In Critique of Pure Reason Kant describes in detail what aspects of our knowledge are a priori, and how they function, but on the empirical aspects he is sketchy and cryptic. Very briefly, our sensations/intuitions are synthesized into perceptions via schemes of productive imagination, including a priori forms of space and time. These schemes are somehow coordinated with concepts of understanding, which allows perceptions to be "brought" under them. The highest concepts, categories, are a priori, and come not even from understanding, but from reason itself. But what about empirical concepts and schemes coordinated with them that are not spatiotemporal, where do they come from? Size and shape are a priori, how about color, weight, firmness... universal gravity?

Kant admits that empirical concepts are "acquired". But how? Do productive imagination and understanding have them on hand when they synthesize perceptions and make judgments, or do they synthesize them by grasping regularities of the "sensory manifold" itself? If it is the former, are they "acquired" over time from what is "in the senses", or is the "sensory manifold" completely undifferentiated, and productive imagination "invents" them (with assistance from understanding) to frame it?

Both answers seem to undermine Kant's construction. If empirical concepts ultimately originate in our mind then the dichotomy between empirical and a priori collapses, they would be as a priori as space, time and the categories. If on the other hand, they are extracted from senses (somehow), then what collapses is Kant's impenetrable wall between appearances and things in themselves. Indeed, the sensory comes from the senses, which are impressed upon by things in themselves without any a priori interference. An alternative to both is to deny that we can isolate the sensory in perception (contrary to Husserl's phenomenology with its "bracketing out" of presumptions, and reaching "to things themselves"), but it is also problematic. Kant has no trouble isolating the a priori, and the rest has to come from the senses. Did Kant indicate somewhere which option he chose and how he dealt with complications?

  • 2
    Where is the quote from?
    – virmaior
    Commented Jul 16, 2015 at 23:10
  • @virmaior Sorry for confusion, it is not a quote. I just highlighted the part where the main questions are located.
    – Conifold
    Commented Jul 16, 2015 at 23:12

2 Answers 2


Basically the schemes are the guides to forming empirical concepts, according to the Critique of Pure Reason. The schemes are not merely coordinated with the categories, they are the empirical contents of the categories. The categories, themselves a-priori concepts, are the possible types of empirical concepts: substance concepts, causation concepts, etc. The schemes are modes of temporal pattern matching within the sensory manifold (not all sensory input is spatial, but it is always temporal). For example: if something always found to endure through changes, it would be a prima facie substance. If events of type A are always found to be followed by events of type B, this is a prima facie case of causation. And so on for the rest of the categories.

Now this representation of a general procedure of the imagination to present its image to a conception, I call the schema of this conception. In truth, it is . . . but schemata, which lie at the foundation of our . . . sensuous conceptions . . . The conception of a dog indicates a rule, according to which my imagination can delineate the figure of a four-footed animal in general, without being limited to any particular individual form . . . time is contained in every empirical representation of the manifold. Thus an application of the category to phenomena becomes possible, by means of the . . . determination of time, which, as the schema of the conceptions of the understanding, mediates the subsumption of the latter under the former. (Critique of Pure Reason, "Of the Schematism of the Pure Conceptions of the Understanding.")

As to the objection, that this "collapses Kant's impenetrable wall between appearances and things in themselves", well, it doesn't seem to me to do so over and above Kant's picture of mere sensation. The sensory manifold is located in time and space, which are, according to Kant, a priori (and therefore, subjective) forms. The specific ways in which the sensory manifold is distributed in space and time are however a posteriori, and therefore prima facie correspond to something in the things-in-themselves. Kant apparently thought that you could not really get any further than that, so that the things-in-themselves still remain unknowable. It may be of interest to note that Schopenhauer, in his version of Kantianism, indeed held that empirical concepts can teach us a great deal about the thing-in-itself.

  • My understanding is that the "sensory manifold" precedes perception, which spatiotemporal schemes are used to frame, what Kant calls sensation and intuition are free of them because in his definition sensibility is passive and not productive. At least that is how Kitcher describes it in Kant's Transcendental Aesthetic. So everything he says about appearances (collections of perceptions), including space and time, does not apply to whatever is sensory in perception, assuming it can be isolated. And then being passive, it is "witnessing" things in themselves, albeit relationally to us.
    – Conifold
    Commented Jul 19, 2015 at 3:33
  • @conifold As you say that we experience the things-in-themselves only relationally, your version is relevantly similar to mine, isn't it? Besides, if sensation were not spatio-temporal, it would have to receive the time and space forms first, and only then the schemes, which rely on time. Commented Jul 19, 2015 at 12:21
  • Kind of, but relational to us is not quite an appearance largely formed by us. Kant was likely right that geometry and kinematics are innate, we still can't imagine non-classical ones, they are understood abstractly. Yet they were extracted from classically framed perceptions, so innate mental framing doesn't amount to a condition of the possibility of knowledge like he thought. I am trying to see if Kant has resources to potentially account for formation of such non-conforming concepts, which would have to be based on pre-framed sensations, in the empirical part of his system.
    – Conifold
    Commented Jul 20, 2015 at 23:44

Probably a good picture to hold in mind is to consider Chomskys notion of a universal grammar, which for him is an inate idea of the mind, ie a priori; but a specific grammar of a language is still acquired empirically, through participation in a language community.

Empirical concepts or rather intellectual concepts don't originate in the mind; but are conditioned by the categories to make them intelligible to the mind/understanding.

  • This is a good analogy but it works equally for empirical and a priori. For example, basic mechanical laws and concepts are a priori to Kant, but it still took Newton and others to acquire them. Chomsky considers universal grammar genetic,l which would have been "a priori" for Kant, but assuming Chomsky's critics are right and grammar is acquired my question is what Kantian mechanism is available for that.
    – Conifold
    Commented Jul 19, 2015 at 3:25
  • Good - it was meant as an analogy; Kants a priori discussion of Newtonian mechanics is of a different type from his discussion of epistemology in the first critique; it's more aligned with Aristotles discussion in his physics, ie from first principles; the discussion I've read by Chomsky on the Kantian a priori links it to the notion of inate ideas - I don't recall him connecting it to genetics per se. Commented Jul 19, 2015 at 13:25

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .