Faith as Kant's Key
to Justifying the Transcendental Perspective
by Stephen Palmquist (stevepq@hkbu.edu.hk)
A purely rational
belief is ... the signpost or compass by which the speculative thinker can
orient himself in his rational excursions in the field of supersensuous
objects. But to the man of ordinary but (morally) sound reason, it can show the
way for both the theoretical and the practical standpoint, in a manner
entirely suitable to the end to which he is destined. This rational belief must
also be made the basis of every other belief—indeed of every revelation.
[Kt20:142]
1.
Faith and Kant's Transcendental Turn
Kant
is sometimes reproved for not having devoted more attention to the question of
how Transcendental philosophy itself is possible [e.g., W9:249-55]. To
compensate for this neglect, Kantian philosophers tend to focus on one or more
of three ways of justifying his Transcendental Perspective[1] (and
particularly his infamous presupposition of the 'thing in itself'[2]). Interpreters
have stressed the role it plays: first, in the overall System to which it gives
rise; second, in Kant's special theory of transcendental idealism; or third,
in relation to his transcendental arguments. Part Three of this book will
provide a version of the first type of justification; a version of the second
type will be given in VI.2; and Appendix V considers the possibility of the
third type. However appealing Kant's theories may or may not turn out to be
once an interpretation of his System brings out their consistency and
coherence (and perhaps even their validity), we must concede that such an
approach on its own could not provide an entirely conclusive justification for
Kant's 'transcendental turn'. In this chapter, therefore, I will touch upon
each of these methods, but only in order to demonstrate their secondary
importance to Kant's own
justification for assuming an unknowable reality as his epistemological
starting point. Though almost always neglected by interpreters, this fourth method, I will argue, is his
ultimate and indispensable key to justifying the Transcendental Perspective,
so it accounts for his failure to justify it explicitly in other ways.
In
Kt2:371 Kant suggests it is occasionally necessary for the philosopher to adopt
'a rational faith which alone may be possible for us, sufficient to our wants,
and perhaps even more salutary than knowledge itself.' Although this would
appear at first sight to be precarious ground on which to build an epistemological
foundation for a philosophical system [cf. S17:241], Kant seems to treat it not
only as sufficient, but as the necessary support for both his theoretical and
his practical standpoints [see e.g., Kt20:142q.a.]. Establishing such an
approach as philosophically valid will not only provide an initial justification
of his starting point, but will also play an important role in Pq20, where I
will explore the theological implications of his System [s.a. Ch. X]. The
purpose of this chapter, therefore, will be to demonstrate that Kant does in
fact appeal to faith as the justifying key to his Transcendental Perspective,
and to defend its legitimacy in this role.
Before
discussing the implications of this rather unconventional claim, we must
understand what Kant means by the word Glaube
(translated as either 'faith' or 'belief'). Schrader may be right in saying
'the term faith is not used in any conventional sense' by Kant [S4:183]; but
this does not justify his charge that it therefore 'serves to confuse the
function of the critical method as applied to moral experience' [183]. For if
we make an effort to understand Kant's own special use of the word and apply it
consistently, then it can serve instead to clarify
the implications of his entire Critical philosophy.
Most
commentators limit the application of all Kant's comments on faith to his moral philosophy, ignoring or denying
their application to his epistemology
[e.g., W16:109]. For instance, his claim to 'deny knowledge [Wissen], in
order to make room for faith'
[Kt1:xxx] is often taken out of its wider context and regarded solely as an
allusion to the way the denials of the Dialectic in Kt1 make room for the faith
of practical reason in Kt4. Kemp Smith's translation perpetrates this
misconception by dividing Kant's one long paragraph at Kt1:xxiv-xxxi into five
paragraphs. Kant's immediate point in making this assertion is indeed to stress
the need for faith in one's practical
employment of the ideas of 'God, freedom, and immortality'; but it is rarely
acknowledged that his general argument in this passage assumes that the same
'attitude of reason' must also hold true in systemt for the thing in
itself.[3] Just as he
regards the faith of practical reason as the legitimate response to our lack
of knowledge concerning God, freedom, and immortality, so also he regards the
faith of theoretical reason as the legitimate response to our lack of knowledge
concerning the thing in itself. The two cases are directly parallel; in fact,
Kant's overall purpose in the paragraph in which the statement occurs is to
demonstrate their similarity, and in so doing to establish the requirement of
faith in the thing in itself for both epistemological and moral reflection.
Admittedly,
the great majority of Kant's references to faith do concern its role in his
practical, rather than in his theoretical, system; and in a few cases he even
seems to deny the validity of applying it to the latter [see e.g., Kt7:
472,475]. The least ambiguous of these apparent denials comes in Kt10:69n
(76-7n): 'The belief of reason ... can never be directed to theoretical
cognition, for there the objectively inadequate holding-to-be-true is mere opinion. It [i.e., faith] is merely a
presupposition of reason from a subjective,
though absolutely necessary practical standpoint.' However, this and all such
statements are not intended to deny the need for rational faith in the thing
in itself; for such faith, though indeed theoretical,
is not intended as a 'theoretical cognition'
of an 'object' [s.e. Kt23:396-7n(173-4n)]. Only when we aim at reaching
empirical knowledge (i.e., only within the limits of systemt) does an appeal
to faith lead to 'mere opinion'. By contrast, the thing in itself is precisely
that sort of subjective 'presupposition of reason' which Kant is here describing
as being supported on the pillar of rational faith.[4] Once this is
fully understood, it becomes unreasonable to regard the thing in itself as a
'wild card', slyly included in Kant's epistemology solely 'for the purpose of
his ethics' [W9:159-60; s.a. 166; M6:144]; for, even though he does not always
state his position explicitly, Kant uses faith in this rational presupposition
to form the very ground in which his System of Perspectives is planted.
Kant
describes faith as 'the moral attitude of reason in its assurance of what is
beyond the reach of theoretical knowledge' [Kt7:471; s.a. 469]. He is most
interested in such 'realities which transcend experience' [L3:44] when they
have their primary positive function in systemp, and therefore
inspire such a 'moral attitude'; nevertheless, he admits that 'even in purely
theoretical judgments there is an analogon
of practical judgments, to the mental entertaining of which the term
"belief" is appropriate' [Kt1:853]. Moreover, Kant says even a belief
in a transcendent reality such as God depends not just on practical faith, but also on 'pure rational faith, because pure reason alone (in its theoretical as
well as practical standpoint) is the source from which it springs' [Kt4:126].
In
both practical and theoretical contexts, faith's 'assurance' can result from
either ordinary 'contingent belief' or special 'necessary belief'. The latter,
strictly transcendental kind of belief, can apply only to situations in which
'I know with certainty [ich gewiss weiss]
that no one can have knowledge [kennen
könne] of any other conditions which lead to the proposed end.'[5] Such
transcendental faith does not entitle us to 'speculative knowledge' [Wissen] concerning the conditions in
question [Kt2:371]; instead, it means the conditions are regarded as sufficient
'only subjectively' [Kt1:850]. For 'rational belief ... is a subjectively
sufficient assent associated with the consciousness that it is an objectively
insufficient assent; therefore ... [it] can never be converted by all the
natural data of reason into knowledge, because the ground of assent in this
case is merely subjective' [Kt20:140-1]. Accordingly, 'the expression of
belief is, from the objective
standpoint, an expression of modesty'—i.e., of an awareness of our
ignorance—yet it is 'at the same time, from the subjective standpoint, an expression of the firmness of our confidence'
[Kt1:855; s.a. 498-9; Kt2:278]. In spite of our ignorance, therefore, the
'conviction' which results 'is often firmer than any knowing. In knowing one
still listens to counter-reasons, but not in belief, because this turns not on
objective grounds but on the moral interest of the subject' [Kt10:72(80)].
Ewing aptly
observes that 'faith does not mean for Kant belief on authority or belief
without ground, but believing what we have adequate grounds for believing but
cannot absolutely prove' [E5:9]. Along these lines, Kant notes: 'Belief is no
special source of cognition. It is a kind of holding-to-be-true with
consciousness of its incompleteness' [Kt10:67n(75n)]. As such, faith always
involves the formation of something like a hypothesis.[6] Kant defines the
latter as: 'The holding-to-be-true of a presupposition as a ground' [84 (92)].
In working with a hypothesis 'we conclude from the truth of the consequence
to the truth of the ground'; and although 'hypotheses always remain hypotheses,
that is, presuppositions whose complete certainty we can never attain',
nevertheless 'the more consequences can be derived from a hypothesis, the
more probable it is' [85(92-3)]. Kant says metaphysics
itself does 'not permit of hypotheses' [86(93)]. But this does not prevent
him from adopting faith in the thing in itself as the starting point of the Critical philosophy; on the contrary, as we shall see, this metacritical
hypothesis enables him to avoid depending on such hypotheses at crucial points
within his three Critical systems
[see Kt20:141-2]. Moreover, the systematic consequences of presupposing the
thing in itself [see Part Three] lend a measure of 'objective validity' [see
note VI.22] to this hypothesis, which initially is supported only by the
subjective validity of faith. Thus theoretical as well as practical faith 'lies
at the foundation of the critical philosophy' [W25:20], and is, in fact, 'the
outlook, the Weltanschauung of the
critical philosophy itself' [W24:249].
In
the remainder of this chapter I will attempt to defend this interpretation of
the role of faith in Kant's philosophy by relating it to three specific issues.
First, I will elaborate on what it means to say the thing in itself must be
accepted on faith [V.2]. I will then contrast this approach with that which
regards transcendental arguments as the only valid method of justifying this assumption
[V.3]. And finally, assuming my interpretation is correct, I will consider the
relationship between theoretical faith in the thing in itself and moral faith
in the ideas of practical reason [V.4]. (In Appendix V I will examine the
legitimacy of the claim made by some Kant-scholars that the thing in itself
could in fact be viewed as knowable without destroying Kant's basic program.)
If I am successful, I will have demonstrated not that Kant believed he had
established his System as undoubtedly
'true', but rather that he knew the truth of Transcendental philosophy can be
fully recognized only by those who choose to follow his initial leap of faith.
2.
Filling the Transcendent 'Space'
In
discussing the concept of 'noumenon'—a term related to, but not to be equated
with, 'thing in itself' [see VI.3]—Kant says:
The relation of sensibility
to an object and what the transcendental ground of this unity [i.e., what the
thing in itself] may be, are matters undoubtedly so deeply concealed that we
... can never be justified in treating sensibility as being a suitable
instrument of investigation for discovering anything save always still other
appearances—eager as we yet are to explore their non-sensible cause. [Kt1:334]
He concludes on this basis that the only
legitimate function of this 'empty' concept is negative: 'to mark the limits of
our sensible knowledge and to leave open a space which we can fill neither
through possible experience nor through pure understanding' [345; cf. A10:46].
Kant
does not explain at this point whether (and if so, how) this 'space' can be regarded by us as filled in any
way. But he leaves no doubt about his position when he states that even in
systemt 'the supersensible was not mere fancy and ... its concepts were not
empty' [Kt4:6]. Furthermore, he says in K2:12.222(Z1:247): 'The sphere of
non-sensible objects is ... not quite empty, though from the standpoint of
theoretical knowledge such objects must be viewed as transcendent.' Since, as I
will argue in VI.3, 'noumenon' is for the empirical perspective what the
'non-sensible cause' of appearances [Kt1:334] is for the transcendental
perspective, Kant must regard this space as being 'filled' in some sense by the
thing in itself. For us the thing in itself is unknowable by definition [see
Ap. V], so it would not fill the space beyond our sensibility with an actual
noumenal object, or with any other knowledge of 'the relation of
sensibility to an object' [334]; it would merely provide us with a label for
the belief that there is a
supersensible 'ground' of the world as we know it. The only problem this raises
is, of course, how to justify such a belief.
Allison
defends a view of the thing in itself in A10:49 which might seem to render the
above interpretation superfluous:
Properly construed, this
problem is not, as is generally believed, that of somehow justifying the
assumption of the existence of unknowable entities. Rather, it is to show the
possibility and the significance of ... considering thing [sic; i.e., empirical objects]
as they are in themselves, which
means as they are apart from the conditions under which we can know anything
about them.
On the whole, Allison's interpretation is a
good one; for Kant is indeed primarily concerned with investigating the
various perspectives from which 'things' (i.e., objects) can be considered. Consequently, a good deal of what he
says about the thing in itself can be interpreted as referring to how the concept of 'things apart from our
knowledge of them' can be analyzed. But this, surely, is not what Kant has in
mind when he talks about the thing in itself as the 'cause' of appearances [as
in Kt1:334]. On the contrary, he is regarding the thing in itself as that
which fills the transcendent space
which is for us unknowable. To regard it in this way is to hypostatize its concept, so that its existence (in some transcendent, but
analogously categorial sense of the word) must be presupposed [cf. 855]. And
any talk which even alludes to such hypostatization of unknowable concepts does require some justification.
Kant
maintains that 'the modest language of a rational belief [vernünftigen Glaubens]' enables us to 'assume ... the existence of something possible and even
indispensable for the guidance of the understanding
and of the will in life'
[Kt2:278e.a.]. The assumption which guides 'the will', or 'practical reason'
[Kt4:15], concerns the existence of
what for theoretical reason are merely 'ideas' (especially God, freedom and
immortality). Likewise, the assumption which guides 'the understanding', or
theoretical reason, concerns the existence
of what would otherwise be a mere concept (viz., the thing in itself). Kant's
treatment of practical reason and theoretical reason is in many respects
parallel, not only because he regards the form of both as arising out of their mutual
dependence on architectonic logic [see III.3-4], but also because their
justification is essentially the same: even if sensibility cannot intuit as an
appearance something which understanding can think as a concept, its
employment in a philosophical system is not forbidden as long as reason
(theoretical or practical) can present itself with sufficient evidence to
assume its existence on faith. Without such evidence, Kant's faith would not be
'rational belief'.
To
adopt faith in the thing in itself is not so much to posit it as a separate
'entity', as to accept that the objects which fill the world of our experience
have some real nature 'in
themselves': it is to believe they are not mere
appearances, for there is something about them which, in principle, transcends
our knowledge. Findlay is largely correct, therefore, to say Kant
plainly does not regard
[this assumption] as involving any difficult, problematic step: it is rather
... a primordial certainty which stands in no need of justification, even if
its lack of empirical content means that it can never be ranked as
knowledge.... [Thus] we cannot but conceive of, and believe in, such
non-apparent objects or aspects of objects, even though we can have no
knowledge of them. [F3:2]
But he errs in saying the thing in itself
'stands in no need of justification'; for Kant does not regard it merely as an
unsupported 'extra' which can be accepted or rejected, and which is peripheral
to the main purposes of Critical philosophy [cf. S17:38-42]. On the contrary,
his entire System is an attempt to provide good
reasons for assuming that the objects of our experience are appearances
which are grounded in the unknowable thing in itself. Yet faith is still
required as the ultimate justification of his presupposition because, as we
shall see in V.3, it is impossible ever to prove
the objective validity of the existence of the thing in itself. But before
examining why this is so, I will touch upon one of Kant's most important
reasons for having faith in the thing in itself.
If
philosophers adopt faith in the thing in itself at the outset of their transcendental inquiries, then they will
find it is not necessary to appeal to faith to justify their empirical inquiries. Kant states this
rather obscurely in Kt1:xxxixn:
it still remains a scandal
to philosophy and to human reason in general that the existence of things
outside us (from which we derive the whole material of our knowledge, even of
our inner sense) must be accepted merely on faith,
and that if anyone thinks good to doubt their existence, we are unable to
counter his doubts by any satisfactory proof.
This passage might at first sight appear to
contradict the claim that faith is Kant's key to justifying his assumption of
the thing in itself [cf. S18:117]. It would do so if 'things outside us'
referred to the thing in itself as regarded from the transcendental perspective [see e.g., S13:3.35]. But this could not
be Kant's intention, since he goes on to give a brief summary [Kt1:xxxixn-xlin]
of the proof he supplies in the Refutation of Idealism section of the Analytic
of Principles, where the third (empirical) stage of systemt is being
elaborated. His empirical idealist opponent doubts the real existence of empirical objects —i.e., objects of
experience, or, phenomena [see VI.3]. So the 'things outside us' in this
context must be phenomena. Moreover,
Kant would never claim to have provided a proof
for the existence of the thing in itself, since he consistently regards it as
unknowable [see Ap. V].
Far
from contradicting my interpretation, Kant's statement confirms it by implying
that philosophers can avoid the 'scandal' of having to take refuge in faith to
justify the existence of phenomena (i.e., 'things outside us' from the
empirical perspective) [as in H21:1.3.7] only by having faith in the thing in
itself (i.e., in the reality of 'things outside us' from the transcendental perspective).
In other words, philosophers will be unable to prove the objective reality of
empirical objects unless they presuppose that these objects are phenomena
which, when regarded transcendentally, are appearances of the unknowable thing
in itself. If they do muster up enough faith to make such a presupposition, the
way is open for a completely adequate solution to this long-standing
philosophical problem. The solution is that outer objects are unknowable from
the transcendental perspective, yet from the empirical perspective 'inner
experience itself [the real existence of which the empirical idealist does not
doubt] depends upon something permanent which is not in me, and consequently
can be only something outside me' [Kt1:xln-xlin]—namely, phenomena. Not only
can phenomena be proved to exist, they can also be regarded as (in principle)
completely knowable by the scientist [see VI.3]. But philosophers who cannot
bring themselves to take the initial step of faith in an unknowable thing in
itself will inevitably confuse the transcendental and empirical perspectives,
and, taking the objects of experience to be things in themselves, will be
unable to prove their existence.
Kant
would surely be disappointed to find that his solution to this 'scandal' has
not prevented it from continuing to spread—indeed, to thrive—in some
philosophical circles to this day. A good example is Quine's declaration that
'in point of epistemological footing the physical objects and the gods differ
only in degree and not in kind' [Q1:41]. He bases this on the rather more
plausible supposition that '[t]he totality of our so-called knowledge and
beliefs ... is a man-made fabric which impinges on experience only along the
edges' [39; cf. IV.4]. But the 'man-made' nature of all knowledge does not
preclude the philosopher from discerning different kinds of knowledge or belief; it does not require him to reduce
all knowledge to mere 'cultural posits', as Quine suggests [41]. Quine is
unable to make any strict distinction between knowledge of 'physical objects'
and belief in the 'gods' for the same reason which led Berkeley to his empirical
idealism: both philosophers confuse the transcendental with the empirical
perspective and consequently fail to distinguish between faith in the thing in
itself and faith in matters of empirical interest.
The
reason philosophers such as Berkeley and Quine refuse to take advantage of
Kant's transcendental faith is obviously that they assume a fundamentally
different approach. But why some philosophers, on the other hand, shun the need
for faith in the thing in itself even though they do accept the bulk of Kant's
System is not quite so evident. Probably the most common example of the latter
is found among those Kantians who attempt to pick up where Kant left off in his
use of 'transcendental arguments' (i.e., proofs by reference to the possibility
of experience [cf. note IV.26]). To the alternative offered by such
philosophers, therefore, we shall now turn our attention.
3.
Transcendental Arguments or a Concession to the Skeptic?
Kant
himself never uses transcendental arguments on behalf of the thing in itself;
instead, he limits their use to determining what must be presupposed to be true
about the phenomenal world in order
for our experience of it to be possible [e.g., Kt1:38,A402]. Paton thus has
good reasons for suggesting that the validity of presupposing the thing in itself
'is not considered by Kant to be in need of proof', and that such a proof would
have seemed 'ludicrous' to him [P2:1.70; s.a. M8:56; I1:756]. But this is not
entirely fair, since Kant does occasionally supply rather brief arguments for
his presupposition, such as that if experience has bounds, then 'that which
binds it must lie quite without it' [Kt2:360]. Such comments do provide a
possible basis for attempting on Kantian grounds to construct a transcendental
argument for the necessary existence of the thing in itself.
Chipman
constructs such an argument by paying close attention to Kant's own statements.
He gathers together a number of fragmented bits of reasoning and forms them
into a single, surprisingly complete, transcendental argument:
(a) The objects of
experience exist only as spatio-temporal appearances,
(b) Insofar as we judge that
there are appearances, we must be prepared to judge that there are things
appearing,
(c) To speak of a thing
appearing is, by implication, to make use of the concept of a
thing-other-than-as-it-appears,
Ergo,
(d) Objective experience
requires us to postulate things-other-than-as-they-appear, or things in
themselves. [C8:491]
Although Chipman thinks Kant's theory of the
thing in itself is 'consistent', he admits that even this reconstructed
argument remains 'inconclusive' [489]: 'Kant has not succeeded in establishing
that we must postulate things in themselves as a necessary condition for the
possibility of objective experience, but ... their possibility must
nevertheless be countenanced' [498].
Walker
attempts to compensate for the inconclusive nature of such reconstructed forms
of Kant's position by straying rather further from Kant's own methods. He
maintains that, once we recognize that 'Kant was simply mistaken' to think the
thing in itself must be completely unknowable, the way will be open to
construct transcendental arguments which can establish some highly significant
conclusions about its nature [W3:3]. But, as I will argue in Appendix V, his
position is radically incoherent both in itself and as a basis for
interpreting Kant. For if his transcendental arguments prove anything, they
yield conclusions only about the phenomenal world—conclusions which Kant
himself has already established.
Philosophers
who attempt to construct a transcendental argument for the validity of
presupposing the thing in itself apparently fail to see the reason why 'the
formal character of transcendental arguments does not interest [Kant]'
[K1:421]; as a result, like 'the critic who dissects a transcendental argument
[i.e., one which Kant does give] with merely formal instruments', they 'may
miss the heart of the matter' [423]. It is indeed unfortunate that Kant himself
does not present his argument as clearly as Chipman has shown to be possible.
But it is not surprising that even in its clearest form it is not undeniably
conclusive. For Kant never intends his comments to provide anything but good
reasons for assuming the thing in
itself as the starting point for the transcendental perspective in systemt.
When
Kant says, for instance, that the presupposition of the thing in itself is
'not only admissible, but unavoidable' [Kt2:314-5], he is implying not that a
transcendental argument could ever prove
it to be necessary, but that it is unavoidable for anyone who wants to ask transcendental
questions (i.e., questions about the subject and/or object of experience, which
assume the transcendental perspective). To ask such questions without
presupposing the thing in itself (as in some versions of phenomenalism) is 'to
be landed in the absurd conclusion that there can be appearance without
anything that appears' [Kt1:xxvi-xxvii]. For, as George rightly says, 'if
Kant's theory of objectivity is correct, it is not for us men to say that there
are objects of representation but no things in themselves' [G8:194]. It may be
that representations are somehow generated out of nothing; but if we wish to
construct a coherent theory of the phenomenal world on the basis of the
transcendental perspective, then it is more reasonable to presuppose that
they are rooted in the thing in itself. Therefore, the 'heart of the matter',
which is likely to be missed by those who wish to prove the validity of Kant's presupposition, is that doubt in its
validity can in the end be countered only by faith.
This
interpretation of Kant's intentions reveals a fundamental point at which he agrees with the skeptic: both Kant and
Hume maintain that there is a sense in which objects are unknowable, so that
even the philosopher cannot claim to have objectively valid knowledge of them.
For Kant this is true only so long as objects are regarded as things in themselves;
but this should not be taken as an attempt at entirely refuting the skeptic.
For his response to being 'awakened' by Hume was not, as Russell suggests, to
conjure up 'a soporific which enabled him to sleep again' [R13:731]. Rather,
having realized that the skeptic's doubts can never be entirely satisfied
merely by means of logical proof and disproof, he formulated the doctrine of
the unknowable thing in itself as his concession
to the skeptic [see e.g., Kt2:20 and Kt4:53].
Transcendental
arguments are designed to convince the skeptic that experience is possible
only if the object is regarded as an appearance conforming to certain synthetic
a priori forms of knowledge, supplied by the subject (namely: space, time and
the categories). When the role of these arguments is taken into consideration,
it becomes evident that Kant's position is not, as Schaper suggests [S3:237],
that we must view the world 'as if' these forms of knowledge apply to objects,
even though we know that in reality they do not, or that objects of experience
'can be taken "as real" for all empirical purposes'; on the
contrary, it is that these forms must really
apply to empirical objects and that such objects must be empirically real in order for experience itself to be
possible. If the arguments are successful, skeptics are still allowed (indeed,
encouraged) to regard the thing in itself as unknowable, but they must now also
admit that empirical knowledge of objects is possible so long as it is regarded
as representational; for such knowledge is not a direct apprehension of the
thing in itself. However, skeptics are not
forced to admit that the representations of experience are grounded in the
thing in itself—a view more likely to be rejected by Berkeley than by Hume [see
VI.2]. Kant does not construct a transcendental argument on the latter point
because he knows the skeptics' acceptance of it depends in the end on their
willingness to adopt some amount of faith.[7]
This
conclusion enables us to accept Hartmann's suggestion that all systems of
transcendental philosophy 'can be explained only from within, not from without'
[H3:249], yet without jeopardizing the success of Kant's attack on Hume's
skepticism. Kant is here attempting not so much to force skeptics (or any other
'outsiders') to relinquish their position, as to integrate the valid insight
acknowledged by the skeptic (that reality in itself is not knowable) with the
valid insight acknowledged by the dogmatist (that reality as we know it is
necessarily characterized by certain unchanging forms), and in so doing to
provide the impartial reader with a coherent alternative to both extremes.
Kant's
success, though by no means complete, can be measured by the fact that
philosophers who refuse to adopt faith these days are generally not so
skeptical (or perhaps, not so forthright) as to admit to drowning their
despair in food, backgammon and merriment, as Hume did [H21:1.4.7]; instead
they adopt one of several agnostic positions, from which they either ask
transcendental questions without committing themselves to belief in the thing
in itself (as in phenomenalism), or else simply ignore transcendental
questions altogether (as in pragmatism). Kant would have little to say to the
latter group: perhaps only that their pragmatism itself only makes sense within
the broader context of Critical philosophy. But to the former group he would
question the authenticity of their supposedly 'transcendental' questions by
pointing out the incoherence of adopting a 'phenomenal' view of the world
without assuming something transcendent which necessarily limits our knowledge.
Unfortunately, it is unlikely that either group would be anxious to listen.
4.
Theoretical Faith and Practical Faith
Now
that I have clarified the role of faith as Kant's key to justifying the
transcendental starting point of his
System, I shall touch upon its role as the hypothetical goal of systemt. It is important for us to consider the latter not
because it is essential to the justification of his transcendental turn [cf.
note V.6], but because such practical
faith is often regarded as the only way in which Kant refers to faith [see
e.g., W24:13-25]. We must therefore conclude our discussion by examining how
practical faith in the ideas of reason differs from theoretical faith in the
presupposition of the thing in itself.
An
idea of reason is a concept which serves to unify our interpretation of
experience by providing a rational explanation of what the unconditioned ground
of some kind of experience might be [Kt1:383-6]. Speculative metaphysicians
make the mistake of believing they can reason their way to knowledge of such
unconditioned objects. But Kant denies this possibility, insisting that,
because an idea is a mere concept, with no corresponding intuition, it can
never constitute knowledge of the unconditioned. Instead, it can be employed
effectively only by those who are willing and able to adopt a kind of
hypothetical faith in the practical value of its use [499-500].
Kant
says practical faith 'refers only to the guidance which an idea gives me, and
to its subjective influence in that furthering of the activities of my reason
which confirms me in the idea, and which yet does so without my being in a
position to give a speculative account of it.'[8] Another, better
known description of what is involved in adopting such faith is that it entails
viewing certain objects of experience 'as
if they had their ground' in the object represented by an idea [709], even
though the reality of this object is not objectively knowable. In suggesting
this interpretation, Kant is intimating that we should regard the idea not
merely as a 'philosophical fiction',[9] but as our best
approximation (or hypothesis) of
what the unconditioned object really would be, if we were able to gain
knowledge of it. Because its speculative employment can never be legitimate,
such hypothetical faith in the idea is required in order to justify its
practical employment.
How,
then, are the ideas of reason and the faith they require related to the thing
in itself and the faith it requires? Some of the claims Kant makes in
discussing the ideas can be applied directly to the thing in itself as well. He
says, for example, that through the ideas 'we really know only that we know
nothing [about transcendent reality]' [Kt1:498]; and the same can be said of
the thing in itself [cf. 45]. Moreover, Kant's references to the thing in
itself in his discussion of the ideas often imply that the two notions are
inextricably related [e.g., 593-4; Kt2:353-4]. But the closeness of their
relationship should not lead us to agree with those interpreters who propose
that the thing in itself should actually be included as an idea of reason.[10] To do so would
be to ignore the difference between Kant's transcendental and hypothetical perspectives
[see IV.3]. For when the thing in itself is mentioned in the context of a
hypothetical perspective (such as Kant adopts whenever he discusses the ideas),
it always refers to what the thing in itself would be if it were an unconditioned object of knowledge; and as we will see in VI.3, this is precisely what he
means by the (technically more accurate) term 'positive noumenon'. The ideas
are concepts of what might be true about the thing in itself as positive noumenon, based on what we
actually encounter in experience; but they are regulative, rather than
constitutive, of experience. By contrast, the thing in itself as such is a transcendental
presupposition referring to that which is ultimately
constitutive of (at least) the matter of experience, but which is not
considered in experience to be
either regulative or constitutive. When this distinction is taken into
consideration, it becomes obvious that Kant's reference to the thing in itself
as such from the transcendental
perspective is quite distinct from his connection of the thing in itself as noumenon with the ideas from the
hypothetical perspective.
The fact that Kant positions the thing in
itself and the positive noumenon at opposite ends of the spectrum of knowledge
indicates that the sort of faith employed in each case is at least in some
respects different. In its transcendental use it is directed towards a single,
necessary, theoretical presupposition: the whole realm of transcendental
reflection is closed to the philosopher who is unwilling to adopt faith at this
point. In its hypothetical use, on the other hand, it is directed towards a
variety of practical presuppositions, the denial of which may or may not
affect the coherence of the overall System. But apart from the difference in
the object to which it is directed and in its relation to that object (as
determined by its position in the unfolding of systemt [see VII.2-3]),
the kind of faith required at these two points is really the same. It is the reasoned
decision to treat as true a presupposition which cannot be verified
objectively, but which is suggested by the objectively known facts to be the
best (or perhaps, the only) choice available which will enhance the unity of
the systematic context (i.e., of the perspective) to which it belongs. When
Kant adopts rational faith in the thing in itself, it generates a systematic
movement of thought through the whole spectrum of human knowledge, which comes
full circle to rest where it began, with faith in the unknown, considered now
as the tool of practical reason.
Numerous
implications follow from this understanding of the role of faith, particularly
in the areas of moral philosophy and philosophical theology. The most obvious
example, however, concerns those philosophers whose ultimate goal is to reach a
certainty which is in no sense dependent upon faith. Philosophers of all sorts
can readily be found who tend towards this attitude: Heidegger can serve as a
typical case. He describes philosophy as a science of 'thinking' which
excludes all faith, relegating the latter to theology [H10:10,27,205-6]. But
if I am correct in my interpretation of Kant's justification for the type of
reflection which Heidegger himself practised, and if Kant is correct in putting
it forward, then philosophers who push aside faith in this way are in danger of
either letting their overconfidence in the powers of unaided human reason lead
them to make unwarranted speculations (e.g., about 'Being'[11]), or letting
their lack of confidence prevent them from making any significant affirmations
at all [see Kt20:145-7]. For there would then be an important sense in which
faith is not only relevant, but vital to the task of the philosopher (as well
as that of the theologian). As Bradley puts it, 'philosophy demands, and in
the end rests on, what may fairly be termed faith. It has, we may say, in a
sense to presuppose its conclusion in order to prove it' [B25:15].
Another
example of how this perspectival view of the role of faith is applicable to
theological issues is the potential it has for resolving the age-old
controversy over the question of the relationship between faith and reason.
With all his 'enlightened' emphasis on the superiority of reason, Kant saw that
transcendental reflection must be preceded by faith if it is to be consistently
and coherently employed. Yet at the same time, he saw that reason precedes
faith when it is viewed from the empirical perspective. Theologians, I suggest,
could apply this perspectival understanding of faith to their own interests by
arguing that religious faith likewise precedes reason transcendentally (e.g.,
when considering the necessary conditions for salvation), even though reason
precedes faith in empirical matters (e.g., when considering the significance of
some specific religious activity). Of course, such an attempt to integrate the
two alternatives into a single consistent solution would have to be discussed
in greater detail than is possible here before any firm conclusions could be
established [but see Pq20].
Other
examples could be provided, but are peripheral to our purpose in this chapter,
which has been to clarify the nature of Kant's emphasis on faith. Although
Kant's Transcendental Perspective can, as we have seen, be partially justified
in other ways as well, our conclusion has been that faith is Kant's key to its
ultimate justification. Thus we can say with Weldon that Kant is 'entitled to
his view that the nature [indeed, the very 'existence'] of things in themselves
inevitably remains for us a matter of belief, not of scientific metaphysical
knowledge' [W16:11]. Only when we follow Kant in replacing our natural tendency
towards dogmatism or skepticism with a rational faith in the legitimacy of the
transcendental presupposition will we be in a position to understand, and so
also to judge the validity of, his Critical philosophy. Just as practical faith
in the ideas can 'bring unity into the body of our detailed [i.e., empirical]
knowledge' [Kt1:675], so also we may find that such theoretical faith can bring
unity into our understanding of the nature of human knowledge in general. For
in both cases, as Kant says in Kt43:222(82), 'faith ... provides ... a clear
light to enlighten philosophy itself.'
[1] On the general nature of the Transcendental Perspective, its epistemological status, and the role it plays in Kant's System, see II.4, III.1 and IV.3.
[2] Some commentators have criticized Kant for referring to both a 'thing in itself' (singular) and 'things in themselves' (plural) [e.g., S13:3.37]: since the terms are supposed to denote an unknowable reality, how can we know whether there are many or just one? Ewing's reply to such a criticism is that this problem is 'rather a defect of European languages than a philosophical inconsistency' [E5:191]. Although this may be true to an extent, I would add that this concept has to be expressed in a singular or plural form because the reality to which it refers is, for us, an abstraction from empirically knowable objects, which are represented in singular or plural terms. (In support of this Allison cites Prauss' claim [in P12:20; s.a. P13:222-39] that Kant intends the phrase 'thing in itself' (Ding an sich) to be an abbreviated form of the more accurate phrase 'thing considered as it is in itself' (Ding an sich selbst betrachtet) [A9:317n]; thus 'thing' and 'things' both refer—Allison says directly, but I will suggest indirectly [see VI.2]—to the object or objects of ordinary experience [cf. G12:771].) Accordingly, Kant's usage reflects not so much a 'defect', as a potentially (but not necessarily) misleading ambiguity in the meaning of his terms. In order to minimize confusion, I will limit myself wherever possible to the singular form. (This seems particularly appropriate since Kant uses the closely related term 'transcendental object' in the singular everywhere in Kt1 except one passage [A394], where he slips into the plural three times.) The phrase 'thing in itself' will therefore be short for something like 'the thing (reality) which is independent of me (transcendent) and which may or may not be composed of "things"'.
[3] Vleeschauwer mentions that in Kt1 'Kant had expressed his faith in the existence of the transcendent in the theoretical necessity of its participation in the Critical synthesis as a limiting concept' [V4:191]; but he does not develop the implications of this fact. Barker also admits 'it is implausible to suppose that [Kant] himself regarded his moral philosophy as providing the main basis for saying there are things-in-themselves' [B2:285]. But instead of offering an alternative justification, Barker argues that the thing in itself is a groundless assumption, resulting from Kant's acceptance of two contradictory theories of perception [283,289]. In VI.2 I will attempt to demonstrate how the principle of perspective provides for the compatibility of these two theories.
[4] Paulsen supports this position by arguing that the thing in itself, 'that crux interpretum ... is not the central principle of [Kant's] system; it is a self-evident presupposition' [P4:153]. As such, faith in the thing in itself is quite similar to what Peano called a 'primitive concept', inasmuch as 'it is assumed and affords a basis for demonstration although it is not itself demonstrated' [S14:175]. According to Stebbing, the whole of science is based upon faith in such 'unprovable' fundamentals [402].
[5] Kt1:852; s.a. Kt10:65-70(72-8). Kant's technical word for empirical or objective 'knowledge' is Erkenntnis (sometimes translated as 'cognition'). Wissen, by contrast, is a more inclusive term, used to refer to cases in which 'the holding of a thing to be true [Fürwahrhalten] is sufficient [for a given person] both subjectively and objectively' [Kt1:850]. Kant's use of the verbal forms of both words in the same sentence at this point is therefore a potential source of confusion for the English reader, since the two are not distinguished by Kemp Smith. Wissen is a more general term, which includes both Glaube and Erkenntnis as mutually exclusive types.
Wiebe makes a distinction in W18:524 between 'doctrinal' and 'moral' faith, which is virtually the same as the distinction I am making here between theoretical and practical faith. As he explains, quite rightly, in W18:522-3: 'Faith, ... in the Kantian view of things, is not outside the realm of reason, but is rather one aspect of reason, as is knowledge.' Unfortunately, Wiebe does not explain how Kant's notion of theoretical faith serves as the foundation for the Transcendental Perspective.
[6] The hypothetical character of faith will be discussed further in V.4 [s.a. IV.3]. Kant alludes to the close link between the hypothetical perspective and the practical standpoint [see IV.4 and Ap. II] in Kt69:297: 'faith ... is ... an assumption or presupposition (hypothesis), which is necessary only because an objective practical rule of conduct that is necessary supports it. Such faith is assent to a theoretical proposition, e.g., there is a God, through practical reason.' He then clarifies that an 'opinion' can become a 'belief' if, on its basis, a person 'must decide what to do' [298], because belief always 'has in itself a moral worth, for it includes a free commitment.' In this sense, Kant's presupposition of the thing in itself could be regarded technically as an opinion until we discover in systemp that it demands moral commitment as well as recognition of theoretical ignorance. However, in light of Kant's doctrine of the primacy of practical reason [see VIII.1], it would be better to think of practical reason as itself providing the initial basis for theoretical faith.
[7] A transcendental argument for the existence of the thing in itself would be analogous to an ontological argument for the existence of God [cf. X.2]. Kant rejects both for much the same reason (viz., '... to make room for faith' [Kt1:xxx; s.a. 642-8]). This similarity reveals the inadequacy of Cameron's attempt in C2 to force Kant into a dilemma on this subject, whereby it would be necessary for him either to accept the possibility of an ontological argument or to reject his own theory of the transcendental object (which Cameron wrongly regards as 'extentionally equivalent' to the thing in itself [C2:355n; but cf. VI.2]). Cameron's error is to assume with little or no evidence (a misreferenced quote on the empirical certainty of one's own existence [supposedly from Kt1:356]) that Kant is committed to the view that the transcendental object 'exists' in the same sense that empirical objects exist.
[8] Kt1:855. Although Kant often uses the term 'speculative' in a way which appears to be synonymous to his use of 'theoretical', there is technically an important difference between these two terms. Theoretical knowledge is any legitimate knowledge which arises out of the synthesis of concepts and intuitions, as described in systemt. Speculation, by contrast, is the illegitimate attempt to gain knowledge without securing an intuitive content for the concepts (or ideas) being employed.
[9] S3:233. In K2:11.278n(Z1:180n) Kant says: 'When ... I seek the unconditioned and the highest ground of the totality of the series [of conditioned appearances], it is as if I were looking for .' Although she does not refer to this text, Schaper uses the same comparison in her attempt to prove in S3 that Kant intends his ideas of reason to be nothing but useful 'fictions'. However, Kant applies the analogy of 'impossible roots' only to the 'cosmological ideas' (i.e., the Antinomies), not to all ideas of reason [K2:11.278(Z1:180)], and it seems likely that the 'something similar' to which he refers is simply the fact that in both cases the process in question implies a self-contradiction. In other respects the analogy would not hold. For mathematicians know they are working with a fictitious hypothesis, to the extent that they label it an 'imaginary number'. Philosophers, on the other hand, do not know the extent to which the ideas they employ accurately represent the unconditioned; but they may have reasons for believing they do, otherwise they could not be regarded as valid hypotheses. Kant says in Kt1:397 that, even in their 'pseudo-rational' (i.e., speculative) use, the ideas 'are not fictitious, and have not arisen fortuitously, but have sprung from the very nature of reason.' Only if the object represented by an idea is hypostatized as a real object of empirical knowledge would Kant regard it as 'a mere fiction' [608].
[10] E.g., S3:238. Schaper strays even further from Kant by combining this approach with the interpretation of ideas as 'useful fictions' [see note V.9], so that the philosopher would be required to know the thing in itself does not exist, yet to make use of it anyway. This stands in sharp contrast to Kant's more modest approach of adopting rational faith in the validity of presupposing its reality from the transcendental perspective [cf. I1:756]. Schaper's belief that things in themselves 'have to remain fictions' in order to 'make sense' [S3:236] results from her tendency to regard the thing in itself as an empirical, rather than a transcendental, presupposition [see VI.2-3].
[11] Van de Pitte fails to distinguish Kant's transcendental and empirical perspectives when he says: 'For Kant, Being is—and remains—transcendent, and thus is not a presupposition of his transcendental account' [V3:1027]. From the empirical perspective, Being is not transcendent, but immanent: it consists of everything that 'exists' in the ordinary world. From the Transcendental Perspective, on the other hand, Being is indeed transcendent, and so does not actually enter into the realm of the transcendental. But for this very reason it is necessary to presuppose its reality as the foundation of the Transcendental Perspective. For Kant's entire System is based 'on the assumption of a fundamental reality, subject to a rule of necessary existence, without which nothing could possibly happen and even nature would not be possible' [I1:757; cf. Kt1:280].
Send comments to the author: StevePq@hkbu.edu.hk
Back to the Table of Contents for this book.
Back to the listing of Steve Palmquist's published books.
Back
to the main map of Steve
Palmquist's web site.