Navegación – Mapa del sitio

Reason and Agency in Kant and Fichte

Michael Vater


This paper explores the question of the unity of Transcendental Idealism at the end of Eighteenth Century German philosophy, given that it circulated in different versions, Kant’s Critique [of humans’ rational powers] and Fichte’ System of Science [Wissenschaftslehre]. Both thinkers take the transcendental turn. They base conceptual investigations not on facts or empirical evidence, but on the possibility of a situation; they are idealists since they look inward to the spontaneity of the agent/knower for explanation, not the environment, stimulus, or sensory given. Reason can fathom only what it has constructed.

Inicio de página

Texto completo

  • 1 I use agency to translate Fichte’s Agilität, although the term is closer to Handeln, which suggests (...)

1This paper explores the question of the unity of Transcendental Idealism at the end of Eighteenth Century German philosophy, given that it circulated in different versions, Kant’s Critique [of humans’ rational powers] and Fichte’ System of Science [Wissenschaftslehre]. Kant invoked the normative character of reason to address two issues in the manifest image of humanity: the presumed objectivity of cognition and the universality of behavioral rules. Fichte invokes the spontaneity of agency to sculpt a global account of human existence, one that makes activity fundamental and cognition derivative.1 It is difficult to imagine the uproar this caused to the pre-Darwinian mindset of Fichte’s readers and students; recall how Goethe has Faust read the Bible’s creation story as: “In the beginning was the deed.” Yet Fichte’s insurgent activism is but a potentiation of Kant’s Copernican Revolution—solving the problem of the objectivity of empirical knowing through the intellectual powers of the subject—or ‘Copernicus2’. Both thinkers take the transcendental turn. They base conceptual investigations not on facts or empirical evidence, but on the possibility of a situation; they are idealists since they look inward to the spontaneity of the agent/knower for explanation, not the environment, stimulus, or sensory given. Reason can fathom only what it has constructed.

2The differences between Kant’s and Fichte’s system of critical idealism I consider fall into those explicitly noted by Fichte and those that deal with the form of a philosophical system or its method. Fichte’s major criticisms of Kant are:

  • Kant critiqued reason’s cognitive and legislative powers, using items borrowed from experience, whereas he (Fichte) constructs a system of human capacities based on the internal evidence of freedom or agency (FTP 80; WLnm-K, GA IV/3, 325).

  • Though Kant says that freedom is the center of Critique and that practical reason is primary, he only explains cognition—or representation, as in Reinhold’s Element Philosophy (FTP 162-63; WLnm-K, GA IV/3, 362-63).

  • Retaining a substance ontology while advancing a phenomenalistic theory of cognition commits Kant to dubious entities like things in themselves, and forces him to separate the sensible from the supersensible (FTP 271-72; Wlnm-K, GA IV/3, 430).

3The context of these criticisms is as important as their substance: Fichte voices them as a confederate of Kant who wishes to correct his flaws and set a higher standard of consistency for transcendental idealism. Accordingly, he tries to fashion the fragment into a system, make reason consistently practical (or practical-theoretical), and banish the traces of Kant’s commitment to the ‘subjects’ and ‘objects’ of the manifest image—making the empirical world of interacting subjects and objects the end-point of a transcendental construction, not its starting point. Fichte’s systematization is generally faithful to Kant’s views on cognition and morality, but as in Spinoza’s reformulation of Cartesian dualism, the clarity and logical rigor that formalization imports produces something at least optically different. Kant’s informal rhetoric imports the manifest image, with its quotidian persons and things, into theory even when he wants to explore impersonal cognitive mechanisms or pre-personal sorts of intelligence and agency.

4I see two other points of differences between Kant and Fichte which follow from their decisions about the starting- and end-points of philosophical theory.

  • Kant’s approach to philosophy is to offer a synoptic description from the third-person stance; it tacitly depends on a conventional view of things and persons. Fichte’s construction is genetic, not argumentative; it offers a temporal narrative, really a performance--a product of poesy, not of prose. In Fichte’s view, Kant “considers the I only as ordering the manifold, and not as producing the same” (FTP 261-62; WLnm-K, GA IV/3, 424).

  • Both thinkers claim to use ‘synthetic method’. Kant expounds the synthetic nature of consciousness in a stepwise or combinatory manner, so his synthetic method is ascending. For Fichte, synthesis is present from the start in its most radical form, antinomy or the direct joining of opposites. Conceptual analysis unravels the contradiction, so for him synthetic method is descending (FTP 394-96; WLnm-H, GA IV/2, 232-233; WLnm- K GA IV/3, 486-87).

Transcendental Idealism

  • 2 To Jacob Sigismund Beck. December 4, 1772” in: Kant, I., Correspondence, tr. & ed. By A. Zweig. Ca (...)

5Fichte acknowledged Kant’s lead in pointing to the resolution of philosophy’s chief problem: the justification of the claim of empirical cognition to objectivity, or in Fichte’s terms, establishing the “necessity of our representations.” In the struggle to elaborate his position, Kant encountered persistent misunderstanding of his two key arguments, the ideality of space and time, the formal frameworks of intuition, and the mind-furnished character of the concepts and judgments that supported meaningful predications about the structure of appearance. Goaded by critics’ misapprehension of the first Critique as a metaphysical idealism akin to Berkeley’s,2 Kant rewrote sections of the work to show that only through the supposition of integrative work in the knowing subject could sensations be combined into perceptions of properties, and properties be combined into judgments about objects. Mind furnishes the formal glue that holds together the data furnished by the world--taking ‘mind’ and ‘world’ in a weak sense where they indicate only the poles of spontaneity and passivity inside a representation (or figurative mental event). Kant used reworked portions of the Analytic to make clear that synthetic judgments require only potential integration of their contents under a logical subject (an empty “I think”) and that such judgments have purchase only when applied to experience. He used the reworked Preface to make the subtler point that his idealism is purely hypothetical and heuristic: mind can understand only what conforms to its logical intent, or what it has ‘constructed’. Just as Ptolemaic and Copernican astronomies rely on similar observations, but reach less or more accurate predictions depending on whether they assume the earth or the sun as the center of the planetary system, so commonsense philosophy and transcendental idealism reach more or less reliable judgments about empirical reality depending on whether they call on logic or on ‘self-standing’ facts to explain the ‘conformity’ of objectivity and subjectivity that truth involves. Although everyone starts by supposing subjects depend on objects for truth, no one can find a path from a supposed ‘outer’ to ‘inner’ to guarantee the transfer of information, and philosophers are beset by doubts about sensory inputs and concept formation—until a Copernican reversal of frameworks shows that working knowledge is a tango of mind and world, not an immaculate perception delivered by Instagram (KrV, B xvi-xxii).

6Kant did not set out to prove our knowledge is certain, for no idea is clear enough, and no impression vivid enough to breach the wall between ‘inside’ and ‘outside.’ Subject and object form a categorial dyad, everywhere together because everywhere apart; no single point permits a leap from one order to the other, since they occur together or not at all. What is needed to quiet the skeptic and settle the debate between empiricists and rationalists is an argument that knowledge is reliable or affords accurate predictions--one that shows how reports about what happens are not just habitually believed, but operatively true. Understanding authorizes justified beliefs about experience or what goes on in the world. It also authorizes, thought Kant, our moral convictions—the conviction we that should have moral convictions, and even identical convictions about what should and should not be done, and what can or cannot be permitted.

7So, for Kant reason’s business is to warrant what is true and what is obligatory, each of which compels our attention because it involves lawfulness (and so universality and necessity). Kant’s Critique comes down to two theses, held to be constitutive of reason and not at all discretionary:

    • 3 Kant also points to reason’s economic function in discussing the subjective principle that counsels (...)

    Reason, in its logical or ordering function, compels our belief in the objectivity of experience because it has assembled the elements of cognition (intuitions and concepts) according to the sole conceivable lawful patterns.3

  1. Because it is authoritative, reason does not inspect experience for patterns of useful or beneficent conduct, but merely frames a formal standard for whatever patterns might be suggested. It legitimates only such laws that can command everyone’s obedience.

8In the first case, reason’s power is logical, the ability to combine or synthesize representations a priori or on its own authority; its most powerful concepts are that of causality and causal interaction among substances. In the second, its power to authorize conduct is freedom or causality from pure concepts, in a nonsensible order (CPrR 183-84). The reality of freedom is not explicable, but co-extensive with one’s perception of the moral law or with one’s sense of citizenship in a moral order. A third Critique explores cases where reason seems to advise an integration of knowledge or a harmonization of interests wider than what it would compel in the name of truth or duty, and this power is reflection.

9As a self-proclaimed discoverer of new lands, Kant had to elaborate the foundations of Criticism gradually, a task that consumed more than twenty years. It was not the task of amplifying, emending, or extending transcendental philosophy that made the task difficult, but its central discovery: while intellect (reason as understanding) can comprehend only what can fit into the explanatory architecture of causal explanation, freedom (causality in another order) is completely incomprehensible to it. And while the moral point of view urges that duty and responsibility presume freedom--the ability to assess or conform conduct to some concept of conduct rather than brute response to sensory stimulus--there is no way of demonstrating that any rational being acts in this way. Reason’s chief stumbling block in the theoretical domain, the antinomy that advances contradictory valid arguments for both freedom and determinism, becomes its signal achievement in the practical. Despite the susceptibility of every agent’s conduct to analysis in terms of motivation or self-interest, practical reason addresses commands to an agent whose deed is supposed to result in something other than what can appear in the sensible world. Therefore, if reason is to speak in one voice, it must yoke together both logic and freedom (CPrR 215-216).

Kant on the Root of Cognition

  • 4 4Fichte is the beneficiary of Karl Reinhold’s definition of presentation as both the unit of cognit (...)

10Kant and Fichte assume that human cognition and the moral regulation of behavior both involve mind’s activity or the effects of it, not just a passive assimilation of content or rules furnished from elsewhere. Kant tends to subsume action to cognition, figuring both the works of nature and the acts of free beings as specimens of causality, while Fichte offers a more provocative construction of cognition based on original or unconstrained activity. Kant starts with a basic, but ready-at-hand item (the presentation) where activity and product are jointly present, but only as past static factors rather than dynamic tendencies, sidestepping the more basic question of how a mental state can depict or stand for something other than what it is.4 By contrast, Fichte tries to describe how a mental state comes into being with its polar features of spontaneity and passivity and the logical capacity to merge them into a determinate state. This happens in a move that is both first act and first acquisition of content: an inchoate agent doubles its features—what it is and what it is not—through an unconstrained (or free) choice and makes the logically first choice of deciding to be something rather than nothing, thus moving from indeterminacy to a determinate state. All further content (and knowing of content) evolves through repetition of this basic choice or act. In Fichte’s language, Kant’s account starts with fact, his with act.

11It will take me effort to flesh out the comparison. As I view it, the two thinkers defend the same formal structure (synthetic a priori judgment) at the basis of experiential cognition, but they do it in different philosophical styles. Kant offers an essentialist or account inscribed in a synoptic table of rational capacities. Fichte offers a functionalist account, embedded in a Bildungsroman narrative which pretends to be the autobiography of reason. Let us start with Kant.

12Like most contemporary writers, Kant uses the philosophical treatise to convey his views. It combines patches of exposition prefaced by definitions with specific arguments; the ensemble functions as one argument. Kant insists that one can assess this philosophy only by looking to the quality or explanation (or coherence) it achieves, for its hypothesis that mental processes are ingredient in the constitution of objects as objective runs contrary to the complex of beliefs about persons and things we call the manifest image. A transcendental supposition cannot be supported by specific points of evidence. The Transcendental Aesthetic and Logic describe the role of sensation and concept in cognition from a formal point of view. It is only in view of their relational properties that these items can be combined or interrelated in judgment; any specific sensation will have idiosyncratic content, likewise any specific empirical concept. If cognition rests on combination—either of sensations in perceptions or various perceptions and concepts in predication--it must be the mind-furnished elements (relational frameworks) that permit their sorting, classifying, and identifying. The objects of experience present themselves as finished and self-subsisting only through mind-furnished forms of intuition and conceptual judgment. There is no other way of explaining the fit of our senses and our logic to the world experience presents.

13The Analytic of the Critique is generally thought to have four stages, with treatments of:

  • the relational continua (space, time) that frame sensory content and allow sensations to be located,

  • the properties of concepts that make perception possible and patterns of judgment that make predication possible,

  • a first pass at unifying intuitions and concepts using schemata or concepts patterned to the durations of various parts of the perceptual process,

  • and a retrospective enunciation of principles for empirical cognition.

14The ensemble makes the argument that we understand a world of objects only by viewing the work of intuition, imagination, and judgment as informing by the synthetic formal (or framing) activity of mind. This is the tool-kit that reason brings to experience, with whose help a reliable world of objective knowledge can be assembled. Since reason is both architect and contractor on this project, the resulting artifice is not only objective but normative. For Kant, knowledge is not aggregation of fact, or a history of trial and error attempts to find regularity and significance in life (myth, ritual, and magic); it has necessity or normative traction because it is founded on patterns of thinking that could not be thought otherwise. I have set out the Analytic’s argument as a list not to summarize its contents, but to bring Kant’s narrative strategy to light. Although cognition is through and through synthetic—structured by the unifying and concretizing work of transcendental imagination—Kant’s exposition is a linear narrative, like the conventional examples of viewing a house or a boat coming down river which are offered in the Analogies.

  • 5 5The second text speaks of the five-fold synthesis as “expanding this original synthesis ‘from the (...)

15Fichte has identical beliefs about cognition, but starts from the existing synthesis of deed (or agility) and fact in consciousness, analyzes that synthesis into its opposing factors, and then further widens the contrast into an absolute antithesis or antinomy (FTP 371, K 477;FTP 427-28, K 500).5 These expository strategies look different, but each does the only thing that can be done to explain something: display the totality of conditions that allow it to arise as a unique item--Kant’s as the anatomist of human cognition, Fichte’s as the field biologist.

16But discussing the levels of synthesis or Kant employs says little about the content that is integrated. Each item that is sensed, perceived, conceptualized or judged must have properties which make it amenable to such treatment: logical and epistemic characters that (a) make this intuition or concept a case of knowing, and (b) permit its integration into a more complicated instance form of cognition. All content must be alike; it must have structure, and commensurate structure, to fit into the imaginative (or judgmental) calculus.

  • 6 6Throughout four decades of lecturing on logic, Kant consistently described the Vorstellung as the (...)

17For Kant, the generic unit of cognition is the Vorstellung, a placing of something before the mind or a mental presentation.6 At the generic level, ‘mind’ means only a joining of something known to a knowing. The unit of cognition is already complex, composed of formal and material elements; the formal element is credited to spontaneity (forming, unification), the material element to passivity (raw content, or other presentations taken as content). Even at the basic level, a presentation is a self-enclosed and self-generated little world, with subjective and objective poles or directions, themselves not entities but shadows of the constitutive features of spontaneity and passivity. There is no ultimate or atomic presentation; whatever is at hand as a presentation is a presentation of presentations, already figured or unified into a cellular or self-sufficient cognition-entity. Any more structured cognition will be a dense constellation of representations, and as the description of cognition shifts from sensations to perceptions and from perceptions to concepts and judgments, the more spontaneity comes to the fore and the more form predominates over matter or content.

18Current research in artificial intelligence and neuroscience makes similar assumptions, viewing intelligence as a function of iterated simple cognitive competencies, not a single power of consciousness. To read Kant in this light helps banish the whiff of substantialism or anthropomorphism that lingers around his account of cognition. If Critique takes away the possibility of knowing supersensible objects, it equally removes the possibility of knowing sensible subjects or objects in themselves, outside the continuum of experience and the multiple voices/visions that constitute it. On any idealistic account, Critical or metaphysical, we can with certainly know only how we know, not what.

19Kant is generally satisfied to treat a presentation as an entity, to call it an ‘appearance’ but figure it as a thing, and so to view the relation between it and whatever grounds it as a real relation, not one merely ideal or embedded in the project of cognition. Because he does not plumb the structure of presentation, Kant tends to treat the presentation as a ready-made thing and falls into the idea that there must be a thing behind appearance, a thought-thing behind an appearing thing--a transcendental object behind appearances. In the manifest image, it’s a useful mistake to think that what appears in the mirror is an image of me, but complete folly to think there is another person behind the glass.

  • 7 7Kant deemed Reinhold’s reconstruction of Critical Idealism murky, “weighted down with obscure abst (...)
  • 8 8 In the Logic lectures, Kant leaves the matter of the presentation undefined (with formal features (...)

20Reinhold attempted to think his way to the bottom of Kant’s epistemology in his New Essay on the Human Capacity of Representation. Inspecting the presentation’s elements, he identified two chief features, one whereby it offers some contents (a given that is also a manifold)) and another that does the depicting (a spontaneous unification of that manifold).7 These features are functions of the free-standing presentation, quasi-logical or -mathematical functions, not linked to a thing-in-itself or subject-in-itself. This cognitive atom, like the Leibnizian monad, has the powers of signification and combination—signification (depiction) arising from the passive, material element, combination arising from the spontaneous, formal element.8 Although Reinhold tries to explain the increase in logical power as one ascends from presentation to intuition, to concept, to judgment (all iterations of the same structure), one can see another route: to isolate the mark of spontaneity and figure it not as just a component feature of a presentation, but as an agency or ‘mover’ that constructs the presentation.

  • 9 9 See “Declaration concerning Fichte’s Wissenschaftslehre. August 7, 1799,” Correspondence, 559-60.
  • 10 See Kant’s brief distillation of the whole cognitive process—from representation, to intuition and (...)

21Fichte pursues such an agent-centered approach, particularly in the second iteration of his Jena transcendental philosophy. Kant dismissed his first attempt as “mere logic,” perhaps misled, as many readers were, by the Grundlage’s initial presentation of consciousness’s synthesis of subjectivity and objectivity as ‘principles’.9 Spontaneity was but one of the marks that Kant ascribed to the presentation, but it is predicated of the whole of the monadic thought-entity, taken as finished and self-sufficient. In the Transcendental Logic, the conceptual facet of cognition is put in terms of logical properties and processes, not of agents and products. Imagination, for instance, is a global feature of cognitive processing, unification of content or distillation of a representative instance from an unspecified multitude.10 In the A-Deduction, it is parsed into three episodes of functions: perceptual apprehension or synopsis content, reproduction of temporally past content in the current flow of experience, and recognition of the concept in concept-formation. It is not altogether clear whether there is a single feature in the three that makes them ‘imagination,’ or whether it is only family resemblance that unites these ways of taking what is multiple as one or distilling many cases into a representative instance. It is not even clear that imagination is a specific mental function, or where it finds its place in Kant’s general definition of cognition as a unification of intuition and concept. Perhaps it is the ‘image’ in the English translation’s rendering of Einbildungskraft that confuses English-speakers; Einbildung seems to directly indicate logical unification—or the ability to simultaneously be on both sides of a significant difference.

22I think the fact that the rewrite of the B-Deduction simply omits reference of imagination and speaks instead of already-unified presentations being accompanied by (or able to be accompanied by) an abstract mark of judgment, the representation “I think” is telling. Kant’s transcendental story need not be anthropomorphic, though it is indeed mentalistic. His use of the phrase: “the I, he, or it in me that thinks” (KrV A364/B404) says plainly enough he is talking about logical functions—and that the minimal language of mental functions is sufficient for transcendental thinking. Impersonal functionalist language seems appropriate for mere thinking: it matters not a whit whether you or I or a logic bot parses a tautology or produces a formal proof of a theorem of the predicate calculus. The logical subject is a pretty boring fellow. The display of objects for a subject is just the pinnacle of a stack of logical operations, the projection of a logical, not a Cartesian, theater. It will take a complete account of consciousness such as Fichte crafts to see whether the logical spectator can be attached to agent who has learned responsible behavior. The question may have more than anthropological significance. The normative or law-bound procedures of cognition may carry one form of necessity, and the moral necessity to performing the duty that one can intuits in one’s situation another. Kant presumed that reason is one, its legislation uniform, its decrees equally compelling in logic and life. Fichte set out to prove it.

Fichte’s Account of Agency

23One can appreciate Fichte’s desire to continue Kant’s transcendental philosophy and how he wishes to alter it by considering two criticisms he advances early in the nova method lectures:

  • Kant’s Critique probed the fitness of finite reason, established its competence in empirical cognition and moral legislation, forbade its extension to the supersensible and stipulated that philosophy’s task was to systematically display of reason’s contents. But Critique could only be prologue to System, whose real task is to present reason’s evolution in its own voice (FTP 80, WLnm-K, GA IV/3, 325).

  • Kant’s asserted the primacy of practical reason, or the strength of reason’s interest in resolving the antinomy of freedom and determinism in favor of freedom (FTP 162-63, WLnm-K, GA IV/3, 371). But in doing so, he only added another sort of reason to theoretical reason, one that took the coin of moral necessity as equal in value to truth or epistemic necessity.

24To overcome these deficiencies, Fichte sees he must abandon the philosopher’s armchair and find a stage where reason can display its basic function as an actor or agent and still serve as its own commentator. Spinoza’s and Reinhold’s efforts to formalize philosophical argument inspire his attention to detail and interconnection, but Fichte’s literary friends who played with the walls convention had erected between author, artistic depiction, and audience may also have nurtured the idea of producing reason’s autobiography. Only as an agent can reason have a career or appear as actor in a narrative.

25It is beyond the scope of this paper to discuss Fichte’s entire construction. I will cover the basic story of agency’s evolution into subject-objectivity and mark the points where Fichte’s view agency undergirds objectivity seems to coincide with Kant’s transcendental account of cognition. If that can be made clear, it will be clear that Kant spared himself effort in speaking from the armchair and introducing just a few transcendental somersaults into the manifest ontology, while Fichte takes on a huge project in promising a fully scientific account of the functions of finite reason (FTP 83-84). I will discuss only three episodes: (1) agency’s first gig, (2) how it paints its habitat, or acquires an array of material qualities, and (3) its split into subject and object.

The First Move

  • 11 Fichte, J. G., The System of Ethics According to the Principles of the Wissenschaftslehre, ed. & tr (...)

26Fichte rarely speaks of reason as the agent in Foundations of Transcendental Philosophy, but typically of the I’s self-positing, i.e., of action, agent, and agency. Freedom effects its self-substantiation; it is not a property or qualifier of a substance (FTP 83-85, WLnm-K, GA IV/3, 326-328). I use the term agency to translate Agilität, employed by Fichte to indicate (i) the mover or agent, (ii) its power to change state, from indeterminacy to determinacy, and (iii) to reflect upon or react to that change. At the most basic level, the germ of freedom or rational life in principio is a movement that initiates activity, but is reflexive or self-evaluative, and so is responsible for what is done, not just as efficient cause but as chooser. From its bare first moves, Fichte’s account of reason integrates freedom and cognition: the agent’s first act is to qualify itself, i.e., move from indeterminacy or lacking all quality to determinacy or having the bare quality of having some quality. In the germinal world of agency, the first item of information—the being-something (along with the awareness of that being-something) that will evolve into a cognition of a complexly qualified something--is the first choice. One might think of Fichte’s first item as Hamlet the transistor, pondering “to be (charged) or not to be, that is the question.” Others have more artfully spoken of this situation, at once very simple and very complicated, as “the original duplicity of intelligence and will”.11 A recent critic characterizes Fichte’s ethics as more existentialist than Kantian, but I think Fichte knots together “whether to be?” and “how to be?” from the first twitch of the string.

27The difficulty of Fichte’s first move is not only that it has theoretical and practical directions, it is enunciated in a narrative so thin as to be barely intelligible, but so basic that it undergirds all further predication of properties and powers, or qualities and action: there is an initial movement of intuition from lacking any character to having one--from indeterminacy to determinacy--and a simultaneous awareness of that process --a slide from that lively state of intuition to its congealed result, concept (FTP 139, WLnm-K, GA IV/3, 359-60). The fixed character or concept that results from this first movement does not exhaust agency, but it constrains further movements and accretion of characters. It’s as if the logical and moral germ of the finite rational being were a self-assembling set of toy blocks. Just as every block can be joined to every other because they are identical in structure, every move or ‘action-block’ that agency deploys will be like the first, a concept-property. Anything it does and any further quality it takes on will have the same logico-practical character: it becomes x from being non-x because nothing prevents it, but once it has become x, it is a fixed character or concept. All marks, character, qualities, properties and psychological/moral items such as striving, feelings, desires, deliberations, choices etc. are compounded of such self-qualifying choices. Not until the system reaches complicated levels like personal choice in social or intersubjective contexts does the self-constricting finitude of original agency become apparent. Prospectively, choice seems open or unconstrained, but retrospectively one must lie in the bed one has made for oneself.

Feelings and Qualities

28Kant was horrified by initial reviews of the Critique of Pure Reason that imputed a sort of ontological idealism to him, with mind in some way responsible not just for logical form, but sensory qualities as well. Idealists are aware that logic is a matter of relations, not of predicates or functions that are so related, and Kant viewed the framework properties of cognition as extensions of logical relations. The a priori appears to be a self-sufficient domain. While he could remedy the reviewer’s misunderstanding by underscoring the hypothetical character of the transcendental postulate, he still was left with no account of the ‘thisness’ of sensory qualities other than the logical place-holder of a variable, with the predicate or function specified from elsewhere.

29Fichte was not content to leave idealism a half-theory and so threw himself into the task of deducing qualities, material and formal, on the basis of agency or determination via choice. He found two ways of characterizing qualities or properties that were of philosophical interests. (1) Kant had suggested the first feature-- that perception simultaneously indicates an object’s quality and one’s awareness of it—by calling the process an intuition (Anschauung), a term that connotes the subject’s activity, not a passive reception. Two centuries of attempts by physics to discriminate primary and secondary qualities formed a consensus that perception involves changes in both the perceiver’s body and mind, or that a ‘quality’ involved a registering a change of state in both. Fichte underlines the subjective or active element in intuition by using the term feeling (Gefühl) to indicate what is sensed as well as the change in awareness by which it is registered. Initially, this seems an odd move; as we usually talk in the manifest image, perception mostly captures properties said to be external, so we pay little attention to our awareness of it. This certainly is the case for visual qualities like brightness or color, and for acoustic, gustatory and olfactory qualities which we project onto the object; only in touch, where what is perceived (pressure on the skin or temperature) is displayed in or on the body, do we call those qualities ‘feelings’. Or if what we perceived is a quality of mind or indicates a change in mental state, that too is deemed feeling. No matter how we usually speak, a philosophical (wissenschaftlich) account must indicate that what is given in the perception is given only in the giving-and-taking of awareness: all sensible qualities are things that affect us, for each one involves a change in mental awareness. Even the properties we impute to things are (mediated by physiological and psychological factors) felt items, modulations in our awareness (FTP 176-77, WLnm-K, GA IV/3, 377).

30(2) We do not have feelings one by one, or additively, but a perceptual field or drive is an orientation to a system of feelings or logico-perceptual options. One can see three (according to Goethe), five (Buddhists), six, or seven (Newton) basic colors or roughly 17 billion in an 8-coded web display. One can hear high or low sounds, for short or long intervals. One can discriminate salty, sweet, sour, bitter and umami on the tongue; hot, cold, soft, or hard on the skin. The nose, I fear, is not so logical. On this physiological-psychological basis, Fichte asserts that having a quality is a case of choosing a quality, or making a determinate choice from an array of determinate possibilities. Bearing a property is an upshot of an evaluation and choice; being aware of a property is a function of paying attention to how one is affected (FTP 178-180, WLnm-K, GA IV/3, 378-79). Even before consciousness arrives on the scene, is embodied and is placed in space and time through causal efficacy, there is an interactive aspect to perception and all forms of higher-order cognition. What we see logically depend on what we don’t see, or have in some sense decided not to see. The focal or aleatory character of cognition arises from the logic of characters (and predicates): to see a is not to see non-a (FTP 134, WLnm-K, GA IV/3, 357). It is only at a higher and voluntary level of consciousness where Spinoza’s ‘all determination is negation’ becomes confining or problematic.

From Subject-Objectivity to Subjects and Objects

31As far back as the 1770 Dissertation, Kant offered several piecemeal arguments for the ideal nature of space and time, claiming they are the formal features of the sensible world, independent of the categories of quantity. If abstracted as continua, they provide the content for formal operations (judgments) of mathematics and the constructions of geometry. Neither here nor in the Transcendental Aesthetic of the Critique does Kant offer a single sustained argument comparable to that of the Logic. Treating space and time as ideal allows him to put their purely relational structures in the foreground; they are continua of sorts, but, viewed in abstraction from quantitative concepts, vague ones. In the manifest image, we commonly take space and time as containers of indefinite proportions, ‘showrooms’ for the display of discrete sensible items. Kant seems to support this ‘aquarium’ image. These frameworks, even if not compounded from granular lengths or durations, seem to subsist on their own, indifferent to objects ‘placed’ inside them. No property or operation ties things to spatial-temporal frameworks, though the formality of space and time connects to quantitative categories in a way that allows us to quantify and measure the spread of perceived things in judgments. Perhaps this sensible-conceptual connectivity that joins one form of relation to another is the work of imagination, e.g., the productive imagination of the A-Deduction, but this does not tell us much, since Einbildungskraft is just unification (Einbildung).

32In Fichte’s version of Transcendental Philosophy, there is no being that is not a seeing, and no seeing that is not an exercise of reflexive agency. The bundle of feelings the agency has constructed to this point is not an amalgam of properties or perceptions, just a direction or striving toward what emerges as the functions of intelligence and willing. At a preliminary stage, agency’s product must endow itself with the sensible stretch of spatial and temporal properties; only then can it take on the relative independence of being a body (a substance in causal interaction with others) or being a will (an actor in an interactive setting, whose agency is elicited by another). But the closer that agency gets to actualizing its key value of self-sufficiency, the more its function is hemmed in by finitude and its being/willing reduced to modalities of being-alongside-another. Agency’s logical program, to move from being nothing to being something, results in determinate modes of being and willing spread over multitudes of individuals. And freedom (or the coin of agency) gets restricted when it is realized, its normative force diluted in that it encounters limitation instead of self-sufficiency. We cannot adequately address the issues we mention here—why reason must be finite, why freedom must be intersubjective, and why self-limitation as a substitute for self-sufficiency lies at the foundation of both law and morality. When it arises, agency seems to be mine or my deed, a setting sail on an ocean of freedom, but when it finds appropriate ‘somethings’ to do and to be, it trades freedom for determinateness, and, like the modern consumer, finds itself smothered in the ‘stuff’ of finitude.

33On Fichte’s reading, agency or subject-objectivity does not merely insert itself into pre-existing temporal or spatial structures. He gives them a four-step derivation, where the contrast between free and reflective activity (or determining and determined agency) drives the deduction. (1) Agency expresses itself as a dance of activity as such, pure space, and free activity—figuratively expressed as the act of “drawing a line in space.” Where no dimension is given and the range of motion is in no way restricted, all dimensions—the three of conventional perception and the higher-order spaces that current physics entertains—are possible. But there is no extension and no direction until a first action and a first dimension are established--inscribing a line in space, or extending activity until it has a dimension. Fichte’s reasoning seems arbitrary; perhaps it is a rebuttal of Kant’s distinction between figurative synthesis, the unfolding of continuous phenomena, and inner sense, the flow of consciousness (KrV B 154-55). (2) In an intensification of its double activities, agency posits itself as an activity fitted to objectivity (the subject, schematized) and an opposed object. Activity directed to the subject concentrates itself into a point or ‘place’, and further free acts extend the spatial spread, so matter-space become an infinitely divisible continuum. If that is freedom of willing, a corresponding freedom of intellect collects feelings as properties of the object. Thus, space is filled, or there is matter. (3) A further potentiation puts objects in space. Free activity posits the rational being--“the practically striving being “-- as located in one place or working from a point of view. All objects gain location, but only relative to the striving-and-perceiving subject, and distance between subject and object becomes a function of the activity between them. (4) The activities mentioned, subjective and objective, are translated into two forms of efficacy or force (FTP 238-46, WLnm-K, GA IV/3, 412-416). We previously mentioned the concentration of inner force which is intuited as place, the origin of location. On the objective side, physical force extends or extrudes the single point into a continuous series, each member determined by its prior—the arrow of time. As so we have space and time as functions of freedom, agency, willing, causality—all forms of activity; quite different from Kant’s formal fishbowl.

34We cannot follow Fichte’s further steps in detail, but we can mention two general directions. In a fourth section of his deduction, Fichte defined space materialized as the condition of body, and body as the sole way that force can be expressed or will actualized. When agency is made plural or fragmented into individual agents, bodies are the sole vehicle that allows agents to interact. Shared or inter-subjectivity becomes a matter of causal interaction among bodies, with an inferential decoding of the bodily gestures and collisions between bodies mediating the freedom-and-limitation of living under law and discerning the moral duties of one’s physical and social situation (FTP 302-304, WLnm-K, GA IV/3, 444-45). A concluding section, oriented towards the object-body, outlines the parameters of agency in the body: articulation, the delegation of specific functions to specific organs or limbs, and organic functioning. The organism is the living embodiment of normativity, and a clue for interpreting the social and moral interactions of individual rational beings (FTP 458-62, WLnm-K, GA IV/3, 514-516). This is truer for us now that we have replaced ideas of top-down coordination for evolutionary or bottom-up design.

Is the I a Mirror or an Eye?

  • 12 “To Jakob Sigismund Beck. September 27, 1791,” Correspondence, 391-92.
  • 13 See Foucault, F., The Order of Things: An Archaeology of the Human Sciences, a translation of Les m (...)

35We have looked at Kant’s and Fichte’s versions of transcendental philosophy in some detail. The first offers a comprehensive investigation of the rational powers from the outside, an objective description one that lends itself to synoptic presentation in a table. Kant thought this way of stating his key moves and results in a table original and quite apt.12 A later-day Kantian, Michel Foucault, seizes on the table to epitomize the intellectual thrust of the Enlightenment.13 In contrast, Fichte offers a ‘genetic deduction’, a construction that threads the conditions of consciousness one by one on the string of agency, and so produce a living image of consciousness in the process of constructing itself. But aside from these differing modes of presentation, the systems are quite similar. Fichte puts the basic difference this way:

The I, as described by previous philosophers, is a mirror. But a mirror does not see, and this is why these philosophers are unable to explain “seeing” or intuition. All they posit is the concept of mirroring. This remark reveals the basis of all the errors of other philosophical systems—the Kantian system included. The error can be rectified only by means of a correct concept of the I. The I of the Wissenschaftslehre is not a mirror; it is an eye […] Everything we see, we see within ourselves. We see only ourselves, and only ourselves as acting, only as passing from what is determinable to what is determinate (FTP 151-52, WLnm-K, GA IV/3, 365-66).

36If we view intellect’s function as primarily cognitive, we might be satisfied with the regularity and completeness of Kant’s tabular view of reason. If we view its function as both cognitive and practical from the very start, Kant’s fourfold tables of judgments, categories, schemata, principle and antinomies will not satisfy, for I cannot find the rational subject inside the lists of cognitive concepts and functions, much less the agent who devises and executes a life-plan or acknowledges normative constraints on her behavior.

37There is another difference I can see, it is minor compared to the difference Fichte sees between fact and act: one will get different views of reason’s authority (normativity) if one describes it as the necessity derived from legislation or as the constraint imposed by having a goal--or by the ultimate goal, self-sufficiency (SE 58-61). In both cases, reason’s work is directive or order-imposing, categorical or compelling in domains like logic (formal and transcendental), law and morality, hypothetical or advisory if a matter of imposing a means-to-end order upon one’s conduct. Reason is fundamentally housekeeping, an order imposed by effort and requiring effort if it is to be maintained. But there is a vast difference between the outward-directed ordering of subsuming judgment and the self-direction of reflexion (SE 107, 112). It is the very nature of the I to be self-reverting activity, the seed of both will and cognition (GA I/4: 272-73).

  • 14 See Wood, A. W., Fichte’s Ethical Thought, Oxford, New York: Oxford University Press, 2016, pp. 151 (...)

38It is beyond the scope of this paper to compare Kant’s and Fichte’s moral theories, but I should mention that while both appeal to duty as the prompt for ethical conduct, Kant’s version of duty is tied to his concept of positive freedom or conduct authorized by universal law, while Fichte seems to understand the prompt for moral conduct situationally, involving the constraint that another will signifies for mine, or the force of the interpersonal ‘summons’ that tells one to become what one must become (FTP 351-53, 451-53; WLnm-K GA IV/3, 467-79, 511-512 ).14

After Enlightenment: Metaphysics in a Minor Key

39If skeptics and empiricists were not convinced by the argument of the Transcendental Analytic, Kant’s arguments in the Transcendental Dialectic about the unthinkability of God, immortality and freedom did not persuade the rationalists. While we tend to read Critical Idealism as having settled certain problem, Kant’s contemporaries saw him as stirring the pot of current debates. Reinhold’s initial appreciation of transcendental philosophy lavishes praise on Kant for having revived religious debates by stirring up genuine discussion, as if the anatomist of pure reason was really the religious provocateur Jacobi always hoped to be. Kant was clear, however, that he had at least cleared away the ruins of past mistakes and thus made a genuine metaphysics possible.

40What Fichte’s recasting of transcendental philosophy signaled, however, was the necessity of redoing intra-terrestrial or empirical metaphysics, for though Critique told us we have to reconfigure all our knowledge as sciences of appearances, not of noumena or ultimate items, by foregrounding Aristotelean relational categories an implicit ontology of things was left in place, and the manifest image of a world of self-contained subjects and objects continued to sound, but in a minor or phenomenological key. Fichte found he had to abandon the manifest image and for scientific purposes construct a free-standing account of a world of agency in which discrete rational beings were the precipitates of an ocean of swirling acts and intentions, a world where the most important cognitive act is not the welding together of classes and concepts in the categorical judgment but the “hovering before the mind” or entertaining of goals in imagination that mediates every step of the I’s evolution. The perspective of agency softens the contours of the world of appearances and exorcises the ghost of realism that suggests picture theories both for logic and perception.

41If there has been progress in metaphysics since Kant, it had been in revision of the categories of substance, essence, and causality as employed even in empirical contexts, and their replacement with ‘less constructed’ ideas such as functions in a field, minimal conditions for naming something x, and clouds of conditions, most of them necessary, few sufficient. Evidently, we cannot do without terms for what is happening or tends to happen, for figuring out minimum standard for naming something or consequently treating it a certain way—when the question arises, e.g., whether animal intelligence is an oxymoron or the beginning of a case for vegetarianism, or for discriminating between statistical correspondence and clinical efficacy in testing medical treatments or pharmaceuticals. We can now appreciate exactly what we want to know when we ask about entities, properties, and conditions. The older language of substance, essence, and cause assumed that what played these roles was more unitary, independent, unchanging, identifiable, and predictable than what makes something a discernible entity, or minimally defines its properties, or states the conditions for it to arise. Fichte’s revision of transcendental idealism was bold, probably too obscure in its statement to be paradigm-altering, but certainly fertile in its basic suggestion that what goes on is more fundamental than what we call things and that we close off our own possibilities when we transform a world of lively goals, acts, desires and decisions into a drab market of things and properties.


42FTP = Fichte [J. G.]. Foundations of Transcendental Philosophy: (Wissenschaftslehre) nova methodo (1796/99). Ed. & tr. Daniel Breazeale. Ithaca/London: Cornell University Press, 1992.

43CPrR = _____. Critique of Practical Reason, In Practical Philosophy. Tr. & ed. Mary J. Gregor. Cambridge, New York, Melbourne: Cambridge University Press, 1996.

Inicio de página


1 I use agency to translate Fichte’s Agilität, although the term is closer to Handeln, which suggests some deliberation or intention informs the act, while agility connotes speed, flexibility, and immediacy. Strictly speaking, Fichte’s account shows how agility transforms itself into agency.

2 To Jacob Sigismund Beck. December 4, 1772” in: Kant, I., Correspondence, tr. & ed. By A. Zweig. Cambridge, New York, Melbourne: Cambridge University Press, 1999, p. 445.

3 Kant also points to reason’s economic function in discussing the subjective principle that counsels the reduction of a multiplicity of rules to a single principle. It prescribes no laws for objects, but just does housekeeping [Haushaltung] of the resources [Vorrathe] of our intellect [Verstandes] (KrV A306/B262).

4 4Fichte is the beneficiary of Karl Reinhold’s definition of presentation as both the unit of cognition and the ‘nest’ of consciousness, whose qualities include material and formal elements, the former just given, the later spontaneously produced, joined in a synthesis that imposes unity on a manifold. See Reinhold, K. L., Essay on a New Theory of the Human Capacity for Representation, transl. by T. Mehigan & B. Empson, Berlin/New York: De Gruyter, 2011, pp. 119-137.

5 5The second text speaks of the five-fold synthesis as “expanding this original synthesis ‘from the inside out’.”

The first speaks of it a “synthetic period” or paragraph. This may seem confusing, if one expects synthesis to involve a temporal progression or narrative, but Fichte’s procedure is more geometrical or geographical, providing a spatial rendition of the logical space that conditions and what arises form them inhabit.

6 6Throughout four decades of lecturing on logic, Kant consistently described the Vorstellung as the minimal unit of cognition, referred either to subject or object, and incapable of definition. All higher capacities involve it: perception, comparison and contrast, conscious acquaintance, understanding, insight and comprehension. See Kant, I. Lectures on Logic, transl. & ed. by J. M. Young. Cambridge/New York/Melbourne: Cambridge University Press, 1992, pp. 103-104, 107; 440; 466; 569-70. The Critique reflects this general view (KrV A 333-34, B 390-91), but notes that the spontaneity of representing subject can be represented only passively, as the flow of inner sense (KrV B 67-69).

7 7Kant deemed Reinhold’s reconstruction of Critical Idealism murky, “weighted down with obscure abstractions,” even as he encouraged Beck to undertake a similar endeavor. See “To Jakob Sigismund Beck. November 2, 1791,” in Correspondence, 394.

8 8 In the Logic lectures, Kant leaves the matter of the presentation undefined (with formal features providing a gesture or direction toward ‘subject’ or ‘object’).

9 9 See “Declaration concerning Fichte’s Wissenschaftslehre. August 7, 1799,” Correspondence, 559-60.

10 See Kant’s brief distillation of the whole cognitive process—from representation, to intuition and concept, imagination and the synthetic unity of consciousness in “To Jakob Sigismund Beck. January 20, 1792” in Correspondence, 400-401.

11 Fichte, J. G., The System of Ethics According to the Principles of the Wissenschaftslehre, ed. & tr. By D. Breazeale & G. Zöller, Cambridge, New York, Melbourne: Cambridge University Press, 2005, pp. 84-85. See also Zöller, G., Fichte’s Transcendental Philosophy: The Original Duplicity of Intelligence and Will, Cambridge, New York, Melbourne: Cambridge University Press, 1998.

12 “To Jakob Sigismund Beck. September 27, 1791,” Correspondence, 391-92.

13 See Foucault, F., The Order of Things: An Archaeology of the Human Sciences, a translation of Les mots et les choses, New York: Vintage, 1973, pp. 74-76. Foucault argues that the table was the distillation of the intersection of three forms of ordering: mathesis, taxonomy, and genesis.

14 See Wood, A. W., Fichte’s Ethical Thought, Oxford, New York: Oxford University Press, 2016, pp. 151-52.

Inicio de página

Para citar este artículo

Referencia electrónica

Michael Vater, « Reason and Agency in Kant and Fichte », Revista de Estud(i)os sobre Fichte [En línea], 16 | 2018, Publicado el 01 diciembre 2018, consultado el 21 abril 2019. URL :

Inicio de página


Michael Vater

Marquette University

Inicio de página

Derechos de autor

© EuroPhilosophie

Inicio de página
  • OpenEdition Journals