Kant’s Transcendental Idealism (and illusions, phenomena/noumena, and real vs logical possibility)

Just the other day SelfAwarePatterns published an excellent entry on the overuse of the phrase “illusion” in explaining away phenomena, and as I agree with the dangers of hasty reductionism, I chimed in with a Kantian-inspired comment on the objective reality of “the ways things appear” being guaranteed by a necessary connection to “the ways things are in themselves”. Following that encounter I recalled how much Kant would disagree with my statement, but at the same time, how necessary something like this position is to making Kant’s view work. This is one of the main criticism’s of Kant’s epistemological theory of Transcendental Idealism, and today I want to consider if such a criticism can be weathered.

1. Kant’s Transcendental Idealism and the Distinction Between Noumena and Phenomena
In the Transcendental Aesthetic Kant argued that space and time are added to what is perceived by the understanding, such that, there is a really existing thing that is being perceived, but how it appears to the subject is structured by the understanding. But if this is the case, then what we have access to in our minds is different than the real objects of the world, though they do exist. Kant terms the appearances of objects that we have access to from sensible intuitions, ‘phenomena’, and those things that we do not have access to from sensible intuitions, ‘noumena’. While phenomena are the way objects appear through the effects of the sensibility, noumena are the way objects are without being shaped by the sensibility. Noumena are, then, the objects as they really are, and what Kant calls a thing-in-itself. Kant points out that if sensible intuitions give us phenomena, then another faculty of intuition must exist to give us knowledge of noumena. This leads Kant to posit that as noumena can only be known by intellectual intuitions, noumena are intelligible entities impossible of being known by the sensibility.

Kant posits that we cannot have objective knowledge of noumena precisely because noumena are not given by sensible intuitions. In the Transcendental Deduction, Kant found that the experience of objects depends on the transcendental synthesis of imagination to unify manifolds of intuition into one sensible object, and that this synthesis is made possible by the categories reveals that the categories make experience possible.

In the schematism Kant endeavors to explain how the categories in fact apply to appearances. We must spend some time tending to this issue as it is crucial to Kant’s conclusion that we cannot have knowledge about noumena. Kant posits that there must be a transcendental schema that connects the pure concept to the appearance. A transcendental determination of time is homogenous with the categories because time is necessary for a manifold of intuitions to be unified by a category. Moreover, the transcendental determination of time is homogenous with appearance because time is added to all appearances by the understanding. As time is therefore homogenous with both the categories and appearances, the transcendental determination of time must be the transcendental schema that makes possible the application of the pure concepts to appearances. In other words, the schemata are the temporization of the categories so that they are homogenous with appearances.

That the categories must be temporalized through the schemata in order to apply to appearances is of great importance, as the consequence is that the categories only have use when applied to what is given through the sensibility. Kant notes this in positing, “the schemata of the pure concepts of understanding are thus the true and sole conditions under which these concepts obtain relation to objects and so possess significance. In the end, therefore, the categories have no other possible employment than the empirical” (Critique of Pure Reason, 186). If it not the case that the categories have any use when not in relation to sensible intuitions, and noumena are not known from sensible intuitions, then the categories cannot be used to make claims about noumena.

However, the categories are what make objectivity possible, such that, if the categories cannot be applied to noumena, then there can be no objective truths about noumena. On the connection between the categories and objectivity, Kant argues that objective scientific and mathematical knowledge is possible in virtue of the synthetic principles of understanding, which are the result of the schematized categories. The synthetic principles of pure understanding are divided into two categories, the mathematical, which are necessary for experience, and the dynamical, which emerge upon reflection on experience. Observations that fit with these principles have objective validity because they are synthetic a priori truths derived from the a priority of the categories, which are universal in human experience. On this Kant writes, “the table of categories is quite naturally our guide in the construction of the table of principles. For the latter are simply rules for the objective employment of the former” (Critique of Pure Reason, 196). The synthetic principles of pure understanding, then, allow humans to have objective scientific knowledge by allowing that the various intuitions of a thing can rightly be brought together in one appearance (axioms of intuition), and that changes in what is sensed about this appearance can be measured in degrees (anticipations of perception). Together, the synthetic principles of mathematical employment just discussed enable scientific inquiry with objective answers. Therefore, the categories cannot be applied to noumena because the categories only have meaning when temporalized to fit with sensible intuitions. Kant’s argument for this is that, “the principles of pure understanding are only of empirical, never of transcendental employment, [so] that outside the field of possible experience there can be no synthetic a priori principles” (Critique of Pure Reason, 265). Thus, as it is impossible to apply the synthetic principles of understanding to noumena, and the synthetic principles of understanding are what make synthetically derived objective knowledge possible, objective knowledge cannot be derived about noumena, and so we cannot even speak about noumena other than to say that they must exist.

Thus, Kant’s transcendental idealism requires that there are entities existing outside of the mind so that there may be something given to the mind through the sensibility, and that the sensibility shapes the entities so that what is given to the mind is the appearance of an object, and not the thing-in-itself. The benefit of Kant’s system is that we can have objective scientific and mathematical knowledge about the way things appear to us because the understanding and the sensibility in conjunction shape appearances, a universal feature of human experience. The drawback is that we cannot know or speak about noumena.

2. The Standard Objection to Positing the Existence of Noumena
In line with the criticism of other commentators, A.C. Ewing argues that Kant’s Transcendental Idealism is problematic because in order to establish a distinction between noumena and phenomena, categories must be attributed to noumena. Moreover, in order for there to be this distinction, it seems that noumena must cause the sense data that we then shape into appearances. On the issue of applying categories to things in themselves, Ewing writes that,

This is not an adventitious slip but necessary to his philosophy. For (1) things-in-themselves must be subject at least to the category of “reality”; [and] (2) we cannot even think of physical objects as appearances of things-in-themselves without thinking of them as determined by the latter, which involves the category of causality, and it is clear that Kant’s whole conception of appearances presupposes the action on us of a thing-in-itself as a part cause of our representations accounting for the empirical element in knowledge (A Short Commentary on Kant’s Critique of Pure Reason, 187-188).

This criticism is quite serious for Kant’s Transcendental Idealism. We have seen that Kant posits that the categories only have meaning when applied to empirical experience, and as noumena are not known from empirical experience, it follows that we cannot speak of noumena even to suggest that they exist, because such talk is meaningless. Yet if this is the case then the distinction between noumena and phenomena crumbles. As this distinction is necessary for the possibility of objective knowledge about how things appear to us, without the division of noumena and phenomena we are thrown back into skepticism about objective knowledge. Thus, we should be motivated to find a way to posit the distinction between noumena and phenomena without applying categories to noumena.

3. Robert Adams’ Solution to The Standard Objection
Robert Adams argues that Transcendental Idealism avoids the problem of applying the categories of ‘reality’ and ‘causality’ if noumena are taken to be what Kant calls ‘problematic concepts’. Kant writes that a problematic concept is a concept wherein, “the objective reality of a concept cannot be in any way known, while yet the concept contains no contradiction” (Critique of Pure Reason, 271). Taking noumena as a problematic concept in no way stretches Kant’s position, as Kant himself posits that ‘noumena’ is a problematic concept because it is not contradictory. If Adams can show that the objections to Kant’s positing of noumena arise because they ignore the fact that noumena are problematic concepts, then it seems that the standard objection is avoided.

Adams posits that Kant’s use of problematic concepts is to establish a distinction between real possibility and logical possibility. A concept has real possibility if it can relate to an object and have meaning. In order to be logically possible, a concept need only be non-contradictory; it need not be possible to apply that concept to an object. Problematic concepts, then, are concepts that are known to be logically possible, but could be either really possible or not really possible. Adams elucidates the definition of a problematic concept by writing that for a problematic concept, “we do not know that an object for the concept is not possible. Indeed it is only the real possibility of such an object that is unknown to us. We do know that it is logically possible, inasmuch as there is no contradiction in the concept” (“Things in Themselves”, 816). Thus, problematic concepts are non-contradictory concepts that either do or do not correctly relate to an object.

Understanding noumena as problematic concepts enables Transcendental Idealism to avoid attributing categories to noumena because logical possibility does not require that noumena are real, such that, noumena are not subject to the category ‘reality’. To claim that noumena must be really possible in order to hold the distinction between noumena and phenomena requires attributing the category of ‘reality’ to noumena. But as Adams has shown that noumena can be considered as logically possible without passing judgment on their real possibility, it follows that we can have the concept of noumena without attributing ‘reality’ to noumena.

Against the second part of the standard objection – that the category of ‘causality’ must be attributed to noumena – Adams responds that, “the concept of noumenal causality should be viewed as a Kantian problematic concept” (“Things in Themselves”, 820). Doing so entails that noumenal causality is logically possible, but that it is unknown whether noumenal causality has real possibility. In other words, by understanding noumenal causality as a problematic concept, it need not be the case that the categories be applied to make a judgment on the possibility of noumenal causality, because the categories are only needed to judge the real possibility of a concept, not the logical possibility of a concept. Thus, understanding problematic concepts as making a distinction between real and logical possibility allows for the postulation of noumena without necessarily applying categories to them.

4. Conclusion
If Adams’ understanding of ‘problematic concept’ is correct then Kant’s theory does not attribute categories to noumena, and thus, Kant’s theory is not inconsistent after all. As it seems that Adam’s conceptualization of ‘problematic concept’ is a fair reading of Kant’s work, it seems that this is the correct way to respond to the standard objection. That the postulation of noumena is only a postulation of the logical possibility of noumena allows the distinction between noumena and phenomena to stay in place, and thereby enables the possibility of objective knowledge of phenomena.


, , , , , , , , , , , ,

  1. #1 by guymax on March 16, 2014 - 6:35 am

    Thanks. Some nice points. Kant always seems almost spot on to me but not quite.

    I feel that if we interpret Kant to be saying that phenomena are emergent from or underpinned by a phenomenon that is not an instance of a category (except for the category ‘not an instance of a category’) then we end up with the Buddhist view. I’m not alone in thinking this, and in consciousness studies this would become ‘relative phenomenalism’, a view that is consistent with Kant as long as we give him a certain interpretation.

    For this view (as far as I can make out) the noumenon would be real, not subject to the categories, and prior to time and space, but there would be only be one and it could not be said unequivocally to ‘exist’, (I.e there would be a difference between ‘exist’ and ‘real’) So we end up with Tao as the uncategorized phenomenon that would be the noumenon for all phenomena except itself, while Tao itself would be a phenomenon without a noumenon.

    Or something like this.

    • #2 by ausomeawestin on March 16, 2014 - 1:38 pm

      Very interesting, thanks for sharing your knowledge on this topic, I’ve been interested in similarities between Western and Eastern thought, particularly the theory that commonalities in the metaphysical ideas of the Greeks and Indians are due to similarities in the grammar of their respective languages.

      This theory of the Tao is interesting in its use of Kantian concepts, but I cannot see how Kant could accept these claims, first and foremost the claim that there could be uncategorized phenomena; Kant would say that that is a contradiction in terms. I’m no died-in-the-wool Kantian but I do see that as an appropriate commitment.

      • #3 by guymax on March 16, 2014 - 6:30 pm

        I think it is clear that Kant proposed a phenomenon beyond the categories, and there could only be one for logical reasons. What is not clear to me, at least, is whether he thought this was a ‘real’ phenomenon (and what he meant by ‘real’). ,

        Here are a couple of relevant comments you may like. The first is Edward Barkin from an essay ‘Relative Phenomenalism’ in JCS. The second is Korner from his book on Kant. I added them to my collection because they seem to clarify Kant’s view. Or maybe they just reinforce my view of Kant. I can’t be sure which it is.

        ‘The history of Western philosophy is filled with discussion, in one guise or another, of what is often called the ‘transcendental’ subject and object. The terms invoke the idea of a hidden self behind the phenomenal self and a hidden object behind the phenomenal object. Although Kant positioned the transcendental ‘things-in-themselves’ as methodological concepts rather than as metaphysical entities, the tendency since Kant has often been to reify them and then debate their objective existence’

        ‘In the Analytic of Concepts Kant has drawn a sharp distinction between the ‘I think which must be capable of all my presentations,’ thereby giving them synthetic unity, and the empirical, introspective, self which is itself a presentation. To be truly a priori rational psychology must have for its subject the former, i.e. the self of pure self-consciousness. This however is not, according to Kant, an object of experience and so of the applicability of the Categories. It is not an instance of any Category.’

        This last point would be what brings Kant into line with Lao Tsu and the Buddha. The categories must have an origin beyond the categories, or, as one mystic puts it, ‘beyond the coincidence of contradictories’.

        It may be the most important idea in philosophy. It is, of course, one that cannot be countenanced in the ‘western’ tradition, for which Kant is supposed to be founding father. I think he might have liked the Tao Te Ching better.

      • #4 by ausomeawestin on March 16, 2014 - 8:08 pm

        Thanks for the interesting quotations, they make for an fascinating interpretation of Kant. However, I really do not think Kant proposed phenomena beyond the categories. Phenomena are appearances filtered through the sensibility and the understanding, and Kant argued that the categories apply to all experiences because the categories make experience possible. His argument for this, known as the transcendental deduction, begins from the plausible premise that there is self-attribution of representations, which is what Kant terms apperception. His next step is to show that this makes the representations objectively valid because they are structured through a process of transcendental apperception. The next move is to show that apperception by itself does not show how the manifold of representations are brought together as being for the one consciousness of the experiencing subject, and that the explanation lies in those logical functions of judgment that are the categories. However, the categories are only visible in sensible objects, such that the categories provide knowledge through experience. There is a necessary combination of the manifold of sensible intuition which Kant terms the transcendental synthesis of imagination. The transcendental synthesis of imagination imposes time consciousness by projecting imagined representations that are unified with present representations in one sensible object. For Kant, this means that the transcendental synthesis of imagination must be the foundational synthesis that makes all synthesis possible – including the original synthetic unity of apperception. Kant concludes that as the transcendental synthesis of imagination makes experience possible and the transcendental synthesis of imagination depends on the categories for the process of unifying not present representations with presentation representations, it follows that the categories make experience possible and apply to all phenomena. But perhaps I am misunderstanding what you mean by “phenomenon beyond the categories”?

      • #5 by guymax on March 17, 2014 - 6:04 am

        AuesomeA – There is no ‘reply’ option on your latest comment, so unfortunately this will appear out of order. I’m replying to your post that ends…

        ” … it follows that the categories make experience possible and apply to all phenomena. But perhaps I am misunderstanding what you mean by “phenomenon beyond the categories”?

        Thanks for explaining Kant’s view so well. I am no expert on him. I’m not too worried about the final details of his view since I don’t think it is correct. But I do think he came very close to being correct. To me it seems that Hegel gives a more complete view, but that Kant got nearly all the way there before him.

        The categories would make experience possible for my view also, but there would be a subtlety. The ultimate phenomenon would be real, even though, as per Kant, it would lie beyond the categories and beyond experience. An experience requires an experiencer, and this is two things, and if there are two things then neither can be beyond the categories. (Since otherwise they would be indistinguishable).

        So while I would agree that the ultimate cannot be experienced, I would not agree that it cannot be verified or known. It could be known by what Kant calls something like ‘non-intuitive immediate knowledge’ or, in other words, ‘knowledge by identity’.

        This is not sensory knowledge. In the sciences ’empiricism’ relates exclusively to sensory information and this means that the word cannot be used to describe what we learn when we ‘know thyself’ as advised by the Oracle at Delphi. So Kant would be correct that this is non-empirical knowledge. It would not be theoretical knowledge either.

        Hegel seems to go a little further than Kant when he defines this ultimate (and verifiable) phenomenon as a ‘spiritual’ unity’. This would be the basis for all empirical phenomena. The multiplicity of space-time phenomenon and space-time itself would be emergent.

        It is this final move that would (to me) align Kant’s phenomenon that is ‘not an instance of a category’ with Tao, Nibbana and so forth, and with the Buddhist idea of ’emptiness’. I don’t think Kant quite made it this far but he seems to have came very close, and so provides a useful bridge between western and eastern philosophical thinking.

        For instance, in advaita Hinduism the word ‘advaita’ means ‘not-two’, It refers precisely to the idea that the space-time universe reduces to and is emergent from a unity. A unity is not an instance of a category and cannot be known by the senses. It cannot be said to exist or not-exist, being transcendent to these categories, but it can and usually is said to be real. Indeed. it is usually said to the only phenomenon that is truly and independently real. It would be the only phenomena that remains uncategorised for Buddhist phenomenology. I feel that Kant helps us understand this idea.

        In short, I think Kant got into a bit of a muddle on this but that he came closer to deducing the facts than the vast majority of ‘non-mystical’ philosophers. I would, however, bow to your expertise on the details of his views.

        Thanks for an interesting discussion.

      • #6 by ausomeawestin on March 17, 2014 - 6:57 pm

        No need to apologize for the flow of comments, I have noticed that this can be a problem during long discussions, but I can certainly follow your argument.

        I think I’m beginning to understand your proposal, but let me make certain. Your idea is that there is an ultimate phenomenon that is the sum total of the phenomenal realm, and the ultimate phenomenon, as ultimate, cannot be experienced, though there can be knowledge of it. But what kind of knowledge can be had of an entity not bounded by the categories? The categories make quantification and qualification possible so what kind of objective and valuable information can be had about the ultimate phenomenon that is not already provided by Kant’s concepts of the noumenal realm?

        It seems to me you have set out on the optimistic project of attaining objective knowledge of the way things really are, and to do so you have posited that the way things really are constitutes the entirely of phenomenal experience thereby entailing that we have epistemic access to ultimate reality in the realm of appearances. This is an interesting move, but I think so long as the ultimate phenomenon is beyond the categories it is beyond knowing in any significant way.

        Thank YOU for this excellent discussion, it has been my pleasure.

        (P.S. I am far from an expert on Kant, but thanks.)

      • #7 by guymax on March 18, 2014 - 5:26 am

        AA – “The categories make quantification and qualification possible so what kind of objective and valuable information can be had about the ultimate phenomenon that is not already provided by Kant’s concepts of the noumenal realm?”

        This is a very difficult question to answer and I’m not going to try a proper answer here. A couple of points though. Kant’s concept of the noumenal realm is not knowledge but a concept and a theory. The knowledge of which I’m speaking is not conceptual or theoretical. The distinction between phenomenal and noumenal would be replaced by a distinction between conventional and ultimate. There would not be a realm to which we have no access. Rather, there would be a realm that we do not usually notice because we usually direct our attention to the input from our senses (inc. our mind) and do not examine who or what is doing the observing. The key point would be that this ultimate phenomenon or state could not be known as something other than ourselves, (where we are the knower and separated from the known), but only as a realisation of identity.

        It is relevant here that the Sufi sage Al-Halaj was crucified not for saying ‘I know the truth’, but for saying ‘I am Truth’. The difference would be completely crucial. In theistic terms we might say that ‘God’ watches every sparrow because by reduction that is what every sparrow is.

        AA – “It seems to me you have set out on the optimistic project of attaining objective knowledge of the way things really are, and to do so you have posited that the way things really are constitutes the entirely of phenomenal experience thereby entailing that we have epistemic access to ultimate reality in the realm of appearances.”

        Yes, something like that. The universe would be a unity, all division would be emergent and all sentient beings would have immediate (one might say unavoidable) access to ultimate reality even while alive if they did but know it. This is roughly what Buddhists mean when they say that all sentient beings have ‘Buddha-nature’.

        I won’t bang on about mysticism here despite the temptation. I was just trying to make the link between Kant, who calculated what was true, and the Buddha and Lao Tsu, who went and found out. They all arrive in a very similar place, which suggests that logic and experience are in full accord. But Kant stopped short, it seems to me, so ends in a slight muddle of theoretical concepts.

        All the same, Kant seems to have a more profound understanding of some of these things than many who come after him, so I’m a fan. It seems an important insight to see that a phenomenon not subject to the categories is required to explain the categories. The resulting cosmological model is described by George Spencer Brown in his book ‘Laws of Form’, where this uncategorised ultimate phenomenon is likened to a blank piece of paper, and the birth of the space-time universe occurs with the making of the first distinction, thus the breaking of a symmetry. I like to think that Kant would have recognised this view as being consistent with his own ideas, but we’ll never know.

      • #8 by ausomeawestin on March 18, 2014 - 11:39 am

        Very interesting, thanks for responding to my questions and for the fascinating discussion. I’m glad we can agree that Kant is a philosopher worthy of being a fan of.

      • #9 by guymax on March 19, 2014 - 7:14 am

        Yes, good chat. Pity Kant isn’t around to take part. Over and out…

  2. #10 by SelfAwarePatterns on March 16, 2014 - 10:12 am

    Thanks for the link and compliment! An interesting post. It reminds me of the divide between scientific instrumentalists and realists. Instrumentalists believe that we only know empirical evidence, sensory impressions, and can’t know the reality behind. Realists believe that empirical evidence does reflect actual reality. Not sure if this is the same divide as noumena and phenomena, but they seem closely related.

    All of this also reminds me about something I’ve pondered from time to time. How do we know something is objective rather than subjective? I’ve started a post on it a couple of times but realized each time that I didn’t have a firm grasp on it yet.

    • #11 by ausomeawestin on March 16, 2014 - 1:54 pm

      An interesting connection! Yes I think instrumentalists seem to be of a Kantian mindset, though, based on your description, not going as far as him in accepting rationalistic knowledge.

      You pose a fascinating and difficult question. Objectivity cannot just be determined by “mind-independence” as it seems there could be objective answers to questions about concepts that are in some sense dependent on minds — morality comes to mind. That everyone, with full knowledge, would come to the same answers, as a “mind-dependence” theory of objectivity, cannot be right either as it still seems that objectivity is more independent of our opinions than this would entail. Is the criterion for objectivity somewhere else along this spectrum are on a different scale entirely?

    • #12 by guymax on March 18, 2014 - 5:50 am

      Nice point sap. There does seems to be a relationship. But it is possible to reject Instrumentalism and realism (defined as you do above) for a third view. This would by my choice.

      And I think you’re right to be confused by the difference between ‘objective’ and ‘subjective’ knowledge, or between our knowledge of ‘subjective’ and ‘objective’ phenomenon. The distinction breaks down at the limit. Hence for Aristotle true knowledge would be identical with its object, and not a relationship between subject and object.

      • #13 by SelfAwarePatterns on March 18, 2014 - 8:53 am

        Thanks guymax. I’m curious what third view you see between instrumentalism and realism. Most scientists are realists, probably because it’s more motivating for them to see what they’re doing as exploring reality. But I think most know that it pays to look at things as an instrumentalist from time to time.

  3. #14 by guymax on March 19, 2014 - 7:27 am

    You say that instrumentalists believe that we only know empirical evidence, sensory impressions, and can’t know the reality behind, while realists believe that empirical evidence does reflect actual reality. I find these long-running two-sided debates interesting because almost invariably there is another option that we must ignore in order to keep arguing.

    The third option would be to say we can know more than sensory impressions, and thus can know the reality behind sensory phenomena, and also that sensory phenomena are emergent, not truly real but contingent, having only a relative existence. This would make both instrumentalism and realism (as defined) wrong. Both would represent a rejection of the perennial philosophy.

    • #15 by SelfAwarePatterns on March 19, 2014 - 1:18 pm

      Thanks for the clarification. I agree that sensory phenomena are emergent (of course I suspect that ultimately everything is). And I think we can also know things innately or by a priori reasoning. Not sure if you mean either of those or something else.

  4. #16 by guymax on March 19, 2014 - 3:52 pm

    More or less that. I’d just leave out a priori reasoning since I’d set the bar higher for knowledge. I’d want to argue that there must be one phenomenon that is not emergent, but would agree that it cannot be a ‘thing’. I like your blog by the way. Great post on conscious intention. It’s better (and shorter, which may not be a coincidence) than anything else I’ve read in the wake of Libet’s results.

  1. Kant’s Critique of Descartes (and time, immediacy, and the external world) | ausomeawestin
  2. Two Solutions re-Causation | The Leather Library
  3. Wedgwood on Moral Knowledge and Moral Epistemology pt. II | ausomeawestin

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: