Name: Empiricism and Subjectivity
Author: Gilles Deleuze

About the Work

Empiricism and Subjectivity: An Essay on Hume’s Theory of Human Nature is Gilles Deleuze’s study of David Hume, translated into English by Constantin V. Boundas and published by Columbia University Press in New York as part of the “European Perspectives” series in social philosophy and cultural criticism, edited by Lawrence D. Kritzman and Richard Wolin. The series exists to make available interdisciplinary work by leading European thinkers—classic texts alongside outstanding contemporary ones—so as to shape major intellectual controversies and facilitate historical understanding; its other titles include Theodor W. Adorno’s Notes to Literature and Julia Kristeva’s Strangers to Ourselves. The volume opens with a preface by Deleuze, followed by the translator’s acknowledgments and Boundas’s introduction, “Deleuze, Empiricism, and the Struggle for Subjectivity.” Its core is six numbered chapters—”The Problem of Knowledge and the Problem of Ethics”; “Cultural World and General Rules”; “The Power of the Imagination in Ethics and Knowledge”; “God and the World”; “Empiricism and Subjectivity”; and “Principles of Human Nature”—and it closes with a conclusion titled “Purposiveness,” followed by notes and an index.

“We are habits, nothing but habits—the habit of saying “I.”

The book is not a neutral exposition of a historical figure. It is a philosophical intervention, written by a philosopher who treats the history of philosophy as a workshop in which concepts are forged. Deleuze’s 1989 preface imagines a history of philosophy that would list only the new concepts a great philosopher created, and it assigns Hume three such creations: the concept of belief, placed where knowledge had stood; the real meaning of the association of ideas, understood as a practice of cultural and conventional formations rather than a theory of the human mind; and the first great logic of relations, in which all relations—not only matters of fact but relations among ideas—are external to their terms. The book is thus at once a reading of Hume and a statement of Deleuze’s own philosophical commitments, and it belongs to the earliest stratum of his work.

Context

Empiricism and Subjectivity is Deleuze’s second book-length publication, initiated in 1952 and brought out in the “Epiméthée” collection of Presses Universitaires de France. A year earlier, with André Cresson, Deleuze had released David Hume: Sa Vie, son oeuvre, avec un exposé de sa philosophie, whose chapter “Complément sur l’oeuvre” contains a less elaborate version of the Hume discussion. In 1972 Deleuze returned to Hume in a chapter for the Histoire de la philosophie edited by François Chatelet for Hachette Littérature—an abbreviated version of the book with no significant departure from its major points. He did not revisit Hume again, except for reminiscing references.

That early date matters for understanding the book’s place in Deleuze’s development. Boundas observes that Hume is absent from the memories and tributes of A Thousand Plateaus, but argues that an argument from silence would miss the point: the intensive encounter with Hume gave Deleuze a decisive preference for empiricism against all forms of transcendental philosophy. The empiricist principle of difference and the theorem of the externality of all relations strengthened Deleuze’s choice of a minoritarian discourse and fed the problematic of paratactic serializations, consolidating his opposition to the petitio principii of theories that endow the transcendental field with subjective—egological and personological—coordinates that ought instead to be accounted for. Boundas cites Deleuze, writing with Claire Parnet in Dialogues, on Hume displacing empiricism and giving it a new power: a theory and practice of relations, of the AND. The theory and politics of paratactic discourse may evoke The Logic of Sense, Kafka: Toward a Minor Literature, and A Thousand Plateaus, but the quotation implicates an earlier segment inscribed with Hume’s name—one that can be considered alongside the Bergson and Leibniz segments to chart Deleuze’s nomadic image of thought.

The translator’s introduction also sketches possible series in Deleuze’s thought: a Nietzsche-Foucault series, asking whether a dynamic genesis of subjectivity can be built where the subject is the folding and internalization of Outside forces without falling into a philosophy of interiority; and a Nietzsche-Klossowski series, asking whether the subject can be thought through inclusive disjunctions and simultaneously affirmed incompossible worlds. These series would have run along their own lines of flight without forming any planes of consistency among them, were it not for Deleuze’s concepts “chaosmos” (chaos plus cosmos) and the “cracked I,” which as portmanteau words circulate through the series and enable the inclusive, disjunctive affirmation of all of them. Chaosmos—the becoming-world—posits the constitution of the subject as a task, and guarantees that the constituted subject emerges not as a substantive hypokeimenon but as an always already cracked I. The germs of these series appear in the early Empiricism and Subjectivity, mostly a segment of the Hume-series crisscrossed by segments of others. It treats the structure-Subject via anticipation and invention, and the actualization of the Subject in concrete, changing circumstances. It is coordinated by the question “how does the mind become subject?” and weaves subjectivity through belief, anticipation, and inventiveness.

Structure of the Book

The book’s arc is governed by a single question—how does the mind become a subject?—and it answers that question twice over, once for knowledge and once for ethics, before drawing the two answers together. The opening chapter, “The Problem of Knowledge and the Problem of Ethics,” sets the terms: the mind is a collection without an album, a play without a stage, a flux of perceptions, and the problem is how this collection comes to be qualified as a subject. The second chapter, “Cultural World and General Rules,” turns to the moral and social domain, where the naturally given elements—sympathies—exclude one another and must be integrated by invented rules. The third chapter, “The Power of the Imagination in Ethics and Knowledge,” examines the imagination as the medium in which passions reflect themselves and become general rules, and it distinguishes the rule of taste from the rules of property and sovereignty. “God and the World” extends the analysis to the largest fictions, the ideas of Subject, World, and God, which the same critique that disciplines the mind ends up producing as incorrigible and constitutive. The fifth chapter, “Empiricism and Subjectivity,” draws the philosophical moral: empiricism is the theory of the externality of relations, and the subject is constituted inside the given rather than grounding it. The final chapter, “Principles of Human Nature” gathers the principles of association and of passion into a single account of human nature as the real object of science. The conclusion, “Purposiveness,” closes the argument by asking after the finality of the whole—what the system of human nature is for, and how its parts cohere.

Inside the Book

The Series, the Preface, and the Translator’s Frame

The volume belongs to Columbia University Press’s “European Perspectives” series in social philosophy and cultural criticism, edited by Lawrence D. Kritzman and Richard Wolin, which sets out to bring leading European thinkers—classic and contemporary—into interdisciplinary circulation so as to shape major intellectual controversies. Alongside titles such as Adorno’s Notes to Literature and Kristeva’s Strangers to Ourselves stands Empiricism and Subjectivity: An Essay on Hume’s Theory of Human Nature, translated and introduced by Constantin V. Boundas. The contents run from Deleuze’s preface through the translator’s acknowledgments and introduction, “Deleuze, Empiricism, and the Struggle for Subjectivity,” into six chapters—on the problem of knowledge and the problem of ethics; cultural world and general rules; the power of the imagination in ethics and knowledge; God and the world; empiricism and subjectivity; and the principles of human nature—closing with a conclusion titled “Purposiveness,” notes, and an index.

Deleuze’s 1989 preface imagines a history of philosophy that records only the new concepts a great philosopher created, and credits Hume with three. First, Hume established belief and installed it in the place of knowledge, laicizing belief: knowledge becomes a legitimate belief, and the question becomes what conditions legitimate it—a sketch of a theory of probabilities. If thinking is believing, thought has less to fear from error than from illusion, since illegitimate beliefs surround it like a cloud; an art with rules must sort legitimate beliefs from the illusions accompanying them, and here Hume anticipates Kant. Second, Hume gave the association of ideas its real meaning, making it a practice of cultural and conventional—not contractual—formations rather than a theory of the mind; association exists for the sake of law, political economy, aesthetics. Deleuze’s illustration is the question of whether shooting an arrow at a site suffices to own it, or whether one must touch the spot with one’s own hand: a question about the correct association between person and thing. Third, Hume created the first great logic of relations, showing that all relations—among matters of fact and among ideas alike—are external to their terms, constituting a multifarious world of experience on the principle of the exteriority of relations. We begin with atomic parts, but these carry transitions, passages, tendencies circulating between them, giving rise to habits. To the question “what are we?” Deleuze offers Hume’s answer: we are habits, nothing but habits—the habit of saying “I.”

Boundas’s acknowledgments thank Jacqueline Code, Allison van Rooy, and Réal Fillion for earlier drafts, Susan Dyrkton for editorial advice, Marg Tully for typing, Trent University for a 1989–1990 sabbatical, and Francois Laruelle and Anne-Françoise Schmid-Laruelle for hospitality in Paris, with a heartfelt dedication to Linda Carol Conway. His introduction opens from Deleuze’s remark, in Dialogues with Claire Parnet, that Hume displaced empiricism and gave it a new power: a theory and practice of relations, of the AND. The politics of paratactic discourse evoke The Logic of Sense, Kafka, and A Thousand Plateaus, but the quotation implicates an earlier segment inscribed with Hume’s name, which Boundas sets beside the Bergson and Leibniz segments to chart Deleuze’s nomadic image of thought. Empiricism and Subjectivity is Deleuze’s second book-length publication, begun in 1952 and issued in 1953 in the “Epiméthée” collection of Presses Universitaires de France; a year earlier he had published, with André Cresson, David Hume: Sa Vie, son oeuvre, whose “Complément sur l’oeuvre” holds a less elaborate version of the discussion. In 1972 he returned to Hume in a chapter for François Châtelet’s Histoire de la philosophie, an abbreviated version with no significant departure. He has not revisited Hume since, save in reminiscence; Hume is absent from the tributes of A Thousand Plateaus, but an argument from silence would miss the point, for the intensive encounter gave Deleuze a decisive preference for empiricism against every transcendental philosophy. The empiricist principle of difference and the theorem of the externality of all relations strengthened his choice of minoritarian discourse, fed the problematic of paratactic serializations, and consolidated his opposition to theories that endow the transcendental field with subjective coordinates it should instead account for.

Empiricism as the Theory of External Relations

A proper definition of empiricism must respect an irreducible dualism between things and relations, atoms and structure, perceptions and their causes. Empiricism is thus the theory of the externality of relations; any theory deriving relation from the nature of things is resolutely nonempiricist. Deleuze’s commitment rests on the conviction that relations are syntheses whose provenance cannot be explained by the representationalist matrix of idea/atom or mind/collection of atoms: relations are the effect of the principles of human nature, and those principles constitute the subject at the same time as they constitute relations. The essay captures Hume’s dual strategy of atomism (the different, the disparate) and associationism (serialization, parataxis). Atomism is the theory of ideas insofar as relations are external to them; associationism is the theory of relations insofar as they are external to ideas and depend on the principles of human nature. Rather than pulverizing the given, Hume studies the mechanism by which atoms fit into a structure. So long as the mind is a collection of atoms in motion, mover and motion indistinguishable, like moving images without a frame, atomism cannot suffice for a science of humanity; that science arises only once the mind is naturalized by associative principles, and the subject is constituted inside the mind as the product of principles transcending it.

The doctrine of the externality of relations, rooted in atomism and introducing associationism, serves the critique of phenomenology and the quest for Bergson’s elemental world through two shared premises: the principle of difference and the serialization or compossibility of different elements. The given is a collection of ideas separable because different, different because separable—which requires that the mind be neither Subject nor Mirror of Nature. No impression is adventitious; all are in some sense innate. Before the subject is constituted, no principle of organization rules the mind; only the indivisibility of impressions interests Hume, licensing the principle of difference and guaranteeing that the mind’s constants will be indivisible atoms. Empiricism is therefore not a philosophy of the senses but of the imagination, and “all ideas derive from impressions” is a regulative principle keeping us within atomist difference. Difference alone does not make an empiricist philosophy: difference and repetition must relate chiastically. From differential perceptions a subject is born inside the given, and imagination becomes a faculty; terms are related and serialized. When a law of reproduction forms under the impact of the principles of human nature, the subject comes to be and begins to transcend the mind—but repetition cannot occur without difference. The principles may be the necessary condition for relations in general, yet particular relations and actual subjects require concrete, different circumstances as sufficient conditions; circumstances define passions and give direction to interests, since affectivity and circumstance go together. Given the primacy of practical interest, the principles of passion are indispensable for concrete associations and so for the constitution of the subject.

Deleuze’s choice of empiricism displaces dialectics. The principle of difference prevents the closure threatened by sublation; hypotactic subsumptions give way to paratactic conjunctions, arborite constructions to the strategy of the AND. Repetition—time and habit—holds paratactic series together, making possible their convergence and compossibility as well as their divergence and resonance, displacing the dialectical labor of the concept and thwarting negation’s mobilization for allegedly superior synthesis. This is a choice for a critical but nontranscendental philosophy. Transcendental philosophy begins from a methodologically reduced field and asks how there can be a given, or how a subject can give itself the given; Hume’s empiricism asks how a subject can be constituted inside the given. The subject is a task to be fulfilled. Empiricism generates a critique of rules by means of rules: extensive rules are criticized and rectified through corrective rules, both originating in habit—though habit is not solely the product of experientially ascertained repetition of similar cases. The task is the constant correction of the imagination by the understanding: habit extends imagination’s range but also corrects judgment’s accuracy. Critique must discipline the anticipating subject, focus it on objects determined by the understanding and the weight of observed repetitions, and educate its moral activity, the act according with the intensive integration of disparate sympathies. Yet a paradox lies at empiricism’s heart: the same critique that disciplines the mind and prompts rejection of imagination’s fictions also leads it to the biggest fiction—Subject, World, and God—turning these into incorrigible, constitutive ideas.

Reason, Feeling, and the Two Senses of Impression

Hume’s philosophy, on this reading, is not chiefly a critique of relations but of representation, since a representation can never present a relation. Rationalism turns representation into a criterion, lodging ideas inside reason and demanding that they stand for what experience cannot constitute: the generality of the idea, the existence of the object, and the content of “always,” “universal,” “necessary,” “true.” In doing so it transfers mental determinations onto external things and strips philosophy of any intelligible account of practice or of the subject. The mind is not reason; reason is an affection of the mind, an instinct, habit, or nature — “nothing but a general calm determination of the passions, founded on some distant view or reflexion.” Reason is a kind of feeling, and the theory of reason accordingly moves from a skepticism of reason to a positivism of feeling that includes reason as a reflection of feeling in the qualified mind.

A crucial distinction separates atomism from associationism, and two senses of “idea” and “impression.” In one sense we have no idea of necessity; in another we do. Despite texts that make impressions of sensation and reflection look homogeneous, their difference is one of nature: impressions of sensation merely give the mind its origin, while impressions of reflection qualify the mind and are the effect of principles within it. The point of view of origin — every idea derives from and represents a prior impression — matters less than usually supposed; it only grants the mind a simple origin and frees ideas from having to represent things or resemble them. What truly matters is reflection, which qualifies the mind as subject. Empiricism’s essence lies not with the atom but with association: it raises not the origin of the mind but the constitution of the subject, understood as the effect of transcending principles rather than a product of genesis. The difficulty is relating these two meanings, and this is the same antinomy of knowledge that defines the problem of the self. The mind is not subject; it is subjected. Once constituted under principles, the mind apprehends itself as a self — yet how a mere collection of ideas can be grasped as a self, how “I” can be said, remains opaque. The self must be both collection and disposition, mind and subject: an incomprehensible synthesis. Hume states two principles he cannot reconcile — that all distinct perceptions are distinct existences, and that the mind never perceives any real connexion among them — while hinting a solution may exist.

Human Nature, Belief, and Sympathy

Human nature is the real object of science, and it presents two modalities: effects of association and effects of passion, yielding the system of understanding and the system of passions and ethics. A parallelism runs between them — belief corresponds to sympathy — and just as association fixes a necessary generality in the mind, the passions supply a constancy that makes practical and moral life, and history itself, meaningful. Without this double movement there would be no human nature; imagination would be mere fancy. The relation of motive to action is of a piece with causality, so history must be construed as a physics of humanity. Everywhere in Hume the only possible theory is a theory of practice: the calculation of probabilities and general rules in understanding, general rules and justice in morality.

These correspondences merely distribute philosophy’s results; we must still ask which domain determines the other. Hume is above all a moralist, political thinker, and historian, and the Treatise’s opening system of understanding raises the problem of reason — but that problem needs a motive. Reason becomes a problem only because a domain escapes it. Hence the decisive sentence: “‘Tis not contrary to Reason to prefer the destruction of the whole world to the scratching of my finger.” Reason does not determine practice.

Sympathy, Partiality, and the Problem of Society

Sympathy is as much an obstacle to large societies as egoism, since both are rooted in partiality rather than a neutral standpoint. The individual always belongs to a clan or community; family, friendship, and neighborliness are sympathy’s natural determinants. Because passion’s essence is partiality, not egoism, sympathy does not transcend particular interest — so noble an affection, instead of fitting men for large societies, is almost as contrary to them as the narrowest selfishness. No one has the same sympathies as another, and this plurality of partialities breeds contradiction and violence. What changes, depending on whether society is viewed from egoism or sympathy, is the very structure of society: egoisms need only be limited, but sympathies must be integrated inside a positive totality. Contractarian theories fail because they define society negatively, as limitations on interests. What nature actually contains are families — social units that exclude one another, partial rather than composed of parts — so contradiction explodes inside nature. Society’s problem is therefore integration, not limitation, implying a positive moral world brought about by invention.

That world is not a moral instinct. It affirms its reality when conversation replaces violence, when property supersedes greed, when the same approbation attaches to the same qualities in China as in England, when sympathy varies without esteem varying. Esteem integrates sympathies and grounds justice; it is not the product of an imaginary voyage, since no real sentiment arises from a known imaginary interest. The moral problem is how to extend sympathy from mutually exclusive real sympathies to a real whole including them. Here nature and morality fail to correspond: the moral world’s reality requires constituting a whole, and this is artificial, not natural. Justice’s rules, universal and inflexible, cannot derive from nature or any natural motive. All elements of morality are naturally given but impotent to constitute a moral world; one can only invent the whole. Justice is a rule, a law of construction, a means — the moral problem is one of schematism, referring natural interests to the political category of the whole. The moral conscience is a political conscience: true morality is politics, and the true moralist is the legislator. Legislation is the great invention; the true inventors are Romulus and Theseus, not Asclepius and Bacchus.

General Rules: Form, Content, and Correction

A system of directed means, a determined whole, is a general rule, with two poles: form and content, conversation and property, customs and stability of possession. To be in society is first to substitute possible conversation for violence, requiring particular sympathies to be artificially exercised beyond their natural limits. The rule determines a stable, common point of view independent of our present situation.

Institutions as Inventive Systems

Institutions are positive, inventive, cultural systems of indirect means — not contracts or obligations. Outside the social lies lack or need; the social is profoundly creative. Convention must not be confused with contract: languages are established by conventions without any promise, and society is a set of conventions founded on utility, not obligations founded on contract. Law is not primary; it presupposes an institution it limits. The legislator is first the one who institutes. The nature–society problem inverts: no longer rights versus law, but needs and institutions. Utility — the relation between institution and need — is fertile, and Hume’s general rule is an institution. Nature and society form an indissoluble complex, yet society cannot be reduced to nature. A drive is satisfied inside an institution: sexuality in marriage, greed in property. But the institution is not explained by the drive; it is an oblique, indirect system of means that satisfies while constraining. The difference between instinct and institution is that an institution exists when the means of satisfaction are not determined by the drive itself. Birds build alike; men build differently. If nature is resemblance, history is difference. Neither private nor public utility explains the institution; reason and custom do — imagination, “the more frivolous properties of our thought and conception.” The javelin planted in an abandoned city’s gates is answered by relations between drive, circumstance, and imagination. The laws of association are both most frivolous and most serious; ethnography returns us to associationism. Imagination is not essentially active but resonates, and the institution is the figure it traces. Because humans lack instincts, they have liberated imagination’s formative power. Nature and culture are one insofar as each satisfies the other, two insofar as the latter is not explained by the former.

Justice’s “schema” and “totality” are justified: the general rule never names owners. Stability of possession divides into immediate possession, occupation, prescription, accession, succession; mobility of persons is regulated by transfer and promise. Three dimensions of the rule emerge: establishment, determination, correction. Yet sympathy, through general rules, gains constancy and uniformity but loses vividness.

Economy, the Table of Rules, and the Power of Imagination

Halévy distinguishes three currents — natural fusion of interests in ethics, artificial identification in politics, mechanical identity in economics — but the economy’s mechanics is no less artificial than legislation. Commerce is an institution presupposing property, yet needs no legislator; property, by contrast, presents a problem of quantity, since goods are scarce, calling for a legislator and state. Money’s quantity does not act on its own; Hume’s economic essays show that effects attributed to money’s quantity depend on other causes.

The Fancy Behind the Rules

The work of the understanding, the text insists, is specious and merely plausible; however much it is disguised by the colouring of the imagination, the fancy still pokes through. Behind the determined content of the rules of property and sovereignty, fancy declares itself ever more clearly in favour of the weaknesses of those rules and of their mutual oppositions. This is why there are trials, and why juridical discussion can be endless. In the case of occupation—the city and the javelin—the dispute proves impossible to decide, because the whole question hangs upon the fancy, which here possesses no precise or determinate standard upon which to give sentence. The historian is left perplexed, and his perplexity links up with the philosopher’s scepticism and completes it. The determination of the rule must therefore be corrected; it must become the object of a second reflection, of a casuistics and a theory of the accidental, filling the gap between the principles of the understanding and the new domain where fancy applies them.

The illusion of the fancy is the reality of culture. From the standpoint of the understanding it is an illusion, yet it asserts itself in a domain where the understanding cannot and should not seek to dissipate it. The necessity of an action, as the understanding conceives it, is neither a quality of the action nor of the agent but of the thinking being who considers it; since we, as agents, cannot feel any necessity in performing it, we inevitably believe ourselves free. The illusion is no less real than the understanding that denounces it: culture is a false experience, but also a true experiment. The understanding may exercise its critique only if we unduly transform the powers of culture into real entities and give real existence to general rules. Otherwise it can do nothing; it lends its principles of association so that the world of culture may be determined, then corrects the extension those principles assume and composes an entire theory of the exception—though the exception itself forms part of culture.

The Passions and the Imagination

At the core of the problem lie the relations between the passions and the imagination, and the determination of these relations constitutes the true originality of the theory of passions. The question is what simple relation between imagination and the passions permits the latter to develop a complex effect inside the former. The principles of the passions, like the modes of association, transcend the mind and fix it; unless nature had given the mind some original qualities, it could never acquire secondary ones, having no foundation for action. But the qualities of the passions do not fix the imagination as the modes of association do. Association gives ideas possible reciprocal relations, while the passions give those relations a direction and a sense, attributing to them a reality, a univocal movement, and hence a first term. The self, for instance, is the object of pride and humility in virtue of a natural and original property conferring a tendency upon the imagination; the impression of the self focuses the mind. If a person is my brother I am his likewise, yet though the relations be reciprocal they have very different effects on the imagination: it passes easily from the farthest to the nearest, from my brother to me, but not from me to my brother. Men are principally concerned with objects not much removed in space or time, and the imagination tends to move from present to future—we advance rather than retard our existence.

Both kinds of affection, relation and passion, situate themselves vis-à-vis each other: association links ideas in the imagination, while the passions give a sense to these relations and so provide the imagination with a tendency. The passions need the association of ideas, and conversely the association presupposes the passions; ideas get associated in virtue of a goal or purpose only the passions can confer. We associate our ideas because we have passions, and there is a mutual implication between them. Hume observes that the association of ideas in knowledge and the association of impressions in the passions very much assist and forward each other. The imagination follows the tendency the passions give it; the relation they suggest, becoming univocal, has been made real—this is the simple effect of the passions on the imagination. But the imagination is also that in which the passions, with their circumstances, reflect themselves through the principles of association, constituting general rules and valorizing things very distant, beyond the tendency of the imagination. This is the complex effect: on one hand the possible becomes real, on the other the real is reflected.

The passage then asks whether we can solve the problem of the self by giving a sense to Hume’s hope, and answers that we can state what the idea of subjectivity is. The subject is not a quality but the qualification of a collection of ideas. To say the imagination is affected by principles amounts to saying that a given collection is qualified as a partial, actual subject. The idea of subjectivity is from then on the reflection of the affection in the imagination and the general rule itself.

Habit, Language, and Illegitimate Belief

Habit by itself can feign or invoke a false experience and produce belief through a repetition not derived from experience, yielding an illegitimate belief or fiction of the imagination. The custom of imagining a dependence has the same effect as the custom of observing it; imagination is not fixed by habit without simultaneously using habit to pass off its own fancies, transcending fixity and going beyond experience. Such illegitimate beliefs, though inevitable, form the set of general, extensive, excessive rules Hume calls nonphilosophical probability—”An Irishman cannot be witty, a Frenchman cannot have solidity.” The understanding cannot count on nature for the immediate determination of the laws of its legitimate exercise; these laws are products of correction and reflection, forming a second series of general rules. Only when the understanding resumes the act of belief and holds it within the limits of past experience are legitimate conditions recognized, forming rules of philosophical probability or the calculus of probabilities. The extensive rules of the passions in the moral world must likewise be corrected, because causality already has a fanciful, extensive use on its own level; the understanding can correct the passions’ extensive rules and question morality only because it must first correct the extension of knowledge itself.

Illegitimate beliefs and nonphilosophical probabilities have two sources: language and the fancy. Language produces belief by substituting spoken repetition for observed repetition and the hearing of a word for the impression of a present object, explaining credulity, education, eloquence, and poetry—the philosopher believing occult words have secret meanings, the liar believing his repeated lies. Words produce a phantom of belief, making a severe critique of language necessary. The fancy confuses essential and accidental, since counterfeit belief depends on the present temper of the person; it interprets accidental circumstances as repetition, as with a vertiginous man whom depth and descent strike so strongly that contrary circumstances of support cannot destroy their influence.

In both understanding and morality the imagination is essentially exceeding, but with a difference: when knowledge is exceeded, no positivity of art remains, only errors and lies, so correction becomes denunciation of error via a calculus of quantities. Extensive rules translate the impossibility of preventive reflection on the principle, since imagination passes naturally from one object to another before reflection. Habit cannot invoke experience without falsifying it, necessitating ulterior reflection as correction, subtraction, second rules, or criterion for the quantified distinction between general and accidental. Philosophical probability maintains belief within the understanding’s limits, ensuring conformity between habit and experience; reasoning must be born of habit obliquely. Experience is partes extra partes; transferring past to future, every past experiment has equal weight, only superior number tipping the balance. Belief is an act of imagination whose idea finds content and vividness in the greatest number of similar parts. Both extensive and corrective rules oppose each other yet share their origin in habit; following general rules, though unphilosophical, corrects all unphilosophical probabilities. Adequation between habit and experience is a scientific result to be obtained; corrective rules recognize causality in nature’s detail, denouncing illegitimate beliefs. Habit has opposite effects on imagination (extension) and judgment (correction).

Religion as the System of General Rules

Religion serves as the example of all general rules: four kinds—extensive and corrective of the passions, extensive and corrective of knowledge. Religion participates in both knowledge and passion, with two poles: polytheism, whose source is the diversity of passions, and theism, whose source is the unity of nature’s spectacle via resemblance and causality. Religion is a system of extensive rules, not itself a passion or instinct; it is a subject of historical study. Polytheistic gods echo the passions; religious feeling confuses accidental with essential, arising from life’s events, contradictions, alternation of happiness and unhappiness, hopes and fears, and strange encounters—defining superstition and idolatry. Idolaters lead artificial lives, making essence of the extraordinary; they are mystics, fanatics, superstitious, for whom moral acts are not enough.

The Given and the Subject

The given is defined by two objective characteristics: the indivisibility of an element and the distribution of elements—atom and structure. Against the charge that Hume’s atomism reduces the whole to a mere sum of parts, the text cites Laporte: the parts taken together are defined by their mode of temporal appearance, an objective and spontaneous mode owed neither to reflection nor construction. Perception, as Hume says of space, consists of parts so situated as to afford notions of distance, contiguity, length, breadth, and thickness.

What, then, is meant by “the subject”? The imagination, formerly a collection, becomes a faculty; the distributed collection becomes a system. The given is taken up again by a movement that transcends it; the mind becomes human nature. The subject invents and believes; it is a synthesis of the mind. Three problems follow: the characteristics of the subject in belief and invention; the principles by which the subject is constituted and which transform the mind; and the stages of the synthesis and the system. The mind, once studied in relation to itself, the sense organs, and time, now becomes—as subject—duration, custom, habit, and anticipation. Anticipation is habit and habit is anticipation: the thrust of the past and the élan toward the future, two aspects of one fundamental dynamism containing most characteristics of the Bergsonian durée. Habit is the constitutive root of the subject; the subject is the synthesis of time—present and past in light of the future.

Invention, Property, and Belief

In invention, each subject transcends its partiality by instituting rules of property enabling agreement among subjects. Hume returns to a simple juridical theory: each man expects to conserve what he possesses. The principle of frustrated anticipation plays the role of the principle of contradiction in the logic of property—a principle of synthetic contradiction. States of possession become titles of property only through the dynamism of habit and anticipation; prescription is the privileged example, where the state of possession is itself time and nothing else. Property produced by time is not real in objects but the offspring of sentiments.

Belief is a vivid idea connected by causal relation to a present impression—an idea felt rather than conceived. The causal relation is that propensity, which custom produces, to pass from an object to the idea of its usual attendant; custom is the great guide of human life. The synthesis posits the past as a rule for the future. Memory alone does not bring about a synthesis of time; it reproduces the structures of the given. Habit belongs to the subject; recollection is the old present, not the past. The past as such is not given but constituted through a synthesis. The causal relation’s duality: experience manifests multiplicity and repetition of similar cases, affecting the span of the past; habit incites movement from one object to a second, organizing time as a perpetual present. The understanding reflects time as a past entity; the imagination reflects time as a determined future filled with anticipation; belief is the relation between these two dimensions.

The Organism and the Passions

Under principles, the organism takes on a dual spontaneity—of relation (animal spirits running into contiguous traces) and of disposition. Impressions of sensation form the mind; impressions of reflection constitute the subject. Some passions (hunger, thirst, sexual desire) have specific bodily dispositions; others (pride, humility, love, hatred, joy, sadness) require external objects. Disposition incites the appearance of an idea of the object corresponding to the passion. Comparing subject and mind leads to the second problem: the principles of human nature—association and the passions—transform the mind into a subject. Relations are external to ideas; this is the fundamental proposition of empiricism, common to James’s pluralism and Russell’s realism. Even relations seemingly dependent on ideas are external, contrary to Kant’s reading of Hume on mathematics.

Two Kinds of Relations

Ideas can be considered collectively or individually—either in the determinable collection where their modes of appearance place them, or in their own characteristics. This gives rise to two kinds of relations, both equally external to the ideas. The first kind—spatial and temporal relations—presents a variable object in relation to the totality in which it is integrated. One might object that the mind already supplies distance and contiguity; granted, but the mind was merely giving a matter, not actual principles. The question becomes: under which influence, external to the mind, do they become a relation? The originality of the relation appears in identity, where the relation is a fiction, and in causality, where it is transcendence. Relations of the second kind—resemblance, proportions, degrees—relate only the characteristics of ideas considered individually. They cannot change without a change in the ideas, yet they are no less external: resemblance between particular ideas does not explain that resemblance is a relation. The relation always presupposes a synthesis that neither the idea nor the mind can account for; it designates that particular circumstance in which we may think proper to compare two ideas. “To think proper” is a normative expression; the problem is to find the norms of this judgment and of subjectivity, leading toward Hume’s voluntarism.

Association and Its Insufficiency

These principles are first those of association: contiguity, resemblance, and causality—now given a different meaning than when they were mere examples of relations. Relations are the effect of the principles of association, which naturalize and give constancy to the mind. Contiguity addresses the senses, causality time, resemblance imagination. Their common point is designating a quality that leads the mind naturally from one idea to another—a quality not of the idea but of human nature; only human nature is qualificatory. The effects of association are complex ideas: relations, substances and modes, general ideas. This intimacy between complex ideas and the subject is presented in language. Yet association is insufficient: it makes relations possible and accounts for immediate relations, but not that A=C or that distance itself is a relation. Hume calls what association explains a natural relation, and what it does not suffice to explain a philosophical relation. Nature is natural, easy going, immediate; in meditations it loses force and vividness. Resemblance, Hume claims, does not always produce a connexion of ideas: when a quality becomes very general and common to many individuals, it prevents the imagination from fixing on any single object.

Most objections against associationism amount to this: the principles explain at best the form of thinking in general, not its particular contents—only the surface of consciousness. Bergson and Freud converge here. Bergson’s famous passage is quoted: between any two randomly chosen ideas there is always resemblance and even contiguity, so discovering a relation between two successive ideas in no way explains why one evokes the other. What we need to discover is how a choice is effected among an infinite number of recollections resembling the present perception, and why only one emerges into consciousness. Hume thought of it first. In his work, association accounts for habits of thought, everyday notions of good sense, and complexes of ideas corresponding to the most general and constant needs common to all.

The Question of Empiricism

Only one kind of objection to a philosopher is worthwhile: one showing that the question raised is not a good question, that it should be raised differently or better. Kant’s criticism of Hume is the example. Philosophical theories involve psychological and sociological factors, but these bear only on the question’s motivation, not its truth. We cannot object that Hume pulverized and atomized the given; the only issue is whether his question is the most rigorous possible. Hume posits the question of the subject thus: the subject is constituted inside the given. Relations are external to ideas; atomism and associationism are merely implications developed from this question.

The classical Kantian definition—that knowledge not only begins with experience but is derived from it—is unsatisfactory: knowledge is not the most important thing for empiricism, but only the means to some practical activity; and experience for Hume lacks the univocal, constitutive aspect we give it. Experience has two senses, in neither of which it is constitutive. In the first—a collection of distinct perceptions—relations are not derived from experience; they are the effect of the principles of human nature, which within experience constitute a subject capable of transcending experience. In the second—various conjunctions of past objects—principles do not come from experience, since experience itself must be understood as a principle. Hume compares reason to a wonderful and intelligible instinct in our souls carrying us along a train of ideas; this instinct arises from past observations, but no ultimate reason can be given why past experience produces such an effect any more than why nature alone should.

Hume never showed interest in problems of genesis or purely psychological problems: relations are not the product of a genesis but the effect of principles. Empiricism is not geneticism; like any other philosophy, it opposes psychologism. Perhaps “the given” is better suited, but the given also has two meanings: the collection of ideas and experience are given; but in this collection the subject which transcends experience and the relations which do not depend on ideas are also given.

The Two Ways the Mind Is Fixed

Principles fix the mind in two very different ways. The principles of association establish natural relations among ideas, forming inside the mind a network like a system of channels: one idea naturally introduces another, and imagination becomes reason, the fancy finding constancy. Yet Hume remarks that were the mind fixed only in this way, there never could have been morality. This is the first argument that morality does not stem from reason. Relation must not be confused with direction: relations establish a to-and-fro movement between ideas, each leading back to the other, and being external to their terms they cannot determine priority or subordination. Action, however, tolerates no such equivocation — it needs a starting point, an origin, something beyond which we need not go. Relations make action eternally possible but cannot account for its actual performance. There is action only through sense or direction, and morality, resembling action, circumvents relations. It is not morally the same to be mean to someone who was good to me and kind to someone who wronged me, even though the relation of contrariety is identical — a radical difference between morality and reason. Causality, to be sure, contains within its synthesis of time a principle of irreversibility, and is privileged among relations; but the real question is which effect interests me and makes me seek its cause. “It can never in the least concern us to know, that such objects are causes, and such others effects, if both the causes and effects be indifferent to us.”

Centers of Fixation and the Principles of the Passions

The mind must therefore be determined in another way. The principles of the passions designate certain impressions, rendering them the ends of our activity — no longer placing bounds around the mind but nailing it down; no longer fixed relations but centers of fixation. Within the mind are impressions called pleasures and pains, but that pleasure is good and pain bad, that we tend toward one and away from the other, does not inhere in them; this is why principles belong here. This is the primary fact beyond which we need not go: if you ask a man why he hates pain, he can never give a reason — it is an ultimate end, never referred to any other object. By making pleasure an end, the principles of the passions give action its principle, making the prospect of pleasure a motive. Thus we find the link between action and relation: the essence of action lies in the nexus between means and end, and to act is to assemble means to realize an end. This nexus includes the causal relation, since means are causes and ends effects, and causality enjoys considerable privilege. A merchant desires the sum total of his accounts only to learn what sum will have the same effects in paying his debt and going to market as all the particular articles together; abstract reasoning never influences action except as it directs judgment concerning causes and effects. But for a cause to be a means, the effect must interest us — the idea of the effect must first be posited as an end, and the subject must tend to achieve it. The relation of means to end is not merely causal but a kind of utility: the useful is defined by its disposition to promote a good, and a cause is a means only for a subject that tends to achieve its effect. These subjective tendencies are the effects of the principles of affectivity, impressions of reflection and the passions. The useful is also a tendency to promote a good, or a quality considered in relation to agreeing circumstances; for human qualities like anger, prudence, audacity, and discretion can be understood generically, as possible universal responses, or differentially, as character traits that may or may not agree with circumstances — and from the latter point of view they are useful or harmful. The best character, were it not too perfect for human nature, is that which is swayed by no temper but alternately employs enterprise and caution as each is useful; Fabius was cautious, Scipio enterprising, and both succeeded because the situation of Roman affairs suited each.

Conclusion: Purposiveness

The concluding pages distinguish two operations of mind: from circumstances and relations the former leads us to discover what is concealed and unknown; after all circumstances and relations are laid before us, the latter makes us feel from the whole a new sentiment of blame or approbation. A circle illustrates the circle: as an object of knowledge it is a relation of parts, the locus of points equidistant from a center; as an object of aesthetic feeling the figure is taken as a whole to which the mind reacts according to its natural constitution. Hume holds that the rules of the understanding are ultimately grounded in the imagination, and likewise the rules of the passions; in both cases the fancy lies at the foundation of a world — the world of culture and the world of distinct and continuous existence. In the schematisms of morality and knowledge there are excessive and corrective rules, but they do not relate to each other in the same way. The excessive rules of knowledge openly contradict the principles of association, and correcting them means denouncing their fiction; a distinct and continuous world is the general residue of this fiction, situated at a level making correction impossible. Excessive moral rules constrain the passions and sketch a wholly fictitious world, but this world conforms to the principles of the passions, frustrating only the limiting character of their effect. Fiction integrates into a whole those passions that excluded each other as particular interests; it establishes, along with the general interest, an adequation of the passions to their principles, of effects taken together to their cause, and an equality between the effect of the principles and the principles themselves. Hence a harmony between fiction and the principles of the passions. The problem of the relation between the principles of human nature and the fancy can be resolved only from the perspective of the relation between principles themselves. In knowledge we must believe in accordance with causality but also in distinct and continuous existence; human nature does not let us choose, though they are contradictory from the standpoint of association — because these principles do not themselves contain the secret of human nature. Association is for the sake of the passions. The principles act separately within the mind yet constitute a subject functioning as a whole; abstract ideas are subjected to the needs of the subject, relations to its ends. “Intentional purposiveness” names the unity of a subject functioning as a whole. To understand associationism as a psychology of knowledge is to lose its meaning; it is the theory of all that is practice, action, morality, and law.

The Subject as Product and Transcendence

The two aspects of the subject are one and the same: the subject is the product of principles within the mind, but also the mind transcending itself. The mind becomes subject by means of its principles, so the subject is at once constituted by the principles and grounded in the fancy. In itself the mind is not subject: it is a given collection of impressions and separate ideas, impressions defined by vividness, ideas as reproductions. Thus in itself the mind has two fundamental characteristics: resonance and vividness — recall the metaphor likening it to a percussion instrument. It becomes subject when its vividness is mobilized so that the impression communicates it to an idea, and when all parts taken together resonate in producing something new. Belief and invention are the two modes of transcendence, related to the mind’s original characteristics; they present themselves as modifications caused by the principles, or as effects of the principles within the mind. We should not ask what principles are, but what they do: they are not entities but functions, defined by their effects, constituting within the given a subject that invents and believes. To believe is to anticipate, communicating to an idea the vividness of the impression attached to it, transcending memory and the senses. For this there must already be relations between ideas — that heat and fire are conjoined — implying experience as a principle, resemblance, and contiguity; and seeing fire at a distance we must believe there is heat, implying habit. The given will never justify relations between its separate parts, nor the transition from one part to another. One may clearly conceive a body falling from the clouds, resembling snow in all other respects, yet having the taste of salt or feeling of fire; no proposition is more intelligible than that all the trees will flourish in December. Not only does the subject anticipate, it conserves itself, reacting by instinct or invention to every part of the given; the given never joins its separate elements into a whole. As we believe and invent, we turn the given itself into a nature. Here Hume’s philosophy reaches its ultimate point: Nature conforms to being, human nature conforms to nature — but in what sense? Inside the given we establish relations and form totalities, but these depend not on the given but on the principles we know; they are purely functional. The functions agree with the hidden powers on which the given depends, though we do not know these powers. This agreement between intentional finality and nature is the book’s final word.

Main Arguments

The book’s central claim is that empiricism, properly understood, is not a philosophy of the senses but a philosophy of the imagination, and that its defining thesis is the externality of relations. Relations are not qualities of ideas; they are not given in the terms they relate. They are the effect of principles of human nature that transcend the mind and act upon it. This is why empiricism is marked by an irreducible dualism between things and relations, atoms and structure, perceptions and their causes, relations and their causes. Atomism is the theory of ideas insofar as relations are external to them; associationism is the theory of relations insofar as they are external to ideas and depend on the principles of human nature. Any theory that derives relation from the nature of things is resolutely nonempiricist.

From this it follows that the mind is not originally a subject. It is a collection of separate, individual items—a theater without a stage, a flux of perceptions. The place is not different from what takes place in it; representation does not occur in a subject. The imagination is not an agent or a faculty that does anything; nothing is done by the imagination, everything is done in it. As the place of ideas, the fancy is a collection of separate items; as the bond of ideas, it moves through the universe engendering fire dragons, winged horses, and monstrous giants. Its depth is delirium, change, and indifference. By itself it is not nature but mere fancy; there is no constancy or uniformity in the ideas we have, nor in how they are connected—only chance makes up this connection. The generality of an idea is not a characteristic of the idea and does not belong to the imagination; it is a role any idea can play under the influence of other principles.

Those other principles are the principles of association—contiguity, resemblance, and causality—and they transcend the imagination while finding in it their term and object. They are qualities unifying ideas, not qualities of ideas themselves. Through belief and causality the subject transcends the given—I believe in what I have neither seen nor touched—but only because the subject is first, inside the mind, the effect of principles transcending and affecting the mind. Before belief, all three principles must organize the given into a system, imposing constancy on the imagination, which otherwise would never be a human nature. Causality alone can make us affirm existence and believe, conferring solidity and objectivity on the idea of the object; the other two principles share the common role of fixing and naturalizing the mind, preparing and accompanying belief. The special ground of empiricism is that nothing in the mind transcends human nature, because human nature in its principles transcends the mind; nothing is ever transcendental. Ideas are connected in the mind—not by the mind.

The effect of association appears in three ways: sometimes an idea takes on a role and becomes capable of representing all ideas with which, through resemblance, it is associated, yielding the general idea; at other times the union of ideas acquires a regularity it did not have, so that nature in a manner points out to everyone those simple ideas most proper to be united into a complex one, yielding substance and mode; finally, one idea can introduce another, yielding relation. In all three cases the result is the mind’s easy passage from one idea to another, so the essence of relations becomes this easy transition. The mind, having become nature, has acquired a tendency.

But the mind cannot be activated by the principles of nature without remaining passive; it only suffers the effects. This is the antinomy at the heart of the problem of the self. When the subject is constituted in the mind under the effect of principles, the mind apprehends itself as a self, for it has been qualified. But if the subject is constituted only inside the collection of ideas, how can that collection be apprehended as a self, how can I say “I,” under the influence of those same principles? We do not understand how to move from dispositions to the self, or from the subject to the self. The self must be both a collection of ideas and a disposition, mind and subject; it is an incomprehensible synthesis, tying together origin and qualification without reconciling them. Hume states two principles he cannot render consistent: that all our distinct perceptions are distinct existences, and that the mind never perceives any real connexion among distinct existences. He adds that a solution may be possible.

The ethical problem is structurally parallel but distinct. The building blocks of morality are naturally given, but they tend to exclude one another. The mind experiences sympathy naturally, but sympathies are partial, limited, narrowly focused. An individual always belongs to a clan or community; family, friendship, and neighborliness are the natural determinants of sympathy. Because the essence of passion or particular interest is partiality rather than egoism, sympathy does not transcend particular interest; the sense of duty follows the common and natural course of our passions. Sympathy is no less opposed to society than egoism, since so noble an affection, instead of fitting men for large societies, is almost as contrary to them as the most narrow selfishness. No one has the same sympathies as another, and the plurality of partialities produces contradiction and violence. This is nature’s course, at which level there is no reasonable human language, since every particular man has a peculiar position with regard to others, making reasonable converse impossible if each considers characters and persons only from his own peculiar point of view.

What changes, depending on whether society is viewed from egoism or sympathy, is the sense or structure of society itself. Egoisms would only need to be limited, but sympathies must be integrated inside a positive totality. Hume’s critique of contractarian theories is that they present an abstract and false image of society, defining it only negatively, as a set of limitations on egoisms and interests, rather than as a positive system of invented endeavors. Hence the importance of the reminder that the natural human being is not egoist. What is found in nature, without exception, are families; the state of nature is always already more than a simple state of nature. The family, independently of all legislation, is explained by the sexual instinct and by sympathy—sympathy between parents and of parents for their offspring. Society begins as a collection of families, but a collection of families is not a family reunion: families are social units that are not added to one another but exclude one another; they are partial rather than made up of parts. The parents of one family are always strangers to other families, so a contradiction explodes inside nature. The problem of society is therefore not limitation but integration—integrating sympathies means making sympathy transcend its contradiction and natural partiality, implying a positive moral world brought about by positive invention.

The moral world is thus not reduced to a moral instinct or to the natural determinations of sympathy. It affirms its reality when contradiction is effectively dissipated, when conversation is possible as an alternative to violence, when property supersedes greed, when the same approbation is given to the same moral qualities in China as in England, and when sympathy varies without a variation in esteem. Esteem is the factor integrating sympathies and the foundation of justice; this uniformity of esteem is not the result of an imaginary voyage transporting us in thought to remote times and lands to constitute possible kin, peers, and relatives, since no real sentiment or passion can arise from a known imaginary interest. The moral and social problem is how to go from real sympathies that exclude one another to a real whole including them—how to extend sympathy.

There is thus a lack of adequation between nature and morality: the reality of the moral world requires constituting a whole, a society, an invariable system, and this reality is not natural but artificial. The rules of justice, by their universality and absolute inflexibility, cannot be derived from nature nor be the direct creation of a natural inclination or motive. All elements of morality are naturally given but impotent by themselves to constitute a moral world; partialities or particular interests cannot be naturally totalized because they are mutually exclusive. One can only invent a whole, since the only invention possible is that of the whole. Justice is not a principle of nature but a rule, a law of construction, organizing within the whole the elements including the principles of nature; justice is a means. The moral problem is the problem of schematism—the act by which natural interests are referred to the political category of the whole or to the totality not given in nature. The moral world is the artificial totality wherein particular ends are integrated and added to one another, or the system of means allowing my particular interest and the other’s interest to be satisfied and realized. Morality may be thought as a whole in relation to parts and as a means in relation to ends. The moral conscience is a political conscience: true morality is politics, just as the true moralist is the legislator. The moral conscience is a determination of the psychological conscience, apprehended exclusively in the aspect of its inventive power. Legislation is a great invention, and the true inventors are not the technologists but the legislators—not Asclepius and Bacchus but Romulus and Theseus.

A system of directed means, a determined whole, is called a rule or norm; Hume calls it a general rule. The rule has two poles: form and content, conversation and property, a system of customs and stability of possession. To be in society is first to substitute possible conversation for violence: the thought of each represents in itself the thought of the others. This requires that particular sympathies be transcended in a certain way and surmount the corresponding partialities or contradictions they generate among people, or that natural sympathy be artificially exercised outside its natural limits. The function of the rule is to determine a stable and common point of view, firm and calm, independent of our present situation.

Institutions, on this account, are positive, inventive, and cultural systems of indirect means, not contracts or obligations. The usual problem is reversed: outside the social lies the negative, lack, or need, while the social is profoundly creative and positive. Convention remains important in Hume’s work, but convention must not be confused with contract. Placing convention at the base of the institution means the system of means is indirect, oblique, invented—cultural. Languages are gradually established by human conventions without any promise. Society is thus a set of conventions founded on utility, not obligations founded on contract. From a social point of view, law is not primary; it presupposes an institution that it limits. The legislator is not first the one who legislates but the one who institutes. The nature–society problem is inverted: no longer rights versus law, but needs and institutions. Utility—the relation between institution and need—is a fertile principle; Hume’s general rule is an institution.

Nature and society form an indissoluble complex, but society cannot be reduced to nature. Humanity is inventive, yet inventions remain inventions. A drive is satisfied inside an institution—specifically social, not governmental, institutions. In marriage sexuality is satisfied; in property, greed. The institution is a designed system of possible satisfaction, a model of actions. But this does not mean the institution is explained by the drive. The institution is a system of means, but oblique and indirect; it satisfies the drive while constraining it. Many forms of marriage or property are possible. The difference between instinct and institution: an institution exists when the means by which a drive is satisfied are not determined by the drive itself or by specific characteristics. Birds of the same species build nests alike; men in different times and places frame houses differently, drawing a like inference from generation and property. If nature is the principle of resemblance and uniformity, history is the scene of differences. The drive is general and does not explain the particular. The establishment of the rule concerning stability of possession, though useful and necessary, can never serve any purpose while it remains in general terms. Utility does not explain the institution: private utility does not, since the institution constrains it; public utility does not, since it presupposes an entire institutional world it cannot create. What explains the institution is reason and custom; elsewhere Hume says imagination, “or the more frivolous properties of our thought and conception.” The example of planting one’s javelin in the gates of an abandoned city is not answered by drives and needs but by relations between drive, circumstance, and imagination. The javelin is the circumstance. The laws of association regulating imagination are both most frivolous and most serious. The drive does not explain the institution; what explains it is the reflection of the drive in the imagination. Ethnography brings us back to associationism; among primitives many prohibitions and prescriptions are explained through a vague association of ideas—and this is not true only for primitives. Associations are vague in being particular and varying according to circumstances. Imagination produces extremely diverse models: when drives are reflected in an imagination submitted to association, institutions are determined by figures traced by drives according to circumstances. Imagination is not essentially active but rings out and resonates. The institution is the figure.

Hume assigns feeling a double function: feeling posits ends and reacts to wholes; these are one—there is feeling when the ends of the drive are also the wholes to which sensibility reacts. Wholes are formed when the drive and its ends are reflected in the mind. Because human beings lack instincts, they have liberated the formative power of imagination and placed drives in immediate relation to it. Satisfaction of human drives relates not to the drive itself but to the reflective drive. This is the meaning of the institution in its difference from instinct. Nature and culture, drive and institution, are one insofar as one is satisfied by the other, but two insofar as the latter is not explained by the former.

Regarding justice, “schema” and “totality” are justified, since the general rule never indicates particular persons and does not name owners. Justice never regards fitness or unfitness of objects to particular persons; the general rule that possession must be stable is applied by other general rules extending to the whole society, inflexible by spite or favour. The rule is established by interest and utility and determined by imagination; it does not determine real people but is determined and modified in statements reflecting situations and possible circumstances. Stability of possession is divided between diverse rights: immediate possession, occupation, prescription, accession, succession. The lack of adequation between real persons and possible situations may itself be considered a circumstance or situation. The mobility of persons is regulated by agreed-upon transfer when the object is present or particular, and by the promise when the object is absent or general. Three simultaneous dimensions of the general rule: its establishment, its determination, its correction.

A difficulty remains: sympathy, through general rules, has won constancy, distance, and uniformity of true moral judgment but has lost in vividness what it gained in extension. The passage then asks whether we are capable of solving the problem of the self by giving a sense to Hume’s hope, and answers that we are capable of stating what the idea of subjectivity is. The subject is not a quality but rather the qualification of a collection of ideas. To say the imagination is affected by principles amounts to saying that a given collection is qualified as a partial, actual subject. The idea of subjectivity is from then on the reflection of the affection in the imagination and the general rule itself.

Key Concepts

The externality of relations. The book’s master concept. All relations—not only matters of fact but relations among ideas—are external to their terms. We start with atomic parts, but these have transitions, passages, tendencies circulating from one to another, giving rise to habits. This is what Deleuze calls Hume’s first great logic of relations, and it constitutes a multifarious world of experience based on the principle of the exteriority of relations.

Belief. Hume established the concept of belief and put it in the place of knowledge; he laicized belief, turning knowledge into a legitimate belief, and asked about the conditions legitimating belief, sketching a theory of probabilities. If the act of thinking is belief, thought has fewer reasons to defend itself against error than against illusion; illegitimate beliefs surround thought like a cloud of illusions. Here Hume anticipates Kant, and an art with rules is required to distinguish legitimate beliefs from accompanying illusions.

Association. The association of ideas is a practice of cultural and conventional formations—conventional instead of contractual—rather than a theory of the human mind; association exists for the sake of law, political economy, aesthetics, and so on. Deleuze illustrates with the question of whether shooting an arrow at a site suffices to become its owner, or whether one must touch the spot with one’s own hand—a question about the correct association between a person and a thing for ownership.

The principles of human nature. Association (contiguity, resemblance, causality) and the principles of passion. These principles transcend the mind and act upon it; they select impressions of sensation, designate them as candidates for association, and constitute impressions of reflection. In cognition they designate impressions and organize the given into a system, bringing constancy to the mind and naturalizing it; they form habit, establish belief, and constitute the subject as anticipating.

Impressions of sensation and impressions of reflection. Despite texts rendering them homogeneous, the difference between them is a difference of nature. Impressions of sensation are only the origin of the mind; impressions of reflection are the qualification of the mind and the effect of principles in it. The point of view of origin—every idea derives from a preexisting impression and represents it—does not have the importance attributed to it; it merely gives the mind a simple origin and frees ideas from representing things and from understanding resemblance. The real importance lies with impressions of reflection, which qualify the mind as subject.

The general rule. A system of directed means, a determined whole; Hume calls it a general rule. The rule has two poles: form and content, conversation and property, a system of customs and stability of possession. The real definition of a general rule is that it is a passion of the imagination. There are determining rules and corrective rules; the former are more extensive than reflective, the latter more reflective than extensive because they correct the extension of determining rules, presenting themselves as general rules concerned with the accidental and exceptional. The exception becomes, through habit and imagination, the object of experience and knowledge—casuistics.

The rule of taste. How does feeling overcome inconstancy and become aesthetic judgment? Passions of the imagination require neither efficiency nor adaptation of their object; they are moved by degrees of liveliness inferior to belief and independent of real existence. Virtue in rags is still virtue; a deserted fertile soil leads us to think of possible inhabitants’ happiness. Taste is a feeling of the imagination, not the heart. What grounds a rule is the distinction between power and its exercise, which only the imagination can bring about, reflecting passions and object, separating them from actuality and recuperating them as possible. Aesthetics envisages things under the category of power or possibility.

The cracked I and chaosmos. Portmanteau concepts that circulate through the series of Deleuze’s thought and enable the inclusive, disjunctive affirmation of all of them. Chaosmos—the becoming-world—posits the constitution of the subject as a task, and guarantees that the constituted subject emerges not as a substantive hypokeimenon but as an always already cracked I.

The self as incomprehensible synthesis. The self must be both a collection of ideas and a disposition, mind and subject. It ties together origin and qualification without reconciling them. Hume states two principles he cannot render consistent: that all our distinct perceptions are distinct existences, and that the mind never perceives any real connexion among distinct existences.