Kant's position on things-in-themselves is often described Socratically, of them we know only one thing, that they are. However, in an old but apparently still popular history of philosophy book I found a different sounding take:

"In the Critique of Judgement Kant recognizes as conceivable two basic types of intelligence: the archetypal or perceptive (intuitive) intelligence; and the ectypal, discursive or constructive intelligence... things-in-themselves may be conceived as realities open to the contemplation of the archetypal intelligence. Thus beauty and the realization of value may in the end characterize concrete ultimate reality more adequately than do the categories of the understanding. This is based on the notion that things-in-themselves are to be known, if at all, as wholes whose real nature may be manifest only in their integrity, and can not be perceived seriatim [in sequence] through their several parts or incomplete aspects... Our sense of beauty, our respect for moral worth and our recognition of living teleology present themselves as disparate and fleeting glimpses of things-in-themselves. But we can not be quite sure of their credentials".

Kant's standards for knowledge are quite high (see How did Kant define knowledge?) so he may not have considered glimpses with uncertain credentials to be "knowledge". But aside from wording the quote implies that (in a colloquial sense) we get to know a lot more about things-in-themselves than that they are. Fichte, Goethe and Schelling later endorsed such holistic "intellectual intuition" of morality and aesthetics, and Nietzsche mocked late Kant for falling back into "dogmatic slumbers" with noumena. But is the above quote a reasonable reading (perhaps one of several) of the third Critique and other late works, or is it more of reading Fichte and Schelling into Kant? More speculatively, is there evidence that this was the private picture that animated Kant's thought, even if he would disown it as a speculation "officially"?

EDIT: In addition to Förster's book pointed out by Philip some essays in Fichte, German Idealism, and Early Romanticism are a relevant read. Estes's in particular quotes Fichte, and we find out where the above quote got its "glimpses":"It is only through the medium of the ethical law that I catch a glimpse of myself and insofar as I view myself through this medium, I necessarily view myself as self active". But Fichte only walked through the door opened by Kant himself already in the second Critique:"The consciousness of this fundamental law may be called a fact of reason, since one cannot ferret it out from antecedent data of reason, such as the consciousness of freedom for that is not antecedently given), and since it forces itself upon us as a synthetic proposition a priori based on no pure or empirical intuition... one must note that it is not an empirical fact but the sole fact of pure reason, which by it proclaims itself as an originating law". Estes concludes:"By defining all intuition as sensible, he precluded both things-in-themselves as objects of knowledge and intellectual intuitions as modes of consciousness. As a result, Kant struggled to describe self-consciousness and moral consciousness. Allowing them as intellectual intuitions would violate the limits of knowledge he imposed but denying them rendered the knowledge he desired impossible".

  • 2
    The quote seems to refer to § 77, 8th paragraph of CJ, Ak. 407-8. In there all terms (archetypal, intuitive - ectypal, discursive) are mentioned. The other term is understanding in English (intellectus in Latin), not intelligence. Another example of bad translation.
    – Philip Klöcking
    Commented Jan 24, 2016 at 2:53
  • What tangled webs we weave, when first we practice to conceive. Practice makes perfect!
    – Scott Rowe
    Commented Aug 30, 2022 at 10:15

4 Answers 4


TL;DR: No, he did not!

To be precise, things-in-themselves may be objects of thought, i.e. abstract concepts of the realm of logic, and therefore concepts of transcendental philosophy as logically necessary conditions of the possibility of experience. But they cannot be objects of knowledge, i.e. things that are particular objects of experience subsumed under concepts (i.e. at least the categories).

That is why there is a quote in his Opus posthumum that states that the very concept of thing-in-itself is self-contradictory.


Regarding direct quote from the Critique of Judgement

Kant used the intuitive understanding as a limiting concept [Grenzbegriff] to understandings like ours and always disjunctive to it, like the holy will vs. a finite (albeit, in abstraction, pure) will like ours in his practical philosophy. From the Critique of Judgement, §77, paragraph no. 8 (Ak. 407.13-408.23, quoting the paragraph as a whole, bolded by me), Cambridge Edition:

In fact our Understanding has the property of proceeding in its cognition, e.g. of the cause of a product, from the analytical-universal (concepts) to the particular (the given empirical intuition). Thus as regards the manifold of the latter it determines nothing, but must await this determination by the Judgement, which subsumes the empirical intuition (if the object is a natural product) under the concept. We can however think an Understanding which, being, not like ours, discursive, but intuitive, proceeds from the synthetical-universal (the intuition of a whole as such) to the particular, i.e. from the whole to the parts. The contingency of the combination of the parts, in order that a definite form of the whole shall be possible, is not implied by such an Understanding and its representation of the whole. Our Understanding requires this because it must proceed from the parts as universally conceived grounds to different forms possible to be subsumed under them, as consequences. According to the constitution of our Understanding a real whole of nature is regarded only as the effect of the concurrent motive powers of the parts. Suppose then that we wish not to represent the possibility of the whole as dependent on that of the parts (after the manner of our discursive Understanding), but according to the standard of the intuitive (original) Understanding to represent the possibility of the parts (according to their constitution and combination) as dependent on that of the whole. In accordance with the above peculiarity of our Understanding it cannot happen that the whole shall contain the ground of the possibility of the connexion of the parts (which would be a contradiction in discursive cognition), but only that the representation of a whole may contain the ground of the possibility of its form and the connexion of the parts belonging to it. Now such a whole would be an effect (product) the representation of which is regarded as the cause of its possibility; but the product of a cause whose determining ground is merely the representation of its effect is called a purpose. Hence it is merely a consequence of the particular constitution of our Understanding, that it represents products of nature as possible, according to a different kind of causality from that of the natural laws of matter, namely, that of purposes and final causes. Hence also this principle has not to do with the possibility of such things themselves (even when considered as phenomena) according to the manner of their production, but merely with the judgement upon them which is possible to our Understanding. Here we see at once why it is that in natural science we are not long contented with an explanation of the products of nature by a causality according to purposes. For there we desire to judge of natural production merely in a manner conformable to our faculty of judging, i.e. to the reflective Judgement, and not in reference to things themselves on behalf of the determinant Judgement. It is here not at all requisite to prove that such an intellectus archetypus is possible, but only that we are led to the Idea of it,—which contains no contradiction,—in contrast to our discursive Understanding which has need of images (intellectus ectypus) and to the contingency of its constitution.

As you can see this could refer to "pure intuitions" (or intellectual intuition) of things-in-themselves ("of a whole as such"), but is only ascribed to understandings fundamentally different from ours. It would also be coherent to your quote as only archetypal/intuitive understanding would include intuitions of things-as-themselves as it literally produces exactly what it thinks by thinking it (basically Gods understanding), so that it automatically intuits them as a whole.

Further evidence

This is consistent to all his earlier writings and his opus posthumum:

Ak 21:75 (manuscripts opus posthumum, first convolute):

Wenn die Grenze der Transsc. Philos. überschritten wird so wird das angemaßte Princip transscendent; d. i. das Object wird ein Unding der Begriff von ihm wiederspricht sich selbst: denn es überschreitet die Grenzlinie alles Wissens: das ausgesprochene Wort ist ohne Sinn.

My translation (as it is not included in Förster's Cambridge Edition) would be:

If the borderline of transcendental philosophy would be crossed, the assumed principle became transcendent; i.e. the object becomes a no-thing [ein Unding], the concept [Begriff] of it contradicts itself: Because it crosses the borderline of all knowledge: the uttered word is without meaning.

In fact here he says that as soon as we try to go further than a priori principles as found in the three critiques and go for transcendent instead of immanent concepts (see Prolegomena, 4:373, fn.), we transcend all knowledge ("es überschreitet die Grenzlinie alles Wissens [borderline of all knowledge]"). The object of speculations like that would become a no-thing [Unding]. That doesn't mean that we couldn't contemplate on them. But all contemplation then would never be in the realm of knowledge, but mere speculation.

That is only consequentia, because the difference between object [Gegenstand] and thing [Ding] is the one between the table of judgements (necessary conditions of reference to objects as such [Gegenständen überhaupt]) and the table of categories (necessary conditions of reference to things as such [Dingen überhaupt]) in the Critique of Pure Reason (A70-71 and A79-80 respectively).

The latter only makes sense for objects of intuition [Anschauung]; concepts [Begriffe] and intuitions are interdependent and both necessary for constituting things, i.e. objects of knowledge (remember the famous quote about concepts and intuitions in CPR, B75). Therefore the proclaimed contradiction in itself: As soon as intuitions are involved, we can schematize, and only then knowledge is possible. And these intuitions cannot be intellectual, because our understanding simply is not capable of it.

I actually think this is the real reason for his needing of a type of the moral law in the second critique, because for the moral law itself, we have no scheme and therefore no (not even practical) knowledge. On the other hand, the moral law itself is arguably the one exception from the rule, as it seems to be some kind of intellectual intuition that cannot be further processed by our understanding and because of that evokes these feelings of reverence.

Regarding Fichte, Schelling and Hegel

Most contemporary philosophers only learned from CPR through Jacobi's reading of the A edition (which wasn't too accurate) or by the B edition, because there have only been 1000 copies of the A edition and the first edition that featured both was published in 1838. The readings and (mis-)interpretations of all three critiques by Jacobi, Fichte, Schelling and Hegel are perfectly described in Eckart Förster's The 25 Years of Philosophy. For a detailed answer on their objections and their ways to finally get access to things-in-themselves I would recommend reading the book as a whole.

One thing you can see from this answer already is that they obviously have to step into speculative idealism, which for Kant would by definition be transcendent, and try to reconcile this with the possibility of knowledge. This is imho exactly what Fichte tries in his Wissenschaftslehre, Schelling with his philosophy of identity in My System of Philosophy, and Hegel in his Science of Logic and even in his earlier Phenomenology of Spirit.

Regarding what "Thing-in-itself" [Ding an sich] actually means for a native speaker

For native speakers, it is quite obvious that "Ding an sich" means "The thing as it is in its full essence/being". "An sich", as an idiom, means "how it is for real". Substitute would be "eigentlich", translation "actual, rather, intrinsic, real".

Regarding the reality of things in themselves

Kant has a clear account of what is real and what's not:

The possibility of experience is therefore that which gives all of our cognitions a priori objective reality (CPR, A 156|B 195)

This now has two different layers:

  1. Something can directly be a thing, i.e. object of experience, and therefore real. This is clearly not the case for things in themselves as I have shown.
  2. Something can be a necessary condition of the possibility of objects of experience (either in general or in particular). This is what Kant thinks is true for things in themselves and the categories (and freedom! in practical philosophy).

That means that things in themselves are definitely 'real', but only in an indirect way, as their immanence is only through logical means. This also means that although we have to assume their reality because of the way we think, we cannot possibly be sure of it, i.e. achieve no knowledge about it.

Kant can therefore be perfectly fine with the idea that we are brains-in-a-vat or that we completely make up reality and are solitary in the universe. But I think he would point out that we have no evidence for that whatsoever and it is rather unlikely.

Conclusion and further remarks

So, "Ding an sich" is not even a 'thing', it is as mere negative object of thought as something outside of thought, totally independent from our mind. To be honest, not even modern sciences can ever be able to describe anything independent from our conceptual (or theoretical) grasp of it. People like Popper, Kuhn and Putnam were pointing that out rather recently.

Therefore, while we are certainly able to gain knowledge about aspects of its reality as we conceive it and strive for getting closer to a full understanding, it lies in the very concept [Begriff] of it that we cannot get knowledge about it as the holistic unity thought under this concept.

Or, to put it in another way, we may even achieve a point were we know all aspects of an object that would constitute the thing-in-itself (although we cannot possibly know wether we already did), but we will because of our finite understanding never be able to think it as a whole, only as aspects put together. Every account of an object "as it is for real" would therefore necessarily be reductionalistic.

This is exactly the point of Hölderlin's little fragment Judgement and Being: As soon as the disparation of object and subject occured and the self as self-conscious and subject of knowledge is in place (making judgements), there is no way to take a step back and conceive the unity of being again. This could only be conceived by a pure, non-judging intuition that contradicts the possibility of self, self-consciousness, and knowledge by definition. The one thing he can explain better than Fichte with this account is how the self comes into being, because in Fichte it literally is out of the blue as there is nothing there before the first act of positing the self (see e.g. Henrich, The Course of Remembrance, 74-76, Henrich/Pacini, Between Kant and Hegel, 279-295).

  • What I am trying to figure out is how he internally viewed noumena, as imaginary as-ifs "completing" phenomena, mere idealizations, or what we would call hypothetico-deductive guesses about underlying reality inspired by glimpses of archetypal intelligence? If it is the latter then in hindsight even geometry and logic fair no better, nothing can clear Kant's maximalist bar. We are out of luck with his Knowledge anyway, but what was his faith? The CJ passage alluded to seems less bold than the quote implies, but was he moving in Spinoza's direction just w/o calling it "third kind of knowledge"?
    – Conifold
    Commented Jan 23, 2016 at 23:59
  • @Conifold: Finally found it at the end of that paragraph has been Ak. 408 - pretty clear that he distinguishes disjunctive concepts here all the time and that only archetypal/intuitive understanding would include intuitions of things-as-themselves as it literally produces what it thinks by thinking it (basically Gods understanding).
    – Philip Klöcking
    Commented Jan 24, 2016 at 2:34
  • @Conifold: After all, I'm pretty happy with the answer. Let me know if anything is left unclear.
    – Philip Klöcking
    Commented Jan 24, 2016 at 3:07
  • Vow, thanks. You were right about the moral law, Kant's "sole fact of reason" is where Fichte caught the "glimpse of himself", I edited refs into the question. I also removed now irrelevant comments.
    – Conifold
    Commented Jan 25, 2016 at 21:31
  • 2
    @Conifold: Did a bit of revision, you may find some points interesting for your own readings.
    – Philip Klöcking
    Commented Nov 17, 2016 at 17:43

While I am no Kant scholar, my understanding is that the noumenon or thing-in-itself (which are not exactly the same, but close enough) might best be defined as "a shadowy object of heated scholarly controversy, refutation, and resuscitation."

The main divide seems to be whether they are treated as ontologically distinct from phenomena (for example, as a "cause" or ground) or as a "way of thinking about" an object. Gerold Prauss and Henry Allison are among the many scholars who prefer this rendering and treat "ding an sich" as a shortening of "ding an sich selbst betrachtet." Thing considered as it is in itself.

The necessity then arises from the idea that we are already considering the thing under conditions of sensibility as, in some way, an "appearance" organized into related appearances, thus always suggesting another way it might be considered, i.e., apart from this condition. Why bother? Because, I would guess, Kant is interested in the "conditions of sensibility," which must be studied by a sort of triangulation between "appearances" and that aspect of the thing which is "not an appearance."

On this interpretation, which makes sense to me, the main mistake would be to think of the noumenal as ontologically privileged or "more real." To me it seems precisely the point of Kant's "critical" method that we drop the whole idea of things being "more real" or "less real." However, I don't think Kant's treatment lends itself to clear agreement on this, especially in the case of the self and its freedom.

That said, I'm not sure what to make of the quote. I'm not sure how the "archetypal" and "ectypal" distinctions relate, and the reference to glimpses of "concrete ultimate reality" sounds more Platonic than Kantian. Or as if we were glimpsing the object's "deeper reality," like the submerged hindquarters of a mermaid.

  • 2
    The impression Critique of Pure Reason gives is that Kant saw things in themselves as unknown unknowns. What is new to me in Critique of Judgement §77 is that not only does he think we know exactly why they are unknowable (because they are wholes while our understanding is discursive), but even describes the kind of understanding it would take to know them. Kant raises this "intellectual intuition", ostensibly to reject it, way too many times. Fichte, Schelling and Schopenhauer must have sensed where his heart was, and ran with it as the highway to things in themselves.
    – Conifold
    Commented Jan 31, 2016 at 23:07
  • Interesting, thanks. Kant wrote a lot and one can hardly expect it to be totally "coherent," which means to be closed and dead. So interpretations breath more life into his work, and his perhaps unresolvable views on the "in-itself." I am now veering back to an interpretation of Kant as a response, above all, to Newton. He was still completely within the Newtonian world, as was Hume. So the "in-itself" might be seen as a version of the "limit" paradox in calculus, and run all the way up to problems like "dark energy." Anyway, any theory of "knowledge" must involve an "unknowable." Commented Feb 1, 2016 at 1:48
  • Well, Marburg neo-Kantians (Natorp) did interpret object of knowledge as a limit, "regulative ideal", but they explicitly saw it as a break with Kant, I think. On Kant's relation to Newton I found Friedman's Kant and Exact Sciences very helpful, also Schonfeld's Philosophy of Young Kant thedivineconspiracy.org/Z5263E.pdf
    – Conifold
    Commented Feb 3, 2016 at 1:24

I would read Kant's ethical work. This is because Kant makes some actual positive claims about things in-themselves (noumenal realities) such as god, freedom, immorality of the soul.

While he denies in the critique of pure reason any meaning to these things from arguments of speculation (without reference to the world of possible experience those things are meaningless), he says we can "know" we are moral and free from the categorical imperative through practical reason.

The reason for this is analysis of practical reason which differs from speculative reason. For one thing the categorical imperative only consists of universalizing as a law of moral nature. its formal reality and its objective reality are the same so they must be considered "valid". The next issue is that of meaning, the will takes an end which is empirical for kant as it gives content to practical reasoning (the will). For Kant, knowledge or cognition (depending on which translation ur using of the critique or pure reason) must have both spontaneous and receptive (both understood through a concept and have content from empirical intuition). So we have satisfied what knowledge is, a desire provides the empirical content and its is understood through the universality requirement. So we know we are moral and since morality presupposes freedom, we know we are free by entailment.

god and immorality of the soul are "regarded" as true by acts of will. Basically u need to presuppose a god that grants moral luck and immorality of the soul so as to achieve happiness. This however is a prudent imperative so only applies to those with that conception of self-love. So god and immorality of the soul arent "proved" as true like we proved human freedom. Though practical reason we know that man is homo numena and homo phenomena (a being that introduces causality into the sensible world and is part of that sensible world). We are autonomous and we know all the moral "truths" that come from the analysis of practical reason.

  • Hello, and welcome to Philosophy.SE. Would you mind improving the formatting by using paragraphs? As of the content, I broadly agree. What you allude to has been discussed in the question and my answer already though. As Conifold quotes from the second critique in reference to the Moral Law: "it forces itself upon us as a synthetic proposition a priori based on no pure or empirical intuition" (Ak. 5:31). We get a "glimpse" of the noumenal only negatively (void of content) since our concept of freedom is negative (ibid) just as the concept of a pure will (our noumenal part in a sense) is.
    – Philip Klöcking
    Commented Oct 1, 2021 at 8:51
  • I broke it up into paragraphs, sorry about that. Counter to what he says in the critique of pure reason and the beginning of the critique of practical reason, he gives a kind of positive account of freedom as autonomy. Maybe its just my interpretation, but when he gives his reciprocal hypothesis (categorical imperative implies freedom and freedom implies CI) its a positive definition of freedom, namely conformity to self law so autonomy. Commented Oct 2, 2021 at 14:32

What provides us with representations of spatial locations? This is not a complete answer, but it at least suggests there is a question. https://plato.stanford.edu/entries/kant-spacetime/notes.html This passage:

One of the clearest recent discussions of intuition, sensation and the representation of space in the first Critique can be found in Allais 2009. She argues that intuition provides us with representations of particulars--of things with spatial locations--independently of concepts, urging that we distinguish this claim about representation from the famous Kantian thought that we cannot achieve cognition without both intuitions and concepts (Allais 2009, 390). According to Allais it is intuition that presents us with the locations of things.

Unfortunately I cannot access Allais's paper, and I don't see how intuition can actually provide the locations.

  • How does this answer the question at hand even partially?
    – Philip Klöcking
    Commented Sep 17, 2022 at 18:38

You must log in to answer this question.

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