Immanence and Difference: Toward a Relational Ontology

zyxwv zyxw zyxwv zyxwvu zy The Southern Journal of Philosophy (2002) Vol. LX Immanence and Difference: Toward a Relational Ontology Robin Durie Staffordshire University Le p l u s interessant dBs lors est la manihre dont Spinoza utilise et renouvelle les notions de distinction formelle et d’univocitd Deleuze, Spinoza et le probleme de l’expression ...posse existere potentia est. Spinoza, Ethics I, 11 The branch of theoretical physics t h a t seeks to unify t h e theories of general relativity a n d quantum mechanics h a s recently received a notable impetus from the research of Lee Smolin and Julian Barbour. What characterizes their distinct positions is an absolute commitment to the ‘relational approach’ that has its provenance in Leibniz and that was taken up and developed by Mach a n d t h e n Einstein. Both Barbour a n d Smolin have striven to demonstrate t h a t “space and time are nothing but aspects of relationships.”’ What is perforce lacking from their accounts is a relational ontology. There have been a number of renowned attempts in the history of philosophy to generate a n ontology of relations, such as those of Aristotle in the Categories and Husserl in the third of his Logical Znuestigations. However, these attempts are limited in scope-they are precisely ontologies of relations rather than relational ontologies, which is to say t h a t the ontology of relations forms part of a more general ontology that determines in advance the nature of the specific ontology of relations. Our aim in the following paper is to provide a first indication of t h e form t h a t a relational ontology might take. zyxwv Robin Durie is a Senior Lecturer in Philosophy at Staffordshire University. His main research interests are i n phenomenology, recent French philosophy, and the Philosophy of Time. He is the editor of Bergson’s Duration and Simultaneity (Manchester: Clinumen Press, 1999) and of the collection Time and the Instant (Manchester: Clinamen Press, 2000). 161 zyx zyxw zyxwvu Robin Durie In developing such an ontology, two possible courses could be followed: one, governed by Aristotle’s dictum in the Metaphysics t h a t “Being is said i n many ways [To de on legetai men pollachosl” (1003*33),would lead to a n ‘equivocal’ ontology; while the other could be summed up by Deleuze’s proposition t h a t “There h a s only ever been one ontological proposition: Being is univocal.”2 The dominant ontologies in the history of philosophy, such as those of Plato and Descartes, have been dualist and have therefore followed the first course. This tradition h a s always been accompanied by a vociferous critical response that has tended to undermine the plausibility of dualist ontologies from the very moment that each new incarnation is expressed. However, those ontologies that are in turn nominally univocal have nevertheless tended to take for granted the basic assumptions of equivocal ontologies, while seeking to provide a univocal explanation of them. If, with the work of physicists such as Barbour a n d Smolin, we a r e witnessing a genuine paradigm shift in physics, then, as Thomas Kuhn has argued, there is a concomitant need for a shift or revolution i n t h e “imagination” of the “world” within which this scientific work is carried Such a revolution consists, we propose, in t h e formation of a genuinely univocal ontology, one whose principle is to begin from relations as such, from which the sense of entities as emerging from these relations is then derived, rather t h a n beginning from static conceptions of, for instance, ontological regions and then seeking to account for how these might ‘interact’. One resource for developing such a univocal ontology is the philosophy of Spinoza. However, i n and of itself, Spinozist ontology harbors a series of fundamental problems. These problems can be shown to arise from a tendency on Spinoza’s part, and on the part of his commentators, to begin from a fixed conception of the elements of his ontology and then to seek to conjoin them within a univocal ontology. This tendency is delineated by Deleuze when, having laid particular stress throughout Expressionism in Philosophy: Spinoza on t h e necessary interdependence of substance and modes,4he proceeds to suggest in Difference and Repetition, published at the same time as Expressionism and Philosophy, that there “still subsists a n indifference [indiffe‘rence]between substance and its modes: Spinoza’s substance appears [apparaitl independent of t h e modes, while the modes are dependent on s ~ b s t a n c e . ”As ~ we shall demonstrate, what Deleuze proposes is an ‘explication’ of Spinoza t h a t seeks to think of the elements i n Spinoza’s ontology as emerging from a prior relation, and only in consequence of this process t o think their self-identity, a n identity that would thus be the result of a genesis. Such an explication would therefore necessarily remove Spinoza from the “worthy company” of “Plato, Descartes, Newton, Locke and Quine.”6 162 Immanence and Difference zyxw I n so explicating Spinoza, Deleuze is not being willfully perverse-quite the contrary. The fundamental relation that, according to Deleuze, underpins Spinoza’s ontology is t h a t of expression. We shall show t h a t t h e logic of the relation of expression, when conceived prior to those objects that it determines, in fact enables Deleuze to propose a series of interpretations of Spinoza that do not fall prey to those paradoxes that result from the thinking in which substance remains indifferent to its modes. We shall also show that these interpretations are i n no sense piecemeal responses b u t r a t h e r form part of a coherent presentation of Spinoza’s univocal ontology, the coherence of which derives from the logic of expression.’ Beyond this, however, the goal of our reading is to bring to the fore the true nature of the logic of the relation of expression. The nature of the relation of expression remains, to a large extent, implicit in Deleuze’s text. In order to disclose the nature of the relation of expression, it will be necessary to pay specific heed to the motivation for the precise means by which Deleuze explicates Spinoza. As we shall show, t h i s motivation is Bergsonist. The clue that we shall follow is Deleuze’s repeated recourse to the notion of a “theory of distinctions” at work in Spinoza’s text. We have already shown in a previous paper how Deleuze’s appropriation of Bergson is governed by the “philosophy of difference” that, as he demonstrates, underpins Bergson’s work.8 What is significant in this demonstration is the extent to which “difference itself ... has a nature, finally t h a t it will deliver Being to US."^ I t is t h e possibility t h a t difference will “deliver Being” which provides the motivation for Deleuze’s Bergson-inspired explication of Spinoza. We shall show t h a t the nature of the relation of expression is i n fact differential: the elements of Spinoza’s ontology emerge from, are expressions of, differential relations. What is significant about this differential way of characterizing the logic of the relation of expression is that it makes clear how the various elements in the Spinozist system and their interrelations form part of a coherent, immanent, univocal ontology. The keystone t h a t supports Deleuze’s demonstration of the univocity of Spinoza’s ontology is the double expressivity of attribution. We shall show that attribution possesses this double expressivity to the extent t h a t i t is differential-at one and the same time and by the same “movement,” differentiating the essences of substance and differentiating the modes from substance. Thus, the univocal relational ontology toward which we are working in this paper is, in the final analysis, a differential ontology. I n t h e reading t h a t follows, we shall focus on four key moments i n Deleuze’s Spinoza interpretation, moments at which the differential sense of the relation of expression and the priority of the relation to the relata that emerge from the relation a r e revealed i n their fundamental importance: the zyxw 163 Robin Dune zyxw zyx zyx zy zyxwvu sense pertaining to the distinction between the attributes of substance, the specifically dynamic nature of the attributes themselves, t h e question of what t h e immanent sufficient reason for the being of things (modes) is, and the conceptual elements involved in conceiving of the being of modes from the perspective of immanence or univocity. 1. In a letter dated 24 February 1663,Simon de Vries writes to Spinoza that a passage in a n early draft of the Ethics t h a t he has to hand indicates that Spinoza “seems to suppose that the nature of a substance is so constituted t h a t i t can have many attributes.” De Vries protests t h a t “if I were to say t h a t each substance has only one attribute, I could rightly conclude that where t h e r e a r e two different a t t r i b u t e s t h e r e a r e two different substances.”lo De Vries’s argument is valid a n d strikes at t h e h e a r t of Spinoza’s alternative to Cartesian dualism. The argument is derived from paragraph 5 3 of Descartes’s Principles of Philosophy: “To each substance there belongs one principal attribute.... A substance may indeed be known through any attribute at all, but each substance has one principal property which constitutes its nature and essence, and to which all other properties are referred.”’l An attribute is a principal attribute when it constitutes the essence or nature of substance without presupposing any other property. If a n a t t r i b u t e does not presuppose any other property, then we can say that it may be conceived independently of any other property and, thus, t h a t this attribute is really distinct from any other attributes. But if “to each substance there belongs one principal attribute,” then by t h i s preceding argument, we can deduce t h a t every “principal attribute” constitutes a really distinct substance, whereby “Two substances are said to be really distinct from one another when each of them can exist without t h e other.”12 Briefly stated, t h e separability of a t t r i b u t e s entails t h e separability of the substances whose essence they constitute. The foundation stone of Spinoza’s monistic response to Cartesian dualism must therefore be the demonstration t h a t really distinct (principal) attributes can belong to a single substance.13 It is only after establishing this conclusion t h a t Spinoza will be able to go on to deduce that just one substance exists. And indeed, t h i s is how t h e order of demonstration proceeds i n Part I of t h e Ethics. Thus, we find Spinoza concluding in the Scholium to Proposition 10: Although two attributes be conceived as really distinct, that is, one without the help of the other [unum sine ope alterius], still we cannot deduce therefrom that they constitute two entities, or two different [diversas] substances. For it is in the nature of 164 Immanence and Difference zyxwvu zyxwvu zyxwv substance that each of its attributes be conceived through itself, since all the attributes it possesses have always been in it simultaneously, and one could not have been produced by another; but each expresses the reality or being of substance. So it is by no means absurd to ascribe more than one attribute to one substance. ( E I. 10s)” It is subsequently in Proposition 14 that Spinoza reaches a point at which he is able to assert t h a t “There can be, or be conceived, no other substance but God” ( E I. 14). Two difficulties pertain to the position enunciated in the Scholium to Proposition 10: How is Spinoza able to demonstrate that one substance can have many attributes? And how are we to make sense of the notion of a single substance having many attributes? Spinoza’s answer to the first question can be reconstructed from Propositions 8 and 9 and Definition 6. Taken together, these assert that “every substance is necessarily infinite” ( E I. 8), t h a t “the more reality or being a thing has, t h e more attributes it has” (E I. 9), and that God is “an absolutely infinite being, that is substance consisting of infinite attributes” ( E 1. def 6 ) . Thus, if we consider two substances, both of which are infinite, then if one substance is “absolutely” infinite whereas the other one is not (it is, we might say, only “infinite i n its kind”-cf. t h e Explication of Definition 6 ) , t h e n the former substance will have “more reality” than the latter substance and, hence, “more attributes.” But, as is established by Proposition 11, there is such a n absolutely infinite substance, namely God, and therefore God must have at least more than one attribute (in fact, as Definition 6 asserts, since God is absolutely infinite, h e must have a n infinity of attributes). Thus, the existence of God already proves that it is possible for substance to have more t h a n one attribute. And i n t h e Proposition immediately following t h e Scholium i n which Spinoza argues t h a t it is not absurd to ascribe more than one attribute to one substance, h e puts forward a number of arguments to demonstrate that God does indeed necessarily exist (E I. 11). Granted this argument, why do we need to pose the second of our questions above? On the one hand, it remains the case t h a t , as Edwin Curley notes, many Spinoza scholars simply don’t find this argument “satisfact~ry.”’~ On the other hand, there is the problem of what would be entailed by the notion of one substance with many attributes. Recalling Spinoza’s definition of a n attribute-that is, “that which the intellect perceives of substance as constituting i t s essence [tanquam eiusdem essentiam constituensl” ( E I. def 4)-it would appear t h a t t h i s position entails our conceiving of t h e essence of substance as being plural. On this basis, we would say that zyx zyxw 165 z Robin Dune each attribute constitutes an essence of substance and that substance itself is the unity of each of the essences constituted by each of its attributes.16What is problematic about this position is that it encourages us t o think of substance as a complex whole, or aggregate, consisting of really distinct essential parts.17 But if the parts really are distinct, then we tend t o assume that the aggregate could be broken up into these parts. However, this whole line of thinking is explicitly denied by Spinoza: Proposition 13 demonstrates that “Absolutely infinite substance is indivisible,” and goes on t o argue that any attempt t o think a part of substance involves a contradiction (E I. 13cs).’* Why does this problem appear so intractable? Stated in this form, it bears an instructive resemblance t o the alternatives explored in the first two hypotheses of the second part of Plato’s Purmenides, which show that conceiving of the One either as a whole without parts or as a whole with parts leads t o irreconcilably contradictory consequences. Now, one of the recurrent lines of argument in Deleuze’s work is that the problem of the One and the Many is, t o use a Bergsonian formulation, a “false problem.” So why has the Platonic problem of the One and Many, first encountered in the Purmenides, arisen in this Spinozist context? Is it because we are trying t o conceive substance, attributes, and essence “indifferently,” or independently, from one another? Are we beginning from the principle of the identity of substance (as Deleuze suggested t h a t Spinoza’s text encourages us to) and then trying t o understand its diversity derivatively from this identity rather than seeking to understand how the “identity” or unity of substance is “generated from difference? As we have already underscored, Spinoza asserts that every substance is infinite ( E I. 8 ) , that God is an “absolutely infinite being, that is substance consisting of infinite attributes” ( E I. def 6), and that each of these attributes express an “eternal and infinite essence” ( E 1. 11). For this reason, Deleuze, citing Merleau-Ponty’s article on Spinoza in his Les Philosophes c&!bres, argues that Spinoza’s thinking sets out “from infinity” (EPS 28/SPE 22). In starting from infinity, however, rather than from, say, the principle of clear and distinct ideas, it remained incumbent upon Spinoza t o demonstrate the “power and actuality” of this positive infinity. Deleuze’s claim is that it is the logic of expression that enables Spinoza to achieve this and that it does so “by introducing into infinity various distinctions” (ibid.). It is in this way, therefore, that expression constitutes the point of departure for Deleuze’s explication of the Ethics.” Spinoza’s monism follows from two premises: one attribute cannot belong t o two or more substances, but, on the contrary, two or more attributes can belong t o one substance. Deleuze’s argument is that, in order t o make sense of these premises, zyxw zyxwvu zyxw zyx 166 Immanence and Difference zyxw zy zy Spinoza needs to put forward a “theory of distinctions” (cf. EPS 37, 66, 182lSPE 31, 57, 165). Such a theory is, Deleuze suggests, foreshadowed in the implicit question to which Proposition 4 responds, namely, “how two things, in the most general sense of the word, can be distinguished” (EPS 28lSPE 22). The answer Spinoza gives is t h a t “Two or more distinct things [res distinctz~]are distinguished [distinguunturl from one another either by t h e difference [diversitatel of t h e attributes of substances or by the difference [diversitatel of the affections of substances” ( E I. 4). The way in which Deleuze interprets this claim stems from a n ambiguity t h a t Descartes notes i n paragraph 56 of the Principles, where he writes that “by mode we understand exactly the same as what is elsewhere meant by a n a t t r i b u t e or quality.” But i n t h e previous paragraph, Descartes writes that “we should regard the duration [and the order or the number] of a thing simply as a mode under which we conceive t h e thing.” The general notion of attribute, Descartes concludes, therefore covers both modal diversification and qualitative differentiationz0(cf. EPS 29lSPE 23). Spinoza of course rejects the Cartesian identification of attribute a n d mode. The Cartesian clue that Deleuze will nevertheless import into Spinozism is the possibility of understanding attributes as qualities and, hence, t h e difference between attributes a s qualitative difference. Following Proposition 4, therefore, we claim that there are two types of difference: difference by attribute and difference by mode. Following Descartes, however, we infer t h a t modal distinction would be a numerical distinction. And if modal distinction is numerical, then we may infer from Descartes’s discussion in the Principles that distinction by attributes could only be qualitative difference. This way of understanding the difference between attributive distinction and modal distinction, namely as the difference between qualitative and numerical (quantitative) distinction, underpins, Deleuze implies, t h e reasoning in the second Scholium to Proposition 8. He therefore concludes that “On the one hand, one deduces from the nature of numerical distinction that it is inapplicable to substance; on the other, one deduces from the nature of substance its infinity, and thus the impossibility of applying to it numerical distinctions. In either case, numerical distinction can never distinguish substances” (EPS 33lSPE 26).z1Thus, t h e first eight propositions of the Ethics deal with numerical distinction and its inapplicability to substance (there are not two or more really distinct substances with the same attribute). Propositions 9 and 10 then serve, as we have seen, to establish that two or more attributes can pertain to t h e same substance. But since numerical distinction is never applicable to substance, what is the nature of the distinction between these attributes? It is, as we have seen, a real distinction ( E I. 10s). Thus, numerical 167 zy z zyxw zyxw Robin Dune distinction must be differentiated from real distinction: ”Numerical distinction is never real; then, conversely, real distinction is never numerical” (EPS 34/SPE 27). But we have also just found that Deleuze wants to argue that the two ways i n which things can be distinguished a r e either numerical (modal) or qualitative (attributive). Thus, we see how Deleuze establishes that real distinction should be recast as qualitative distinction (EPS 38/SPE 3 2 ) and t h a t the nature of the real distinction between the different attributes of one substance and, hence, of the essences of that substance is qualitative. The fundamental difference between qualitative and numerical distinctions begins to suggest why it is that substance might well possess a n infinity of really, t h a t is to say qualitatively, distinct essences while still maintaining its indivisibility, if we accept that division is an essentially quantitative or numerical process. What therefore is substance that possesses an infinity of really distinct attributes and hence an infinity of really distinct essences? If we accept that attributes are qualities, then “There is one substance per attribute from the viewpoint of quality, but one single substance for all attributes from the viewpoint of quantity.” Substance is, therefore, a purely “qualitative multiplicity’’ (EPS 37/SPE 30). The multiplicity is, in itself, already both one and many. If the recourse to the notion of the qualitative multiplicity could thus enable us to conceive of substance as a “multiplicity” of really distinct attributes and, hence, really distinct essences, then we would have taken a first s t e p i n overcoming t h e problem of t h e one a n d the many to which scholars have traditionally represented Spinoza as succumbing.22 The introduction of the notion of the qualitative multiplicity into Spinoza’s text represents Deleuze’s first explicitly Bergsonian explication of S p i n o ~ aBut . ~ ~should we accept that attributes a r e qualities and t h a t attributive distinction is qualitative distinction? And even if we do, does this really help u s i n our attempt, as we asked above, to conceive of a single substance having many distinct attributes? The point is, as Pierre Macherey underscores (referring to EPS 66 & 359n28lSPE 57), that there is no textual justification within the Ethics itself t h a t specifically allows t h e interpretation of attributes as qualities.24 I n fact, when we understand how Deleuze conceives of a qualitative multiplicity, we attain a much clearer insight into the strategy by which he is striving to secure a n immanent, univocal ontology from Spinoza’s text. Bergson’s notion of a qualitative multiplicity is grounded i n t h e notion of t h e multiplicity [MannigfaltigKeitl that he derives from the German mathematician Bernhard Riemann.25 Riemann’s aim was to develop a general sense of the notion of magnitude from which the specific multidimensional magnitudes-such as the threedimensional magnitude of our space-could be derived. The zyx zy 168 Immanence and Difference zyxw zy multiplicity served t h a t function. By multiplicity, Riemann understood a province of objects, or elements, t h a t are indeterminute i n themselves; thus, the elements of a multiplicity are not governed by any transcendental principle, in the way that the elements of a set are determined in advance by a defining property. Rather, the elements come to be determined by the operations to which they are subject, that is to say, the relations into which they may enter. Like the elements of the multiplicity, these relations are indeterminate with regard to their content. They are, however, formally determinate and thereby enable certain connections to be established between the indeterminate elements within the multiplicity. Under these circumstances the elements come to be determined. Since neither the being nor the nature of the elements of a multiplicity a r e determined by a principle transcending t h e multiplicity, there is no need to account for the means by which One and Many interrelate. However, as we noted, the multiplicity is, for Riemann, the conceptual means with which to envisage different possible kinds of magnitude. Now, the notion of magnitude would appear inevitably to imply quantity, and we have been stressing t h a t Deleuze is working from a distinction between qualitative and numerical, or quantitative, differences. This is where Bergson’s precise way of utilizing Riemann’s notion becomes significant. For Bergson employs the notion of the multiplicity to conceive of a reality t h a t i n itself remains unquantifiable. I n Bergson’s thought, what Riemann defines as a “continuous” multiplicity is also conceived as a qualitative, or heterogeneous, multiplicity. We can trace t h e means by which Bergson reconceives t h e “continuous” and “discrete” multiplicities of Riemann on the basis of difference back to a “theory of distinctions” parallel to that which functions in Spinoza. According to the first chapter of Time and Free Will, what distinguishes the two multiplicities is that in the case of the ”discrete” or numerical multiplicity, the nature of the formally determinate relations pertaining to the multiplicity are such as to determine the elements as “homogeneous,” which is to say that their division results in a change in degree rather than in kind (division leads to the quantitatively “more or less”). In the case of the “continuous” multiplicity, the elements are determined by relations of “heterogeneity,” such that “division”-or more precisely, “differentiation”-leads to a change i n kind or quality.26From this perspective, i t would therefore be possible to respect Spinoza’s prohibition on the division of substance tE 1. 13cs), while at the same time arguing that when substance diferentiates, it precisely changes in kind. Also from this perspective, we can apply Deleuze’s definition of the qualitative multiplicity-as t h a t which differs in and from itself ‘-to Spinozist substance itself. By borrowing the notion of the qualitative multiplicity from Bergson, Deleuze is thus able to propose an interpretation of Spinoza’s doctrine of a 169 z Robin Durie single substance with a multiplicity of attributes and, hence, a multiplicity of elements, which avoids the pitfalls associated with the problems of the one and the many and enables him t o flesh out the bones of the proposed theory of distinctions, that is, the difference between real, or qualitative, and numerical difference: zyxwvu zyxw zy zy zyxwv The new s t a t u s of real distinction is fundamental: a s purely qualitative ... real distinction excludes any division.. .. That real distinction is not and cannot be numerical appears t o me to be one of t h e principal themes of the Ethics. The new theory of distinctions h a s a s its fundamental principle the qualitative s t a t u s of real distinctions. Detached from all numerical distinction, real distinction is carried into the absolute,28and becomes capable of expressing difference within Being, so bringing about the restructuring of other distinctions. (EPS 38SISPE 31-2) 2. How is the absolute rendered as a qualitative multiplicity, as difference-in-itself? In the context of Spinoza’s ontology, this issue is expressed in the questions of what attributes are, what the nature of their relation is t o substance, and, ultimately, of just why there are attributes-for it is attributes, Deleuze argues, that are the “dynamic or genetic elements” of substance (EPS 8OISPE 69). This way of responding to Spinoza once again serves t o furnish an answer t o another problem t h a t , as Jonathan Bennett writes, “has plagued Spinoza scholars for c e n t ~ r i e s . This ” ~ ~ is the problem, as Bennett states it, of “what an attribute is.” According t o Bennett, the problem bequeathed by Spinoza turns on the matter of whether attributes are genuinely different from essences or whether, as seems to be suggested by Definition 4, this is a difference that resides only in the intellect’s perception of substance. Take away the notion of the perception of the intellect, and what we’d be left with is a principal property, or attribute, which just is the essence or nature of the substance (cf. 0 53 of Descartes’s Principles). It seems t o us that this problem exists, if it exists at all, only at the epistemological level. It may well be that intellectual perception in and of itself would be unable t o distinguish attribute from essence. However, the issue here is first and foremost ontological rather than epistemological. If we compare Definitions 4 and 6, we see that whereas only the former refers t o intellectual perception, both specify a relation between attribute and essence: Definition 4 says that attributes constitute the essence of substance, while Definition 6 says that each attribute expresses eternal and infinite essence. Now it is notable, indeed remarkable, that Bennett, in his citation of zyxw 170 zyxwv Immanence and Difference Definition 4, elides without explanation the word constituens. This may well be because h e is taking a n epistemological perspective and is interpreting “constitutes” as something like “represents,” thereby implying that Spinoza is saying something akin to “an attribute is the way an essence is represented to, o r for, the intellect.” But from a n ontological perspective, such a reading is inadmissable. Indeed, the substitution of expression for constitution makes this quite clear.30But accepting this point, it still leaves the question of the nature of the ontological relation between substance, attributes, and essence t h a t is designated by Spinoza’s use of the notions of constitution and expression. Two aspects of t h e first definition enable u s to begin to construct an answer to this question. On the one hand, the most fundamental notion t h a t Spinoza seeks to define, the notion from which the whole of the Ethics follows, is “that which is self-caused [causarn suil.” The notion of cause employed here can be understood to imply some sort of event. On the other hand, i t must be borne i n mind t h a t the existence of t h e self-caused, that is, the existence of substance, follows from its essence: “By t h a t which is self-caused I mean t h a t whose essence involves [inuoZuitl existence.” But what is the essence of substance? I t is that which is constituted by, or expressed by, the attributes of substance. Substance, as Definition 6 says, consists [constat]of attributes, each one of which constitutes the essence of substance, the essence from which the existence of substance follows. But t h e attributes of substance a r e not simply identical with substance. Each attribute is, as the second Letter (to Oldenburg) makes clear, “conceived through itself, and i n itself,” i n keeping with substance ( E I. def 3), b u t Spinoza does not go on to say, as he does of substance, that the attributes exist of themselves or t h a t their existence follows from their essence.31 Substance is, therefore, neither simply identical to its attributes nor to its essence(s); equally, attributes are not identical to essences. And yet, on the other hand, substance consists of its attributes, and its existence follows from the essence that these attributes constitute. How are we to understand these differences that seem not to differ at all? The clue is to be found in the first aspect of Definition 1 to which we drew attention. The difficulty we have in conceiving the differences beween the three elements of what Deleuze calls “the first triad of substance” (i.e., substance, attribute, essence), is of the same nature as the difficulty we had in trying to think beyond the paradigm of the One and the Many when conceiving of the multiplicity of substance. The difficulty stems, that is to say, from beginning with the principle of the fixed, or static, identity of each element of t h e t r i a d a n d t h e n seeking to understand the relations and differences between the elements derivatively rather than, as was suggested before, determining zyxwvu zy 171 z zyxw zyxw Robin Durie the identity of each element from the determinate relations into which they enter. What the notion of self-causation in Definition 1 prompts us t o do is t o consider the nature of the relations between the elements of the triad genetically and t o see the elements of the triad as being, in some sense, “generated” in and by these genetic relations. From this perspective, we can begin t o see the significance of Deleuze’s claim that attributes are the dynamic or genetic elements of ~ u b s t a n c e . ~ ~ If we ask what the “genesis”of the triad of attribute, essence and substance is, in what the “event” of self-causation consists, the answer with which Deleuze provides us is expression. We have seen already that attributes express essence, and as Deleuze reminds us, “the essence of substance has no existence outside the attributes that express it .... What is expressed has no existence outside its expressions; each expression is, as it were, the existence of what is expressed” (EPS 42lSPE 34). But the nature of the expression of essence is itself quite specificas Definition 4 makes clear, essence is expressed as the essence of substance (ibid.). In this way, we see how the relation from attribute to essence to substance is constituted: the existence of essence follows from attributes t o the extent that they express essence. This means t h a t each essence only exists in the attribute that expresses it, and thus, through expression, the essences are rendered really distinct in and through attributes. But the really distinct essences exist as essences of substance. Equally, it is clear that substance, as the multiplicity, is different from each of the essences in their qualitative distinctness, while yet being nothing other than these essences. And it is precisely the attributes that constitute this difference between substance and its essences, t o the extent that attributes render the essences in their real, qualitative distinctness. But we also underlined t h a t the existence of substance follows from its essence. Thus, substance’s existence requires the constitution of essences, or as Deleuze says of substance, its “existence necessarily follows [ddcoulel from the essence thus constituted” ( E P S 4 3 l S P E 3 5 ) . And so in order that substance exist, it must be rendered as a multiplicity of essences, which is t o say, rendered as differentin-itself, by attributes. I t is as difference-in-itself t h a t substance is causa sui. It is evident that if we were t o consider each of the elements of the triad of substance, attribute, and essence independently and then were t o t r y t o construct these relations, we would encounter insurmountable paradoxes-for instance, the condition of the possibility of essences, namely substance, owes its own possibility to these very essences. Taking our cue from Riemann’s way of understanding the multiplicity, however, we should instead understand essences as being “determinations” that follow from the specific nature of the determinate relations zyx 172 zyxw zy zyxw Immanence and Difference of expression. And t h e element of t h i s specific relation is attribution, not attributes considered as fixed or static entities: Attributes are for Spinoza dynamic and active forms. And here at once we have what seems essential: attributes are no longer attributed [attribuk], but are in some sense ‘attributive’ [attributeur].Each attribute expresses an essence, and attributes i t to substance. All the attributed essences coalesce in the substance of which they are the essence. (EPS 45fSPE 36) If we understand t h e basis of Spinoza’s ontology from t h e perspective of the determinate relations t h a t determine the nature of the elements of the triad of substance, attribute, and essence, then what we find is a logic of dynamic expression. What is this logic of expression? Any event of expression is made up of certain fundamental components: what expresses itself, the expression i n which i t expresses itself, and what is expressed through, or by means of, the expression. This is the formally determinate nature of the relation that is expression. This formally determinate nature thus determines substance as “what expresses itself” (in attributes), attributes as “the expression” (the “dynamic or genetic element of expression”), and essence as “what is expressed by, or through expression.” And finally, since expression is always expressed of something, we close t h e circle by saying t h a t “essence is expressed of substance” a n d t h a t substance’s existence follows from t h i s expression of its essence.33Our understanding of each of the elements i n these terms thus follows from our understanding of t h e determinate relation of expression that so determines them. The elements of the triad of substance, attribute, and essence do not first exist and then enter into relation with each other-a way of thinking mired i n transcendence and that, as we have seen, renders the basis of Spinoza’s ontology unthinkable; rather, they are the result of a genesis, a genesis whose nature is manifested i n t h e logic of dynamic expres~ion.~~ zyxw zyxwv zyxwv 3. But we can go further in our explication of the nature of the formally determinate relation of expression. For, as we have seen, t h e genetic way of understanding a t t r i b u t i o n is irreducibly interwoven with the conception of the existence of substance as a (qualitative) multiplicity of really distinct essences. We have also seen t h a t i t is a t t r i b u t i o n t h a t qualitatively differentiates essences a n d thereby renders substance as difference-in-itself. Thus, it is t h e differential nature of the formally determinate relation of expression t h a t enables u s to develop a genetic understanding of attribution. It remains to be seen whether the determinate relation that is 173 zyxwv zyx Robin Dune zyxwv zyxwv zy zyxw zyxw productive of modes is also differential and, if so, i n what way. Attribution, as the element of expression, is the milieu of the genesis of substance as difference-in-itself and, t h u s , of substance as causa sui. But this still leaves a series of difficult problems for a n ontology of immanence: “Why does God produce anything a t all [Pourquoi Dieu produit-ill?” (EPS 99ISPE 88). What is the immanent sufficient reason for the being of things (modes)? Once again, the precise nature of the relationship between modes a n d substance h a s provoked a deal of consternation amongst Spinoza scholars. This difficulty is summed up i n recent English language criticism, where the focus of attention has been on whether o r not the relation should be understood as inherence. Curley adduces the language of Proposition 16: “From the necessity of t h e divine n a t u r e there must follow [sequi debent] infinite things in infinite ways” (cf. E I. 28d), Proposition 18: “God is the immanent, not the transitive, cause of all things,’’ and Proposition 24: “The essence of things produced [ p r o d u c t a r u m ]by God ...”- to argue t h a t Spinoza understands t h e relation between substance a n d mode as causal.35Bennett points to Spinoza’s definition of a mode as an “affection of substance’’ as evidence that he shared the contemporary orthodoxy that a mode is a “property” and argues that we should take Spinoza t o mean by a modal property something like the fact that, at a certain moment, the universe comes to be qualified in a certain way, rather in the way that, at a certain moment, my face blushes. In this way, he is able to account for the evident fact that Spinoza denied the independent existence of modes.36Finally, there are those, such as John Carriero, who argue t h a t t h e assimilation of modal dependence to causal dependence “disappointingly flattens Spinoza’s position” and that there is enough evidence within Aristotle and his medieval followers to suggest that efficient causation and inherence are not mutually incompatible, and thus that it does more justice to Spinoza to interpret modal dependence as i n h e r e n ~ e . ~ ~ Deleuze’s response to this problem proceeds in a different register to each of these alternatives. The critical reason for this is t h a t each of the proposals repeats t h e “indifference” of substance to modes to which he draws attention in Difference and Repetition. In each case, substance’s self-causation and the production of modes are separate events. It would have to be by a separate act that God created modes, a n act, that is t o say, of a wholly different nature. But in this case, the reason for the being of God and the reason for the being of modes would be different. Substance and modes are, in the first place, conceived in their separateness by the commentators and that is why they have so much difficulty i n explaining t h e n a t u r e of t h e i r relation. But the principle underpinning Deleuze’s conception of 174 z zyxw zyxw zyxw Immanence and Difference Spinoza’s immanent ontology is that it is a n ontology of univocity: “If one is to give a name t o [Spinoza’s] method, as to the underlying theory, it is easy t o recognize here the great tradition of univocity. I believe that Spinoza’s philosophy remains in part unintelligible i f one does not see in it a constant struggle against the three notions of equivocation, eminence and analogy” (EPS 48-91SPE 40h3* Deleuze’s alternative approach gains coherence once again by turning on the logic of expression. Moreover, it again focusses on the role of the attributes. Just as the attributes are the milieu of expression, and thus of the event of substance’s self-causation, so also they are the milieu of the expression of modes (cf. the Corollary t o Proposition 25: “Particular things are nothing but affections of the attributes of God; that is, modes wherein the attributes of God are expressed in a certain and determinate way”). But, most importantly, Deleuze claims that the essence of substance and the modes is expressed in one and the same way, and in this way, we gain a first sense of why, in Deleuze’s interpretation, substance is not indifferent t o modes: “The attributes are, according t o Spinoza, univocal forms of being which do not change their nature in changing their ‘subject’-that is, when predicated of infinite beings and finite beings, substance and modes, God and creatures” (EPS 49lSPE 40). This univocity of attributes is inseparable from the logic of their expres~ivity,~~ and it is the logic of their expressivity that provides us with the key to understanding the sufficient reason why God produces a modal universe (EPS lOOlSPE 88). First, as we have seen, the expressional relation within the triad of substance, attribute, and essence is that substance expresses itself in attributes, attributes are the vehicles of expression, and the essence of substance is expressed by attributes. We argued previously that attribution therefore amounted t o an expressive genesis, the constitution of the nature of substance from which its existence follows. Furthermore, we argued that, through this genesis, substance was constituted as a differential multiplicity, as difference-in-itself. Deleuze’s proposition is t h a t this expression of substance has a second aspect, that the event of the constitution of substance through expression is doubled by “a second degree of expression” (EPS 101ISPE 87). In making this claim, Deleuze is seeking t o provide a ground in the determinate relation of expression for two interrelated arguments developed by Spinoza: that “in the same sense that God is said t o be self-caused he must also be said t o be the cause of all things” ( E I. 25s, emphasis added),40and that “God acts by the same necessity whereby he understands himself; that is, just as it follows from the necessity of the divine nature (as is universally agreed) that God understands himself, by the same necessity it also follows that God acts in infinitely many ways” (E 11. 3s). 175 Robin Durie zyxw Both of these arguments explicitly deny t h a t God causes modes in a n act separate from that in and by which he causes himself. The second argument further makes clear t h a t God’s act of causing the modal universe follows from t h e same necessity as that by which he understands himself and that his understanding of himself follows from t h e necessity of his divine nature, that is, the necessity of his essence. Thus, from the necessity of his essence, i t follows t h a t God both understands himself and acts in infinitely many ways (i.e., produces the modal universe). What is the nature of God’s essence such that his understanding of himself and his infinite activity both necessarily follow from this essence? The logic of expression clarifies these questions. For essence is expressed as the essence of substance to substance. I n justification of this claim, the following argument of Deleuze is of the utmost importance: zy zyxwvu One distinguishes in an expression (say, a proposition) what it expresses and what it designates. What is expressed is, so to speak, a sense that has no existence outside the expression; it must thus be referred to an understanding that grasps [saisit] it objectively, that is ideally. But it is predicated of the thing, and not of the expression itself; understanding relates it to the object designated as the essence of that object. (EPS 62lSPE 53) zyxw This argument shows once again how the logic of expression underpins t h e claim t h a t t h e essence of substance h a s no existence outside of the attributes in which it is expressed, to t h e extent t h a t t h i s essence corresponds t o t h e sense of a n expression. But i t is also the case t h a t for a n expression to “function,” that is, for it to have sense, it must be referred to a n understanding that comprehends the e x p r e ~ s i o nThat . ~ ~ is why the essence of substance can only be expressed to substance. Equally, however, the understanding is required i n order to relate this sense to the object of which it is designated. Once more, therefore, i n order t h a t the essence expressed by the attributes of substance be expressed as the essence of substance, i n order namely t h a t substance be t h a t of which the sense is designated, substance qua understanding must relate essence t o itself as essence of itself. The logic of expression therefore reveals the necessary role played by God’s understanding of himself in the event of the constitution of his essence, that is, of his s e l f - ~ a u s a t i o n . ~ ~ Why does God produce as he understands himself? Or to pose the question i n another form, how is God’s self-understanding also productive? In order to answer these questions, we need to determine how God understands himself. We can infer from the preceding argument-that the expression must be referred to a n understanding t h a t comprehends it-that what is understood by God is t h a t i n which he is expressed, zy z 176 zyxw zy zyxw zyxw zyxwvu Immanence and Difference namely his infinite attributes (which in turn constitute God’s infinite essence). If God understands himself as an infinity of attributes and exists as the infinite essence constituted by these attributes, then he must also produce in or by these infinite attributes (according t o the reasoning both of E 11. 3s and E 1. 16: “From the necessity of the divine nature there must follow infinite things in infinite ways [modis]”[referring to E I. def 61). But we have seen that the act of production is not separate from the act of self-causation. Therefore, the attribution by which God understands and constitutes himself must be the same as the attribution by which the infinity of modes is produced. But a mode is defined as t h a t which exists in something else ( E 1. def 5). Deleuze therefore infers that: Attributes are univocal and common forms, predicated, in the same form, of ... products and producer, formally constituting the essence of one, formally containing the essence of the other [emphases added]. The principle of necessary production thus reflects a double univocity. A univocity of cause: God is the cause of all things in the same sense a s he i s cause of himself. A univocity of attributes: God produces through and in the same attributes that constitute his essence. ( E P S 103ISPE 90, translation modified) This univocity of the attributes in turn implies that attribution is itself double. And once again, it is t o the logic of expression that Deleuze turns in order to account for this doubleness of the attributes. The move made by Deleuze at this juncture comprises the most audacious wager in his attempt t o secure an immanent ontology of univocity from Spinoza’s text. The stakes are absolute: if he is unable t o demonstrate that there are two orders of attribution and that both orders “function” univocally, the whole project collapses. And what Deleuze proposes is remarkable. He seeks t o show that this conception of the lynchpin of Spinozist immanent ontology derives its sense from the logic of expression understood in a strictly linguistic manner.43Let us consider an expression (or proposition) P,, with sense S,, designating an object 0,. S, does not collapse into P,, which expresses it; nevertheless, it cannot exist outside of P We can say, therefore, that S, “subsists” in P,. Nor does collapse into 0,, though it is expressed of 0,.44 Now, it is a notable fact of the relation between the expression and its sense that P, cannot state S,-I can state that such and such is the case, but I cannot, in the same expression, state the sense of this expression. However, this sense can itself become a designated object 0 for a subsequent expression, P (with its own sense, S,, which, again, cannot be stated in P f. But the state of affairs 0, about which P, is expressed is in fact S,.46 If di 177 z zy zyx zyxw zyxwv Robin Durie we apply this analysis t o the triad substance, attribute, and mode, we would say that an attribute (P,) expresses a sense (S,) that subsists in the attribute-namely, the essence-which is designated of substance (0,-which in this case happens t o be what expresses itself in the attribute). But S,, the sense expressed by the attribute, t h a t is, the essence, can itself become the object designated (i.e., 0 ) in a second expression (PJ. What is the nature of this second?expression? The essence of the attribute, the sense t h a t subsists in the attribute a s expression, comes t o be designated in the second expression that has its own “new” sense ( S J . This new expression is the mode, the sense of which designates the essence of t h e attribute. But the key point t o be noted is that it is the sume sense, that is, the same essence that is expressed in each case, now as a sense subsisting in the attribute and constituting (the essence of) substance, now a s the object designated by the mode, the modal essence contained in the a t t r i b ~ t e As .~~ Deleuze writes: Each attribute is ... a n expression with a distinct sense; but all attributes designate substance a s one and the same thing. The traditional distinction between t h e sense expressed and t h e object designated ( a n d expressing itself in t h i s sense [qui s’exprime]) t h u s finds in Spinozism a n immediate field of application. The distinction necessarily generates [fondel a certain movement of expression. For the sense of a n initial proposition must i n its t u r n be made t h e designatum of a second, which will itself have a new sense, and so on. Thus the substance they designate is expressed in t h e a t t r i b u t e s , attributes expressing a n essence. Then the attributes a r e in their turn expressed: they express themselves in modes which designate them, the modes expressing a modification. ... The expression, through its own movement, generates [engendre] a second level of expression. Expression h a s within it t h e sufficient reason of re-expression. This second level defines production itself God is said to produce things, a t the same time t h a t his attributes find expression [s’expriment].( E P S 104-5lSPE 92-3) zyxwvu This second level of expression thus reproduces the first in the following manner: “Substance expressed itself in attributes, each attribute was an expression, the essence of substance was expressed. Now each attribute expresses itself, the dependent modes are expressions and a modification is expressed” (EPS 11OISPE 97). Moreover, just as the essence expressed has no existence outside of the attribute in which it is expressed, so the modification expressed has no existence outside of the modes in which it is expressed. And finally, the doubleness of attribution t h a t generates the production of modes reveals 178 zy zyxwv zyxw Immanence and Difference why God produces the modal universe in the same sense that he causes himself, if by this latter we understand t h a t genesis whereby his essence is constituted by attributive expression. 4. We have argued t h a t a t t r i b u t e s render substance as a difference-in-itself. They do so by constituting the essence of substance, from which substance’s existence follows. We have also argued that due to the twofold expressivity of attributes, the cause of the existence of substance and the cause of the production of modes is one and the same. In rendering substance as difference-in-itself, how is the essence of substance determined by attribution? With Proposition 34 in mind (“God’s power is his very essence” [ E I. 3411, Deleuze writes of t h e attributes t h a t they “seem in all this to have a n essentially dynamic role. Not that they are themselves powers [puissances]. But t a k e n collectively, they a r e t h e conditions for t h e attribution to absolute substance of an absolutely infinite power [puissance] of existing and acting, identical with i t s formal essence” (EPS 9OISPE 79). But if attributes determine t h e essence of substance as power, then, if we are to adhere to the principle of univocity, we would also have to say that the modal essence expressed by attributes is also power. Equally clearly, however, we cannot say that the power of modal essences is the same as the power that is the essence of substance, since modes are precisely not causes of their own existence. The final step in our reconstruction of a n immanent ontology of univocity based on the determinate relation of expression therefore consists in determining how the production of modes is governed by the logic of expression a n d how a n d why t h i s differentiation between t h e n a t u r e of the essence of substance and modes occurs during the production of modes. Specifically, we wish to determine the role of difference in the production of modes. For, since t h e expressivity of attribution consists i n t h e differentiation of substance, and the self-expression of attributes has the same sense (though modified) as the attributionally expressed essence of substance, we would be led to infer that the production of modes must itself be differential. To the extent that God is the cause of his own existence, that his existence follows from his essence, God’s essence is power, since, as Spinoza writes, “to be able to exist is power [posse existere potentia estl” ( E I. 11, third proof,). On the model of the determinate relation of expression we have been detailing, power is thus the sense expressed by the attributes. Given the doubleness of attribution, power then becomes what is designated or expressed by the modes which themselves are expressed i n t h e self-expression of attributes. Power is designated or expressed by modes precisely as a modification. What is the nature of the modification of power expressed by modes? This 179 zyxw zyxwvu Robin Dune question returns us to the theory of distinctions-how do the essence of substance and the essence of modes differ, such that we can say both that the power of modes is part of the power of substance ( E IV. 4p; EPS 92/SPE 811, thereby maintaining the principles of univocity and immanence and that, nevertheless, modes are different from substance? Two clues point, once again, to attributes as the source for a n answer t o t h i s question. On t h e one hand, as we have argued, attribution determines substance as a qualitative multiplicity to the extent that it renders it differential, as differencein-itself. On the other hand, it is the double expressivity of attributes t h a t accounts for the modal productivity of substance. In expressing the essence of substance, attributes render God’s power qualitatively multiple and, hence, different without numerically dividing substance. A mode is a modification of a n attribute of substance and, as such, has no existence outside of t h i s attribute. When attributes express themselves, that is, when the sense or essence of the attribute becomes the object expressed or designated by a mode, they do not qualitatively differentiate essences. How then do they express? Bearing i n mind once again the principle of univocity pertaining to attributes, we should say t h a t their expression amounts to a differentiation, just as their expression of substance rendered substance a differential qualitative multiplicity. What is differentiated is the essence of substance, which changes from being the sense subsisting in the attributional expression to being what is designated or expressed by the modes. But what sort of differentiation can essence undergo? It cannot, as we have said, be a qualitative differentiation, since this has been expressed by the first ‘level’ of attribution. We recall, however, that the theory of distinctions retrieved from Spinoza by Deleuze established t h a t what differs from qualitative distinction is numerical distinction. Thus, the second level of attribution can only express a numerical differentiation of essence. A modal essence of power is therefore a part of the essence of substancewhile yet remaining different to the essence of substance as power-to the extent that it is a degree, OF quantity, of power. As Deleuze writes: zyx zy The production of modes does, it i s true, take place through differentiation [diff.renciation]. But differentiation is in this case purely quantitative. If real distinction i s never numerical, numerical distinction is, conversely, essentially modal.. .. Attributes are so to speak dynamic qualities to which corresponds the absolute power of God. A mode is, in its essence, always a certain degree, a certain quantity of a quality. ... God’s power expresses or explicates itself modally, but only in and through such quantitative differentiation. (EPS 183fSPE 166) 180 zyxw zyxwvu zyxwv Immanence and Difference I n turn, modes of the same attribute differ from each other quantitatively by t h e degree of power of t h e essence they express (EPS 183!SPE 167). What we are arguing is that the twofold but univocal event of attribution is differential i n each of i t s aspects. The twofoldedness of differentiation 'follows' the twofoldedness of attribution. First, attribution is differential to the extent that it renders substance as difference-in-itself. But attributes also express themselves, and this self-expression is also differential. The nature of this second differentiation is crucial. We have argued that the sense of a differential multiplicity, of differencein-itself, is that it is indivisible as such or, more strictly, that its division is in fact a differentiation, since it leads to a change in kind.47Thus, as substance is rendered differential by attribution, it is, by the same event of attribution, differentiated. This is the second level of attribution. In this differentiation, substance of necessity changes i n kind without thereby becoming separate from itself. This change in kind consists in the qualitative multiplicity becoming a quantitative multiplicity. But the differential relations that pertain to this multiplicity a r e different from those pertaining to t h e qualitative multiplicity. That is to say, division of the quantitative multiplicity no longer leads to a change in kind. Rather, distinctions in the quantitative multiplicity express differences of degree, that is t o say, differences in quantity. And these differences in quantity are to be understood as expressing different degrees of God's essence, that is, degrees of power.48 The final stage i n t h i s interpretation is t h e n to find a differential basis for the nature of the distinction between the essence of modes and the existence of modes, a distinction that arises as a consequence of the fact that, unlike substance, the essence of a mode does not entail or involve the existence of t h a t mode.49Modes a r e quantities and t h u s can only differ quantitatively. But modes of the same attribute differ according to degrees of quantity, so the difference between modal essence and existence cannot be a matter of a straightforward difference of degree. Rather, Deleuze proposes that this difference should be understood as a difference between the two possible types of quantity, namely, intensive quantity and extensive quantity, each of which can, in turn, be divided into degrees ( E P S 1911 SPE 173-4). The grounds for this argument a r e again situated i n the theory of distinctions. Modal essence is what is expressed by the second level of attribution. J u s t as the essences of substance subsist i n t h e a t t r i b u t e s t h a t express them, s o also modal essences subsist in the attributes by which they are expressed. As a modification of the attribute, the modal essence cannot differ qualitatively from the attribute; equally, since the modal essence does not yet exist separately in the existent mode, it 181 Robin Dune h a s no extrinsic difference from t h e a t t r i b u t e i n which i t subsists. What then is t h e n a t u r e of t h e difference t h a t distinguishes the modal essence from the attribute and from other modal essences? I t is a matter of quantity, as we have seen; since it cannot be a matter of extrinsic difference, the difference between attribute and modal essence and between modal essence and modal essence must be intrinsic. How might we conceive of intrinsic differences of quantity? Here, once again, Deleuze subjects Spinoza to a certain explication, introducing into the text a mode of expression borrowed from Duns Scotus but found also in Bergson. Let us take the example of a color, such as whiteness: zyxw zyx zyxwvu whiteness h a s various intensities; these a r e not added to whiteness as one thing to another thing ..., its degrees of intensity are intrinsic determinations, intrinsic modes, of a whiteness that remains univocally the same under whichever modality it is considered. This seems also t o be the case for Spinoza: modal essences are intrinsic modes or intensive quantities. An attribute remains a s a quality univocally what i t is, containing all the degrees that affect it without modifying its formal reason. Modal essences a r e t h u s distinguished from their a t t r i b u t e a s intensities of i t s quality, and from one another a s different degrees of intensity. (EPS 196-7lSPE 179-80) We now have in place the first two elements of a tripartite distinction. Substance is an indivisible qualitative multiplicity. Modal essences are indivisible quantitative multiplicities. They are indivisible because they are intensive, intrinsic (EPS 1971 SPE 180). We can surmise, then, that the final element will be a quantitative multiplicity t h a t is extrinsic and thus divisible. And indeed, this is how Deleuze characterizes the multiplicity of the modal existents.50 In the discourse on physics appended to Proposition 13 of Part I1 of the Ethics, Spinoza makes clear that the cause of a mode’s existence is another mode, which also exists.51But, as Deleuze asks, in what does the existence p e r se of the mode consist (EPS 201ISPE 183)? A mode exists, Deleuze argues, if it has a “very great number of parts,” external both to one another and t o the mode’s essence, “corresponding t o the essence or degree of power of the mode” (EPS 202lSPE 20252).These parts correspond to the extensive quantity of the attribute and a r e actually divided into the infinity of parts in which the existence of the mode consists (EPS 205lSPE 187). The way in which the extrinsic modes a r e distinguished from, and related to, one another is through motion and rest, as Spinoza makes clear in his The relation between modal essence a n d t h e existence of modes is once more one of expression. The intensive modal essence expresses itself in a relation t h a t , Spinoza’s zyxw 182 zyxw zy zyxwv zyx Immanence and Difference physics argues, takes the form of a mechanical law. A mode comes into existence as and when “an infinity of extensive parts are actually determined to enter into this relation,” that is, are caused to enter into such a relation by another existent mode (EPS 208/SPE 191). In other words, if a great number of parts enter into a formally determinate relation, they are determined by t h i s relation, which is itself t h e expression of a modal essence, and hence a r e determined as the expression of this modal essence. From t h i s it is clear t h a t , while t h e modal essence determines the existent mode, it is not thereby t h e cause of the mode-it is always the case that one mode causes another to enter into the formally determinate relation-and thus the existence of the mode does not follow from its essence. It is also clear that, while the essence of substance is power and the existence of substance follows from this power, the power constituting the essence of modes is only a degree of the infinite power of substance and therefore cannot involve the existence of modes. Thus the power of the mode can only be expressed as a striving, a conatus, to maintain itself in existence. 5. Deleuze turns to Bergson in order to explicate Spinoza so that he can secure the univocal, immanent ontology that, he argues, was already the goal of Spinoza’s philosophy in the first place. While this explication is guided by the principle of expression, we have argued that the genuine sense of the explication is to be discerned i n the priority accorded to relations, specifically the differential nature that constitutes the determinate form of these relations. It is from this perspective that Deleuze is able to develop Spinoza’s ontology in such a way that the charge of indifference that he leveled against Spinoza can be avoided. It is the very same movement of attribution-understood as the genetic milieu of t h e relation of expression-which both differentiates the essences of substance, thereby constituting substance as a qualitative multiplicity or difference-in-itself, and differentiates the modes from substance, and in turn from one another. I n the reading t h a t we have developed, we have ourselves sought to explicate Deleuze. While it is apparent that Bergson is a decisive influence on the way Deleuze interprets Spinoza, it remains the case t h a t the specifically differential nature by which we have characterized t h e logic of t h e relation of expression-a differential nature whose provenance is precisely Bergsonian-remains at most implicit within Deleuze’s text. However, in arguing t h a t Deleuze is able to secure a univocal, immanent ontology from Spinoza through a differential conception of the relation of expression, we have ourselves been able to delineate a relational ontology t h a t is fundamentally differential. It is just such a differential relational ontology that 183 zyxw zyxwvu Robin Dune in turn could ground the relational approach relied upon by the physics of complexity and the theoretical developments in quantum gravity proposed by Lee Smolin, Julian Barbour, and their ilk. Notes Lee Smolin, The Three Roads to Quantum Gravity (London: Weidenfeld and Nicolson, 2000), 120. See also Smolin, The Life of the Cosmos (Oxford: Oxford University Press, 1997);Julian Barbour, The End of Time (London: Weidenfeld and Nicolson, 1999); and Robin Durie, ed., Time and the Znstant (Manchester: Clinamen Press, 20001, especially chs. 5 and 6 . Gilles Deleuze, Difference and Repetition, translated by P. Patton (London: Athlone, 1994), 35lDiffkrence et rkpktition (Paris: Presses Universitaires de France, 1968), 52. Thomas Kuhn, The Structure of Scientific Revolutions (Chicago: University of Chicago Press, 1970), 7. Gilles Deleuze, Expressionism in Philosophy: Spinoza, translated by M. Joughin (New York: Zone Books, 1990);Spinoza et le probltme de Z’expression (Paris: Les Editions de Minuit, 1968). (Cited hereafter as EPSISPE). See for example EPS 95lSPE 84. This passage is mistranslated by Patton as “there still remains a difference between substance a n d t h e modes” (Difference a n d Repetition, 40lDiffkrence et rkpktition, 59). Cf. Jonathan Bennett, “Spinoza’s Metaphysics,” in The Cambridge Companion to Spinoza, edited by D. Garrett (Cambridge: Cambridge University Press, 1996), 72. We employ the term “explication” to capture something of what is a t stake in Deleuze’s interpretative methodology. Of course, this term already plays a key function in Deleuze’s work, and we have allowed t h i s to guide our use of t h e term. Explication t h u s designates a process of unfolding of what had previously remained “implicated,” or enfolded. In a parallel manner, one could speak of modes a s being the products of t h e explication, o r unfolding, of substance. What is important to bear in mind is that in such a process, what is unfolded does not simply remain identical to what was enfolded-that is to say, the process of explication “makes a difference.” In this way, Deleuze’s reading of Spinoza goes beyond mere commentary. In fact, it is noteworthy that this reading is guided by a series of problematics generated by Spinoza’s text itself. Bearing this in mind, we might note that Martin Joughin’s otherwise excellent translation of Spinoza et le probleme de l’expression is marred only by the decision to retitle t h e work Expressionism in Philosophy: Spinoza. What is partially lost in this change is a sign of the means by which Deleuze’s text makes i t s way through Spinoza’s Ethics, t h a t is, by way of problems. The force of his “expressionistic” reading is to be felt in the way in which the unifying logic of expression enables both a coherent reading of the whole of the Ethics and the development of a series of responses to problems t h a t have accreted through generations of Spinoza scholarship. If, a s he constantly reminds us, the Ethics is in fact two books, made up of the continuous series of propositions and demonstrations supplemented by the discontinuous series of polemical scholia, t h e n Deleuze’s own exposition could be said to have two 184 z Immanence and Difference aspects, a continuous exposition of the text supplemented by a series of more or less polemical engagements with the history of Spinoza scholarship. Yet the unity, and the radicality, of Deleuze’s text is to be discerned in the fact that, far from these two aspects of his reading being separate, it is the explication to which he subjects Spinoza’s text that, ultimately, enables him to sustain the coherence of his reading and, hence, of the immanent ontology that emerges from it. In “A Nearly Total Affinity,” Len Lawlor argues that one significant difference between the modi operandi of Derrida and Deleuze in their respective readings of Husserl and Bergson is that the former “always seeks contradictions” where t h e l a t t e r “always seeks consistency” (Angelaki 5.2, August 2000, 7 0 ) .This is a typically perceptive observation. What we are suggesting here is that the consistency in the case of Deleuze’s reading of Spinoza is to be found in the reading and that the condition of this consistency is the explication to which he subjects Spinoza’s text. It is interesting that even in his Kant book, where he confronts the philosopher who is perhaps most antithetical to his own spirit, Deleuze’s reading is nevertheless governed by a n extraordinary consistency, namely, the attempt to trace through the whole of Kant’s critical corpus a n immanent “genesis of common sense” amongst the faculties, t h a t is, to read Kant’s philosophy a s “immanent Critique” (Deleuze, Kant’s Critical Philosophy, translated by H. Tomlinson and B. Habberjam [London: Athlone, 19841, 3 a n d 24lLa Philosophie critique de Kant [Paris: Presses Universitaires de France, 19631, 7 and 3 6 ) . I t is only in t h e very l a s t pages t h a t the possibility of such a reading is shown to be betrayed by a n irreducible transcendentalism in Kant’s thought. * See Durie, “Splitting Time: Bergson’s Philosophical Legacy,” Philosophy Today 44 (Summer 20001, 152-68. Deleuze, “Bergson’s Conception of Difference,” translated by M. McMahon, in The New Bergson, edited by J. Mullarkey (Manchester: Lanchester University Press, 1999),42/La Conception de la Diffbrence chez Bergson,” in ktudes Bergsoniennes 4 (1956):79. lo Letter 8, in Spinoza, The Letters, translated by S . Shirley (Indianapolis: Hackett, 1995), 90. l1 The Philosophical Writings of Descartes, translated by J. Cottingham, et. al. (Cambridge: Cambridge University Press, 19851, 210. l2 Descartes, Reply to the Second Set of Objections, Definition 10, Philosophical Essays and Corespondence, edited by R . Ariew (Indianapolis: Hackett, 20001, 363. l3 Spinoza brooks no argument with Descartes’s conviction t h a t thought and extension are really distinct attributes. l4 Spinoza, Ethics, translated by S . Shirley (Indianapolis: Hackett, 1992)lEthica in Opera, edited by C. Gebhardt (Heidelberg: Carl Winter, 1925). l6 Edwin Curley, Behind the Geometrical Method (Princeton: Princeton University Press, 1988), 28-9. l6 Although Definition 6 , which reads “each [attribute] expresses eternal a n d infinite essence [quorum unumquodque teternam et infinitam essentiam erprimitl” does not of itself license such a reading, the scholium to Proposition 10 does. There, Spinoza slightly amends Definition 6 , writing of the attributes of God t h a t “each expresses a certain eternal and infinite essence [quorum unumquodque teternam et zyxwvu zyxwvu zyx zyxw 185 zy zyxw zyx zy zyxwv zy zy Robin Dune infinitam certam essentiam exprimit].”I t should be noted in passing t h a t Deleuze imports t h e wording of t h i s Scholium into h i s [inaccurate] citation of Definition 6 in the first lines of his Introduction to Expressionism in Philosophy (EPS 13/SPE 9 ) . This is not a n unimportant solecism, a s is made clear by t h e lines immediately following in the text, which lay specific emphasis on the distinctness of essences constituted by each distinct attribute. l7 Cf. Jonathan Bennett, A Study of Spinoza’s Ethics (Indianapolis: Hackett, 1984), 64. Cf. Curley’s discussion of this point in Behind the Geometrical Method, 28-9. These issues a r e also discussed in Alan Donagan’s “Essence and the Distinction of Attributes,” Spinoza: A Collection of Critical Essays, edited by M. Grene (New York: Anchor, 1973), 164-81. Donagan analyzes two solutions to the problem, advanced by Wolfson and Gueroult. Wolfson argues that Spinoza intends that the difference between attributes is not a distinctio realis but only a distinctio rationis. Before Spinoza, both Aquinas and Moses Maimonides had argued t h a t , since God must be conceived a s “simple,” attributes cannot really be distinct. As Donagan shows, there is ample textual evidence to refute this line of interpretation. Gueroult’s argument is t h a t t h e essences expressed by each really distinct attribute a r e themselves really distinct, but a s such “must not be confounded with t h e divine essence itself.” This is because t h e essence of God is absolutely infinite, whereas the essences expressed by attributes are only infinite in their kind. Gueroult’s argument is, therefore, that the essence of divine substance is a n infinite “set” constituted by “an infinity of essences of substances, each infinite in its kind” (Donagan, 173, 177). Once again, however, the textual evidence of Definition 4 indicates that Spinoza in fact holds that a n attribute is that which the “intellect perceives of substance a s constituting its essence,” and thus t h a t “each divine a t t r i b u t e express[es] t h e same divine essence” (Donagan, 177). l9 We suggested above that the course charted by Deleuze through the Ethics is determined by a more or less polemical engagement with a series of problems t h a t have troubled the tradition of Spinoza scholarship. This proposal is confirmed a t the outset when the claim t h a t the role of expression a t the s t a r t of the Ethics is to introduce distinction into infinity is given the following context by Deleuze: “What is the character of distinction within infinity? What sort of distinction can one introduce into what is absolute, into the nature of God? Such is the first problem posed by the idea of expression ...” (EPS 28ISPE 22) . And such is the problem we have been outlining thus far. 2o Descartes, Principles of Philosophy, 211. 21 Numerical distinction cannot establish t h e distinctness of substances for two interconnected reasons: on the one hand, the need to distinguish numerically between substances would presuppose substances with the same attribute-and a common attribute precisely does not allow for the distinction between substances; but on the other hand, “substance is always prior to its affections” ( E I. 5d). 22 In Difference and Repetition, Deleuze writes: ‘“Multiplicity’, which replaces t h e one no less t h a n t h e multiple, is t h e t r u e substantive, substance itself” (182/Diflbrence et rbpbtition, 236). 23 See, for instance, Henri Bergson’s Time and Free Will, translated 186 zyxw zyxw zyx zyxw zy Immanence and Difference by F. L. Pogson (London: Macmillan, 1910), 105; Guvres, edited by A. Robinet (Paris: Presses Universitaires de France), 70, for his introduction of the notion of the qualitative multiplicity. 24 Macherey also claims that there is precious little employment of the notion of quantity in the Ethics either. Pierre Macherey, “The Encounter with Spinoza,” in Deleuze: A Critical Reader, edited by P. Patton (Oxford: Blackwell, 1996), 150-1. 26 Riemann, Gesammelte mathematische Werke und wissenschaftlicher Nachlass (Leipzig: Teubner, 18761, 254 ff. (“On t h e Hypotheses which Provide the Grounds for Geometry,” translated by W. K. Clifford, Nature 8:183). 26 Thus, for instance, the factors 2 and 3, of 6, remain the same when 6 is divided by one or the other. On the other hand, if one were to t r y to divide out t h e emotion of happiness from a stream of consciousness, then the whole consciousness would be fundamentally changed, since t h e emotion pervades t h e organic whole of consciousness (Time and Free Will, 9-10 and 100-01; &uvres, 10-11 and 67-8). Bergson thus concludes: “What we must say is that we have to do with two different kinds of reality, the one heterogeneous ... the other homogeneous” (Time and Free Will, 97; G u v r e s , 66); a n d subsequently, that the qualitative multiplicity “might well be nothing but a succession of qualitative changes, which melt into and permeate one another, without precise outlines, without any tendency to externalise themselves i n relation to one another, without any affliation to number: it would be pure heterogeneity” (Time and Free Will, 104; Guures, 70). 27 “Bergson’s Conception of Difference,“ 47-8; “La Conception de la Diffkrence chez Bergson,” 88. See also ch. 2 of Deleuze’s Bergsonism, translated by H. Tomlinson and B. Habbejam (New York: Zone Books, 1991); Le bergsonisme (Paris: Presses Universitaires de France, 1966). At stake in the attempt to think “difference-in-itself“ is the possibility of conceiving a pure difference that does not reduce to the standard markers of derivative, external, difference-contradiction or negation -i.e., difference from a n always already fured identity (ibid., 49/90). In a valuable discussion, Michael Hardt notes the affiliation of this phrase to Deleuze’s early essay on Bergson, where he writes: “To think internal difference as such, a s pure internal difference, to reach the pure concept of difference, to raise difference to the absolute, such is the direction of Bergson’s effort” (Deleuze, “Bergson’s Conception of Difference,” 49I“La Conception de la diffbrence chez Bergson,” 90; cf. Hardt, Gilles Deleuze: An Apprenticeship in Philosophy [Minneapolis: University of Minnesota Press, 19931, 61). 29 Bennett, “Spinoza’s Metaphysics,” 85. 30 A remark by Andre Darbon, cited by Deleuze i n a different context (though one t h a t rejoins our discussion of the previous section), inadvertently lends support to this viewpoint. “To explain the unity of substance, Spinoza tells us only that each attribute expresses its essence. The explanation, far from being any help, raises a host of difficulties. In the first place, what is expressed ought to be different from what expresses itself ...” ( k t u d e s spinozistes [Paris: Presses Universitaire de France, 19461, 117). 31 Spinoza, The Letters, 61. 32 Clearly, the inspiration for seeking to make sense of the triad of attribute, essence, and substance in this dynamic way is Bergsonian zyxw 187 zyxw zyx zyxwvu Robin Durie once again. It is Bergson, after all, who writes that “the permanence of substance [is], in my eyes, a continuity of change” (The Creative Mind, translated by M. L. Andison [New York: Citadel Press, 19921, 88; (Euvres, 1328). 33 One further component of the relation of expression will become significant a t a subsequent point in the argument: t h a t which t h e expression is expressed to (namely, a s we shall see, a n understanding t h a t relates t h e expressed sense to t h e object designated by t h e expression). 34 The Bergson-inspired genetic reading of Spinoza t h a t we a r e arguing Deleuze proposes at this point represents a considerable challenge to the history of Spinoza scholarship. I t is not without reputable precedent, however. As Pierre Macherey notes, both Lewis Robinson and Martial Gueroult “have emphasized the ‘genetic’ and not ‘hypothetical’ nature of the first propositions of the Ethics,” and he goes on to argue that their interpretation entails that “the causa sui is nothing but the process within which substance engenders itself on the basis of the ‘essences’ that constitute it, on which its existence is established: this movement leads to the moment in which it produces substance, a s the product of i t s activity, a s t h e result of i t s own deternunation” (Pierre Macherey, “The Problem of the Attributes,” translated by E. Stoltze, in The New Spinoza, edited by W. Montag and E. Stoltze [Minneapolis: University of Minnesota Press, 19971, 75-7). 35 Edwin Curley, Spinoza’s Metaphysics: An Essay in Interpretation (Cambridge, Mass.: Harvard University Press, 1969), 19. 36 Bennett, “Spinoza’s Metaphysics,” 67. 37 John Carriero, “Mode and Substance in Spinoza’s Metaphysics,” in The Rationalists: Critical Essays on Descartes, Spinoza and Leibniz, edited D. Pereboom (Maryland: Rowman and Littlefield, 1999), 139. 38 In Difference and Repetition, Deleuze writes: ”There has only ever been one ontological proposition: Being is univocal” (35/Difkrence et rdpdtition, 52). 39 “The univocity of attributes merges with their expressivity: attributes are, indissolubly, expressive and univocal” (EPS 59lSPF 50). 40 This argument will be explicated through reference to t h e understanding of God’s essence a s power ( E I. 34). We return to this point when we come to pose the question of the being of modes in our final section. 41 This argument is repeated in the theory of expression [Ausdruckl adumbrated in the first chapter of Husserl’s first Logical Investigation, translated by J. N. Findlay (London: Routledge and Kegan Paul, 1970); Logische Untersuchungen, Husserliana 19/1 (The Hague: Martinus Nijhoff, 1984). In his later work, Husserl will designate sense with the notion of the noema. The noema is precisely the object as it is intended by consciousness, a s it appears in and to consciousness, the sense it has for consciousness. 42 Deleuze reviews this argument a t the s t a r t of chapter 6 (EPS 101ISPE 9), where he emphasizes that the understanding neither, by this argument, constitutes the essence of substance nor “attributes” the attributes of substance. Rather, the understanding is that to which the attributes are referred, that which, precisely, understands what is predicated of substance itself. I n this way, t h e apparent priority accorded to the attribute of thought-to the extent that the possibility of the function of expressing is necessarily dependent upon a n under- zy 188 zyxwv zyxwvu zyxwv zyx Immanence and Difference standing t h a t comprehends t h e expression-is shown to have no ontological significance. 43 It should nevertheless be emphasized t h a t what is at stake in this wager is not so much the linguistic provenance of the analysis but rather the light i t sheds on the nature of the formally determinate relation under scrutiny. Cf. Deleuze, Logic of Sense, translated by M. Lester with C. Stivale (New York: Columbia Press, 19991, 2llLogique d u sens (Paris: Les Editions de Minuit, 1969), 33. 46 Deleuze, Logic of Sense, 28-9ILogique d u sens, 41. 46 Although the same sense is expressed in each case, what is of significance is t h a t modal expression expresses a sense t h a t is, precisely, modified. We shall examine the specific n a t u r e of t h i s modification and its differential basis in the final section of this paper. 41 This argument parallels the point made by Bergson in Time and Free Will t h a t the indivisibility of the heterogeneous multiplicity actually entails that any division would result in a change in kind to the multiplicity (Time and f i e e Will 101; (Euures 68). 48Deleuze provides a summary of t h i s argument from t h e perspective of expression a t the end of the second part of Expressionism in Philosophy (EPS 185-6lSPE 168-9). Clearly, the nonexistence of the mode does not therefore entail the nonexistence of the modal essence; cf E I. 8, s2 and 11. 8 , enunciation and corollary. In passing, we should note here Pierre Macherey’s criticism that through emphasizing the distinction between quality and quantity, Deleuze is foisting an oppositional mode of thinking onto Spinoza and thereby reproducing the very “binary machines” of which he is so critical elsewhere (“The Encounter with Spinoza,” 150-1). Macherey is, to be sure, a perceptive reader of Deleuze, but this observation entirely overlooks the theory of distinctions and the logic of expression that we have argued a r e so fundamental in Deleuze’s reading. As we have demonstrated, expression names the formally determinate differential relations t h a t determine the n a t u r e of the elements of Spinoza’s ontology. Modes are precisely not opposed to substance, just as quality is not opposed to quantity. Quantity is first a n internal differentiation of quality. The fact that the intensive multiplicity “overlaps” with the qualitative multiplicity in being indivisible, while also overlapping with the extensive multiplicity through being quantitative, should also make clear that Deleuze’s reasoning here is not oppositional. 61 This is first demonstrated in E I. 8s2, where Spinoza concludes t h a t “the cause of t h e existence” of a mode “must necessarily be external” to the mode (emphasis added). 62 Referring to “The idea which constitutes the formal being of the human mind is the idea of the body, which is composed of a very great number [plurimis ualdel of composite individual parts (Postulate 1: The human body is composed of very many individual p a r t s of different natures, each of which is extremely complex [ualde compositum])”(E 11. 15d). 63 Cf E 11. Axiom 1 , Lemma 2 proof, and Lemma 3 statement and proof. zyxwvu 189