Kant has shown us what can and what cannot be known by man. What remains to be done, even after Kant, is to show how man came to believe that he could know so much more than he can know, and this will have to be shown by a Critique of Language
"Language," Hamann writes, "is not only the foundation for the whole faculty of thinking, but the central point also from which proceeds the understanding of reason by herself." And again: "The question with me is not, What is Reason? but, What is Language? And here I suspect is the ground of all paralogisms and antinomies with which Reason has been charged." And again: "Hence I feel almost inclined to believe that [our] whole philosophy consists more of language than of reason, and the misunderstanding of numberless words, the prosopopoeias of the most arbitrary abstraction, the antitheses; nay, the commonest figures of speech of the sensus communis have produced a whole world of problems, which can no more be raised than solved. What we want is a Grammar of Reason."A blogger asks:
Is that a fair summation for that time and could we argue that Wittgenstein has now completed the task?If Wittgenstein completed that task, and Müller does not state precisely what he means (he attaches a footnote to the first quote, citing some lectures of his which seem to be extremely difficult to find, especially without the benefit of a subscription to obscure journal databases), Kant himself does not appear to have agreed with Hamann about the source of reason's confusion. In fact, seeking the confusion and error engendered by the inappropriate use of reason in language merely shifts the level of critique; it adds no particularly valuable insight beyond that offered by a critique of reason. The remaining question is this: Why is reason compelled to answer questions that apply to objects outside the proper domain of reason itself?
What Müller probably envisioned was a critique examining not just reason's capacity for making judgments (Kant had already provided that), but a deeper critique going a level below that of reason, examining the ability of human language to connect words in certain ways. The possible ways in which the human mind can coherently connect words to form ideas explains the possible ways that judgments can be formed, so that the power of reason itself depends, as a prior condition for its existence, on the power of language. The rules of language would make up Hamann's "Grammar of Reason" in the same way that the rules of judgment comprise logic. Shifting levels only relocates the proper area to explore and (hopefully) to discover the source of metaphysical error. Further, hoping for a perfect avoidance of all error, and a concomitant rejection of all metaphysical speculation, was neither desirable nor possible for Kant. Metaphysics, while not possible as a science, functions as a storehouse of useful concepts of ideal (though never quite real) perfections.
Language, like reason, is led so easily into error not because of its deficiency, but because of its power. Language must be powerful enough to allow its users to make an essentially infinite number of sentences in order to allow meaningful communication among sophisticated organisms. Increasing levels of abstraction in thought are reflected in increasingly abstract and universal words. The advantage of such abstraction is comprehensiveness of explanation and concision; the disadvantage is vagueness and distance from the concreteness of experience. Language consists in an array of shortcuts and abbreviations (as universal quantification, for instance, stands for an innumerable amount of individual predications), but it often confuses its user into thinking that the shortcuts are not shortcuts at all, not abstractions or artificial limitations on (or extensions of) reality but instead independently significant concepts. This kind of thinking led Plato to reify universals - the abstraction "man" grew from a convenient shorthand for a collection of traits and a name of a class of individuals into a thing in its own right. The Form of Man then, by giving reality to the individual instantiations of the form, became regarded as ontologically prior to individual men, and thus more real than any man ever could be.
As with language, so with reason. Reason, in order to make sense of physical change, assumes an underlying unchangeable material persisting through the changes. Something must be assumed to endure because "Nothing comes from nothing, and nothing perishes into nothing," but change is real; therefore, change must consist in alterations of the characteristics of a permanent something, and not in real creation or destruction. Berkeley rightly criticized this concept as artificial, unnecessary, and dubious, for the simple reason that this material substance could quite literally be shown to have no inherent characteristics. Characteristics are always things the substance acquires, and the unlimited capacity for this featureless substance to be affected and altered was its entire existence. Never hot, never cold, because hot and cold are merely modifications of the permanent substrate, it was the exact center of the onion. Qualities were merely the layers of that onion, and when peeled away, they exposed an utter emptiness at the center. Properly, then, substance was nothing, a convenient fiction but a dangerous dogma.
While Berkeley had his own reasons for denying the reality of material substance, it is not hard to find other mischief perpetrated by the dogmatic acceptance of substance as a real metaphysical object. Substance is permanent; substance is a thing that is never a predicate of anything else, but always the subject. The Cartesians seized upon this property of substance, permanence, the one thing it never lost in the unending flux of physics, to establish the immortality of the soul. Soul is mind; mind is always the subject of thought, never the modifications, never the specific things thought; therefore, the mind (and the soul, since they are identical) is a substance; further, substance is permanent; therefore the mind is immortal. Here is a grand metaphysical claim, and a great comfort to humanity, derived purely from concepts of reason!
A critique of the concept of substance refutes the conclusion that the immortality of the soul can be inferred from a simple law of reason, and it unmasks the subtle leap from thought to reality. Substance is a property of reality insofar as reason understands that reality. The mind must posit a permanent substrate to provide the enduring subject on which we hang the countless predicates experienced in the world. If substance did not exist, then objects could literally wink in and out of existence, without any sort of law governing their creation and destruction; further, any hope of doing such mundane tasks as balancing a chemical equation would be futile. That the mass of smoke and ash produced from the combustion of wood should equal the mass of that wood assumes that the smoke and ash are products of a certain operation upon the wood; if, instead, smoke and ash pop into existence ex nihilo, it is not apparent a priori that they bear any relation to the wood at all. Now attempt to search for that thing that is responsible for the conservation of mass and energy, and you will find it in no experience. It does not come from experience; it is not an inductive generalization from a number of experiences, but it is a concept that gives coherence to experience that otherwise would be rather random. That thing which makes the matter and energy endure in their proper proportion is the idea of substance.
Substance is merely an idea (more properly, following Kant's careful selection of terminology, a concept), not a thing conceived of as being a real object. Substance exists because it is a shortcut for the omnipresent permanent thing that exists in all instances of change. Substance itself is nothing. Things-in-themselves are neither permanent nor transitory, not relating to one another in time at all. Berkeley is right - substance is not a metaphysical object - but he is also wrong - substance is as real a concept as any other we have. It simply is not a concept applicable to things-in-themselves.
If substance is not an object of metaphysics, why was anyone ever led to believe it was? Two possible explanations come to mind. First, reason was used carelessly by those rationalists who elevated substance to such a status. If those people had only thought more clearly and been more precise in their inferences, they would have avoided error. A Critique of Language would cure the corresponding linguistic misuse by dissecting words and by showing the proper and improper ways of combining those words into meaningful sentences. Just as "The soul is substance" is a nonsense sentence that confuses a concept of experience with a property of things-in-themselves, it is a linguistically deficient sentence that uses an ambiguous connector to relate two ill-defined terms. If only "soul," "is," and "substance" were properly understood, and their relations clarified, we could avoid misusing them like that.
Kant himself had a different explanation in mind for erroneous metaphysical speculation, and in fact viewed metaphysics as absolutely indispensable for reason:
Whatever has its basis in the nature of our powers must be purposive and be accordant with their correct use - if only we can prevent a certain misunderstanding and thus can discover these powers' proper direction. Hence presumably the transcendental ideas will have their good and consequently immanent use, although when their signification is misunderstood and they are taken to be concepts of actual things, they can be transcendent in their application and can on that very account be deceptive.What is the good use to which these ideas can be put? The ideas set up models of reality (of the totality of reality) that allow the whole world to be understood, even if not in concrete detail (because our lives are far too short and our physical limitations too great to comprehend every feature of every event through the entire history of the universe), at least as a coherent, rational whole. The mind itself can be understood, not as a metaphysical object actually known through pure concepts, but as a condition of all possible experience that must be assumed to be permanent, numerically one, &c. for the myriad aspects of experience to be unified in one perceiving and cognizing subject. Finally, the idea of a supremely perfect, necessary Being allows us to assume a world created intelligently and imbued with purpose.
The transcendental ideas allow us to approach reality as if the "I" were a Cartesian ego, as if the world were a whole whose parts and thoroughgoing relations were at least capable of being utterly known by us, as if an all-powerful, perfect God created the world and everything in it. That is how man came to believe he could know more than he could know; not because he can know everything (such is the flawed ideal of the old metaphysics) but because he can and must believe more things than he can affirmatively know. Error creeps in when the things naturally believed are taken to be things known in perfect detail, in themselves.
Someone ought at this point to object that Kant's reason for recognizing value in the transcendental ideas may not be valid. That Kant said this or that is a good answer for addressing a matter of textual interpretation, but it resolves the philosophical problem only if we agree with Kant (and, incidentally, all this implies that Müller either did not regard Kant's reason as satisfactory [probable] or did not even understand what Kant's reason was [unlikely for someone who had to have translated and, consequently, read the entire Critique of Pure Reason]). An alternative view, entirely consistent with the Critique's skepticism about metaphysical speculation, is to reside comfortably in Humean doubt. Comfortably not in the sense that such skepticism is easy to handle, because, after all, following Hume would require us to reduce our importance as selves virtually to nothing by admitting that the self is nothing more than a locus of ideas. The soul, far from being the most important thing in the world, would be an empty place where some things happened to occur, but which itself could serve equally well as the vacuum it is. Lacking souls entirely, we should not be surprised to learn that immortality is impossible, for if all that exists is material, then all that exists is perishable; the empty "I" persists in exactly the way space exists; having no reality and thus no existence in the first place, it will keep on being nothing indefinitely. God being a concept not derived from matters of fact and not posited as a mere idea must be a nonsense concept with no real object, and besides that, the idea that one perfect Being exists as the cause for a thoroughly imperfect world is an untenable strain on logic at best. What is comforting in all this agnosticism is the awareness that these confusing, metaphysical things cannot be known by us, so that even if they did exist, we could never know. We can stop short a painful, fruitless journey by answering reason's questions with categorical denial. This denial leaves more time for living, less time for worrying about fairy tales.
Certainly skepticism, whether dogmatic (denying that metaphysical objects exist at all) or Pyrrhonean (neither affirming nor denying metaphysical claims), is a solution to the Transcendental Dialectic. It is unsatisfactory insofar as it denies the value of faith in the transcendental ideas when concrete knowledge is not possible. There are really three issues here relating to the use of transcendental ideas which critical philosophy itself recognizes can never be known scientifically. First, the ideas may encourage good behavior and thus have value for a healthy moral life. Second, the ideas may provide rough outlines of at least some of the properties of the metaphysical existences (the noumena) the illusions of which they are (even a mirage discloses something). Third, the ideas may not give knowledge themselves but may suggest connections among phenomena that disclose deeper truths about those phenomena that reason, unaided by the ideas, would not be capable of discovering. The first and third potential beneficial effects of the ideas have textual support in Kant; the second does not seem to have such support. Whatever support these interpretations have, they must stand up to our critique and the critique of all philosophy if they are to have anything other than historical value.
On to the first purported benefit. Objects in experience cannot act freely. Every event in nature is the effect of a prior cause in nature. Ethics seems to have no application to these occurrences. Because human actions are phenomenal, they must be the result of the operation of physical laws. Telling physical law to stop so that we can do the right thing is impossible. However, if the self is an immortal soul created by God, and which has spontaneous causal efficacy, then ethics is possible in the first place. The noumenal self must be capable of acting freely, and must be the metaphysical "cause" of the actions of the empirical self, if moral responsibility can be reconciled with natural causality. The soul must be a person, identical through time, in order for it to be morally responsible. If the soul now differs from the soul a second ago, then the soul now cannot be responsible for what the previous soul did. So far, so good.
Whether the assumption of the soul's immortality is really necessary for the moral life is trickier. Kant certainly thought so. He saw it as unbearably unjust that sinners in this life are often happy and saints often miserable. Because this result is incompatible with justice, the soul must be capable of receiving punishment or reward after the death of the body. Because that punishment or reward can only come from an infinitely wise and just lawgiver, God must fill that role. If the "right" thing is whatever produces happiness, though, then no injustice exists. Only if correct moral conduct is based on something other than pleasure or happiness is a happy sinner a troubling possibility. Further, because the transcendental idea of immortality cannot be known, perhaps the person who rejects it as if it were false is much better off than the one who accepts it as if it were true, because he can do as he wishes without fear of future punishment. The transcendental idea is a comfort only if we hold to a system of morality that does not equate virtuous conduct with conduct that will produce happiness. If we were utterly convinced that such a morality must be true on other grounds (as Kant did, though a discussion of that will take us too far afield in this post), then the idea of immortality is a comfort and seems to restore justice.
Another benefit of an assumption of immortality of the soul is it allows us to obey a command otherwise impossible: to be perfect. At any instant, a human cannot be perfect. If, however, he improves himself daily, by gaining knowledge and by purifying his will, he will constantly grow closer to perfection. Asymptotically approaching perfection, he will be perfect at infinity; that is, if the line of his life continues forever, he will eventually attain perfection.
These benefits for the moral life are contingent as benefits on our adopting a certain ethical outlook. Kant found this outlook to be the only correct one and correspondingly needed to assume the transcendental ideas in order to construct a reality where that outlook was consistent with the facts. Even those who disagree with Kantian morality should understand the very real benefit transcendental can bestow in ethics: if an ethical theory is held to be true, but for some reason assumes the existence of things which are not possible objects of experience, transcendental ideas can fill in those gaps. Obviously, the theory must rest on firm ground prior to having unprovable ideals posited as bases for the reality of the theory, and transcendental ideas cannot be used to assume away other difficulties. To take just one example of the positive use of such ideas, what would utilitarianism be without a moral calculus? Yet that such a calculus is possible only upon certain nonempirical assumptions about pleasure. Possibly ethics cannot exist without some use of the ideas.
Second potential benefit: If phenomena are the result of the application of forms of thought to noumena, then surely noumena cannot be completely beyond our grasp. On the one hand, we can restrict the possible properties of noumena by noting what the mind adds to experience, then negating all those things of noumena. If our minds impose space upon objects, then noumena must not exist in space; and if time is a form of inner intuition, not applicable to things as they are in themselves, then time must not apply to noumena. This negative definition of a noumenal reality is not knowledge of noumena itself, but it can guide metaphysics by closing off certain avenues of inquiry and trimming the maze of possible paths in order to simplify the search.
On the other hand, that noumena are capable of being understood in some way indicates that they in themselves have some property that is expressible in terms we understand. That noumena can be organized temporally, even though such organization transforms them into something other than they are in themselves, does indicate at least that they are susceptible of existing in time. Further, it is not quite clear what application of intuition and concepts to reality does to change them. Perhaps space, time, and the categories are like silhouettes obstructing a light source. The light emanating from the source is the noumenon; the light that isn't obstructed by the silhouettes is the phenomenon. Thus the phenomenon is exactly like the noumenon in the parts that we are capable of detecting, and the only change to it is that some parts are beyond our possible comprehension. Similarly, some frequencies of electromagnetic radiation are not visible, though a part of the spectrum is. Another alternative way of explaining the change noumena undergo is analogous to, say, a device that converts the light emitted by the source in the previous analogy into sound. Each frequency of light is replicated by a certain frequency of sound, and the intensity of the light is reproduced as loudness of the sound. Here, unlike in the previous analogy, nothing of the noumenon remains; no light at all reaches us, rather than some. But all the light has been converted into sound, and all of that reaches us. So all of the light reaches us in some form, although, as before, the exact nature of the noumenon itself is not revealed.
This line of reasoning unfortunately will not bear fruit. It seems that noumena must at least be capable of producing phenomena such as we actually experience. There must be something to the noumenon that allows it to appear as this phenomenon and not as another. However, that something in the noumenon must not be identical to the something in the phenomenon. The noumenon must have some property that allows it to appear as spatially extended in experience, but without actually being spatially extended itself, because extension is an aspect of things provided by reason. The noumenon must have something like space that isn't space. What that something is is entirely beyond our ability to know. It seems the transcendental ideas do not actually allow knowledge of noumena, except in a negative sense, by listing all the things they are not.
The third, final possible benefit of transcendental ideas: The idea that the self is actually identical throughout cognition, not merely observed as something that has not yet become something different in all experience we have so far had, is probably indispensible even if it cannot ever be known to be true. The expectation that "I" will not suddenly become another person, or split into several persons, or stop existing while still having thoughts, strengthens the connection of ideas in the mind without allowing strict cognition of any connection. Thus Hume is correct: an identical self is never an object of experience, and can never be an object of experience. This hole in experience is filled by the idea of personhood, the Cartesian ego, which, although it cannot be an object of experience nor derived from experience (such derivation being merely inductive, if possible at all, and thus not capable of giving any idea of absolute permanence), we rely on more than anything we actually perceive, no matter with what vividness and clarity. Although "I think" is an empirically known truth, something more than merely "There is a thing that is thinking" is stated by it. In the Analytic, regarding the transcendental unity of apperception, Kant says:
The I think must be capable of accompanying all my presentations.while in the Paralogisms, he claims "I think" is empirical. Whether or not he recognized a tension between these views (views in the same book, no less), the difference between "There is a thing thinking" and "I think" should provide a clue to the value of idea of personhood. "I" must be a unity for experience to be united spatially (seen as a continuum of objects and not merely as completely discrete things incapable of interacting with each other) and temporally (so that what the ego perceives now and what it perceives a second later are known to be succeeding intuitive "snapshots" of the same ego). This unity is implicit in all judgment, in all experience generally, and is not identical to space, time, or the categories. It exists prior to them, allowing them to form a coherent experience by unifying all the parts of space, all the parts of time, and all the conceptually-known objects and events into one thing, experience generally. What that experience is of is known only as "I".
The idea of the world as a whole allows us speculatively to go beyond any concrete thing experienced and beyond all possible experience in order to consider the whole universe, the whole history of time (from its beginning, if any exists, to its end, if that exists), and the smallest increments of space and time. This has had documented benefits throughout the history of science. Observing the relatively gross effects of, say, radiation on an object, one can speculate that the minute parts of it must be constituted just so in order to have such large effects. Without microscopic observation, it cannot be known what the extremely small particles making up the thing are like. When such observation is impossible (because technology has not advanced to allow such fine observation, or because some sort of unavoidable limit to direct observation has been reached), something can be known that is not experienced. In the case of the unavoidable limit (a limit imposed by the universe itself and not by our own deficiency), the thing known is not a possible object of experience.
To be sure, in the case of indirect observation of minute phenomena, what is known is not precisely the unknowable thing but its effect on things that are knowable. Ultimately, it seems, what we know are our own states, and not any truly existing independent things at all. The Refutation of Idealism has a unique solution to that difficulty:
I am conscious of my existence as determined in time. All time determination presupposes something permanent in perception. But this permanent something cannot be something within me, precisely because my existence can be determined in time only by this permanent something. Therefore perception of this permanent something is possible only through a thing outside me and not through mere presentation of a thing outside me.Though found in the Transcendental Analytic, this argument is right on target about the transcendental ideas. Real, metaphysically distinct things outside myself must exist because the sheer fact of having experience at all presupposes their existence, not merely as objects in experience but as real, independent things. This kind of inference from what is known (inner experience) to what must exist for that known thing to exist itself (a metaphysically distinct world) goes on literally every second of human existence. That it should occur in speculation about the origins of the universe, about time, about the smallest particles possible, &c. is not surprising at all.
The skeptic will ask, "Doesn't this kind of thinking lead to credulity? You say that we can know things we can never directly observe, even though it is that observation which is supposed to bring knowledge. Why not assume magical forces, psychic powers, or any such nonsense?" The use of the transcendental ideas must be subject to limiting conditions, of course. Precisely because they go beyond all possible experience, they must be used carefully and subject always to what can determinately be known. If ideas contradict experience, experience must always win out. If ideas are consistent with actual experience, they have found their proper role, because they can expand knowledge that already exists by suggesting possible connections that experience alone could not find. If, however, ideas are consistent merely with possible experience, then their operation is likely to lead merely to confusion and error. The ideas are not to be invoked in the place of rigorous observation and analysis. They simply urge us to consider everything in the world as related to everything else in a way that reason could discover, if it had infinite time and resources to consider all of universal history as a single object.
If the Supreme Being is the cause of the world, then the world can be expected to be imbued with purpose. This purposiveness is not weak, a mere application of teleological terms to descriptions of events known to occur solely due to mechanical laws of nature (the stone's purpose was to fall to the ground) but strong (the heart's purpose is to pump blood throughout the body). However, it can't be strong enough to contradict reason's laws, which interpret phenomena as interacting according to mechanical laws. Instead, purposiveness's place is in judgments about groups of phenomena, about their relations over time, and about reciprocal interaction. A biological organism exhibits best the utility of regarding the world, and parts of that world, as driven by purpose. The answer "Because its purpose is to carry oxygen to and carbon dioxide away from cells" to the question "Why does blood flow?" is correct in a way that transcends phenomena. The phenomena themselves, and the laws of physics governing them, contain no data whatsoever about purpose. From the fluidity of blood, its chemical composition, its pressure, its temperature, and the motion induced in it by the heart, one never finds attached the idea that it is providing support to a living thing. The mechanical account cannot explain why blood flows instead of doing nothing at all. Explaining the function of the blood as a process in an organism, whereby it tends to sustain the continued arrangement of that organism instead of its dissolution into dead tissue (essentially mere matter), and whereby it receives sustenance itself from the body's other organs, allows biology and anatomy to reach greater insight and thus accumulate more knowledge about reality. Where the purpose of an organ is not yet known, theories about its function operate on the premise that it must have such a function. No one would consider it plausible that an organ is "just there" in the same way that someone observing a particular grain of sand dismisses any teleological explanation of its current position.
Even so, the idea of purpose is not real purpose. When breaking down the components of the human body, or of any creature, we do not expect to find a purpose organ, purpose fluid, or the purpose system. Purpose is not a real thing that will be experienced, but merely a convenient organization method for the mind. As in all cases of ideas that are not universal laws of nature, the principle of purpose can give rise to illusion. The illusion in the case of biology is what Aristotle called the soul and what we might call "living force." Living force, as a metaphysically distinct object of experience, is beyond our capacity to understand. But considering reality as if certain things were imbued with such a force, and understanding the behaviors of those things as effects of the interplay of physical laws and living force, allows biology to break free from being a mere branch of physics and allows it to produce meaningful results.
A fine line exists between believing we know more than we can know and between assuming things we could not know in order to better understand what we do know. If Müller wished to eliminate the Transcendental Dialectic, then he wished to destroy a useful tool. The power to imagine things that do not exist can sharpen our knowledge of the things that do exist. If a Critique of Language would shrink the expressive power of language in order to tiptoe around metaphysical bogeymen, then I hope it never will be done.