever is, is in God, it comes as no surprise that Spinoza thinks God cannot be the transitive cause of the things he causes. But this is about all he says in the Ethics on the subject of immanence. Unfortunately, it doesn’t shed much light on God’s causality. I think we can do better than that.

Steve Nadler has argued that I can’t, consistently with my causal interpretation of the “exists-in” relation, give a satisfying sense to the talk of immanent causation. “An immanent cause,” he writes,

is ordinarily understood to be a cause whose effects belong to or are a part of itself (much as the Cartesian mind can be said to be the cause of its own ideas): it is a cause that brings about some state in or within itself. A transitive cause, on the other hand, brings about effects which are ontologically distinct from itself (as the baseball is the cause of the broken window and the heat of the sun is the cause of the melted ice). It might seem that unless we think of the things causally brought about by God as properties or states of God—that is, unless we adopt the inherence interpretation—we will be unable to explain God’s causation of things as immanent causation, as Spinoza demands.83

Nadler criticizes me for saying that God acts on things other than himself.⁸⁴ If I understand the statement that “whatever is, is in God” [E1p15] to mean no more than that all things are caused by God, where those things are understood to be “other than God,” I provide “too thin an understanding of the way […] things are ‘in God’ to support a meaningful sense of immanent causation.”⁸⁵

Now I think that to understand Spinoza we must conceive of God’s modes as “other than” God. After all, they’re defined as things which exist in another, through which they’re conceived, whereas God is defined as a being which exists in itself, and is conceived through itself. So modes are defined by a property inconsistent with the property by which substance is defined. And the “other” in which they are said to exist is substance. Of course modes must be other than, distinct from, substance.

Still, it’s certainly true that, however distinct from substance modes may be, they must also be intimately connected with substance. The best way to see that connection, I think, will be to consider the notions of internal constitution, essence, and power. “Internal constitution” is not a term in Spinoza’s technical vocabulary, but it’s clear from two important passages—the Physical Excursus following E2p13s and the Preface to Short Treatise Part 2—that Spinoza thought each complex body in nature has a certain internal constitution which makes it the thing it is, and determines that so long as it retains (at least roughly) the same internal constitution, it will persist over time as that individual, with the manifest properties we associate with that individual.⁸⁶ This internal constitution consists in the fact that the individual is composed of so and so many simpler bodies, of such and such size, with such and such motions in relation to one another, motions which have such and such degrees of speed. The internal constitution of a thing is what constitutes its essence, in the sense of “essence” defined at E2d2: it is “that which, being given, the thing itself is necessarily given, and which, being taken away, the thing itself is necessarily taken away.” My internal constitution—the particular combination of properties of my component particles which makes me the individual I am—came into being at a certain point in time, and at some later point will pass away. It is vulnerable to assaults from other individuals, as Spinoza says in E4a1:

There is no singular thing 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. [E4a1]

Essences of this sort—which I would suggest we identify with the actual essences referred to in E3p7—have a finite, temporal existence.

But Spinoza does not always use the term “essence” [essentia] to refer to something which has that kind of fragility. For example, in E1p17s he says that “a man is the cause of the existence of another man, but not of his essence, for the latter is an eternal truth” [C I 427/G II 63, ll. 18–20]. So there’s a sense of “essence” in which the essences of things do not come into being and pass away, but are eternal. Now our model would predict that for any individual there will be a law of nature which says that if an individual has that internal constitution, it will have certain vulnerabilities. For example, if I, constituted as I am, ingest more than 300 milligrams of arsenic in a 24-hour period, I will die very soon. Similarly, failing to drink any water over an extended period would also kill me, though it would probably take longer. These are laws of nature, or at least something very like laws, at least empirical generalizations indicative of the operation of laws whose precise formulation would require more knowledge than I have.87 E3pr has told us that anything which happens in nature must be explicable in terms of the laws of nature. If what I’ve argued above is correct, laws do not by themselves explain particular events like my death. The laws are conditional in form. To draw particular conclusions from them we also need information about my internal constitution (and the internal constitutions of the objects I encounter). The laws and the information about my constitution and those of surrounding objects are separately necessary and only jointly sufficient to explain a particular event.

I suggest that this is the sort of thing Spinoza has in mind when he says in E2p8 that the formal essences of singular things, or modes, are contained in God’s attributes. (It’s clear from what he actually says in that passage that he’s having some difficulty articulating his ideas.) I think we should understand the formal essence of a singular thing as the set of laws which govern the interaction of things with its internal constitution with other things which have different internal constitutions. It’s the internal constitution of arsenic which gives it the power to kill me. But that internal constitution would not have those effects if the laws of nature were different than they are (and if my internal constitution were different than it is).

I have linked the concept of the internal constitution of a thing first to the concept of its essence, understood in two different ways, and finally, to the concept of its powers. Let’s focus now on the concept of a thing’s power. In the demonstration of E4p4 Spinoza writes:

The power by which singular things—and consequently [any individual] man—preserve their being is the power itself of God, or Nature (E1p24c), not insofar as that power [God or Nature’s] is infinite, but insofar as it can be explained through the man’s actual essence (by E3p7). So the man’s power, insofar as it is explained through his actual essence, is part of the infinite power of God or Nature, i.e. (by E1p34), of its essence. [E4p4d]

In E3pr Spinoza tells us that it’s the laws of nature which constitute Nature’s power of acting [C I 492/ G II 138, ll. 12–15]. The demonstration of E4p4 adds that Nature’s power of acting is the power of singular things to preserve themselves insofar as their power can be explained through their actual essence. Here’s how I take this.

What gives a body with a certain constitution its power to persist in existence are laws of nature which say that a body so constituted will need other things of a certain sort to sustain itself (certain nutrients, say, like water), will survive threats to its existence from bodies of another sort (certain bacteria, perhaps), and succumb to other bodies constituted in a different way (certain poisons, for example). Whether the body actually does survive will depend partly on the laws of its nature, and partly on the existence and powers of the surrounding bodies it encounters. There is, in some sense, an element of chance in all this, but the outcome is determined by the laws of nature, operating in a particular set of circumstances.

The laws of nature dealing with a particular kind of body Spinoza calls its formal essence. Those laws are deducible from the more general laws of nature, so they can be said to be “contained in” them. We cannot say that a particular individual—say, a particular man—is a part of God [E1p13s]. But we can say that that man’s power, his capacity to affect others, and liability to be affected by them, is part of the power of God or nature [E4p4d]. And this, I think, is enough to show that modes have a very intimate connection with substance.

My Spinoza thinks there is something eternal and immutable in the world, the laws of nature and the general features of reality they describe, which determine how particular things change from one form to another. It’s reasonable to classify him as a pantheist because he thinks those laws and the features they describe are something pervasive in the world, a part of the world, not something separate from it, but a part whose power is felt everywhere in the world. And he thinks of that eternal and immutable element in the world as God. It’s enough like the God of traditional religion—an uncaused cause of all things, the contemplation of which can be a source of joy to us—to justify our using religious language to describe it and to make it understandable that we experience religious emotions when we contemplate it. It’s not just the uncaused first cause of all things, but a being we can love, once we recognize its existence and role in the world.

Notes

This is a revised version of the paper presented at the Colloque Spinoza France États-Unis, a conference held in Paris in June 2016. That version is available in English on the conference website at www.spinozaparis8.com. A longer version will appear in the French-language conference proceedings. The version published here is longer than either of the preceding two.

1    Edwin Curley, Spinoza’s Metaphysics: An Essay in Interpretation (Cambridge, MA: Harvard University Press, 1969).

2    Yitzhak Y. Melamed, Spinoza’s Metaphysics: Substance and Thought (New York: Oxford University Press, 2013).

3    See, for example, the high praise from Don Garrett, Mogens Lærke, Alan Nelson, and Martin Lin on the back cover of the paperback edition of SMM.

4    Melamed is not the only one to display these misunderstandings. In crucial respects he follows Charles Jarrett, “The Concepts of Substance and Mode in Spinoza,” Philosophia 7, no. 1 (1977): 83–105 and John Carriero, “On the Relationship between Mode and Substance in Spinoza’s Metaphysics,” Journal of the History of Philosophy 33, no. 2 (1995): 245–273. But he develops the case more fully, and thereby creates more possibility for error.

5    W. V. O. Quine, Word and Object (Cambridge, MA: MIT Press, 1960), 59.

6    Neil Wilson, “Substances without Substrata,” Review of Metaphysics 12, no. 4 (1959): 521–539, here 532.

7    Donald Davidson, “Truth and Meaning” [1967], in Inquiries into Truth and Interpretation, ed. Donald Davidson (Oxford: Clarendon Press, 1985), 27 and passim. Davidson’s notes make clear Quine’s influence on him, though it’s not clear Quine intended anything as general or as strong as what Davidson argued for.

8    For a useful survey of the controversy, see Jane Heal, “Radical Interpretation,” in A Companion to the Philosophy of Language, ed. Bob Hale and Crispin Wright (Oxford: Blackwell, 1997).

9    For the full context see http://www.hughhewitt.com/donald-trump-makes-return-visit.

10  See Edwin Curley, “Spinoza’s Geometric Method,” Studia Spinozana 2 (1986): 151–168. Melamed might reply that this article just shows that I had changed my mind about the principle of charity between 1969 and 1986, not that he misunderstood me. (As we’ll see, he is wont to do this.) I don’t think this response would survive a scrupulous reading of my book. See in particular: SMC, 78–80; SMC, 100–101; and SMC, 155.

11  Myself among them. See “Charitable Interpretations and the Political Domestication of Spinoza, or, Benedict in the Land of the Secular Imagination,” in Philosophy and Its History, Aims and Methods in the Study of Early Modern Philosophy, ed. Mogens Lærke, Justin E. H. Smith, and Eric Schliesser (New York: Oxford University Press, 2013). Other targets included Jonathan Israel and (of all people!) Jonathan Bennett.

12  Cf. SMM, 40–48, with the corresponding pages in Yitzhak Y. Melamed, “Spinoza’s Metaphysics of Substance: The Substance-Mode Relation as a Relation of Inherence and Predication,” Philosophy and Phenomenological Research 78, no. 1 (2009): 17–82, here 56–62. The conference at which Melamed presented CI was held in Montreal in October 2011. Hence, the conjecture that it represents a later stage of his thought about these issues. To judge by his website he still regards CI as an important statement of principle. See http://philosophy.jhu.edu/directory/yitzhak-melamed.

13  Edwin Curley and Gregory Walski, “Spinoza’s Necessitarianism Reconsidered,” in New Essays on the Rationalists, ed. Rocco J. Gennaro and Charles Huenemann (New York: Oxford University Press, 1999), 242. When the polemical mood is on him, Melamed can be rather free with his paraphrases. His gloss on the passage quoted reads: “Our default attitude should be such that we try to avoid ascribing radical and implausible views to great dead philosophers” [my emphasis]. So a prescription to exercise caution in interpretation becomes a prescription to avoid certain kinds of interpretation. And a prescription to exercise care in attributing “tremendously implausible” views to past philosophers becomes a prescription to avoid attributing “radical and implausible” views to past philosophers. Having written that Spinoza’s position on miracles is more interesting than Hume’s because it is more radical, and more radical positions are inherently more interesting than less radical positions, I was somewhat surprised to be accused of a bias against radical interpretations. Cf. Edwin Curley, “Spinoza on miracles,” in The Proceedings of the First Italian International Congress on Spinoza, ed. Emilia Giancotti (Urbino: Bibliopolis, 1985).

14  In Curley, “Spinoza’s Geometric Method,” 155–156. The name alludes to an article by Michael R. Ayers, “Substance, Reality, and the Great, Dead Philosophers,” American Philosophical Quarterly 7, no. 1 (1970): 38–49.

15  Melamed’s initial comment on this is to suggest that “no view should be attributed to anyone ‘without pretty strong textual evidence’” [CI, 262]. I think this goes too far. When Descartes writes in the First Meditation that he’s found that the senses sometimes deceive him, this familiar observation is unlikely to make us ask: “Could he really believe that?” Nor should it. But when he writes at the beginning of the Discourse that good sense is the most equitably distributed thing in the world, that’s exactly the question we should ask. I’ve discussed the latter example in Edwin Curley, “Dialogues with the Dead,” Synthèse 67, no. 1 (1986): 33–49.

16  Cf. SMC, 18 and SMM, 40. In emphasizing this passage Melamed is following Jarrett and Carriero. But they thought my mistake was not to deploy the distinction between inherence and predication. I was, of course, aware of that distinction (cf. note 3 on SMC, 161, discussing the controversy between Ackrill and Owen). But I thought I could safely disregard it, since I didn’t think it enabled us to avoid Bayle’s critique. If something inheres in a subject—say, some particular piece of grammatical knowledge—then that particular piece of knowledge may not be predicable of the subject. That is (to deploy Owen’s criterion), we cannot say “Pierre is knowledge of how to form the third person plural present of voir” or “Pierre is a knowledge of that grammatical point.” Still, there will always be a closely related predication we can make. For example, “Pierre knows how to form the third person, etc.” Melamed seems to agree that the distinction is no help in this context.

17  Cf. Pierre Bayle, “Spinoza,” Remarque N, in the Dictionnaire historique et critique. Bayle’s complete Dictionnaire is available online in the University of Chicago’s ARTFL project at https://artfl-project.uchicago.edu/content/dictionnaire-de-bayle. The text is that of the 5th edition, 1740. I translate from and make my page references to this edition, but also give references to three useful volumes of selections: Pierre Bayle, Pour une histoire critique de la philosophie, ed. Jean-Michel Gros (Paris: Honoré Champion, 2001); Pierre Bayle, Ecrits sur Spinoza, ed. Françoise Charles-Daubert and Pierre-François Moreau (Paris: Berg International Editeurs, 1983); Pierre Bayle, Historical and Critical Dictionary, ed. Richard Popkin (Indianapolis: Bobbs-Merrill, 1965). This quotation comes from vol. IV, 261, of the fifth edition [= 569–570 in Gros, 67 in Charles-Daubert and Moreau, and 309–310 in Popkin]. Charles-Daubert and Moreau is particularly helpful because it includes passages about Spinoza in articles from the Dictionnaire other than the article on Spinoza as well as passages from works other than the Dictionnaire, including the correspondence.

18  Not all predicative interpretations make this assumption. Bennett’s doesn’t. By positing that the property would be a different property, he avoids some of the difficulties Bayle encounters. See Jonathan Bennett, A Study of Spinoza’s Ethics (Indianapolis, IN: Hackett, 1984), §23. But this has its own disadvantages, as I’ve argued in Edwin Curley, “On Bennett’s Interpretation of Spinoza’s Monism,” in God and Nature in Spinoza’s Metaphysics, ed. Yirmiyahu Yovel (Leiden: Brill, 1991).

19  See, for instance, TTP, ch. ii, §§24–55, or E2p3s, or Ep. 54 and Ep. 56.

20  Here I concentrate on Melamed’s argument in the first paragraph on SMM, 35. Subsequently he proposes an alternative reading of Bayle, which makes the argument depend on Spinoza’s claim that God is indivisible. I’m not persuaded by his interpretation of Bayle, but I won’t pursue that issue here.

21  Steven Nadler made a similar move in Spinoza’s Ethics: An Introduction (Cambridge: Cambridge University Press, 2006), 78.

22  Negations are trickier. “Peter qua Romanian citizen cannot vote in Polish elections” translates into “It’s not the case that Peter’s being a Romanian citizen entitles him to vote in Polish elections.”

23  On Spinoza’s early education in traditional Jewish theology, and his disillusionment with it, see the Preface to his Opera posthuma, written by his friends Jarig Jelles and Lodewijk Meyer. See the photographic reprint edition of Spinoza’s Opera posthuma, ed. Pina Totaro (Macerata: Quodlibet, 2008), 3. Steven Nadler and I are collaborating on a third volume of The Collected Works which will contain a translation of the OP preface.

24  For example, in Ep. 2, where Spinoza derives the definition of God he will use in the Ethics from the definition of God as a supremely perfect being [C I 165/G IV 7–8]; or Ep. 35, where he insists at length on God’s perfection [C II 27/G IV 182]; or E1p33s2 [C I 436–438/G II 74–76]. In correspondence Melamed writes: “Spinoza rejects the definition of God as supremely perfect in Ep. 60.” True. But it’s one thing to say that perfection should not be used to define God, and quite another thing to say that God is not perfect.

25  C I 428–429/G II 65.

26  See C II 610–612.

27  Clearest, I think, is E1p4d; but see also: KV, I, ch. vii, §10; Ep. 2; Ep. 9; E1p10; and E1p19. Cf. SMC, 16–18. Though I hadn’t known Gueroult’s work when I reached this conclusion in SMC, when I did discover it, I was pleased to learn that he had independently reached the same conclusion. See Martial Gueroult, Spinoza 1: Dieu (Paris: Aubier-Montaigne, 1968), 47–56. And on E1p20c2, see Gueroult, Spinoza 1, 300–308, 344. If Melamed wishes to challenge my conclusions on these points, he needs also to deal with Gueroult’s arguments.

28  See CM II, ch. iv. But that’s not actually the line Melamed takes. He apparently thinks that at this stage of his development Spinoza agreed with Descartes that God is immutable, and held that God cannot have any modes [SMM, 38]. Since Spinoza clearly thinks in the Short Treatise that God has modes (e.g., in KV, I, ch. ii, §21–22), this reading of the CM makes Spinoza vacillate on what we might have thought was a pretty fundamental point.

29  SMM, 39n116. Similarly Nadler, Spinoza’s Ethics, 78.

30  For example, in KV, I, ch. i, §2 & KV, I, ch. i, §9; KV, I, ch. ii, §29; KV, I, ch. iv, §3 & KV, I, ch. iv, §7; KV, I, ch. vii, §6 & KV, I, ch. vii, note a; KV, II, ch. xiv, §4; KV, II, ch. xxiii, §2; and KV, II, ch. xxiv, §2. No doubt it’s fair to discount early texts when they contradict later ones. But when they consistently say the same thing as the later texts, the proper conclusion would seem to be that they articulate a fundamental doctrine.

31  See the argument drawn from Akkerman’s work in C I 405–6.

32  In the discussion of a paper Melamed presented at Princeton in May 2017 he granted that the Theological-Political Treatise counts as a late text.

33  SMM, 17–20. Other Spinoza scholars have also made this move. Cf. Nadler, Spinoza’s Ethics, 83.

34  See my discussion of quatenus in C II 612. Sometimes Melamed shows himself to be aware of this ambiguity. See SMM, 53n. But he doesn’t seem to have considered its implications for his argument here.

35  Though his phrasing varies, that’s the way Melamed generally seems to understand pantheism. Cf. SMM, 10, 17, 20, 25, and 47. And this is a common way of understanding pantheism. See Keith Yandell, “Pantheism,” in Routledge Encyclopedia of Philosophy (Online 2017 Edition), ed. Tim Crane. Available online at: https://www.rep.routledge.com/articles/thematic/pantheism/v-1.

36  See Stuart Hampshire, Spinoza (London: Penguin, 1951), 54.

37  For example, Don Garrett, “Spinoza’s Necessitarianism,” in God and Nature Spinoza’s Metaphysics, ed. Yirmiyahu Yovel (Leiden: Brill, 1991), 191–192, and Michael Griffin, in “Necessitarianism in Spinoza and Leibniz,” in Interpreting Spinoza: Critical Essays, ed. Charlie Huenemann (Cambridge: Cambridge University Press, 2008).

38  Bayle, “Chrysippe,” Remarque S, in Dictionnaire [II, 174; Gros, 126; Charles-Daubert and Moreau, 122; not in Popkin]. It’s striking how similar what Bayle says about Spinoza’s necessitarianism is to what he says about his monism. He treats both as manifestly false, contrary to what is most evident to us.

39  As Griffin makes clear in “Necessitarianism in Spinoza and Leibniz.” Arguably Leibniz is more necessitarian than Spinoza. If you think of possible worlds in abstraction from the God who might create them, there are many. But given God’s existence and nature, only one world is really possible.

40  G. W. F. Leibniz, “Ad sententiam Spinozae de necessitate rerum,” Sämtliche Schriften und Briefe, Philosophische Schriften, VI, iv, #338, 1777. When I first wrote on this topic, this fragment had not been published. I’m indebted to Ursula Goldenbaum for calling it to my attention.

41  See SMC, 83–84, where I made an objection similar to Bayle’s, but did not treat it as decisive, just a reason to wonder how Spinoza might have justified his claim that nothing is contingent.

42  I believe I originally got this idea from Carnap’s notion of a state description and not directly from Leibniz. See Rudolf Carnap, Introduction to Semantics (Cambridge, MA: Harvard University Press, 1942), 101ff. But Leibniz was Carnap’s inspiration, and Leibniz’s discussion of possible worlds at the end of the Théodicée [§§405–417], which gives it a clear temporal dimension, seems a better model.

43  This is, at least, one common way of understanding determinism. Cf. Lawrence Sklar’s article on determinism in A Companion to Metaphysics, ed. Jaegwon Kim and Ernest Sosa, 1st edn (London: Blackwell, 1995).

44  See J. R. Milton, “Laws of Nature,” in The Cambridge History of Seventeenth-Century Philosophy, ed. Daniel Garber and Michael Ayers, 2 vols (Cambridge: Cambridge University Press, 2000), vol. 1, 680–701. Cf. SMC, 79. See also Edwin Curley, “Law of Nature,” in Cambridge Descartes Lexicon, ed. Lawrence Nolan (Cambridge: Cambridge University Press, 2016).

45  See also TTP, ch. vi, §§7–12. I’ve slightly modified the translation in C II.

46  See TTP, ch. iv, §§1–2 [C II 125–126/G III 57–58] where Spinoza defines a “law,” understood in the most general sense, as “that according to which each individual, or all, or some members of the same species, act in one and the same fixed and determinate way.” This definition is intended to cover both human laws and the laws of nature. As examples of the latter Spinoza cites universal statements about all bodies or all human beings. But he explicitly allows for generalizations about some members of a species. And I think his definition should also be understood as allowing statements which identify regularities in the behavior of particular individuals to qualify as laws. More on this below.

47  See, for example, TTP, ch. iv, §§1–2, where Spinoza maintains that the laws of nature follow from the nature or definition of a thing.

48  The most notable defender of a necessitarian view, I think, was William Kneale, Probability and Induction (Oxford: Clarendon Press, 1949). For a mid-century defense of the Humean view which gives a good statement of the arguments then made in its favor, see A. J. Ayer, “What Is a Law of Nature?,” in The Concept of a Person: And Other Essays (New York: St. Martin’s Press, 1963). I believe necessitarian views of laws are more common now. But writing as recently as 1977, Parkinson felt able to dismiss Spinoza’s critique of miracles as depending on a “widely rejected” necessitarian view. See G. H. R. Parkinson, “Spinoza on miracles and natural law,” Revue internationale de philosophie 31, no. 119/120 (1977): 145–157.

49  The best suggestion I’ve seen is Garrett’s: that Spinoza would have thought possible worlds which are not actual were excluded by the principle that whatever series of finite modes exists must express the highest degree of reality and perfection. See Garrett, “Spinoza’s Necessitarianism,” 197. Garrett does not claim that Spinoza ever states this principle, but argues that the connection Spinoza makes between an entity’s degree of reality or perfection and its existence [E1p9 and E1p11] makes it likely that he would have accepted it. I have some difficulty understanding how this constraint is supposed to operate. On a typical theistic view, like Leibniz’s, where an omnipotent personal agent, God, created the world at a particular time in the past, by an act of will which had no preconditions (ex nihilo), then it’s easy to see how God’s choice might have been constrained by his preference for a particular kind of world. But that’s typical theism, not Spinoza. If God’s causality operates in the way I have claimed that it does, that is, if his governance consists in the operation of universal laws of nature, where the causal effectiveness of laws presupposes the prior existence of certain conditions, then it’s difficult to see how Garrett’s principle could determine what happens.

50  CM I, ch. vi [C I 312/G I 246]. If the Metaphysical Thoughts seems a suspicious source to cite, on the ground that it’s just an exposition of Scholastic and Cartesian ideas, we can reply that the Ethics expresses a similar view in E1a6.

51  See Bertrand Russell, The Philosophy of Logical Atomism [1919] (New York: Routledge, 2010) or Ludwig Wittgenstein, Tractatus Logico-Philosophicus [1921], trans. D. F. Pears and B. F. McGuinness, 2nd edn (New York: Routledge, 2014). Recently the term “truthmaker” has gained some currency. This seems to be usefully non-committal about the precise details of the ontology. See, for example, Marian David, “The Correspondence Theory of Truth,” in The Stanford Encyclopedia of Philosophy, ed. Edward N. Zalta (Fall 2016 Edition). Available online at: https://plato.stanford.edu/entries/truth-correspondence/.

52  In thinking of what I was doing as hypothetical in this way, I saw it as analogous to what Wolfson did in Harry Austryn Wolfson, The Philosophy of Spinoza: Unfolding the Latent Process of His Reasoning, 2 vols (Cambridge, MA: Harvard University Press, 1934), which provided my first introduction to the secondary literature on Spinoza, in a seminar Bernard Peach taught at Duke in the fall of 1962. Wolfson seemed to argue as follows: suppose we understand the technical terms of Spinoza’s philosophy (like “existing in”) to mean such-and-such (say, “being in as an individual essence is in its genus”); do we get the right results? The “right” result here is, not necessarily one which makes Spinoza say something true, but one which leads to consequences he would have accepted. For example, Wolfson claimed it as a virtue of his interpretation that it explained why Spinoza would say that substance is prior in nature to its modes. My main objection to Wolfson was not that his procedure was misguided, but that often it yielded the wrong results (defined as results Spinoza would not have accepted), such as the conclusion that “Spinoza’s substance is inconceivable, and its essence undefinable and hence unknowable” [Wolfson, The Philosophy of Spinoza, vol. 1, 76].

53  Cf. Wittgenstein, Tractatus Logico-Philosophicus 2.061, 5.136, 6.37.

54  Notably in Arthur Schopenhauer, On the Fourfold Root of the Principle of Sufficient Reason [1813], trans. E. F. J. Payne (La Salle, IL: Open Court, 1974), which appealed to Spinoza’s frequent use of the expression causa sive ratio—e.g., in E1p11d2—and his use of logical language to describe causal relations—e.g., in E1p16, E1p17s, and E1p21-E1p23.

55  See Kenneth Clatterbaugh, “Cartesian Causality, Explanation, and Divine Concurrence,” History of Philosophy Quarterly, 12, no. 2 (1995): 195–207. Clatterbaugh was moved to attribute this conception of causality to Descartes partly by his treatment of the laws of nature as causes in the Principles of Philosophy II, §37, a text which was also on my mind in SMC. (I was also influenced by Hampshire’s discussion of causality in Spinoza. See Hampshire, Spinoza, 35.)

56  This should not surprise us. If the Short Treatise shows anything, it shows that Spinoza talked about efficient causation in ways that would have seemed very strange to Aristotle. Cf. KV, I, ch. iii [C I 80/G I 34–35].

57  I take the example from J. L. Mackie, “Causes and Conditions,” American Philosophical Quarterly 2, no. 4 (1965): 245–264.

58  Since I devoted an entire chapter of SMC to the concept of God’s causality, and never described God’s relation to his modes as one of efficient causation—indeed, emphasized how unusual the conception of causality I attributed to Spinoza seemed [SMC, 76–77]—I was surprised to read, in an early draft of Melamed’s first chapter, that I interpreted the causal relation between substance and its modes as one of efficient causation. I wrote to Melamed to protest this misunderstanding. So I was even more surprised, when I read the final version [SMM, 4n], to find him rejecting my attempt at correction, and insisting that when I wrote SMC, I did too understand God’s causality as efficient causality. He presented my protest as reflecting a change of mind, not acknowledged as such. I’m not impressed by his attempt to provide evidence for this reading. But I can see that it serves his larger purposes, as will become clear later.

59  My account of Descartes requires an understanding of his creation doctrine not universally accepted. I claim that Descartes held that the laws of nature were necessary truths and that the breakthrough he thought he had achieved in 1630 was to see how the necessity of those laws could be reconciled with God’s creation of them. One motivation for his view that we should think of the eternal truths as like a king’s laws was a concern about explaining the truth of general propositions whose subject terms have no instances. I’ve argued this in some detail in the articles cited below. Here I’ll simply cite the Discours, where Descartes summarizes what he claimed to have accomplished in Le monde: “I showed what the laws of nature were, and without basing my arguments on any principle other than the infinite perfections of God, I tried to demonstrate all those laws about which we could have any doubt, and to show that even if God created many worlds, there could be none in which these laws were not observed” [CSM I 132/AT VI 43]. See Edwin Curley, “Descartes on the Creation of the Eternal Truths,” Philosophical Review 93, no. 4 (1984): 569–597, and Curley, “Law of Nature.”

60  Hampshire, Spinoza, 47. In SMC, 167 I acknowledged my indebtedness to Hampshire in a note. In retrospect I think I should have been more emphatic about his influence, which now seems to me more considerable than I had realized.

61  Believing that the attributes are the primary locus of God’s causality, when I first read Gueroult, I found the following observation very congenial: “L’attribut n’est rien d’autre qu’une substance en chair et en os, révélée dans ce qui constitue sa nature propre.” See Gueroult, Spinoza 1, 47.

62  Another case where I was indebted to Hampshire. Cf. Hamsphire, Spinoza, 86.

63  Cf. SMC, 155–158.

64  Spinoza’s interest in, and sympathy for, this project is indicated by the fact that when he undertook to teach Casearius the Cartesian philosophy he chose Parts II and III of Descartes’s Principles as the place to begin. See Meyer’s Preface to DPP [C I 227/G I 129–130], and Ep. 13 to Oldenburg [C I 207].

65  C I 458-462/G II 97-102. That this idea goes back to the earliest stages of Spinoza’s thought seems indicated by his discussion of the infinite modes in the Short Treatise, where he says that the discussion of motion—named there as the only universal mode we know in matter—“belongs more properly to a treatise on natural science.” I think Ep. 83 shows that in the end Spinoza was dissatisfied with what he had been able to do in physics. I suppose he was right to be, since his initial ambitions seem to have been based on the assumption that physics would be much easier than it turned out to be.

66  This is not to say that Melamed is wrong to argue that there are infinitely many infinite modes in each attribute [SMM, 119–120]. That’s a dramatic reversal of previous thought on the topic, which has generally supposed that each attribute produced only two infinite modes, one immediate, the other mediate. I think the traditional interpretation rests on very slender textual evidence (mainly cryptic statements in Ep. 64), and that Melamed is probably right to draw this conclusion. However: even if there are infinitely many infinite modes under a given attribute, the series of causes leading back from any one infinite mode to the first cause must be finite. Analogy: the axioms of Euclidean geometry have infinitely many consequences. But the deductive path from any theorem back to the axioms is, and must be, finite.

67  See TTP, ch. ii, §§27–28; TTP, ch. vi, §§14–15; and Spinoza’s comments on Boyle’s experiments with nitre in Ep. 13.

68  This is a conclusion I initially reached because I was wondering how to reconcile E1p26, which proclaims that an infinite being, God, is the cause of all things, with E1p28, which proclaims that the cause of any finite thing must be an infinite series of finite things [SMC, 62–64]. I find this idea also in Gueroult, when he talks about the “double determination” of finite modes in Gueroult, Spinoza 1, 338–339. Bennett seems to have essentially accepted this picture in Bennett, A Study, 113. Is it possible to accept this interpretation of Spinoza’s theory of divine causality and yet reject my conclusions about his necessitarianism? Only, I think, by making Spinoza inconsistent on a central point of his philosophy.

69  Melamed writes in similar terms at SMM, 6, and in CI. I don’t understand why he thinks his account of Spinoza’s definition is “more literal” than mine. Perhaps this is an illegitimate extension from debates about whether scripture should be interpreted literally or figuratively.

70  See Wolfson, The Philosophy of Spinoza, vol. 1, 72.

71  DPP1d5 [C I 239/G I 150] and DPP2d2 [C I 262/G I 181]. In SMC I spent several pages calling attention to the two definitions of substance in Descartes [SMC, 6–11], ending with a brief comment that in his geometric exposition of Descartes, Spinoza showed himself aware of the ambiguity in Cartesian usage. Perhaps I should have said that more emphatically.

72  Here again I find a kindred spirit in Gueroult, whose Spinoza was opposed to scholastic conceptions of substance according to which being in se signified only non-inherence. Instead being in se was to be understood as being by itself, a being which could not owe its being to another [Gueroult, Spinoza 1, 44–45.] Gueroult’s Spinoza has completed the metamorphosis of traditional notions of substance which Descartes began, translating the notions of being in se and in alio in terms of causality [Gueroult, Spinoza 1, 63.] I didn’t know his work when I wrote SMC, but I now think the fact that we arrived independently at similar conceptions of the mode-substance relation is significant confirmation of my interpretation.

73  I use the notions of an adequate and a partial cause in senses modeled on those explained in E3d1, which I would gloss as follows: an adequate cause is one whose effect can be deduced from that cause alone; a partial cause is one whose effect cannot be deduced from that cause alone, but can be deduced from it in conjunction with other causes.

74  Lately I’ve also argued for my view by citing the exchange between Spinoza and van Velthuysen, in Ep. 42- Ep. 43, where Spinoza seems to react with great indignation at Van Velthuysen’s accusation that he takes the universe itself to be God [cf. C II 375/G IV 208 ll. 28–35 and C II 388/G IV 223 ll. 22–25].

75  In SMM, 113–114, Melamed poses this as an unsolved problem in Spinoza interpretation, dismissing my solution in a brief note. Readers might wonder whether he gives as good an answer.

76  KV, I, ch. ix, §1; Ep. 64. In Ep. 64 Spinoza also gives “the face of the whole universe” as an example of an infinite mode which follows from the absolute nature of an attribute by the mediation of some infinite mode which follows from the attribute immediately. Some Spinoza scholars take this language to indicate that the whole of nature, identified in E2le7 as an individual consisting of all bodies [G II 101–102], is also an infinite mode in the attribute of extension. Cf. SMM, 136; Garrett, “Spinoza’s Necessitarianism,” 198; and Nadler, Spinoza’s Ethics, 104–108. For an alternative reading, see my note on “face” in C II 629–630, the annotation of Ep. 64 [C II 439], and SMC, 61. Treating the face of the whole universe as identical with the totality of bodies creates the problem of explaining how a whole can follow from the absolute nature of one of God’s attributes (by E1p23) when none of its component parts does (by E1p28d). This is as if we were to say that the owners of a baseball team could put together a team (a whole constituted by the players, coaches and manager, and their relations to one another) without entering into contracts with any of the players, coaches, or manager. (It also seems to me somewhat awkward that some of these authors should treat the whole of nature as identical both with God and with one of his infinite attributes.)

77  I take it that this is the truth which underlies the objections of interpreters like Hegel and Joachim, who hold that Spinoza must deny the reality of the finite. See, for example, G. W. F. Hegel, Lectures on the History Philosophy, trans. E. S. Haldane and Frances H. Simson, 3 vols (Lincoln: University of Nebraska Press, 1995), vol. 3, sect. 2, ch. 1, A2. See also Harold H. Joachim, A Study of the Ethics of Spinoza (Oxford: Clarendon Press, 1901), 107–119.

78  See C I 624, 661 and C II, 613, 666.

79  See Curley, “On Bennett’s Interpretation of Spinoza’s Monism,” 51. Similarly in Edwin Curley, Behind the Geometrical Method (Princeton, NJ: Princeton University Press, 1988), 142, I say that I “must have known” that this evidence existed because SMC frequently referred to other passages near it in the TIE.

80  Variations on this theme also occur on SMM, 25 and 28.

81  For example, that because Leibniz visited Spinoza in The Hague in late 1676, had been prepared for that interview by Tschirnhaus, and had an opportunity to read the manuscript of the Ethics there, and to discuss it with Spinoza, that it’s “very unlikely” he misunderstood Spinoza. Spinoza’s system is not that easy to understand, and Leibniz would have had many issues he wanted to discuss in a limited time. What he wrote about the Ethics later, after he had a copy of the Opera posthuma to consult at his leisure, contains clear misunderstandings. See SMC, 14–18.

82  That “pantheism” is sometimes used thus broadly is supported by the OED and by Keith Yandell’s account in the Routledge Encyclopedia of Philosophy.

83  See Steven Nadler, “‘Whatever is, is in God’: substance and things in Spinoza’s metaphysics,” in Interpreting Spinoza: Critical Essays, ed. Charlie Huenemann (Cambridge: Cambridge University Press, 2008), 61.

84  Ibid., 64, citing Curley, Behind the Geometrical Method, 38.

85  Ibid., 63.

86  See the definition of a (composite) individual at C I 460 [G II 99, l. 26—G II 100, l. 5], and KV, II, pref. [C I 94–96/G I 51, l. 16—G I 52, l. 40]. I say “at least roughly” because the passage in the Short Treatise suggests that the individual may retain its identity, in spite of some variation in its internal constitution, so long as the variation remains within certain limits. This conception of essence addresses only the continuing identity of the individual thing over time. It does not exclude the possibility that there might be two distinct individuals with the same internal constitution. In articulating this view Spinoza declares his commitment to the new, mechanical philosophy, a commitment also shown in his correspondence with Oldenburg regarding Boyle’s experiments with nitre. See Ep. 6, and particularly C I 178–179/G IV 25, l. 5—G IV 26, l. 3.

87  I don’t know much about either my internal constitution or that of arsenic. But I imagine that with information about both these things a very good scientist might come up with a precise formulation of the underlying law.