718 Book Reviews

tical reasoning. For Graham tends to run together questions of agency with
questions of value. In particular, he does not sharply distinguish between acting as part of a group and acting for the sake of a group. Here as elsewhere the
discussion of reasons for action takes place at a distance, via talk of identity.
Graham distinguishes between mere passive identification with a group
simply recognizing that one is a member of some groupand active, practical
identification with collectivities (p. , italics in original), which involves
being motivated by the good of the group, and so involves endorsement of a
practical attitude rather than just recognition of a fact (p. ). But the question of whether membership of a group has any practical import cuts sharply
across the question of whether one identifies with the good of the group. For
an individual can have both self-interested and altruistic reasons for acting as
part of a group without those reasons having anything to do with the good of
that group. And equally she can act as an individual for the sake of a group. We
should sharply distinguish these two ways of identifying with the group.
Thus I have some reservations about Grahams book. The main one concerns its reticence in spelling out the implications of our connectedness for
reasons for action. Nevertheless, a reader interested in collective agency will
find many helpful distinctions and clarifications in this book, and Grahams
project remains interestingif not yet fulfilled.
Department of Philosophy
University of Nottingham
Nottingham NG7 2RD
UK

christopher woodard

Kants Doctrine of Transcendental Illusion, by Michelle Grier. Cambridge: Cambridge University Press, . Pp. . H/b ..

Kant held that human reason is beset with inescapable illusions that ground
classical metaphysicss arguments in favour of the immortality of the soul and
the existence of God, and generate endless debates about the nature of the
world (the totality of physical things). In the Transcendental Dialectic of the
Critique of Pure Reason, Kant undertakes to uncover the origin of what he calls
the illusions of reason, and to denounce the fallacies of the metaphysical arguments grounded on these illusions. One major difficulty that has baffled commentators is that Kant seems to maintain both that reasons illusions are
inevitable, and that special metaphysics (inquiries into the existence and
nature of the soul, the world, and God) can and must be eradicated by the critical denunciation of its fallacious arguments. Is there a way out of this apparent inconsistency?
Michelle Grier argues that there is no inconsistency at all if one distinguishes, as Kant does, between the (inevitable) illusory representations of reason
and the (corrigible) fallacies of judgement by way of which those representa-

Book Reviews 719

tions are referred to objects supposed to have actual existence and to be knowable according to Kants categories. Griers merit is to follow up that distinction
systematically and thus to bring new light on the structure of Kants complex
argument in the Transcendental Dialectic of the Critique of Pure Reason and its
Appendix, On the Regulative Employment of the Ideas of Pure Reason.
The book is in four parts. In part one, Grier shows how the idea of an inescapable, transcendental illusion of reason gradually emerges in the course of
Kants pre-critical efforts to define the methods and concepts proper to metaphysics. In part two, she shows how the critical theme of reasons illusions is
inseparable from Kants new distinction between understanding and reason.
In part three, she provides a detailed analysis of Kants distinction between the
illusions of reason and the fallacies of the metaphysical arguments based on
them, for each of the three ideas of reason: the soul, the world, God. And
finally, in part four, she shows how the inevitability of the illusion is related to
the regulative employment of the ideas of reason as expounded in the Appendix of the Transcendental Dialectic.
Her account of the pre-critical period is understandably selective, since her
goal is only to lay out the main stages and themes that prepare the way for
Kants critical doctrine of transcendental illusion. The story she tells is generally reliable. One surprising exception is her account of the New Elucidation of the Principles of Metaphysics and Morals. She is quite right to stress its
importance. Among other things, it offers Kants first extensive discussion of
the rationalist principle of sufficient reason (which, as she explains, Kant more
modestly calls, after Crusius, principle of determining reason). But Grier is
not quite accurate in her account of Kants distinction between different kinds
of reason or ground (Grund, ratio) and between the different aspects of the
principle of determining reason that, according to Kant, depend on them. For
instance, she reduces Kants specifications of the notion of reason or ground to
two: ratio fiendi (reason of coming to be) and ratio cognoscendi (reason of
knowing), whereas in Kant there are at least four kinds of ground, classified
under two very important headings, which she ignores: antecedently determining reason, and consequently determining reason (identified with the ratio
cognoscendi). The result is that she attributes to Kant a criticism of a supposed
rationalist confusion between ratio cognoscendi, assimilated to logical reason (a
reason for predication contained in the subject of the predication) and ratio
fiendi (assimilated to cause or real ground). But in fact, what Kant calls ratio
cognoscendi is not at all a logical reason in the sense just outlined, and the ratio
fiendi is not what he calls a cause. The notion of cause appears later in the
text, as Kants ratio existendi. Griers simplification of Kants view allows her to
antedate Kants criticism of the rationalist confusion between logical and real
ground (which really occurs systematically, as she herself points out, in the
s). This in turn makes it more difficult to understand the relation between
this early text and Kants later discussion of the principle of sufficient reason in
the Second Analogy of Experience in the Critique of Pure Reason (which,

720 Book Reviews

admittedly, is not the topic of her book). More importantly, I think it leads her
to give the Elucidation a greater role than it deserves as an ancestor of Kants
critical view of the errors of rationalist metaphysics.
I found Griers discussion of the pre-critical period otherwise interesting
and helpful. Especially noteworthy are the sections she devotes to the
Dreams of a Spirit Seer and to the Inaugural Dissertation, where she rightly
emphasizes themes that point the way towards Kants later treatment of the
illusions of reason and of the metaphysical errors to which the illusions contribute.
In part two (Fallacies and illusions in the Critique of Pure Reason) Grier
gives an illuminating account of Kants distinction between the fallacies of the
understanding and the power of judgement, on the one hand, and the illusions
of reason and the fallacies derived from them, on the other. Criticizing the
former is the task of the Transcendental Analytic, criticizing the latter is the
task of the Transcendental Dialectic.
Grier convincingly shows what the Transcendental Analytic owes to Kants
pre-critical reflections on the concepts and methods of metaphysics, even
while moving forward to a specifically critical position. She picks out two
main themes: Kants criticism of attempts to derive knowledge of objects from
merely formal, logical principles; and Kants criticism of attempts to apply the
categories, or pure concepts of the understanding, to knowledge of objects
considered independently of the conditions of our sensibility. She shows how
each theme was foreshadowed in some of Kants pre-critical writings, but also
how they are now unified into a general rejection of what Kant calls transcendental realism, that is, the ontological confusion between appearances and
things in themselves. She is less successful, I think, when she tries to introduce
a new terminology to clarify Kants view. She suggests that Kant accepts what
she calls the transcendental employment of the categories, in the negative
sense while rejecting their transcendental employment, in the positive sense.
By transcendental employment of the categories, Kant means their employment in knowledge of objects independent of the conditions of sensibility. As
Grier correctly recounts, Kant expressly maintains, in the chapter on the Schematism of the Pure Concepts of the Understanding, that no such employment
is possible. Nevertheless, she suggests that he does admit what she proposes to
call a transcendental employment of the categories in the negative sense
that is to say, a purely a priori and intellectual component in the empirical use
of the categories. What he rejects is their transcendental use in the positive
sense, namely, the application of the categories for knowledge of objects that
are not given in sensibility. The distinction she proposes is inspired by Kants
own distinction between noumena in the positive sense (objects of an intellectual intuition, whose notion is empty) and noumena in the negative sense
(objects that are not given to sensible intuition, whose supposition, Kant
maintains, is the necessary counterpart to the characterization of empirical
objects as mere appearances). Her parallel distinction seems to me to be more

Book Reviews 721

confusing than illuminating, and the text she cites in support of her suggestion
seemed to me to undermine rather than support it (p. ; Critique of Pure Reason B).
By contrast, she is most impressive in her analysis of Kants distinction and
correlation between the fallacies of transcendental realism denounced in the
Transcendental Analytic, and the transcendental illusion of reason laid out in
the Transcendental Dialectic. She emphasizes the importance and novelty of
Kants critical view, according to which reason is the source of a specific drive
to push knowledge beyond the limits of any possible experience. And she gives
a clear account of the nature of reasons illusion, which consists in transforming a subjective, epistemic principle (which she names P: to find for conditioned knowledge the whole series of conditions) into an objective principle
(which she names P: If the conditioned is given, the absolutely unconditioned is given) (pp. ). The fallacy occurs when reason, with the collaboration of the power of judgement, tries to apply to an actually existing object the
(illusory, but useful) representation of an objective totality of conditions for
any given conditioned.
Griers account of the illusion is sometimes hindered by unclarities in her
understanding of the relation between categories and forms of judgement: it is
not true, for instance, that to judge categorically just is to take the logical subject within the judgement substantivally (p. ). Kant is very careful to distinguish between the mere logical function of relating subject to predicate in
judgement and the transcendental relation between substance and accident
(see for instance B). Accordingly, she is somewhat obscure in her account
of the transition from each syllogistic form (categorical, hypothetical, disjunctive) to the corresponding idea of the unconditioned of the series of conditions. And I am not sure she sufficiently distinguishes between Kants appeal
to these forms to account for the generation of the illusion, and his subsequent
laying out of a quite distinct fallacy of reasoning grounding the representations
of metaphysical objects under each idea of reason. In other words, I am not
sure she is sufficiently faithful to her own recommendation to distinguish
between the illusion of reason and the fallacious inferences drawn by reason
with the help of the power of judgement (see for instance pp. ).
Grier might respond that this worry is taken care of in the third part of her
book, The Dialectical Inferences of Pure Reason. There she works hard to
show, in the case of each idea of pure reason, how Kant distinguishes between
the natural illusion carried by the idea and the fallacious reasoning by way of
which an existing object is erroneously thought to be determined under it. In
the case of the first ideathe idea of the soulher method allows her to put
to rest some standard charges of inconsistency in Kants exposition, and to give
an interesting account of the nature of the fallacy of reasoning in the first three
Paralogismsthe implicit reasonings that, according to Kant, lie behind the
Cartesian and Leibnizian view of the soul as an immaterial substance. She does
not analyse the fourth Paralogism, whose content, as she points out, differs

722 Book Reviews

from the first to the second edition of the Critique of Pure Reason.
The second idea of reason is the idea of the world, giving rise to the antinomies of pure reason. Grier shows how in all four antinomies, the conflict
between thesis and antithesis is born from the assumption, common to both
sides, that appearances are things in themselves. But the thesis endeavours to
determine them independently of their spatio-temporal framework, whereas
the antithesis takes them to be intrinsically spatio-temporal. Both thesis and
antithesis end up undermined by the objections born from the opposing view.
And both share the (erroneous) assumption that a totality of conditions for
any given appearance can be determined as an object an assumption that
holds true only under the supposition of transcendental realism.
One may find quarrel with some details of Griers account. For instance, I
am not convinced that she is correct in her cursory account of Kants move
from the thesiss argument against the temporal infinitude to its argument
against the spatial infinitude of the world (p. , n. ). And I am puzzled by
her claim that one should distinguish between the antinomy (in the singular)
and the particular cosmological arguments (in the plural) (p. ). What she
designates as the antinomy is not an antinomy at all, but just the paralogism,
born from the error of transcendental realism, that grounds the antinomy (in
the singular), specified according to each category (thus the antinomies, in the
plural). The paralogism, in this case, is (as she correctly cites): If the conditioned is given, then the entire series of all conditions is given; objects of the
sense are given as conditioned; so, the entire series of all conditions of objects
of the senses is given (A/B; see Grier, p. ). In fact, her own account of
the relation between the paralogism and the antinomy (antinomies) is more
accurate later (see p. ) when she recounts Kants solution to the antinomy
(and antinomies).
Grier is certainly correct in challenging the misleading viewmade especially prevalent by Al-Azms authoritative work on the Antinomies (Sadik J.
Al-Azm, The Origins of Kants Argument in the Antinomies, Oxford: Oxford
University Press, ), and shared by many commentators before and after
himaccording to which Kants exposition of the antinomies should be read
as a systematic reconstruction of the LeibnizClarke correspondence, where
Clarke/Newton is supposed to represent the finitist side (thesis), and Leibniz
the infinitist side (antithesis). Kants target, Grier maintains, is a more general
confrontation between what he describes as Platonism (on the side of the thesis) and Epicureanism (on the side of the antithesis). This confrontation pervades all metaphysical thinkingfor instance, both finitism and infinitism are
represented in Leibnizs viewand ultimately concerns the way in which one
should conceive the relation between the world and things in the world on the
one hand, and their spatio-temporal framework on the other.
The third idea of reason is the idea of God. Grier is very clear and helpful in
distinguishing four main steps in Kants generation of the idea itself: () in
applying the principle of complete determination to things in general, reason

Book Reviews 723

presupposes a totality of possible predicates, with respect to which any individual thing is determined either positively or negatively; () this totality of
possibilities is itself supposed to be an individual entity, an ideal of pure reason; () the idea of such an individual object is the idea of a totality of possible
positive predicates, or realities, by limitation of which negative predicates can
also be determined; () as an individual entity, the whole of reality is thought
of as an ens realissimum.
In the course of unfolding the third step, she devotes several pages to criticizing the account I have given of Kants view of the Ideal of Pure Reason
(Batrice Longuenesse, The Transcendental Ideal and the Unity of the Critical
System, in Hoke Robinson (ed.), Proceedings of the Eighth International Kant
Congress, volume , part , Millwaukee: Marquette University Press, ; discussed in Grier, pp. ). She takes issue with my maintaining that there is,
according to Kant, a legitimate critical version of the principle of complete
determination and its application, outlined in the Transcendental Analytic. In
this version, the presupposed whole of reality would be the presupposed whole
of experientially given matter of experience (in Kants words, that which corresponds to sensation) given in the formal intuitions of one space and one
time. I think she is right in pointing out that I do not give full due to the independent origin in reason of the illusory idea of the ens realissimum, or to the
positive role played by this idea according to Kants argument in the Transcendental Dialectic. She is also right in pointing out that I do not sufficiently distinguish the illusion of reason from the fallacy of subreption by which the
existence of the presupposed whole of empirical reality is translated into the
presupposed existence of the ens realissimum itself. But she is mistaken in
some of the views she attributes to me, for instance when she claims that I
identify the principle of complete determination understood in the context of
the Analytic with Kants supreme principle: The conditions of possibility of
experience are the condition of possibility of the object of experience. What I
do say is that the former can be considered as a corollary of the latter. I explain
why this is so in light of the role played, in the constitution of the objects of
experience, by Kants logical forms of infinite and disjunctive judgment (and
the corresponding categories), and the unity of apperception (op. cit., pp.
). Similarly, contrary to Griers charge, I do recognize that Kant endorses the
idea of the ens realissimum. I also express surprise at this fact, and I try to
understand why Kant should hold such a position, which is clearly in tension
with his criticism of the notion of ens realissimum in the Amphiboly of Concepts
of Reflection (op. cit., pp. ).
Grier shows very well how accounting for the illusory idea of an ens realissimum still leaves open the issue of why rational metaphysics should also assert
its necessary existence. But she gives a surprising account of Kants criticism of
the ontological proofthe Cartesian proof of the existence of God from the
mere concept of an ens realissimum. She maintains that according to Kant, the
proof commits the dialectical fallacy of ambiguous middle. She summarizes

724 Book Reviews

the Cartesian proof as follows:

() The ens realissimum is a being that possesses all realities.
() A being that possesses all realities necessarily exists.
() Therefore, the ens realissimum necessarily exists,.
The ambiguous middle is being that possesses all realities. Grier seems to
think that on one understanding of the phrasewhere a being that possesses
all reality means an actual object to which real (determining) predicates can
be synthetically attached (p. ) both () and () would be legitimate.
According to her, the fallacy comes from the fact that in () being that possesses all reality is understood differently, as a purely intellectual notion. But
surely this diagnosis is not correct. Under no interpretation at all can existence
be derived from the characterization of a thing as possessing all realities (all
positive predicates), whatever the nature of these realities, that is, whether
they are understood as sensibly conditioned or not. Kants criticism of the
proof is a criticism of any attempt to derive existence from predication,
whether empirical or purely intellectual.
In part four (Illusion and Systematicity), Grier is home free. She is now in
a position to show how her elucidation of Kants doctrine of transcendental
illusion, and the distinction she has pressed between the illusions of reason
and the errors of judgment on which rational metaphysics is based, give
unprecedented clarity to Kants doctrine of the regulative employment of the
ideas of reason in the Appendix to the Transcendental Dialectic. She shows
that reasons search for the unconditioned is not just added to the independently defined cognitive activity of the understanding in empirical cognition,
but becomes part and parcel of this activity. She also has interesting things to
say about the similarities and differences between the role of P in the theoretical use of reason, and the moral law in the practical use (p. ).
Department of Philosophy
1879 Hall
Princeton University
Princeton, NJ 08544-1006
USA

batrice longuenesse

The Positivist and the Ontologist: Bergmann, Carnap and Logical

Realism, by Herbert Hochberg. Amsterdam and Atlanta, GA: Rodopi Press,
. Pp. vi + . H/b ,, $..

This book contains a detailed examination of Gustav Bergmanns intellectual

journey from Viennese positivism towards a metaphysical realism closer to the
schools of Meinong and Brentano. In tracing Bergmanns development, Herbert Hochberg sets the historical stage for a defence of logical realism, the view
that in opposition to conventionalism aims to articulate what is ontologically