SOURCE: “A Character Sketch of the Ethics” and “The Cast of Spinoza's Mind,” in A Study of Spinoza's Ethics, Hackett Publishing Company, 1984, pp. 7-54.

[In the following essay, Bennett outlines the central theses of the Ethics and positions them in their historical context.]

CHAPTER ONE: A CHARACTER SKETCH OF THE ETHICS

The central topic of this book is Spinoza's one indisputable masterpiece, the Ethics. In my first chapter I shall say what sort of work the Ethics is, what sort Spinoza took it to be, how it relates to the rest of his work, and in what spirit I intend to approach it.

§1. THE PLACE OF THE ETHICS IN SPINOZA'S CORPUS

1. Setting aside Spinoza's grammar of the Hebrew language and his two works on politics, which I do not find helpful in understanding the Ethics, we are left with six substantial items. Four of these were finished or abandoned by the time Spinoza was 31 years old, and the other two occupied him intermittently between then and his death in 1677 at the age of 44.

2. Of the four earlier works, I shall not attend much to the Short Treatise. The manuscript of this, written in Dutch by a hand other than Spinoza's, came to light only in the nineteenth century. It seems clear that it stems from Spinoza somehow; but its status is dubious, its content confused, its fit with the rest of his work uncomfortable. These factors and its probable early date make this work a feeble aid to understanding Spinoza's mature thought.

The Emendation of the Intellect, a treatise about ways of acquiring knowledge, avoiding error, and vanquishing scepticism, is largely epistemological, although its announced purpose is practical:

After experience had taught me that all the things which regularly occur in ordinary life are empty and futile, and I saw that all the things which were the cause or object of my fears had nothing of good or bad in themselves, except insofar as the mind was moved by them, I resolved at last to try to find out whether there was something which …, once found and acquired, would continuously give me the greatest pleasure.

The discrepancy is explained by the fact that Emendation is an unfinished work, presumably abandoned in favour of the fresh start that Spinoza was making at about that time in the Ethics, to show that durable happiness may be reached through the cultivation of intellectual virtues—the emendation of the intellect.

Spinoza's practical or ethical aims in the Emendation contrast strikingly with the project that launched Descartes into his Meditations:

It is now some years since I detected how many were the false beliefs that I had from my earliest youth admitted as true, and how doubtful was everything I had since constructed on this basis; and from that time I was convinced that I must … commence to build anew from the foundation, if I wanted to establish any firm and permanent structure in the sciences.

Leon Roth's observation that ‘Descartes' ultimate interest was in the “true”; Spinoza's, in the “good”’,¹ captures accurately the contrast between their ultimate targets. But we shall see that on the way to his practical conclusions in the Ethics Spinoza traverses oceans and continents of theory—metaphysics, biology, philosophy of mind and of matter—not only because they project into territory that lies directly between him and his practical conclusions, but also because he loved the philosophical problems and his own solutions to them. Much of the present work will be devoted to those discussions, which are some of the most instructive and nourishing things in early modern philosophy.

The Emendation is a risky guide to the thought in the Ethics—not because it is merely a start, but because it is a false start. Let us remember that Spinoza dropped it and started afresh. Still, the work is not entirely negligible by us: although Spinoza left it unfinished he did, late in his life, include it with the works he wanted published posthumously.

3. The remaining two early works, unlike any others described in this section, were published during Spinoza's lifetime. They are Descartes's Principles and, tacked on as an ‘appendix’ to that, Metaphysical Thoughts. The former work claims to present in Spinoza's way the main content of the first two Parts of Descartes's Principles. It throws helpful light on Spinoza's own thinking; but because much of its content is Cartesian doctrine which Spinoza confessedly does not accept, it must be handled warily. So, too, must the Metaphysical Thoughts which, although it has Spinoza more often speaking for himself, carries in its Preface a claim to ‘express the opinions of Descartes' which are not always Spinoza's. These two works were published only at the urging of friends, one of whom wrote the Preface; most of their content was written originally to help a pupil whom Spinoza would trust with Descartes's ideas but not with his own. Nevertheless, just once in the Ethics Spinoza obliquely refers to something in his Descartes's Principles, and in §24 I shall show that this reference is brilliantly illuminating.

4. Of the two works that stretch out to the end of Spinoza's life, one is his Ethics Demonstrated in Geometrical Order, a product of labours which extended, albeit with interruptions, across sixteen years. We have no early drafts of the work—an astonishing fact, as versions of it were being read in Holland and in England many years before Spinoza's death. Still, a book cannot go through as much revision as this one did without leaving traces of how it developed, and some of these will be mentioned in due course. Also, after Spinoza's death most of his works were published both in the original Latin and in Dutch translations; the Dutch version of the Ethics was probably based on an earlier Latin version,² and numerous small differences point to changes in Spinoza's thinking.

Further clues come from the other partly late body of work by Spinoza, namely his letters. We have about fifty of these (the numbering runs higher because the standard edition includes letters to Spinoza); they spread across the years when the Ethics was being written, often replying to questions from friends who were reading drafts of the work. In the first of them (Letter 2), written before Descartes's Principles was published, and at about the time of the Emendation, it is clear that Spinoza has already promulgated certain of the basic doctrines on which the Ethics is founded. So the letters can help us.

Often, however, they are disappointing. When a correspondent confronts Spinoza with a profound, central difficulty in his work, the response is seldom satisfactory. It is often an outright snub, especially towards the end of Spinoza's life; but even in the earlier years, before he became sick, his replies tended to be unsympathetic and evasive. The source of this behaviour is more intellectual than moral, I believe. Spinoza was not at his best in correspondence because his mind, although deep and powerful and tenacious, was rather slow; which may also explain why his best work is the one which occupied him for the longest time.

5. Although Spinoza had at an early stage some of the seminal ideas of the Ethics, I am sure it underwent steady changes, mostly of a deepening and broadening tendency, throughout the last third of his short life. We should take the chasm seriously: on the one hand, the works Spinoza finished before he was halfway through his philosophical lifetime; on the other, the great, baffling masterpiece which he had begun on the edge of the chasm, but which, in its present form, is the fruit of its author's maturity.

§2. WHY IS IT CALLED ‘ETHICS’?

1. The work contains three elements that belong to ‘ethics’ in some normal sense of that term.

One is a thesis in the metaphysics of morals: there are no properties of goodness and badness that states of affairs can inherently possess, and no properties of rightness or wrongness that can inhere in actions.

The second is an account of what is actually going on when the plain man judges things to be good or bad. Ordinary ‘ignorant’ people think that goodness and badness are objectively out there in the world, Spinoza says, and he explains why:

[Good and bad] are nothing but modes of imagining, by which the imagination is variously affected; and yet the ignorant [a] consider them to be chief attributes of things, because … they [b] believe all things have been made for their sake, and [c] call the nature of a thing good or bad, sound or rotten and corrupt, as they are affected by it.

(1 Appendix, at 82/16)

Like much in the Ethics, this is tough and interesting in ways that will be missed by a fast reader who overlooks the word ‘because’ in the passage, or an inexperienced one who does not know how amazingly condensed Spinoza's writing can be. Let us try to understand what he is saying here.

It has the general form ‘a because b and c’, which is ambiguous. I think its meaning is analogous to that of ‘She hit him because he insulted her and she became angry’; that is, b explains c which in turn explains a. Here is how the connections go.

We (b) put ourselves at the centre of the universe, taking it that everything is ‘for us’ because made for us by a loving God. This encourages us not merely to care most about how things suit us, but to think that what really matters about things is how they suit us. (A little earlier, at 81/25, Spinoza said: ‘After men persuaded themselves that everything that happens, happens on their account, they had to judge that what is most important in each thing is what is most useful to them.’) And so we are willing (c) to abbreviate ‘This is good for me’ into ‘This is good’: the ‘for me’ goes without saying, and so it plays no role in our speech or thought, just as the tip of my nose plays no role in my visual field. And this practice of characterising things in the monadic language of property attribution (‘This is good’) leads us (a) to think that there are value properties corresponding to our talk.

On that account of Spinoza's views, he sees an intimate connection between (a) a certain metaphysical belief which he attributes to plain people and (c) the way in which they use words like ‘good’ and ‘bad’. If he thinks that (a) somehow infects (c), so that when uninstructed people call a state of affairs ‘good’ they mean that it has a ‘chief attribute’, i.e., is objectively and inherently good, then he should conclude that all the plain man's value judgments are false. Alternatively, he might hold that when an ignorant person calls something ‘good’ there is an outer crust of metaphysical error and an inner core of meaning that is confused but not downright false. There is no way to choose between these two accounts of his position: what separates them is the distinction between ‘what he means when he says …’ and ‘what is going on in his mind when he says …’, and Spinoza had no such distinction in his conceptual armoury.

Anyway, Spinoza deplores the way in which ordinary value judgments are made, and in Part 4 of the Ethics he offers something better. In brief, he proposes to call ‘good’ whatever is useful to mankind, and ‘bad’ whatever hinders the good (see 4d1,2,p39). This is an improvement on ordinary value judgments in two ways. (i) It makes the relative nature of value judgments fully explicit: we are now to understand that ‘x is good’ really stands in for something of the form ‘x has relation R to y’, so that our apparent subject-predicate language will no longer seduce us into thinking that goodness and badness are properties. (ii) Taking it that ‘x is good’ means something of the form ‘x is conducive or favourable to y’, Spinoza offers us a better y. An ignorant person will call x ‘good’ so long as it suits his mood then, and those three words correspond to three inadequacies in his position. Spinozistic value-judgments on x depend upon whether x is favourable to everyone's interests in the long run. Not just mine but everyone's, not just feelings but real interests, not just for now but for the future. In the later parts of the Ethics, then, where Spinoza puts his own sort of value judgments to work, we are given an account of what is ‘good’ in the sense of being conducive to survival, stability, health, happiness and clarity of mind (Spinoza argues that these stand or fall together).

Why is that better than basing value judgments on one's immediate reaction to the thing being judged? How, indeed, can any total theory of value be better than another? If the judgment that B is superior to A is based on standards established by B, then the judgment is trivial; but if it is based on standards lying outside both A and B, then B is not after all a comprehensive theory of value.

2. The answer lies in a theory of Spinoza's about human motivation. He argues that any person who clears his mind, and looks both a ordinary value judgments and at Spinozistic ones, must be swayed by the latter and not the former. Although Spinozistic judgments are really a manual for survival, happiness, et cetera, they are entitled to be expressed in the language of value because they relate in the right manner to human motivation: they ought to hold sway, in the sense that they must hold sway if our minds are not clouded by error and confusion.³ Spinoza's reasons for holding all this will appear in due course, mainly in chapters 10 and 11.

That, then, is the third of the ethical elements which I mentioned earlier. The first was the negative metaphysical thesis; the second, the description of ignorant people's value-judgments; and the third is an extensive, detailed, substantive morality.

The action-guiding parts of the Ethics are sometimes stated without using the language of value: humility does not arise from reason; a free man does not brood on death; you'll have more power over your emotions if you grasp that whatever happens is inevitable; someone who truly loves God will not try to get God to love him back (4p53,67, and 5p6,19). There is no problem about how nonevaluative statements can influence conduct. We are familiar with that phenomenon: ‘There is a dead elk in the water fifty yards upstream’, ‘The dried fruit is at the bottom of my pack’. But Spinoza would be prepared to associate any part of his manual for survival and happiness with a value-judgment, on the grounds that his sort of value-judgment is the best possible—the most informed and rational and clear-sighted—that could be salvaged from the muddled value-judgments of the man in the street.

Although the later parts of the Ethics are meant to be action-guiding, Spinoza does not emphasize that aspect. When he does use value-language, he does so coolly rather than with urgency. This is partly a matter of personal style, but it also reflects doctrine: Spinoza has theoretical reasons for being suspicious of the very concept of intention or purpose, and thus for playing down his own purpose of influencing his readers. But he certainly has such a purpose. When he refers to ‘one who desires to aid others by advice or by action so that they may enjoy the highest good together’ (4 Appendix 25), he surely means himself.

3. Some systems of morality assume little about human nature. Kant's was one of those and in our own day Here's is another. Spinoza's is at the other extreme: his normative material is a doctrine about certain aspects of human nature. More specifically, in Part 5 and the last nine propositions of Part 4 Spinoza describes the causes and effects of the state of being which he favours—one of pleasure (in a broad sense), rationality, prudence, and stability—while the rest of Part 4, heavily supported by Part 3, presents the contrasted state of anxiety, confusion, inner turmoil, and impulsive rashness. Thus, we are shown first a ‘bad’ way of life, and then a ‘good’ one, in Spinoza's senses of those value-terms.

The bad way of life in Part 4 is dominated by what Spinoza calls ‘affects’. These include all the emotions, and also certain motivational drives such as ambition and the craving for alcohol. In the good way of life of Part 5, the affects are controlled through the use of reason. Thus, Part 4 is entitled ‘Of Human Bondage, or the Strengths of the Affects’, while Part 5 is called ‘Of the Power of the Intellect, or of Human Freedom’. I cannot find any significance in the switch in the ordering.

Are we being offered an intellectualist ethic? Yes, in a way. Spinoza does think that the best way of life involves the guidance of reason, and that this is also the way of knowledge, of understanding, of relative freedom from error; and he would have agreed with Locke that nobody is so enslaved as he who is so in his understanding. But Spinoza seems to value knowledge and understanding less for themselves than for their effects. In the Ethics as in the Emendation the centre of the target is not knowledge but happiness.

4. To achieve his ethical purposes in Parts 4 and 5, Spinoza must convince the reader not only of what the contrasted ways of life are like, but also of certain theses about what causes people to follow one rather than the other. That brings in his Part 3 theory about the causes and workings of the ‘affects’, i.e., those emotional turbulences and dominating drives that are said to bind, agitate, depress, and confuse us.

§3. HOW PARTS 1 AND 2 FIT IN

1. The conceptual repertoire of the later, strictly ethical Parts mainly comes from Parts 1 and 2. These also help to set the scene. The ethic is totally medical or psychotherapeutic: to improve yourself you must understand your mechanism and then intervene in it so as to reduce your propensity for feeling and thinking and acting in ways that make you ill and unhappy. Three of Spinoza's earlier doctrines help to remove possible cramps or inhibitions or distractions from this programme.

If the programme were combined with a belief in objective, intrinsic goodness and badness, it would immediately lose its bearings. For then the pursuit of stability, survival and happiness—which are understandable and amenable to scientific treatment—would be replaced by the pursuit of goodness; and from where are we to get our scientific understanding of that?

In Part 1 Spinoza vigorously denies that there is a personal God. If he had accepted a traditional Judaeo-Christian kind of theology, this would have inhibited his psychotherapeutic programme. There is no straight conflict between the two, but it is hard to see how someone can both deeply believe that he was created by a God who loves him and whom he should love, and think that the principal route to self-improvement is to study one's own pathology and deal with it intelligently.

Third, the strict determinism launched in Part 1 is negatively felt all through the therapeutic material. One need not be a determinist in order to study and manipulate causes; but someone with an interest in self-improvement, if he also thinks he has ‘freedom’ in some radical sense which involves rising above causal influences, will be apt to think that self-improvement should come partly from uncaused moral effort; and Spinoza could only regard that as a distraction. If asked about exhortation to moral effort, he would say: ‘That is one way of acting causally upon a person, but not a notably efficient one.’

2. So much for the rejection of value properties, of a personal God, and of radical freedom. What about more specific doctrines, corresponding to numbered propositions in the early Parts of the work? Plenty of those are invoked in the formal demonstrations later on, but often the invocations amount to little. In 3p6d, for instance, the mentions of 1p25c,34 have a dubious role in the argument. In other cases, Part 1 or 2 is brought in quite genuinely, but only in proving something marginal to Spinoza's ethical position.

I can find only one important specific doctrinal link between Part 1 and the last three Parts; I shall present it in §52.1. There are two solid links between Part 2 and the rest, which I shall now explain.

Part 2 is called ‘Of the Nature and Origin of the Mind’. The most striking thing Spinoza says about the human mind is that it contains representations of whatever happens in the human body, and contains nothing else (2p11,12). Combine this with his doctrine that there can be no causal flow either way between mind and body (1p10, 2p6), and you get a picture of a person's mental and physical aspects running in harness without either acting on the other. That lets Spinoza treat ‘affects’ as phenomena which are at once mental and physical (3d3), without having to discuss how the two aspects are interrelated. His interests in human affects are largely curative, and he has some cures to propose, but he is faced with no problem of the form ‘Should this condition be cured through the mind or through the body?’ His doctrines would allow him to ask ‘Should this condition be cured through psychotherapy or through drugs?’, for he can regard the condition—the affect—as physical, and can view psychotherapy as a physical process, a way of causing medically helpful vibrations in the patient's eardrums, vocal cords et cetera. But the question of mental versus physical therapy would be an absurd one for Spinoza to raise, given his parallelism doctrine in Part 2.

Also, Part 2 advances a view about the nature and sources of error (see for instance 2p35). This hovers about Part 3, and is prominent in 4 and 5, where it generates Spinoza's thesis that Part 4's bad state of being, unlike 5's good one, essentially involves a proneness to error. Not needed for Spinoza's route to happiness, Part 2 is needed for his claim that the route to durable happiness is also the route to secure knowledge. This is the aspect of Part 2 which lies closest to the Emendation.

Those are the only large, integral ways in which Part 2 conditions the ethical parts of the work. But Part 2 also provides a kind of frame for what follows. Spinoza was by temperament a broadener and deepener, and it is typical of him that his account of the human condition should rest on a psychology that is linked to a biology which is based on a physics; and all of this occurs in Part 2.

3. Part 1 is a metaphysic—an account of the basic structure of reality, so general as to be consistent with any physics, biology, psychology, or ethics. But it is solidly relevant to Part 2 through a doctrine in the latter which I have not yet mentioned. Spinoza's view about how people's minds mirror their bodies comes from his more general thesis that the entire physical realm is mirrored, event for event, by a mental realm (2p3,7): without this, Spinoza would have no basis for what he says about the human case in particular. And this is where Part 1 most visibly tries to come to the aid of Part 2.

Spinoza purports to derive the general thesis directly from an ‘axiom’ which stands at the head of Part 1. But that derivation is not what I am looking for, since it jumps across Part 1's entire argumentative content. Of more interest in our present context is the Part 1 doctrine that the only thing that meets the strictest conditions for being a substance, or absolutely thing-like thing, is the whole of reality. This one-substance doctrine is used in support of the thesis that the physical realm is mirrored by a mental realm: the argument for that depends on assuming that the two realms are different ‘attributes’ of a single ‘substance’ (2p7s).

Help for parallelism is indirect help for Spinoza's approach to human emotional health and sickness, in which both mental and physical aspects are acknowledged but the difference between them is not allowed to generate any details. This whole line of support, tenuous as it is, marks the only transmission of weight from the ethical doctrines back down to the argumentative core of Part 1.

4. No careful reader of Parts 1 and 2 could doubt that Spinoza is pursuing them for their inherent interest as well as to help the rest of the work. There is abundant evidence, not only within the Ethics but also from the letters and the early Descartes's Principles and Metaphysical Thoughts, that Spinoza acquired early and never lost a deep interest in metaphysics, epistemology and philosophy of mind. We shall follow these interests through Parts 1 and 2 without caring much about how they bear on the rest.

5. Inevitably, Spinoza inherited most of his problems from previous philosophers. The one who looms largest is his older contemporary Descartes, the only philosopher named in the Ethics. I shall make much of Descartes as a bequeather of problems which Spinoza tried to solve, and of Leibniz as another of Descartes's legatees, tackling many of those same problems in vividly and instructively different ways. Descartes died when Spinoza was eighteen years old; Spinoza died when Leibniz was thirty years old.

Of course the background goes further: there are ancient and medieval philosophers who influenced Spinoza and would have to be considered in an intellectual biography of him. But I am not writing biography. I want to understand the pages of the Ethics in a way that will let me learn philosophy from them. For that, I need to consider what Spinoza had in mind, for readings of the text which are faithful to his intentions are likely to teach me more than ones which are not—or so I believe, as I think him to be a great philosopher. And one can be helped to discover his intentions by knowing what he had been reading, whose problems he had been challenged by, and so on. But this delving into backgrounds is subject to a law of diminishing returns: while some fact about Maimonides or Averroes might provide the key to an obscure passage in Spinoza, we are more likely to get his text straight by wrestling with it directly, given just a fair grasp of his immediate background.

I am sure to make mistakes because of my inattention to Spinoza's philosophical ancestry; but I will pay that price for the benefits which accrue from putting most of one's energies into philosophically interrogating Spinoza's own text. I am encouraged in this by the massive work in which Wolfson places Spinoza in a densely described medieval setting: the labour and learning are awesome, but the philosophical profit is almost nil.⁴ Such philosophically interesting readings of Spinoza as are contained in Wolfson's two volumes could all have been arrived at without delving into the medieval background.

§4. ‘DEMONSTRATED IN GEOMETRICAL ORDER’

1. Much of the Ethics consists of smooth expository prose, in portions of half a page or more in length; these are Spinoza's prefaces, appendices, and scholia. Their literary standard is high: although Spinoza employed a fossilized academic Latin, drained of warmth and colour, his writing could still have virtues of elegance and dry pungency; and some of the prose passages are lively, entertaining and a good source of epigrams.

In contrast with that, the remaining sixty percent of the work consists in a charmless apparatus of ‘demonstrations’. Spinoza himself implicitly apologizes for the prolixity which these force on him (4p18s), but he clearly thought the trouble was worthwhile. His remark that demonstrations are ‘the eyes of the mind, by means of which it sees and observes things’ (5p23s), though vague and metaphorical, clearly assigns them some important epistemic role. Let us see what it is.

When Spinoza says that his propositions are demonstrata, that ought to mean that they are rigorously, deductively proved—that being the main seventeenth century sense of the Latin demonstrare and the English ‘demonstrate’. And the phrase ‘demonstrated in geometrical order’ invites comparison with that most celebrated of all deductive enterprises, the Elements of Euclid.

Euclid began with ‘postulates’ involving spatial concepts, and ‘axioms’ concerning magnitudes in general (not just spatial ones), and some initial definitions. Whatever his own intentions were, his system came to be viewed by many people as a direct convincer—proceeding from unassailable starting points, through irresistible argumentative moves, to theorems which are thus equally unassailable. It did directly convince Hobbes:

Being in a gentleman's library, Euclid's Elements lay open, and 'twas the 47th Proposition of the first Book. He read the Proposition. By God, said he, … this is impossible! So he reads the demonstration of it, which referred back to such a Proposition; which proposition he read. That referred him back to another, which he also read. [And so on, until] at last he was demonstratively convinced of that truth. This made him in love with Geometry.⁵

On the face of it, that is how things are supposed to be in Spinoza's Ethics: the only fresh material that is explicitly added in the later Parts consists of definitions, axioms and postulates; they could all have been given at the outset, so that the entire work would seem to flow from its opening two or three pages; and the material in those pages would all, apparently, be offered as self-evident, undeniable.

2. If the Ethics was intended by Spinoza as a direct convincer of that sort, he must have been madly optimistic about the plausibility of the initial definitions and axioms. I don't believe it. No doubt he thought that the definitions are correct, and the axioms true; but he cannot have expected them to impose themselves on the mind of someone whose course of Spinozistic study was just beginning.

Consider 1d7, for instance: ‘That thing is called free which exists from the necessity of its nature alone, and is determined to act by itself alone.’ If this claims that that is what people in general actually mean by ‘free’, then it is obviously false. It would be true, or anyway unassailable, if it were merely stipulative—an announcement meaning ‘This is how I shall use the term “free” in the pages that follow’. Many of Spinoza's definitions are worded as though they were mere warnings about his own idiolect: ‘By cause of itself I understand …’, ‘By substance I understand …’ and so on. But the wording misleads. Spinoza means his definitions less tamely than that, offering them as saying how he will and how everyone should use the defined words. Although this is not controversial, I shall present a bit of evidence that Spinoza is not offering his key technical terms in the spirit of: ‘This is how I shall use these words, and I have nothing to say in justification of these uses.’

It concerns 1d1, which I shall abbreviate to: ‘By cause of itself I understand that which is F’—never mind what F is. This is used in 1p7d, as follows:

By cause of itself I understand: that which is F.

Substance is cause of itself.

[there4] Substance is F.

That is valid. A definition licenses replacing the definiendum by the definiens, which is just what that argument does. But from where does the second premiss come? Well, 1p6c says that substance is not caused by anything other than itself, and Spinoza assumes that everything is caused by something; from which he infers that substance is caused by itself. That is valid if ‘caused by itself’ is used in its normal meaning. What is not valid is the move from ‘Substance is caused by itself’, understood like that, to ‘Substance is cause of itself’ when this is taken in the sense given in 1d1. An argument using that definition should involve only the sense of ‘cause of itself’ which the definition ceremonially confers on the phrase; and it must not be assumed that only what is ‘cause of itself’ in that sense is ‘caused by itself’ in the phrase's ordinary untechnical meaning.

The only way to clear Spinoza of this charge of malpractice is to suppose that 1d1 is not intended stipulatively, and is rather meant as a substantive thesis linking ‘that which is F’ to at least some aspects of the ordinary meaning of ‘cause of itself’. And that provides some evidence that 1d1 was intended in the latter way, though of course the evidence is not conclusive.

When Spinoza's definitions are not stipulative, we must take them—as he says in Letter 9—to be ‘propositions’ which are acceptable only if they are ‘true’. That kills any attempt to use the Ethics as a direct convincer, for such definitions as 1d1,7 cannot possibly count as unassailable starting points unless they take refuge in a plea of stipulativeness.

Why would Spinoza disguise a proposition as a definition? I think he does this when the proposition is partly about common meanings of terms. This can be seen in 1d7, Spinoza's definition of ‘free’. I think he means this to embody a doctrine which he will defend later on, namely, that the common idea of freedom as involving uncausedness is indefensible, a piece of bad thinking which careful philosophy will rectify. We should take 1d7 as saying: ‘I am going to use “free” as equivalent to …, and if you want to give the word a meaning as close to the ordinary one as possible without falling into error or confusion, you will use it like that also.’

3. What about the axioms? Only someone who approaches Spinoza in a spirit of stultifying piety could regard the axioms as immediately certain, or their contradictories as impossible or unthinkable. When Spinoza says things like ‘Everything which is, is either in itself or in something else’ (1a1), our first problem is to know what he means. He is inviting us to adopt and employ a certain way of thinking and talking, and we cannot know whether to comply until we know what we would be getting into. So we, at least, must regard the axioms as something to be accepted provisionally in order to see how they work out in the sequel. ‘These first definitions and propositions can be properly understood only in the light of the propositions which follow them in the order of exposition; they form a system of mutually supporting propositions.’⁶

Spinoza addressed himself to an audience who were more familiar with his terminology than we are; but I am sure that he offered his axioms to them, too, not as immediately compelling but rather as the first bricks in an edifice which was to be tested as a whole. In one place he says: You won't like what I have just said, but please suspend judgment until I have gone somewhat further (2p11cs).

4. So when Spinoza says that his results are demonstrata—demonstrated or rigorously proved—presumably he is talking about the logic and not the psychology of his procedures. That would be reasonable: if one has a logically valid argument with true premisses and conclusion P, this is a ‘demonstration’ of P even if it would not convince anyone of P's truth because the premisses are not obviously true.

Sometimes, indeed, Spinoza uses ‘demonstrate’ more weakly still. The full title of his Descartes's Principles says that it presents Cartesian doctrines ‘demonstrated in the geometrical manner’; yet Spinoza insists that much of this ‘demonstrated’ material is false. The Preface implies that the work will be based on ‘propositions so clear and evident that no one can withhold his assent from them, provided that he has understood the terms’; but Spinoza did not write the Preface, and that bit of it must be mistaken. It would be much more in character for Spinoza to offer the premisses as partly false—which he must do if he is to reject some of the conclusions while claiming that the deductive moves are valid.⁷ Presumably, then, he is willing to call something ‘demonstrated’ if it is validly derived from false premisses. It is easy to believe, then, that in the Ethics he took his conclusions to be ‘demonstrated’ in the sense of validly derived from premisses which, though true, are not always immediately convincing. (In the seventeenth century ‘demonstrate’ et cetera could mean ‘display’ or ‘exhibit’; but that is irrelevant here, for Spinoza's demonstrations are certainly arguments.)

As for ‘the geometrical manner’ or ‘geometrical order’, this must point to the likes of Euclid's Elements, but it is not clear what its precise implications were in Spinoza's mind, if indeed it had any. He may have been in a slight muddle about the notion of doing things geometrically. Of ‘those who prefer to curse and laugh at the affects and actions of men, rather than understand them’, he says: ‘To them it will doubtless seem strange that I should undertake to treat men's vices and absurdities in the geometrical manner, and that I should wish to demonstrate by sure reasoning things which they declare to be contrary to reason.’ (3 Preface at 138/6.) This seems to refer to that ‘geometrical’ demonstrative procedure which is my present topic; but the sequel shows that Spinoza chiefly has in mind his view that men's affects can be causally explained through the laws of nature, and are therefore fit subjects for scientific study. His point, which would have been better served by a comparison with physics, has nothing to do with handling the affects through an apparatus of axioms and logical derivations.⁸ Spinoza may not have been clear about this in his own mind.

§5. THE HYPOTHETICO-DEDUCTIVE METHOD

1. It is best to view the Ethics as a hypothetico-deductive system—something that starts with general hypotheses, deduces consequences from them, and checks those against the data. If they conflict with the data, something in the system is wrong; if they square with the data, the system is not proved to be right but it is to some extent confirmed. That is a widely accepted story about the abstract structure of scientific method, though in recent years it has been challenged. I shall keep out of that dispute.

In a hypothetico-deductive procedure, there are entailments running downwards, so to speak, and weaker confirmation relations running upwards. The two are related, but not as simply as this: if P entails Q, then Q somewhat confirms P. Let P be ‘My bicycle is rusty and Mt. St. Helens will erupt in 1997’ and let Q be ‘My bicycle is rusty.’ In this case, P entails Q, but it is unsatisfactory to say that Q supports or partially confirms P.

The further requirements for confirmation are probably complicated and may not yet be fully known by anyone. But Leibniz made a strong start on them, which will do for now. He entertained the objection that one should not ‘undertake to discover the unknown by assuming it and then proceeding by inference from it to known truths’—which neatly describes the hypothetico-deductive method—on the grounds that ‘this is contrary to logic, which teaches that truths can be inferred from falsehoods’. In reply he said that this procedure ‘sometimes yields great likelihood, when the hypothesis easily explains many phenomena which otherwise would be puzzling and which are quite independent of one another’.⁹ That is a good nutshell: the hypotheses must explain, and must unify, if they are to gain strength from the truth of their consequences.

2. In recommending that we view the Ethics as hypothetico-deductive, I am implying that Spinoza would have endorsed this. Perhaps an approach to a text could be profitable even though sharply at variance with the author's own view of it; but it is not likely, and I would be uncomfortable with my approach if I thought Spinoza would reject it.

He was certainly no stranger to the hypothetico-deductive method. In the introduction to Part 3 of Descartes's Principles, where he seems to be speaking for himself, Spinoza says that a ‘good hypothesis’ is consistent, simple, intelligible, and ‘permits the deduction of everything which is observed in the whole of nature’. That last requirement is extravagant, and Spinoza would certainly not claim so much for the starting points of the Ethics, which I am taking to be his ‘good hypotheses’; but he would defend them as consistent, simple, intelligible, and entailing a lot of important truths.

3. It sounds as though Spinoza, according to me, thinks he has found a set of sturdy hypotheses—ones we can retain until they fall foul of the data and/or are supplanted by more satisfactory rivals. That was not his attitude. In Letter 76 he says: ‘I do not assume that I have discovered the best philosophy, but I know that I understand the true one’; and that dogmatic tone permeates the Ethics also. However, that can be reconciled with a hypothetico-deductive view of the work, as follows. Spinoza could—and I think would—say that although his system must work on untutored minds in a hypothetico-deductive manner, when the tutoring is completed the reader will see the starting points to be certain, indubitably true, beyond question. Something like that is implied in Descartes's remarks on this topic, late in his reply to the Second Objections to the Meditations, by which Spinoza is known to have been influenced. And perhaps we can charitably take it to be implied when Spinoza, having claimed to know that he understands the true philosophy, adds this: ‘If you ask me how I know this, I shall answer, in the same way as you know that the three angles of a triangle are equal to two right angles.’¹⁰

But there is a problem. What is supposed to be happening to us as we work through the Ethics? We are to entertain its definitions and axioms as hypotheses, to follow through their consequences and find that they square with the data, and to finish up in a state of mind where we see them as self-evident. But how is this change to be wrought in us? It is to be assumed that it is not like swallowing a philosophical tranquiliser, a pill to make sceptical doubts dissolve. What we must demand, and what Spinoza will agree that he should supply, is an account of rational procedures whereby a careful study of the Ethics could convince us of the truth of its doctrines.

We might look for help to a passage in the Emendation where Spinoza says that ‘When the mind attends to [something false], considers it carefully, understands it, and deduces from it in good order the things to be deduced, it will easily bring its falsity to light’ (§61). This shows him as willing to combine a hypothetico-deductive approach with absolute confidence in its results. But why will that method ‘easily bring its falsity to light’? Spinoza's answer to this is supposed to be gatherable from Emendation §§61-65, but I can get nothing useful from that material.

4. Fortunately, we can find something better to say on his behalf. I shall state it abstractly at first.

Suppose we want to test hypothesis H by deducing consequences from it and then observing whether those consequences are true. If we have deduced a number of consequences and found all of them to be true, that may give some support to H. But if there is a rival hypothesis H^* which also has those consequences, our testing of H so far does not support it as against H^*. For that, we must find a test which discriminates between them. Let us suppose that H entails something which H^* contradicts, and that we perform an experiment which shows this proposition to be true. That leaves H still standing while it knocks out H^*. If we test a wide variety of consequences of H, each test that it survives may eliminate rival hypotheses; and a test can eliminate a rival which has never been thought of or formulated, but which exists in logical space and threatens H's claim to truth.

Some philosophers have thought that confirmation is just the procedure of making H ever more secure by reducing the number of unrefuted rivals to it. Others have objected that there are infinitely many hypotheses at the outset, and each observation eliminates only finitely many of them, so that we cannot make the slightest dent in the number of unrefuted rivals to H. If they are right, then a fortiori we could never establish H as our only unrefuted hypothesis.

Spinoza would say that such critics are wrong, that one could be left with only one surviving hypothesis, and that in the Ethics he achieves just that. He holds that most of our general terms are unsuitable for serious theorising, and that only a very limited stock of concepts are fit for use in metaphysics and science; so he could argue that at the outset there is really only a small number of considerable hypotheses to be dealt with, which makes it credible that by the time his hypothesis has shown its paces all the others have fallen by the wayside. That could be grounds for untentative confidence in his axioms and definitions—a system I have called his ‘hypothesis’. For if no other system is true, then Spinoza's must be, he would conclude; for he would never entertain the thought that no system is true, i.e., that reality is not fundamentally orderly (see §8).

I would be willing to argue that a given hypothesis has only finitely many considerable rivals, but I would not use as a premiss Spinoza's doctrine about theoretically respectable general terms. That doctrine, so far from expressing deep insights into reality, merely shows how thoroughly Spinoza was a child of his time. But he does hold it, and I have sought to show how it could help him to maintain his absolute confidence in his results while still permitting us to see the Ethics as working in a hypothetico-deductive manner.

5. Opinion is divided on what the job of an hypothesis is, independently of the debate about what proportion of the rivals can be eliminated. Some philosophers think that an hypothesis cannot do more than ‘save the phenomena’, that is, imply all the lower-level truths and no falsehoods; and they reject as senseless or at least unprofitable the question of whether the hypothesis, as well as doing that, is actually true. That line of reasoning is utterly foreign to Spinoza, and may have been what he meant to reject when claiming for his philosophy not that it is ‘the best’ but that it is ‘true’.

§6. WHAT ARE THE DATA?

1. The hypothetico-deductive scheme has mainly been explored as a rational reconstruction of what natural scientists do: they devise high-level theories, deduce predictions about how experiments will turn out, and then experiment to see what happens. In this scheme of things the ground floor is occupied by particular events in laboratories: that is the touchstone of the truth of the ‘propositions’ and thus of the degree of confirmation for the ‘axioms’ and ‘definitions’, i.e., the theories at the top level of the system.

How can we apply that picture to the Ethics? Against what data are we to check Spinoza's propositions?

Some of the propositions are answerable to empirical fact. When Spinoza says in 4p57 that ‘The proud man hates the presence of the noble’, it is relevant to check this against the actual attitudes of people who are indisputably proud; and Spinoza would not deny this. We shall see in chapter 5 that he at least flirted with the idea that every truth, 4p57 included, is absolutely necessary, so that this is the only possible world. But that does not imply that every truth can be known a priori; and I am inclined to believe that Spinoza always distinguished truths which can be established by conceptual analysis from ones which we can know only by looking at the world. Some of his starting points are evidently offered as things everyone knows from observation—for instance 2a4 and the postulates preceding 3p1.

2. The Ethics is noted for playing down experience in favour of reasoning; but this emphasis does not falsify what I have been saying. In 2p40s2 Spinoza presents three kinds of cognitive process of which the first, much frowned on by him, is said to have two sources of input: hearsay, and experientia vaga. The principal meaning of the Latin vagus is ‘inconstant’ or ‘wandering’, and in Spinoza's use it means ‘random’. In the Emendation he speaks of ‘random experience, that is experience which is not controlled by the intellect’ (§19). Presumably, what is frowned on by Spinoza, as a source of confusion and error, is the jumbled sequence of sensory states which one would gain by roaming the world with no sense of epistemic direction.

This contrasts with the orderly, controlled experience which comes through attending to experiments which have been deliberately contrived to test some prior hypothesis. Such a methodology was memorably advocated by Kant a century after Spinoza's time: ‘Accidental observations, made in obedience to no previously thought-out plan, can never be made to yield a necessary law’, he said, and went on to say that although we ‘must approach nature in order to be taught by it’, we must ‘do so not in the character of a pupil who listens to everything that the teacher chooses to say, but of an appointed judge who compels the witnesses to answer questions which he has himself formulated’.¹¹

3. Spinoza nowhere discusses experientia non vaga. So far as hard textual evidence is concerned, we could attribute to him the view that all experience is vaga.

But that cannot be right. Several times in the Ethics he appeals to something which he says ‘experience shows’; all those could be appeals to unrandom experience, i.e., to the experience of someone who puts to nature questions dictated to him by Spinoza'd philosophy,¹² and it is hard to see what else they could be.

Why, then, is Spinoza so silent about intellectually controlled experience? Perhaps he thought there was nothing to be said about it until more is known about how the senses function. In the Emendation he mentions the senses as needed for knowledge of particular things, but he does not theorize about them because that can be left until ‘the nature of our senses has become known to us’ (§102). That makes best sense if we take Spinoza to mean that a treatment of the senses should be left until more is known about how they work; and that would express a scruple which could also explain the silence about unrandom experience in the Ethics.

4. Much of the Ethics is philosophical rather than scientific, i.e., is answerable to conceptual analysis rather than to empirical observation. These propositions might be evaluated by how well they fit in with material which can be empirically tested; but that is not the only way to do it, for the hypothetico-deductive method can be used where the terminal propositions are closely related not to experience but to logical or conceptual truth or falsehoods.

Much philosophical analysis is like that. For instance, we may evaluate a proposed analysis of propositions of the form ‘If A had happened, B would have ensued’ by deducing from it conclusions of the following form: ‘In a situation of kind K, if A had happened B would have ensued’, which are then checked against the intelligent and careful person's intuitions. If an analysis implies ‘Whatever the actual states of mind and body of the people on the scene, if Sirhan Sirhan hadn’t shot Robert Kennedy at the moment he did, someone else would have shot him a moment later’, that is powerful evidence against the analysis. We look for an analysis which doesn’t entail anything that we and our friends intuitively judge to be false. But usually we cannot find such an analysis, which is one reason why good philosophers sometimes say things which appear to be dazzlingly false. When we cannot rescue all our intuitive judgments, it may be hard to decide which to relinquish. There may be different alternative trade-offs and no short, sharp procedure for choosing amongst them; which is one thing that makes philosophy difficult and wonderful.

In trying to evaluate the Ethics, then, we can appeal not only to how we find our world to be, but also to our sense of which propositions of logic, broadly so-called, are true and which false. When Spinoza offends against this sense, we shall be caught up in the harder business of evaluating the trade-off he offers us, that is, his reasons for saying that we must embrace his strange conclusions to avoid ones which are stranger still.

§7. THE INVALIDITY OF THE DEMONSTRATIONS

1. From the upward relation of support we now come to the relation of entailment which is supposed to run downward through the official demonstrations. Many of these are not valid as they stand, and we must decide what to think about this fact.

Sometimes we should treat a defective argument as a valid one with a premiss left unstated. Sometimes a defective argument of Spinoza's can be non-trivially rescued by adding a premiss which we know, on other evidence, to be accepted by him. We have met one example already, namely, the move in 1p7d from ‘Substance is not caused by anything else’ to ‘Substance is caused by itself’. We can make this valid by adding the premiss ‘Everything is caused by something’, which Spinoza does accept and presumably employs here.

Other invalidities can be repaired by plausibly supposing that Spinoza did not say quite what he meant; as when he infers P from Q on the strength of ‘If P then Q’, where he clearly meant ‘If and only if P, then Q’. Sometimes, again, we have trouble because he doesn’t give a word the meaning we would give to it, as in the use of 1d6 in 1p14d, where Spinoza in effect moves from ‘Infinite[ly many] Fs are G’ to ‘All Fs are G’. Using ‘infinite’ as we do, that is invalid; but we shall find independent evidence that Spinoza gives to ‘infinite’ a meaning different from ours. This argument itself is evidence for that, but is far from decisive. To treat it as decisive would be to adopt the axiom that all Spinoza's arguments are, when properly understood, valid.

2. So there are two sorts of demonstrations that can be regarded as valid but not perfectly realized on the page. Some others, however, cannot be thus rescued.

Still, they are not negligible. Even when an argument cannot plausibly be saved by addition or reinterpretation, its premisses may give some reason for accepting the conclusion. I have no crisp theory about the conditions under which, if P and Q are logically independent, P gives some reason for Q. We might start with the thought that P gives some reason for Q if there is a probably true R such that Q is entailed by (P & R) but not by either conjunct alone; but that cannot be the whole story. Still, even without such a theory we can sometimes reasonably judge that one proposition gives some reason for another, and many of Spinoza's putative demonstrations can be viewed in that manner.

Much of the structure, then, is really hypothetico-supportive, with many downward as well as upward relations being less than entailments. The difference between (weak) confirmations and (strong) entailments does not coincide with the difference between the two directions of flow.

3. The inferential structure of the Ethics is not wholly contained in its demonstrations, officially so-called. Argumentative movement also occurs in the scholia, where Spinoza's comments on his demonstrated propositions sometimes involve drawing conclusions from them. In §3.3 I mentioned the important attempt to bring the one-substance doctrine to the aid of the thesis that the physical realm is paralleled by a mental realm; this attempt is entirely confined to a scholium, 2p7s, making no appearance in the official demonstrations. So we must often consider arguments of the informal kind which most philosophers employ most of the time, as well as ones purporting to be rigorous.

Spinoza sometimes offers more than one demonstration for a proposition, or indicates how a second one could be given; and in yet other cases we shall find alternative demonstrations without help from him. There can be many Spinozistic routes to a single doctrine.

That is a strength in the work, and should not occasion surprise or complaint. What does invite comment is that, of the argumentative routes to Spinoza's conclusions, some which we are left to find for ourselves are stronger than the ones he marks for us. For example, his demonstration that there is only one substance or ultimately thing-like thing (1p14d) is a creaky, leaky affair which cannot be fully salvaged. There are much better reasons than that for accepting the one-substance doctrine—reasons that Spinoza has but leaves unexpressed.

This can be explained. It seems that Spinoza's demonstrations are not meant to indicate the reasons which first led him to his conclusions, or even necessarily the principal reasons which he now has for retaining them. The demonstrations are supposed to give reasons—good enough ones, indeed—for the conclusions; but they are also meant to enable Spinoza to lock his doctrines together into a rigid, logically unbreakable structure. Some of his more powerful reasons are unsuitable for that because they are not sharp and hard enough to be handled demonstratively, even by Spinoza's standards of rigour.

Because there are different routes to a single Spinozistic conclusion, when the inferential thread is broken at a given point we are not always compelled to relinquish everything depending on it. Even if the break is total, leaving not so much as a weak support relation let alone an entailment, there may be other threads to take the weight.

4. I do not credit Spinoza with conceding that some of his demonstrations are not valid, or side with those who hold that the work's deductive surface was not meant to be taken seriously. There is room for debate about why Spinoza chose to present his results in that manner;¹³ but whatever else he meant his demonstrations to do, I am sure that he did offer them as strictly valid, intending them to achieve a tight interlocking of his principal doctrines, and expecting that a proper study of this structure would eventually enable us to respond to it as though it were a direct convincer. I also think that he would agree to its being viewed, first time through, as hypothetico-deductive. But not as hypothetico-supportive: that change in the story is mine, based on my seeing something that Spinoza didn’t see, namely, that many of the demonstrations are invalid.

How could someone as supremely able as Spinoza have been satisfied with demonstrations so many of which are invalid and unrescuable?

To begin with, he was dealing with some of the hardest problems there are, and elementary slips and oversights are easier when the matter is difficult. Also, when an argument depends on concepts which are dizzyingly abstract, basic, general, one is not much helped by a firm, reliable, intuitive sense of how the argument should go.

Second, it is relevant that Spinoza's demonstrations do not represent his own order of discovery. When he ‘demonstrates’ R by deriving it from P and Q, it is not that he accepted P and Q and wondered what they entailed; rather, he accepted R and wondered how best to derive it from a mass of other accepted doctrine including P and Q. In these circumstances, and given strong confidence that it could be done somehow, it is natural that Spinoza should have been uncritical about the derivation he actually came up with. I speak from experience.

Third, Spinoza was not good at close, rigorous reasoning. Leibniz, who is ordinarily charitable about others' intellectual failings, sometimes comments on Spinoza's central demonstrations in terms of casual contempt,¹⁴ and I think that Spinozistic studies are best served by facing the fact that deductive rigour was not Spinoza's strong point. This is not to deny that his thinking was considerable and valuable (it didn’t stop Leibniz from taking him seriously¹⁵), but it is to say that he was not skilled at constructing strictly valid arguments or at criticising invalid ones.

Perhaps this fact about Spinoza's mind is linked with his being somewhat slow. This thought was prompted by the double contrast with the nimble, razor-sharp Leibniz, and is confirmed by my own experience of my contemporaries: those whom I find to be faster on their feet than I am are also those who are better at rigour than I am. I would add that although agility and sharpness are enviable, they are not everything. Spinoza was a greater philosopher, and also a better philosopher, than most of those who could think faster and more tightly than he could.

In addition, Spinoza was interested in logic only for the good it could do him. Had he cared more about it for itself, he might have used it better in doing philosophy, just as a carpenter might be saved from some failures if he chanced to be interested in glue for itself as well as for its service in holding things together. Leibniz was, among other things, a connoissur of glue: his New Essays, for example, are full of his pleasure in logical relations as an inherently interesting topic of study. There is nothing like that in Spinoza.

CHAPTER TWO: THE CAST OF SPINOZA'S MIND

In this chapter I shall present five aspects of Spinoza's thinking which lie deeper than any of his argued doctrines and are so influential in his thought as to deserve special attention.

§8. FIRST ASPECT: RATIONALISM

1. Spinoza is customarily classified, along with Descartes and Leibniz, as a ‘rationalist’. This can mean various things, of which at least three are true of Spinoza.

The term comes from the Latin ratio, meaning ‘reason’. Like all philosophers who have been called rationalists, Spinoza is impressed by what he takes to be the superiority of reason to the senses. This has a large effect on the Ethics, but I shall not discuss it here. It is a matter of explicitly defended doctrine, and can be treated when its turn comes.

Other strands in Spinoza's rationalism concern ‘reason’ not as the name of a cognitive faculty but rather as involved in the notion of a reason for a belief or a reason why something is the case—this notion also lying within the scope of the Latin ratio.

2. Spinoza assumed that whatever is the case can be explained—that if P then there is a reason why P. I call this ‘explanatory rationalism’. It is the refusal to admit brute facts—ones which just are so, for no reason. As between this and the opposing view that there are brute facts, the latter will always have a certain advantage—it can produce various facts and demand ‘How could that be explained?’, and it will not always be easy to answer; whereas the explanatory rationalist cannot make similar trouble for his opponent. If he demands to know how a given fact could be inexplicable, the answer may be that it just is; and that answer, since it leads nowhere, must be safe. I believe that the explanatory rationalist's problems are insoluble, and that his position is therefore untenable; but I admit the discomfort of this stand, and proclaim the attractiveness of the idea that everything can in principle be explained. So I urge that disbelief in explanatory rationalism, though perhaps it may be carried out of our study of the Ethics, is not something to be carried into it.

3. In addition to explanatory rationalism, Spinoza seems to have accepted causal rationalism.

At least since Hume it has been widely thought that what is causally possible is only a fragment of what is logically or absolutely possible: a kind of event which cannot happen, given the causal laws by which the actual world is governed, may nevertheless be logically possible—e.g., fully describable without overt or covert contradiction. It is causally impossible that the moon should spontaneously veer away from the earth towards the sun; but one could tell a consistent story—in fine detail, with every gap plugged, every question answered—in which the moon did just that. Or so Hume implied, and most of us agree. Before Hume, some philosophers distinguished causal from logical necessity; but others did not, and most were unclear or indeterminate about it.

Spinoza does not make the distinction: he thinks that a cause relates to its effect as a premiss does to a conclusion which follows from it. When he speaks of ‘the reason or cause why Nature acts’ (4 Preface at 206/26), he thinks he is talking about one relation, not two.

It is not that he sees logical links as weaker than they are; rather, he sees causal ones as stronger. He sometimes uses the language of causality in discussing logico-mathematical topics, as when he says that ‘the properties of a circle’ should be approached through ‘the cause of a circle’; but by ‘cause’ here he means ‘method of construction’, this being thought of not as a concrete event involving compass and pencil, but rather as a mathematical item, a procedure or method which necessarily exists even if no one ever implements it. (That is my description of what happens in Letter 60, not Spinoza's.) So this passage does not bring into mathematics anything we would call ‘causal’.¹

On the other hand, Spinoza's thinking about the matters which we would call causal is much influenced by his not distinguishing causal from absolute necessity. For example, after describing the effects on behaviour of despondency he adds: ‘These things follow from this affect as necessarily as it follows from the nature of a triangle that its three angles are equal to two right angles’ (4p57s). There are many other such turns of phrase.

Because I think that logical (or absolute) and causal necessity are distinct, I believe that Spinoza was in error about this. But that is not to accuse him of an elementary blunder. It is a theory of ours, these days, that causal necessity is weaker than absolute necessity; it is not trivially or obviously true, nor has it been well supported by arguments based on clearly good accounts of the two sorts of necessity.² So we are not well placed to be condescending about Spinoza's causal rationalism.

4. It is, indeed, virtually forced on Spinoza by his explanatory rationalism. I now explain why.

An F event occurs, and causes a G event. Why did a G event occur just then? Because an F event occurred just before. That throws up two further why questions. (a) Why did an F event occur just before? This promises to launch an infinitely long series of why questions, each answerable by mentioning a still earlier event about which a new why question can be asked. The rationalist may be untroubled by this. He may say that this backward running series is infinite: each of its answers throws up a further question, but each of the questions has an answer; so it poses no challenge to explanatory rationalism, he may say. That is not the end of that issue, however, and I shall return to it in §28. I mention it here only in order to set it cleanly aside. (b) Why does the occurrence of an F event explain the occurrence of a G event? This is the question I want to confront.

The answer will adduce a causal or natural law according to which F events cause G events. Now the rationalist must step with care, for he needs an account of the status of the causal law which will let him cope with the question ‘Why is that law true?’ If he must allow the question, he must think it has an answer, or else his explanatory rationalism collapses. Let us consider his options.

He cannot adopt Hume's view that the causal law states a general brute fact. According to Hume, that F events are followed by G ones is simply the way it goes, and there is no more to be said about it except for some doctrine—irrelevant here—about our psychological attitude to such sequences. On that view, a causal explanation of a particular event handles one brute fact merely by putting it in with countless others. That does not meet the explanatory rationalist's need because it answers a small why question only by leaving a bigger one unanswered and, if Hume is right, unanswerable.

The explanatory rationalist might take the causal law linking Fs to Gs to be stronger than a mere brute fact, though less than absolutely necessary, thus invoking the notion of a causal necessity intermediate between absolute necessity and mere brute-fact generalization. But this familiar and perhaps useful notion will not serve the rationalist's turn. Consider the question ‘Why are F events always followed by G events?’ or ‘Why was the F-G law operative on this occasion?’—either will do. If it is not absolutely necessary that F events are followed by G events, then it is legitimate to ask why the F-G sequence happens when it does. The question is not self-answering. Nor is it met by saying that there is a less than logical necessity at work here, i.e., that F events have some sort of G-making power by means of which they force G events to occur. For we can still ask two questions:

Why does every F event have a G-making power? Why is a G-making power, whenever it is present, operative?

If either is answered by appeal to some deeper-lying causal law, then the question arises again about that; and even those who accept an infinite horizontal series of events will jib at the idea of an infinite vertical series of ever more basic causal laws. So the questioning has to be stopped; and that requires saying that my why questions are both self-answering, i.e., answerable on the basis of logic alone. That is to say that it is absolutely necessary both that every F event has a G-making power and that whatever has that power is followed by a G event. This, however, makes it absolutely necessary that every F event is followed by a G event; that gives absolutely necessary status to the causal law in question, which is to take the causal rationalist line about it.

That is how an explanatory rationalist is pushed towards causal rationalism. Causes help to explain events; but if causal laws are less than absolutely necessary then a causal explanation always leaves something unexplained, and the infinite regress of laws which this generates is not acceptable. I should perhaps mention that 1a3 can be seen as conjoining causal rationalism with a version of explanatory rationalism: causes necessitate, and nothing happens without a cause. But if that is what the axiom is saying, then Spinoza is wrong to say that the first conjunct is the converse of the second.

§9. SECOND ASPECT: THEISM

1. Spinoza was a pantheist, in that he identified God with the whole of reality. Thus he agreed with the atheist that reality cannot be divided into a portion which is God and one which is not. Although pantheist and atheist may seem to be poles apart, with one saying that everything is God and the other that nothing is, in the absence of an effective contrast between God and not-God we should not be quickly confident that there is any substantive disagreement at all.

It depends on what Spinoza means by ‘God’. Is he a genuine theist, or rather a mealy-mouthed or perhaps an ironical atheist? The answer must depend on why he selects ‘God’ as a name for the universe. If his reason hasn’t enough to do with the concept of God as ordinarily understood, then Spinoza's uses of that word do not express a theology, and we can view him as an atheist. Well, then, why does he choose the word ‘God’?

A pantheist might be abundantly a theist because his position is reachable by starting with Christian theology, say, and deleting the non-God part, i.e., contending that the ‘created world’ is an illusion and that all there is is God. But that is not a usual form for pantheism to take, and Spinoza's position is nothing like it. He does not hold that the everyday world is unreal, the only real thing being a God lurking behind the illusion of galaxies and comets and lizards and grocery shops and hopes and fears and headaches. On the contrary, what he calls ‘God’ is the familiar, everyday, natural world. If Spinoza and the atheist each pointed while saying, respectively, ‘That is all God’ and ‘None of that is God’, they would point to the very same world.

Some writers have disagreed with this; for example Caird, who said that Spinoza's metaphysic of God proceeds from his belief in the unreality of the given world.³ But these days no commentator as able as Caird would say such a thing. It has become clear that Spinoza was, in Wolfson's happy phrase, ‘no mystic, no idealist of the kind to whom everything that kicks and knocks and resists is unreal’.

2. Spinoza's main reason for liking ‘God’ as a name for the entire natural world is that the world comes closer than anything else to fitting the traditional Judaeo-Christian account of God. If God is to be infinite, eternal, not acted on by anything else, the ultimate source of the explanation of everything, and not susceptible to criticism by any valid standard, then God must be Nature as a whole; or so Spinoza thinks.

Of course, if God must be a person who is infinite, eternal, et cetera, then Spinoza has no candidate to offer. The most Godlike item he can find is Nature; but it does not completely fit the traditional account of God, and, in particular, it is not ‘a man, or like a man’ (2p3s; see also Letter 23 at IV/150/1). Many theologians in the tradition Spinoza is attacking would agree that God should not be conceived in an anthropomorphic way, and some held that no general term is true, in a single sense, of God and of human beings. But Spinoza wants a larger retraction than that. When he says that God is not ‘like a man’, he thinks he is in conflict even with the theologians who warned against anthropomorphism, and I believe he is right. If asked ‘Are we helped to understand the nature of God by being told that God is, in some broad but not empty sense, a person?’, those theologians would answer Yes and Spinoza would answer No. I know of only one passage in the Ethics that might suggest the contrary. Between 2p13 and 14 Spinoza presents a theory of ‘individuals’, meaning something like ‘organisms’. Near the end of that he says that one individual can be part of a larger one, which in turn is part of a larger, and so on to infinity, so that ‘we easily conceive the whole of Nature to be one individual’ (102/11). If this means that the entire universe is a sort of organism, then it implies that God has something closer to personhood than I have allowed. The remark is suspect, and Spinoza certainly rejects that implication of it, e.g., when he implies that to credit God with ‘will’ and ‘intellect’ in anything like the normal senses of these terms would be as a great a mistake as expecting Sirius (the Dog-star) to bark (1p17s at 62/34). Not that he is well placed to take this line. To argue cogently that the universe is not a person, he needs something he does not have, namely a satisfactory account of why a human being is a person.

It is a nuisance that English has pronouns which are always used for people and not freely used for anything else. We use ‘he’ for God, thinking of God as a person; if we came to think of God as impersonal, we should switch to ‘it’. No such choice faced Spinoza, because none of his half-dozen languages has personal pronouns. In Latin, for example, he had to use a masculine pronoun for God, to agree with the masculine noun Deus; but this masculinity is grammatical only, and implies nothing about the nature of the object. Whereas in English we must use ‘he’ in referring to a man and ‘it’ in referring to a pebble, in Latin a single pronoun serves for both, since homo and calculus are both masculine. As for ‘his’, ‘hers’, and ‘its’: there is a single Latin word for all of these, whatever the gender of the relevant noun.

In translating Spinoza into English, then, it is inaccurate to use ‘he’ and ‘his’ where God is in question. But it would be misleading to use ‘it’ and ‘its’, since that would get the tone wrong, implying that Spinoza keeps reminding his readers that the item in question is not a person. The problem is insoluble for a translator into English; but a commentator can step delicately around it, and I promise that it will give no trouble in this book, except of course in direct quotations from Spinoza.

3. So Spinoza holds that the natural world answers to many traditional descriptions of God; but not to all, and in particular not to the description ‘a person’. Isn’t that sufficient to atheize Spinoza's use of the name ‘God’? It was commonly thought to be so in his own day. Leibniz reports Arnauld's opinion that Spinoza was ‘the most impious and the most dangerous man of this century’, and adds: ‘He was truly an atheist, that is, he did not acknowledge any Providence which distributes good fortune and bad according to what is just.’⁴ If that is all it takes, then Spinoza was an atheist. But there is more to be said.

Spinoza had another reason for using the name ‘God’ for Nature as a whole—namely his view of Nature as a fit object for reverence, awe, and humble love, i.e., for the attitude traditionally reserved for God. While viewing Nature as an inexorably unfolding system of mental and physical events, rigidly controlled by necessary laws, and while having no hope that it could ‘love him in return’,⁵ Spinoza could still be genuinely awed by its extent, grandeur, complexity, and ultimate orderliness. He could thus regard Nature not only as the best subject for the metaphysical descriptions applied to God in the Judaeo-Christian tradition, but also as the best object of the attitudes which in that tradition are adopted towards God alone. In Pollock's words: ‘The Mosaic conception of the one God of Israel wedded to the Lucretian conception … of the one and inflexible nature of things; such is the mood Spinoza would have us bring to the questioning of the world, such the majesty and gravity of nature in his eyes.’⁶

I conclude that Spinoza's position is a kind of theism rather than of atheism. No doubt Santayana is right when he says, in his wonderful essay on Spinoza's religious aspect, that as compared with ‘some other masters of the spiritual life [Spinoza] was more positivistic by temperament and less specifically religious’.⁷ But Spinoza did accept pantheism as a kind of religion, and apparently did not think of himself as an atheist. Although as a young man he was expelled by the Jewish community, he remained permanently preoccupied with God's nature and relation to man. If he was not ‘drunk with God’ (as the poet Novalis said), he was obsessed with God. I am sure that he thought of himself as discovering things about God rather than as revealing that there is no God.

Not that the issue affects my engagement with the Ethics. Even if Spinoza saw his work as growing out of his early religious education, what he says about God in the Ethics is always to be explained by philosophical considerations, never by his having detoured to rescue something from the Judaeo-Christian tradition. He is awesome in his combination of strong religious concerns with great penetration, courage, and imperturbably cold-blooded honesty in his views about God. From my standpoint, which is concerned not with Spinoza's mental biography but with getting his help in discovering philosophical truth, it would make no difference if he saw himself as an atheist.

§10. THIRD ASPECT: NATURALISM ABOUT MANKIND

1. Spinoza could not regard humanity as specially favoured by a personal creator of the universe; so he lacked one common support for the view that humans have features not found elsewhere in the universe. He had indeed no patience with that view. His thinking is firmly grounded in the conviction that there is nothing fundamentally special about mankind as compared with chimpanzees and earthworms and cabbages and rivers; for Spinoza, man is just a part of Nature.

He offers to prove that each person is ‘a part of Nature’ in the weaker sense of always being at the mercy of his environment, never causally insulated from it (4p4). But what I call his ‘naturalism’ is the stronger thesis that what happens in any human being is fully explicable through the same laws that explain everything else. A good part of this is captured by 2p13 which says that our bodies are ordinary material systems, and 2p48 which applies determinism to us. But none of Spinoza's demonstrated and numbered propositions adequately expresses his profound conviction (not that he puts it this way) that the whole truth about human beings can be told in terms which are needed anyway to describe the rest of the universe, and that men differ only in degree and not in kind from all other parts of reality.

None of my formulations of Spinoza's naturalism quite captures it, however, for they all fit Leibniz's theory of man too, yet I would not call it naturalistic. When Leibniz says: ‘The foundations are everywhere the same; this is a fundamental maxim for me, which governs my whole philosophy’,⁸ Spinoza would applaud. But Leibniz's ‘foundations’ are what you might get by starting with mankind and forcing your whole metaphysic into that mould. According to him, the only things that there ultimately are are ‘monads’, sizeless substances, which include human minds; they are all on a continuum with minds or ‘souls’, and that is our main conceptual handle on them. For example, the varying states a monad is in, says Leibniz, are its varying ‘perceptions’ of the rest of the universe. I do not count this as ‘naturalistic’ because although it tells a uniform story about people and pebbles, it starts with people instead of at the other end. That makes my use of ‘naturalism’ a little vague, but not harmfully so.

Spinoza's naturalism about mankind is eloquently expressed in 3 Preface, where he protests against those who ‘conceive man in Nature as a dominion within a dominion’ because ‘they believe that man disturbs rather than follows the order of Nature’. And a little further on he says in effect that there is no pathology of Nature, and that any pathology in Nature must be perfectly natural and fully explicable. He reverts to this theme later on, when theorizing about certain affects. He says that pride and despondency are bad from the standpoint of ‘human advantage’, but adds a warning that that standpoint is a parochial one:

But the laws of Nature concern the common order of Nature, of which man is a part. I wished to remind my readers of this … in case anyone thought my purpose was only to tell about men's vices and their absurd deeds, and not to demonstrate the nature and properties of things. For, as I said in the Preface of Part 3, I consider men's affects and properties just like other natural things. And of course human affects, if they do not indicate man's power, at least indicate the power and skill of Nature, no less than many other things we wonder at and delight in contemplating.

(4p57s at 252/31)

This invites us to extend the range of things we can wonder at and delight in contemplating, by enlarging the scope of our understanding. Anyone will rejoice in the outer appearance of a live butterfly, but even the innards of a dead and dissected one may engender ‘wonder and delight’ as one learns how the organism functions—the complex, orderly processes which constitute the life of the butterfly. Anyone may delight in a fine waterfall, but more humdrum movements of water can also interest someone studying how mountains take water from the sky and then return it. It is a familiar line of thought, which Spinoza invites us to extend to human beings. He wants us to see that we might study the aetiology of pride or despondency—or of cruelty, cowardice, vanity, stupidity, or whatever—and find ‘wonder and delight’ in the intricacy and inevitability of the mechanisms that are involved.

2. He doesn’t hold that we should relate to our fellows only as students of biology and psychology: ‘But I continue to note concerning the affects’, he goes on, ‘those things which bring advantage to men, and those which bring them harm.’ Each human, Spinoza believes, has special reason for concern with things which are ‘good’ or ‘bad’ in the special sense of being conducive or inimical to the welfare of humans; this is because he thinks he can prove that self-interest is inevitable, and further that the real interests of any individual are tied to those of the rest of his species. But this special human concern of ours is, though natural and inevitable, also parochial. It is a domestic matter, with no metaphysical importance.

3. Other philosophers, such as Hobbes and Hume, have tried to carry through a naturalistic account of humanity; but Spinoza may be unique in how thoroughly he abides by his commitment to naturalism, refusing to slip back into treating humans as special in some basic way. Unique, that is, until our century: there are many unblinking naturalists now, because naturalism is easier now than it was in the seventeenth century, partly because of the nineteenth century science which suggests that Homo sapiens is a species of animal and the twentieth century science which suggests that animals are chemical systems.

4. Someone who thinks that the whole truth about humans can be told using only concepts which are necessary to describe the rest of the universe should want to tell enough of the story to satisfy us that the rest could be told without stepping over the line. That is what Spinoza tries to do.

He fails, however, in at least two ways, and probably in three. His account of us is impoverished: it positively excludes some large aspects of human nature, such as our ability to act because we have purposes. Also, some of what it includes is misconstructed out of the basic materials. And—the third possible failure—Spinoza's list of concepts for describing the rest of Nature is probably too long: he credits reality as a whole with a feature that seems to be confined if not to its human then to its animal fragment. I allude, of course, to the doctrine (mentioned in §3.3) that there is a mental realm which precisely mirrors the physical realm.

If Spinoza did not have that doctrine of panpsychism, his naturalism would commit him to some kind of materialism: he would have to show how a mostly nonmental world could generate small pockets of mentality through the same physical laws that reign throughout the rest of Nature. As things are, he lets himself off from that task by introducing mentality as a basic feature of the entire universe. Still, he does not shy away from specific problems about the mind: for example, he does not simply help himself to the concepts of perception, memory, pleasure and unpleasure, and belief; rather, he works hard to earn his right to employ these concepts. Unlike Leibniz, who was catapulted across much of the philosophy of mind by his doctrine that all substances are comparable to human minds, Spinoza keeps his feet on the ground, despite his panpsychism: he travels overland, the hard way, and we can learn a lot by following his steps.

Let me be clear about this. I do say that Spinoza's total naturalistic programme fails at both ends and in the middle; as though he undertook to build a sturdy mansion all out of wood, and achieved only a rickety shack using bricks as well as wood. But his attempt was a work of genius; and a thorough, candid study of it can be wonderfully instructive. The failures have at least as much to teach as the successes, if one attends not only to where Spinoza fails but why. I shall add a few words about this at the end of my next section.

§11. FOURTH ASPECT: CONCEPTUAL MINIMALISM

1. A philosopher embarking on a programme which is ‘naturalistic’ in my sense must start with a small stock of basic concepts—ones which are needed all through the description of the universe—and any others that he needs should be explained through these. Conceptual parsimony and constructive ingenuity, then, will be marks of any serious attempt to give a naturalistic account of the human condition.

Spinoza's parsimony, however, is not purely a result of his naturalism. He likes to work with exiguous raw materials, that being part of what I mean in calling him a conceptual minimalist. The tiny stock of basic concepts in the Ethics reflects Spinoza's intellectual temperament as well as his naturalistic programme.

2. It also has a basis in a doctrine of his which is sometimes wrongly called nominalism. He writes at times as though he were a nominalist, allowing existence only to particulars. In the Metaphyical Thoughts he writes that ‘Universals do not exist, and have no essence except that of particulars’ (I/263/5), and echoes of this linger on in the mature work. Some commentators have held him to this, trying to explain away the strands in his thought which go against it. They have said, for instance, that what he admits into his ontology is not really the universal extension but only the extended realm, which is a vast particular. That interpretation, although wrong, is plausible; but other doctrines of Spinoza's cannot be given plausible nominalistic readings. Most notably, in 2p40s2 he says that reason involves knowledge ‘of the properties of things’, which would be odd (to put it mildly) if he denied the existence of properties. Other facets of his work, too, become strange if they are combined with nominalism.⁹

The chief passage that looks nominalistic is 2p40s1; but most commentators agree that really it is not. Its topic is not the metaphysics of all general terms, but rather the semantics of some of them. Spinoza is arguing that general ‘notions’ of a certain kind are unfit for serious theoretical use because of how they are formed in our minds:

Those notions they call universal, like Man, Horse, Dog, etc., have arisen … because so many images (e.g. of men) are formed at one time in a human body that they surpass the power of imagining … to the point where the mind can imagine neither slight differences of the particular [men], such as the colour and size of each one, etc., nor their determinate number, and all it imagines distinctly is whatever is common to their effects on the human body. … And the mind expresses this by the word man, and predicates it of infinitely many particulars.

(2p40s1 at 121/13)

The ‘imagining’ spoken of here includes perception. When I see you, your body acts causally on mine, creating in it a physical ‘image’ of your body; and the mental counterpart of that is my ‘imagining’ of you. Now, my universal notion of man results from a piling up of these imaginings, and so it will be a fuzzy mess except to the extent that my imaginings of men have been alike. (In fact Spinoza thinks that each separate imagining is a mess, but he is not pressing that point here.) This alikeness, however, depends not on whether the men have been alike but on whether they have had like effects on my body, and that depends on thousands of coincidences.

Spinoza is making the metaphysical point that a ‘notion’ formed in that way cannot be expected to carve up Nature at a real joint.¹⁰ He is also making a point in social semantics, namely, that my universal notion of man is bound to be unlike yours, because my body differs from yours in how it has been affected by men and in which of those effects have stuck in the memory. ‘These notions … vary from one mind to another, in accordance with what the body has more often been affected by and what the mind imagines or recollects more easily.’

This is not a metaphysical view about all general terms, but only an aetiology and a criticism of some of them. It is not nominalism, but rather a rejection for theoretical purposes of sense-based universality. We shall see later that it is accompanied by a welcome for ‘rational universality’,¹¹ that is, for general notions coming from reason rather than from the senses.

And so, as I said, Spinoza has grounds in addition to his naturalism for confining himself to a skimpy stock of basic concepts; for in an enterprise such as the Ethics no weight should be carried by any of the sense-based notions treated in 2p40s1.

3. Spinoza thinks that nobody's general notion of man is fit for serious use. How then can he write his treatise on human freedom and human bondage? Since he did not regard himself as bound by the meanings of words, he could have given ‘man’ and ‘human’ new meanings, divorcing them from the cluster of general notions by which they are backed in ordinary language. In fact, he performs no such manoeuvre; the Ethics contains no redefinition of ‘man’, explicit or implicit; and in §64 I shall explain how Spinoza does get the term ‘man’ into his theorising without putting onto it a load which, according to 2p40s1, it cannot carry.

4. Back now to conceptual minimalism. Although we have found two theoretical grounds for Spinoza's economy in basic concepts, I still see his minimalism more as a fundamental aspect of his cast of mind than as forced on him by other opinions and attitudes.

And a second strand in the minimalism could not come from either of those theoretical sources: I refer to Spinoza's evident liking for drastic simplicity in his conceptual constructions. Of course when philosophers go wrong they usually oversimplify, and Spinoza often goes wrong; but that is not the whole story. Nor is it just that he likes to express himself in brief, pithy utterances, offering the reader dense little wads of philosophy to be chewed on for hours. In addition, he takes pleasure in starting at a dizzyingly abstract level and moving in about two short steps down into the thick of everyday life. I cannot illustrate this before expounding details in the Ethics; but examples will come in due course.

At the end of §10 I spoke of how much we can learn from Spinoza's successes and, especially, his failures. It is his minimalism that makes his work so instructive. If you set a mechanical genius to build an automobile engine out of a Meccano set, you won’t get a working engine from him, but as you watch him fail you will learn a lot about automobile engines.

§12. WHAT IS DUALISM?

1. Because most of Nature seems to be mindless, any naturalist (in my sense of that term) is likely to hold that human beings are fully describable in physical terms—that mental terminology may be brought in through definitions but is not needed on the ground floor.

That was not Spinoza's position, however. He never questioned the truth of a doctrine that I shall call property dualism, which says that the properties of things can be cleanly split into two groups, mental and physical, with no property belonging at once to both groups; this being so understood as to rule out any defining of mental terms through physical ones.¹² When I speak of Spinoza's dualism, I allude to that doctrine which he shares with Descartes. He rejects Descartes's stronger substance dualism, which adds to property dualism the thesis that no one thing can have properties of both kinds—so that a human being is not one thing but two, e.g., a worried mind and an emaciated body.

Spinoza, then, thinks that a person's mental aspects are part of the basic story about her; this, combined with his naturalism, leads him to say that all of Nature has mental aspects. I shall discuss later his case for this bold hypothesis; my present topic is the dualism which lies in its background. I want to get clearer about what the dualism is: my statement of it above, though no worse than most in the literature, is too vague to support a proper study of the Ethics. We need a formulation which is tighter and clearer, yet still in touch with Descartes's thought and with Spinoza's; and it should not make dualism come out as obviously true or as easily demonstrably false.

2. I shall approach dualism by way of a thesis about a split through a certain range of concepts, rather than properties; that is, abstract items in logical space rather than universal items in the contingent world. But because the end-point is a thesis about properties—and indeed about properties which some things could have and some things could lack—many of our concepts can be kept out of range of the split. For example, the split has nothing to do with the concept of existence, for if that corresponds to a property it’s a property which necessarily everything has. Nor do tautologous or contradictory concepts come under the axe.

More generally, our entire apparatus of logical concepts, and everything we can construct out of them alone, should be kept out of reach of the dualist divide. Conceptual dualism is supposed to be a split within a scheme which has some kind of unity, but if the very logic is split there is no unity left. Anyway, who would want a mental/physical dualism that cut so wide and deep that it involved a mentalistic logic and a different physicalistic logic? Let us, then, exclude from the dualist split all our logical concepts and, for similar reasons, our numerical concepts.

There will have to be other exclusions as well. But first let us take a preliminary look at the dualist split.

3. In expounding it, I shall speak of concepts as entailing one another, borrowing this from entailments between the corresponding propositions. Thus, F entails G just in case ‘Something is F’ entails ‘Something is G’, and R entails S just in case ‘Something has R to something’ entails ‘Something has S to something’, and so on. I am speaking only of entailments which hold by virtue of the nature of these concepts. Spinoza has general metaphysical views that imply that whatever is thinking must also be extended, but I do not want to commit him to holding that thought entails extension; and I head off that implication by stipulating that for F to entail G there must be between the corresponding propositions an entailment which holds because of the nature of the concepts themselves and not by virtue of any more general metaphysical thesis. This is unsatisfactory, and I hope it can be improved on, but at present it is the best I can do.

Two bits of technical terminology, to be used briefly and then discarded: two concepts are logically ‘in touch’ if either entails the other; and two concepts are logically ‘connected’ if they are the termini of some sequence of concepts each of which is logically in touch with its immediate neighbours. Thus, cubic is connected with spherical, because each is in touch with shaped. And jealous is connected with aware of danger, because the former is in touch with in an emotional state which is in touch with afraid which is in touch with aware of danger. Notice that the relations I have defined are symmetrical, although they are defined in terms of entailment which is not.

Now, property dualism rests on the idea that all our concepts fall into two great groups—call them ‘categories’—such that each concept is connected with every member of its own category and with no member of the other.

Can we put every concept into one category or the other, apart from those we have set aside as belonging to logic or arithmetic, or as being tautological or contradictory? No. If dualism is to have a hope of being true, we must exclude some more concepts from its range. If F belongs to one category and G to the other, then the concept F-and-G is connected with both, thus contradicting dualism as I have formulated it. For example, the concept happy and strong is logically connected with both the mental and the physical. But it is easy to modify dualism so as to avoid this difficulty: instead of saying that all our concepts can be segregated into the two categories, we can say that they can all be resolved into concepts each of which belongs in one category or the other.

This will trivialize our dualism unless some constraints are laid on the notion of ‘resolving’. I propose a very strong constraint: a concept F can be ‘resolved’ into concepts G₁, …, G_n only if F involves no concepts except G₁, …, G_n and concepts which are explicitly excluded from the scope of our dualism. Thus, our explicit exclusion of logical concepts, on any reasonable understanding of what that means, will cover the relations used in constructing strong and happy and strong or happy out of strong and happy. But there are other combinations—such as strong when happy and strong because happy—which are still making trouble for us. These will easily succumb, however, if the concepts of time and cause are put out of reach of the dualist split.

4. The dualist is certainly under pressure to treat temporal concepts as transcategorial, i.e., to protect them from the split. Because there is no way of resolving the metrical brief or the topological before into components which divide between the two categories, we must either make temporal concepts transcategorial or else assign them en bloc to one category or the other. But either assignment seems arbitrary. How could we decide between treating brief pain as a mixture of physical brief and mental pain, and treating brief storm as a mixture of mental brief and physical storm? It might be argued that our temporal metric is essentially rooted in measures of duration, which must be based on regular kinds of physical process, so that the physical category is indeed the home ground of brief and its kin. I am not convinced that this is right; but in any case it can be no help with topological concepts such as simultaneity and beforeness. One might in desperation suggest that words expressing temporal topology are ambiguous—e.g., that ‘… occurs before …’ expresses a different concept in each category. But then what could we make of temporal relations between mental items and physical ones? And are we to say that the argument ‘The headache preceded the storm, which preceded the flood, so the headache came before the flood’ is fallacious through ambiguity? Surely not!

So there will be trouble if temporal concepts are not pushed out of reach of the dualist split. But we can also give independent reasons for this. Just as the dualist needs a transcategorial logic if his two categories are to be held together within a single intellectual system, so he also needs some nonlogical transcategorial concepts if his categories are to apply to anything he can regard as a single world.¹³ Now, I contend that a multitude of items can be taken as pertaining to a single reality only if they belong to a single temporal scheme: the reality must not only be temporal but must involve one time. For temporal concepts to be transcategorial, they need only be applicable within each category; but to secure the unity of the world, temporal topology must apply between the categories, in such statements as ‘the pain stopped before the sun went down’. I base this on some thoughts of Kant's, and especially his thesis that we cannot envisage a personal inner history which does not occur in a single time-series, and cannot apprehend any reality except by somehow embracing it within our own inner histories. But I cannot follow that up in detail: it ranges too far into unSpinozistic territory. Anyway, even non-Kantians would agree that an intending dualist should keep temporal concepts out of reach of the split.

5. Now for causal concepts. One who holds that physical events alone have causal powers may be able to locate causal concepts purely within the category of the physical, though he may have problems if he allows that mental events, although powerless, can be caused. But Spinoza, for whom the situation is symmetrical with regard to causal concepts, must treat them as transcategorial.

A Kantian justification could be given for this too: Kant argued powerfully that any conceivable reality which we could know about must be held together causally, because temporal order presupposes causal order.¹⁴ It is not maintained, this time, that there must be causal flow between the categories, as there must be temporal relations between them; but at least there must be causal laws operating within each category, and so causal concepts must be protected from the dualist split.

Spinoza, however, would say that he needs no such justification, since for him there is no need to add causal concepts to a transcategorial list which already contains our full equipment of logical concepts. His causal rationalism equates causality with logical entailment, and that was pushed out of reach of the split right at the outset. Oddly enough, if he held causal laws to be general brute facts, he would again be spared from adding causality to his transcategorial list. For he would then be holding that the concept of causality can be analysed in terms of true and universal and many instances and the like, and these all belong to that logical and numerical apparatus which was the first thing to be declared transcategorial.

6. We have been establishing the class C of concepts whose members consist of our logico-numerical apparatus, purely temporal and causal concepts, and nothing else. And the notion of ‘resolving’ which I introduced in 3 now becomes: F can be resolved into G₁, …, G_n if and only if F involves no concepts except G₁, …, G_n and members of C. With that in hand, we can formulate dualism as follows:

All our non-tautological and non-contradictory concepts which do not belong in C can be resolved into ones which fall into two categories, the mental and the physical, such that every member of each category is logically connected (a) with every member of that category and (b) with no member of the other category except through non-logical members of C.

Notice that C is referred to twice—once to keep brief etc. out of both categories, and once to allow brief fight in one category and brief worry in the other, although they are logically connected through brief. Of course brief worry can be resolved into brief and worry, according to my definition of ‘resolved’; but we can’t handle this sort of difficulty by supposing that the members of the categories have been ‘resolved’ down far enough to contain no temporal or causal implications. I am sure that all mental and physical concepts have temporal import, and probably causal import as well.

An untrivial dualism requires a complete list of the kinds of concepts that have transcategorial status—a list such as I have tried to provide in specifying the contents of the class C. If we specified C merely as containing ‘whatever concepts it has to contain to make dualism true’, that would trivialize the position. Is my list of sorts of transcategorial concepts complete? Well, if you want a unified theory about a single world, you will make a pretty good start if you have a unified logic and universal applicability of temporal and causal concepts; but I cannot prove that there are no other concepts that must be transcategorial if dualism is to have a hope of being true, although I do know that my study of Spinoza won’t force any more on us.

For a while I thought that the concept of aboutness must be transcategorial, so that the two-legged concept thinking about a rose could be resolved into thinking and rose, and that this would prevent that concept from counting against my version of dualism. But really there was no need. Although thinking about a rose does in a fashion involve the concept rose, it is not in my sense ‘logically connected’ with it: ‘Something is thinking about a rose’ neither entails nor is entailed by ‘Something is a rose’.

Am I in trouble from the concept ambiguous written word? A written word is physical; ambiguity is having more than one meaning; and many of us think that meanings are mental. It would be miserable if I had to add to my transcategorial list a concept of having which could relate words to their meanings! I think that can be avoided, however. With great labour and at the expense of some complexity we could resolve ambiguous written word into concepts belonging to the separate categories, with the aid of nothing but logical, temporal and causal concepts, i.e., ones already belonging to C. Here I rely on the general approach to meaning taken in my book Linguistic Behaviour.

It has been suggested to me that the concept of part/whole is transcategorial. I think it must be. We have the notion of the parts of physical things, and it is not obvious that we don’t also have the concept of parts of thoughts and sensations; in addition to which we do speak about temporal parts (of events) and logical parts (of propositions or concepts). I would be reluctant to handle all this through a plea of multiple ambiguity. On the other hand, I am not sure that part/whole has to be added to the existing membership of C; perhaps it is already in C, as part of our logical apparatus. My unsureness about this is a sign of unclarity about the scope of the ‘logical’, which brings an admitted softness into my formulation of dualism.

7. Descartes sometimes alludes to transcategorial concepts, giving two laconic lists: ‘substance, duration, number, and other such things’,¹⁵ and ‘being, number, duration etc.’.¹⁶ I have already dealt with being, number and duration, and also with cause which is inexplicably absent from Descartes's lists. There remains substance, about which I should say something.

The root sense of ‘substance’ in early modern philosophy is just that of ‘thing’, or ‘subject of predication’: there are properties or qualities or attributes or modes, and there are substances that have them. When ‘substance’ is used in that manner, the concept of substance is involved whenever there is predication; and that suffices to include substance in that logical apparatus which we have already protected from the dualist split. If the latter is to cut only through properties, then certainly substance must be kept out of its reach somehow, for there cannot be a property of substantiality: if there were, it would be the property of being-a-thing-that-has-properties, which leads to absurdity.¹⁷

In Descartes's hands, and even more in Spinoza's and Leibniz's, the term ‘substance’ takes on a richer meaning. Spinoza, for example, denies the title ‘substance’ to ordinary physical things, reserving it for ultimate, basic, irreducible subjects of predication; and, as we shall see, he thinks that these must be very special indeed. However, the specialness of a ‘substance’ in this rich sense consists in its being logically and causally self-sufficient, and in its lasting for ever; and this is all provided for by concepts already assigned to C. If C needs to be enlarged, therefore, it is not because of anything in Descartes's lists.

8. My formulation of dualism throws no light on what it is for a concept to be mentalistic. It says merely that our concepts can be split into two groups, to which it gives the labels ‘mental’ and ‘physical’; but unless we already have meanings for these, all I have presented is the thesis that some dualism of concepts is true. That is to be expected: the formulation is too abstract to help with any such concepts as those of mentality and physicality. How much this matters for our understanding of Spinoza will appear shortly.

§13. FIFTH ASPECT: DUALISM

1. Spinoza assumed a dualism like the one I have been developing: he thought there was a deep logical split through most of our concepts. That would forbid him to accept philosophical behaviourism, according to which mental concepts are logically derivable from physical ones; and it would rule out phenomenalism, as that allows entailments running the other way.

Dualism as so far formulated implies that there is no logically true conditional whose antecedent uses only concepts from one category and whose consequent uses only ones from the other. (I here ignore conditionals with impossible antecedents or necessary consequents.) For Spinoza, that is equivalent to saying that there is no causally true conditional—no causal law—with the antecedent confined to one category and the consequent to the other. On the reasonable assumption that all mental and physical properties have concepts corresponding to them, it follows that there is no causal flow between the two sorts of property—e.g., a fact about an entity's physical properties can’t be explained by reference to any entity's mental properties.

Spinoza uses the term ‘attribute’ for what is common to all the properties in one group. Thus, mentality and physicality are attributes. In calling them distinct attributes, Spinoza is saying more than is asserted by mere dualism about them, in my sense. Dualism implies that neither is reducible to the other and there is no third item to which both are reducible; but Spinoza holds further that neither is analysable in terms of anything else. According to him, each is a basic, irreducible way of being.

He seems to hold that there are many attributes in addition to these two; indeed, he speaks of ‘infinite attributes’. Fortunately, our two are the only ones that play a significant part in the Ethics, so that ‘dualism’ is an acceptable label for what actively goes on in that work. Furthermore, in developing dualism in §12, I have stated it in terms of a split through ‘our’ concepts, and that keeps my formulation well clear of Spinoza's notion of ‘infinite attributes’; for he does not think that we have concepts of any attributes other than the famous two. Still, because others are being theoretically allowed for, we find Spinoza speaking of the causal quarantine not between the mental and the physical but more generally between attributes.

He states this explicitly in 2p6, according to which all causal flow is confined within the attributes and never goes across from one to the other. That is derived from 1p10 which says that each attribute is ‘conceived through itself’, meaning that you can’t get at it conceptually through anything else, which entails that the attributes are not logically connected with one another. In the shift from 1p10 to 2p6 we can see Spinoza's causal rationalism silently at work.

2. Where I speak of physicality Spinoza speaks of the attribute of ‘extension’. That reflects an important doctrine about the metaphysics of matter, namely, that the fundamental spatial reality is space, and that what we loosely treat as things in it are really intermittent thickenings of it. On this view, the difference between matter and empty space, ordinarily so-called, does not belong to basic metaphysics—it is merely the difference between stony regions of space and airy ones, so to speak. In chapter 4 I shall say more about how this part of Spinoza's doctrine works. In the meantime, we can say that it answers the question of what there is on that side of the dualist divide. The category of what I call the ‘physical’ has in it every property whose corresponding concept includes the concept of extension; and if you want an explanation of ‘extension’, Spinoza and I will refer you to Euclidian geometry.

3. What about the other side of the dualist split? We might try to explain it simply as what is left over when the physical is subtracted;¹⁸ but that would not be acceptable to Spinoza, for he certainly held that there could be more than two attributes, and that forbids us to define ‘mental’ as ‘not physical’. He has no other explanation of mentality either, apparently thinking that this concept lies too deep to be capable of analytical explanation.

We can, however, get some light on his view of the mental by noting that his word for this attribute is ‘thought’. In this, as in his use of ‘extension’ for the category of the physical, Spinoza follows Descartes, largely for Cartesian reasons. It seems odd to us to use ‘thought’—or the Latin cogitatio or the French pensée—to cover the whole range of the mental. It may reflect a tendency to use ‘thought’ more broadly than we would find natural, but it also, more certainly, indicates a narrowed emphasis within the range of the mental. Although Spinoza allows that mental events include sensings and feelings as well as thinkings, his stress is mostly on the latter.

4. I have described Spinoza's dualism as a fundamental aspect of his cast of mind, rather than as explicit doctrine. He has plenty of doctrine based on the view that thought and extension are basic and mutually irreducible ways of being; but that view itself is an undefended assumption. In 3p2 he says that there is no causal flow either way between body and mind, but his argument for that runs back only to 2p6, which says that there is no causal flow between attributes; the argument merely assumes that body and mind pertain to different attributes.

We might see Spinoza as arguing that thought is an attribute, and pointing to a similar argument for extension, in 2p1d. But the argument is no good. It amounts to this: ‘Particular mental events occur within Nature; so Nature must have some attribute within which they all fall; therefore thought is an attribute of Nature.’ This assumes that mental events—thoughts—are a natural class and not a mere ragbag like the class of surprising events or nuisances or obstacles; and it also assumes that there are attributes, i.e., that the notion of a basic way of being is not a mistake. But granting both assumptions, what follows is only that thoughts fall under thought and fall under an attribute; it doesn’t follow that thought is an attribute. For all this argument shows to the contrary, thought could be a special case of something more basic, of which extension is another special case. I don’t think Spinoza was capable of asking whether thought is an attribute of Nature, as an addition to the question of whether thought is instantiated by Nature; and the wording of 2p1 is evidence for that. As I said, his dualism is assumed rather than defended.

5. A few words about property dualism in Descartes. He accepted the underlying concept dualism, but he was not enough of a causal rationalist to infer a causal barrier between the two categories of property. In some of his works, indeed, interaction is freely allowed, and I shall be discussing Spinoza's criticism of that in §32. Usually, however, Descartes was uneasy about allowing causal flow between thought and extension, and many of his followers rejected it outright. This was because they held that causing consists in one thing's giving a property to another, handing over part of its own resources. That is plausible enough if you think of transfers of motion, of heat et cetera, but as a general account of causation it is probably false and certainly disputable. But the Cartesians thought it was true. As Descartes said, ‘Where else can the effect get its reality from if not from the cause?’¹⁹ And from this, together with concept dualism about mind and body, they inferred a kind of causal dualism: a causal law whose instances are transfers must have some concept occurring in both antecedent and consequent, to represent what the cause loses and the effect gains; which implies that the antecedent cannot belong wholly to one category and the consequent wholly to the other. So there can be no causal transfers, and thus no causal influence, between mental and physical.

We find this thought in Spinoza also. In Letter 4 he writes: ‘Of things which have nothing in common between them, one cannot be a cause of the other. For if the effect has nothing in common with the cause, whatever it receives [in the causal transaction] it would receive from nothing.’ Whether this line of thought was at work in his later years, encouraging his view that if one thing has nothing in common with another it cannot be its cause (1p3), I do not know. But that view was certainly supported, in his case, by causal rationalism.

§14. PSYCHOLOGY AND LOGIC

1. Spinoza's dualism involves a logical and causal split between extension and thought—between physics and psychology, as one might say. But there is a way of taking his term ‘thought’ which divorces it from psychology, and makes his position different from the property dualism of Descartes. I don’t quite say that Spinoza conflates these two senses of ‘thought’; but both are perceptibly at work in the Ethics, and are not explicitly confronted and sorted out.

I shall approach the second sense by focusing on Spinoza's Latin word idea, which means something like ‘particular item belonging to the attribute of thought’. We translate it by ‘idea’, and think of it as mental—for someone to have an idea, we assume, is for a mental event to occur or a mental state to obtain. And we take it that my idea of Socrates is a different particular from your similar idea of Socrates: one is an episode in my mind, the other in yours.

But there is a way of taking ‘idea’ in which ideas are not mental at all. In this second meaning, there is an idea of Socrates which you and I might both have before our minds as an object or content of our thoughts. This is to use ‘idea’ to mean what ‘concept’ means in philosophy these days: an ‘idea’ is a logical item, a constituent of propositions, something thinkable by many people. Much of the time Spinoza takes ideas to be propositionally structured, i.e., to be of the form ‘that P’, where P stands for a sentence; and then on the psychological reading an ‘idea’ is a state or episode of believing that P or the like, while on the logical reading it is just the proposition that P.

A single propositional content can occur in beliefs that are not only numerically distinct, because occurring in different minds, but also vastly dissimilar from one another. My belief in determinism is intermittent and tentative, while yours is fervent and continuous, but what we believe is the same.

2. Whether there really are such items as concepts and propositions is an old topic of debate amongst philosophers. A resounding affirmative answer is given in Frege's ‘The Thought’,²⁰ which powerfully argues that if someone believes that grass is green, this involves not just a thinking but something to which the thinking is done. Frege calls it ein Gedanke, which, though inevitably translated as ‘a thought’, is not a thinking but rather the possible object of a thinking, as a football is a possible object of a kicking. Since this item is not a (mental) thinking or a (physical) sentence, Frege says, we must move from dualism to trialism: ‘Thoughts are neither things of the outer world nor [mental things]. A third realm must be recognized.’²¹

Surely he is right that ‘thoughts’ in his sense—propositions and concepts—must be distinguished from both mental items and physical ones. Those distinctions are needed for it to be true that two people share a belief; that there is a concept which Smith did not grasp until last week; that the entailment of Grass is coloured by Grass is green did not come in with the English language; and so on. But that is not necessarily to endorse Frege's making this a matter of basic metaphysics, his postulating ‘a third realm’. The concepts of concept and proposition may be constructible out of mental or physical concepts—e.g., a proposition may be a special class of sentences, or a concept may be a special class of mental episodes. But even if the so-called third realm can be constructed out of the other two, it will not be a construction by simple identification. Whatever the ultimate metaphysics of the matter, we are not entitled simply to ignore the differences between concepts and propositions on the one hand and mental episodes or linguistic expressions on the other; and we can applaud Frege's warning about that, even while suspending judgment on his third-realm metaphysics. When I speak of ‘third realm’ entities, I mean entities of the sort Frege would put into the third realm.

Spinoza does ignore the difference between mental items and third-realm ones. He does not argue that belief-contents are constructs out of believings, and that logic really comes down in a complicated way to psychology. In fact, he says nothing about logic except for the passing remark that logic is normative for psychology, laying down rules for how one ought to think (5 Preface at 277/13). But from time to time he makes his psychology double as a logic as well, taking the term ‘idea’ to stand indifferently for a mental item and for a concept or proposition.

3. If this seems like such a crass error as to make Spinoza unworthy of further attention or my account of him incredible, let me plead in his defence or in mine.

He was not alone in this tendency to conflate logic with psychology. Leibniz was also guilty of a subtle form of it, and it is less subtly present in the work of Locke, who is led by it to put elementary logical truths into the same box as psychological self-descriptions, on the grounds that each can be known by inspecting one's ‘ideas’.²² This should not be confused with Locke's better known conflation of the sensory with the intellectual. He thought that when (a) a sensing occurs, the mind is cognitively related to (b) an inner object, an image or sense-datum or sensory ‘idea’; and that when (c) a thinking occurs, the mind is cognitively related to (d) an intellectual ‘idea’. Now, the conflation of psychology with logic is a smudging of the line between (c) thinkings and (d) concepts or propositions which are their objects or contents;²³ whereas Locke's sensory/intellectual conflation is a smudging of the line between (a)-(b) on the one hand and (c)-(d) on the other. It is extremely unsatisfactory to describe the latter, as I have done, as Locke's propensity to use ‘idea’ to cover ‘both sense-data and concepts’, and then, on the same page and purportedly on the same topic, to speak of his ‘assimilating the sensory far too closely to the intellectual’.²⁴ Those two phrases pick out two entirely different philosophical mistakes: one, shared with Leibniz, about the nature of the subject-matter of logic; the other, of which Leibniz was innocent, about the internal geography of the mind. The difficulty that some of us have had in coming clear about this is some measure of the temptingness of the conflation of (c) the psychological with (d) the logical.

There have been two main alternatives to the view of logic as psychology. One sees it as physics. This view, different forms of which have been defended by Mill and by Quine, has the merit of providing an answer to the question ‘How can anyone have reason to believe that it is necessary that P?’ Locke has an answer to this question, namely, ‘Since the necessity of P stems from relations between our ideas, we can know about it by introspection’; and Mill and Quine have a different answer: ‘For P to be necessary is just for it to be true and scientifically well entrenched; and we can have the same grounds for believing that as we can for any other scientific belief, namely, that it stands up well to the battering it takes as we continue with the scientific enterprise.’ So there is a possible epistemology for logic as psychology, and for logic as physics; but for the currently popular view of logic—according to which necessary truths are about ‘possible worlds’, items belonging to Frege's ‘third realm’ between the mental and the physical—there is no coherent epistemology at all. We speak of being guided by our ‘intuitions’ about what is possible, but third-realmery allows us no explanation of how such intuitions relate to their subject-matter or, therefore, why they should be trusted. This should warn us against being too smugly condescending about the seventeenth century assumption that necessary truths are at bottom truths about one's own mind.

4. Some writers on Spinoza seem not to have noticed that he has logic jumbled up with his psychology. I know of only one, Balz, who has contended that Spinoza is offering only a logic and no psychology at all:

Spinoza's problem is not set in terms of the relation of psychical ideas in a spiritual substance to modes of a physical, extended substance. Nor does Spinoza split existence into halves, things of the mind and things of extension. Idea means for Spinoza what essence had denoted in scholastic philosophy. It is a logical entity, and in explicated form and verbally expressed, it is a definition.²⁵

The flavour of Balz's vigorous working out of this theme is conveyed by his treatment of two terms which are usually thought to be mental, namely anima (soul, but connected with ‘animated’) and mens (mind). According to Balz, one belongs to biology and the other to logic; psychology, in the sense of the theory of what is ‘psychic or spiritual’, is squeezed out.

This is not a mad view of the Ethics. It has in its favour many remarks about ‘thought’ or ‘ideas’ which make better sense as logic than as psychology, and occasional places were Spinoza seems to say outright that ideas are logical entities. In deploying his doctrine about ‘ideas of ideas’, for instance, Spinoza says: ‘The idea of an idea is nothing but the form of the idea …’ (2p21s), and ‘form’ is sometimes a term pertaining not to psychology but to something like Frege's third realm. And there is precedent for this way of using the term ‘idea’: Descartes sometimes adopted it, saying that ‘ideas themselves are nothing but forms’.²⁶ The word ‘idea’ was given to philosophy by Plato, for whom ideas were not mental, but were objective, mind-independent, third-realm entities.

As for doctrines which go better as logic than as psychology: Balz mentions as a prime example Spinoza's doctrine that corresponding to every idea there is an idea of it, and an idea of the latter, and an idea of that, and so on ad infinitum.²⁷ These infinite sequences do indeed seem implausible as psychology and all right as logic: there is the proposition that P, the proposition that the proposition that P is true, … et cetera.

Yet Curley, who construes ‘the attribute of thought’ purely logically through most of Part 2 of the Ethics,²⁸ switches to psychology precisely at the point where ‘ideas of ideas’ come in, taking the latter to be a doctrine about self-consciousness. This is natural too. For although the doctrine itself is easier to swallow when understood logically, Spinoza's justification of it, namely, ‘For as soon as someone knows something, he thereby knows that he knows it, and at the same time knows that he knows that he knows, and so on to infinity’ (2p21s), reads so much like psychology that one cannot think it was intended as logic. If Spinoza meant only to make a logical claim concerning propositions about propositions, how could he hope to convey it in words like ‘as soon as someone knows something’? On passages like this, Balz is prudently silent.

The Balz thesis is not supportable, and in chapter 6 I shall also argue against Curley's restricted version of it. There is too much in the Ethics that must be read psychologically, and Balz is forced to misrepresent these passages.²⁹

Notes

1. Roth, Spinoza, p. 234.

2. For a dissenting view see F. Akkerman, ‘L’édition de Gebhardt de l’Ethique de Spinoza et ses sources', Raison présente no. 43 (1977), at p. 43.

3. For more on this, see Curley, ‘Spinoza's Moral Philosophy’.

4. Wolfson, The Philosophy of Spinoza.

5. John Aubrey, Brief Lives, ‘Thomas Hobbes’.

6. Hampshire, Spinoza, p. 30.

7. This is well argued for in Barker, ‘Notes on the Second Part of Spinoza's Ethics’.

8. On this point, Wolfson is useful: The Philosophy of Spinoza, vol. 1, pp. 45-55.

9. New Essays, p. 450.

10. For more on this, see the note on ‘known through itself’ in the Glossary-Index of Curley's forthcoming edition of Spinoza's works in English translation.

11. I. Kant, Critique of Pure Reason, B xiii. See also F. Bacon, Novum Organum I, 100.

12. This line of thought is developed by Fløistad, ‘Spinoza's Theory of Knowledge in the Ethics’.

13. Wolfson offers four conjectures about that (The Philosophy of Spinoza, vol. 1, pp. 57-9). An interesting further suggestion is propounded by Efraim Shmueli, ‘The Geometrical Method, Personal Caution, and the Ideal of Tolerance’, in Shahan.

14. Loemker, pp. 532, 555.

15. Loemker, p. 663.

16. The opposing view is vigorously argued in A. Wolf's introduction to his edition of The Correspondence of Spinoza (London, 1928).

17. The only argument I know of against causal rationalism is a Humean one given in my Locke, Berkeley, Hume: Central Themes (Oxford, 1971), §61, using a premiss about the nature of absolute necessity which few philosophers today would accept.

18. Caird, Spinoza, pp. 20-25. According to Parkinson's ‘Hegel, Pantheism, and Spinoza’, Hegel accused Spinoza of being committed to believing, and of actually believing, that the familiar given world is unreal. Parkinson defends him against both charges.

19. G. W. Leibniz, Sämtliche Schriften und Briefe II.1 (Darmstadt, 1926), p. 535.

20. ‘He who loves God cannot try to get God to love him in return.’ (5p19)

21. Pollock, Spinoza, p. 143.

22. George Santayana, ‘Ultimate Religion’, in I. Erdman (ed.), The Philosophy of Santayana (The Modern Library, n.d.), p. 588.

23. New Essays, p. 490.

24. For a good survey of them, see Haserot, ‘Spinoza and the Status of Universals’. Friedman, in ‘An Overview of Spinoza's Ethics’ at pp. 89-90, puts clearly the view that Spinoza was a nominalist.

25. This is confirmed in 2p18s at 107/9. In drawing this conclusion Spinoza is implying that what kind or property is picked out by a general term is determined purely by what is in the mind of the term's user. That is probably false, for reasons given sketchily in the New Essays (e.g., p. 345) and re-presented in a more disciplined fashion in our own time by Kripke and Putnam.

26. Haserot's phrase—op. cit., p. 53.

27. To avoid the ugly words ‘mentalistic’ and ‘physicalistic’, I speak of properties, concepts, and words as ‘mental’ or ‘physical’, meaning that they pertain to the mental or the physical.

28. I am here guided by John Passmore, Philosophical Reasoning (London, 1961), ch.3.

29. The most accessible version of the Kantian argument is in my Kant's Analytic (Cambridge, 1966), §54. See also C. Peacocke, Holistic Explanation (Oxford, 1979), p. 186.

30. Meditation III, in Alquié, vol. 2, p. 443. The Latin version says ‘… and any other such things as there may be’.

31. Letter to the Princess Elizabeth, in Alquié, vol. 3, p. 19.

32. The point is developed in my Locke, Berkeley, Hume: Central Themes (Oxford, 1971), §11.

33. That account of the mental is suggested in Matson, ‘Spinoza's Theory of Mind’, p. 51.

34. Meditation III, in Alquié, vol. 2, p. 438.

35. G. Frege, ‘The Thought: a Logical Inquiry’, translated by A. and M. Quinton in Mind vol. 65 (1965), pp. 289-31.

36. Ibid., p. 302.

37. For textual and other details, see the abridged edition of the New Essays (Cambridge, 1982), pp. xxi-xxiii.

38. There is no comparable mistake of conflating (a) with (b). The mistake there is to believe that (b) exists, i.e., that when someone is in a sensory state his mind has an inner object.

39. J. Bennett, Locke, Berkeley, Hume: Central Themes, p. 25.

40. Balz, Idea and Essence in the Philosophies of Hobbes and Spinoza, p. 30.

41. Reply to the Fourth Objections, Alquié, vol. 2, p. 673.

42. Balz, op. cit., p. 71.

43. Curley, Spinoza's Metaphysics; see for example p. 124.

44. See for example Balz's p. 52 for an attempt to give a third realm reading to Spinoza's phrase ‘action of the mind’.

Abbreviations

In the text I use what is becoming the standard method of referring to items in the Ethics—the method according to which ‘2p7s’ names Part 2, Proposition 7, Scholium, and according to which 4d2 names Part 4, definition 2. In a context where there is no possible ambiguity, I sometimes drop initial numerals from references, putting ‘p7s’ instead of ‘2p7s’. In these references a comma means ‘and’. Thus the expression ‘5p23,d’ refers to Part 5, Proposition 23, and its demonstration.

References to passages within long scholia, prefaces, appendixes, et cetera are given through page and line in the Gebhardt edition of Spinoza's works. Thus ‘1 Appendix 80/9’ refers to the passage in the Appendix to Part 1 which begins at line 9 on p. 80 of the Gebhardt edition, volume II. Where a reference of this kind is to any work but the Ethics, it is given with the volume number first—e.g., ‘Letter 4 at IV/14/15’. This system is at last widely usable because Curley's forthcoming edition of the works in English will, happily, provide Gebhardt's page and line numbers in the margin.

In footnotes and Bibliography, I use the following abbreviated names for certain works which are mentioned a number of times:

Alquié: Ferdinand Alquié (ed.), Oeuvres philosophiques de Descartes, vol. 1 (Paris, 1963), vol. 2 (1967), vol. 3 (1973).

Descartes's Principles: Benedict Spinoza, Parts 1 and 2 of René Descartes's Principles of Philosophy, Demonstrated in the Geometrical Manner.

Edwards: Paul Edwards (ed.), The Encyclopedia of Philosophy (New York, 1967).

Emendation: Benedict Spinoza, A Treatise on the Emendation of the Intellect. I have used Curley's forthcoming translation of this work, and my references are based on his section numbering.

Essay: John Locke, An Essay Concerning Human Understanding.

Ethics: Benedict Spinoza, Ethics Demonstrated in Geometrical Order.

Grene: Marjorie Grene (ed.), Spinoza: A Collection of Critical Essays (Notre Dame, Indiana, 1973).

Gueroult: Martial Gueroult, Spinoza, vol. 1 (Paris, 1968), vol. 2 (1974).

Hessing: Siegfried Hessing (ed.), Speculum Spinozanum 1677-1977 (London, 1977).

Kashap: S. Paul Kashap (ed.), Studies in Spinoza: Critical and Interpretive Essays (Berkeley, 1972).

Kennington: Richard Kennington (ed.), The Philosophy of Baruch Spinoza (Washington, D.C., 1980).

Linguistic Behaviour: Jonathan Bennett, Linguistic Behaviour (Cambridge, 1976).

Loemker: Leroy E. Loemker (ed.), G. W. Leibniz: Philosophical Papers and Letters, second edition, (Dordrecht, 1969).

Mandelbaum: Maurice Mandelbaum and Eugene Freeman (eds.), Spinoza: Essays in Interpretation (La Salle, Illinois., 1975).

New Essays: G. W. Leibniz, New Essays on Human Understanding, translated and edited by Peter Remnant and Jonathan Bennett (Cambridge, 1981).

Principles: René Descartes, The Principles of Philosophy.

Shahan: Robert W. Shahan and J. Biro (eds.), Spinoza: New Perspectives (Norman, Oklahoma, 1978).

Treatise: David Hume, A Treatise of Human Nature, edited by L. A. Selby-Bigge (Oxford).

Bibliography

The following list gives the writings on Spinoza which I have found most valuable. Since I have not systematically searched the journals, there is doubtless good material which I do not know; so exclusion does not necessarily express a negative judgment. As regards English translations of the Ethics: the version by Elwes should not be used; the version by White and Stirling, now out of print, is also highly defective; the rendering by Samuel Shirley (Hackett Publishing Company) improves on all its predecessors, and is thoroughly usable for low-pressure study of the work; but for serious scholarly or philosophical research the forthcoming version by E. M. Curley (Princeton University Press) is indispensable. It is less fluent and pleasing than the Shirley version, but it sets a standard of accuracy—a real mirroring of the Latin in the English—to which the other does not aspire.

Albert G.A. Balz, Idea and Essence in the Philosophies of Hobbes and Spinoza (New York, 1967), first published 1917.

H. Barker, ‘Notes on the Second Part of Spinoza's Ethics’, Mind vol. 47 (1938), reprinted in Kashap, pp. 101-167.

Edward Caird, Spinoza (Edinburgh, 1888).

E.M. Curley, Spinoza's Metaphysics: An Essay in Interpretation (Cambridge, Mass., 1969).

E.M. Curley, ‘Spinoza's Moral Philosophy’, Grene, pp. 354-376.

Guttorm, Flφistad, ‘Spinoza's Theory of Knowledge in the Ethics’, Inquiry vol. 12 (1969), reprinted in Grene, pp. 101-127, and in Kashap, pp. 249-275.

Joel Friedman, ‘An Overview of Spinoza's Ethics’, Synthese vol. 37 (1968), pp. 67-106.

Stuatt Hampshire, Spinoza (London, 1951).

Francis S. Haserot, ‘Spinoza and the Status of Universals’, The Philosophical Review vol. 59 (1950), reprinted in Kashap, pp. 43-67.

Wallace Matson, ‘Spinoza's Theory of Mind’, The Monist vol. 55 (1971), reprinted in Mandelbaum, pp. 49-60.

G.H.R. Parkinson, ‘Hegel, Pantheism, and Spinoza’, Journal of the History of Ideas vol. 38 (1955), pp. 449-459.

Frederick Pollock, Spinoza, his Life and Philosophy (London, 1880), reprinted by the Reprint Library.

Leon Roth, Spinoza (London, 1929).

H.A. Wolfson, The Philosophy of Spinoza (New York, 1934).