Visit https://ebooknice.

com to download the full version and

explore more ebooks

(Ebook) The Logical Must: Wittgenstein On Logic by

Penelope Maddy ISBN 9780199391752, 0199391750

_____ Click the link below to download _____

https://ebooknice.com/product/the-logical-must-
wittgenstein-on-logic-4918296

Explore and download more ebooks at ebooknice.com

Here are some recommended products that might interest you.
You can download now and explore!

(Ebook) Logic colloquium 2007 by Françoise Delon, Ulrich Kohlenbach,

Penelope Maddy, Frank Stephan ISBN 9780511787850, 9780521760652,
0511787855, 0521760658

https://ebooknice.com/product/logic-colloquium-2007-1492434

ebooknice.com

(Ebook) A Plea for Natural Philosophy by Penelope Maddy ISBN

9780197508855, 0197508855

https://ebooknice.com/product/a-plea-for-natural-philosophy-50607080

ebooknice.com

(Ebook) What Do Philosophers Do?: Skepticism and the Practice of

Philosophy by Penelope Maddy ISBN 9780190618698, 0190618698

https://ebooknice.com/product/what-do-philosophers-do-skepticism-and-
the-practice-of-philosophy-5716074

ebooknice.com

(Ebook) Wittgenstein On Logic As The Method Of Philosophy: Re-

Examining The Roots And Development Of Analytic Philosophy by Oskari
Kuusela ISBN 9780198829751, 0198829752

https://ebooknice.com/product/wittgenstein-on-logic-as-the-method-of-
philosophy-re-examining-the-roots-and-development-of-analytic-
philosophy-10000982
ebooknice.com
(Ebook) Logic & Natural Language: On Plural Reference and Its Semantic
and Logical Significance by Hanoch Ben-Yami ISBN 9780754637431,
0754637433

https://ebooknice.com/product/logic-natural-language-on-plural-
reference-and-its-semantic-and-logical-significance-12321252

ebooknice.com

(Ebook) The politics of logic : Badiou, Wittgenstein, and the

consequences of formalism by Badiou, Alain; Badiou, Alain; Livingston,
Paul M.; Wittgenstein, Ludwig ISBN 9780415891912, 9781138016767,
0415891914, 1138016764
https://ebooknice.com/product/the-politics-of-logic-badiou-
wittgenstein-and-the-consequences-of-formalism-10564278

ebooknice.com

(Ebook) Logical Forms An Introduction to Philosophical Logic by Mark

Sainsbury ISBN 9780631216797, 0631216790

https://ebooknice.com/product/logical-forms-an-introduction-to-
philosophical-logic-4630016

ebooknice.com

(Ebook) Biota Grow 2C gather 2C cook by Loucas, Jason; Viles, James

ISBN 9781459699816, 9781743365571, 9781925268492, 1459699815,
1743365578, 1925268497

https://ebooknice.com/product/biota-grow-2c-gather-2c-cook-6661374

ebooknice.com

(Ebook) Logic as a Tool: A Guide to Formal Logical Reasoning by

Valentin Goranko ISBN 9781118880005, 1118880005

https://ebooknice.com/product/logic-as-a-tool-a-guide-to-formal-
logical-reasoning-5677508

ebooknice.com
T H E L O G IC A L M UST
T H E L OGIC A L M UST

Wittgenstein on Logic

Penelope Maddy

1
 1
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide.

Oxford New York

Auckland Cape Town Dar es Salaam Hong Kong Karachi
Kuala Lumpur Madrid Melbourne Mexico City Nairobi
New Delhi Shanghai Taipei Toronto

With offices in
Argentina Austria Brazil Chile Czech Republic France Greece
Guatemala Hungary Italy Japan Poland Portugal Singapore
South Korea Switzerland Thailand Turkey Ukraine Vietnam

Oxford is a registered trade mark of Oxford University Press

in the UK and certain other countries.

Published in the United States of America by

Oxford University Press
198 Madison Avenue, New York, NY 10016

© Penelope Maddy 2014

All rights reserved. No part of this publication may be reproduced,

stored in a retrieval system, or transmitted, in any form or by any means,
without the prior permission in writing of Oxford University Press,
or as expressly permitted by law, by license, or under terms agreed with the
appropriate reproduction rights organization. Inquiries concerning reproduction
outside the scope of the above should be sent to the Rights Department,
Oxford University Press, at the address above.

You must not circulate this work in any other form

and you must impose this same condition on any acquirer.

CIP data is on file at the LOC

9780199391752

1 3 5 7 9 8 6 4 2

Printed in the United States of America

on acid-free paper
For my mother, Suzanne Katherine Lorimer Parsons,
with love and gratitude.
 CON TEN TS

Preface ix

Introduction 1
1. Kant on Logic  5
2. Naturalizing Kant on Logic  15
3. The Tractatus 37
4. Naturalizing the Tractatus 49
5. Rule-Following and Logic  63
6. But Isn’t Logic Special?! 81
7. Naturalizing the Logical Must  101
Conclusion 123

References 127
Index 133
 PR EFACE

The aim of this monograph is to examine Wittgenstein’s philoso-

phy of logic, early and late, from the perspective a particular brand
of naturalism called ‘second philosophy’. The Second Philosopher
holds a roughly empirical account of logical truth derived from
Kant: logic is grounded in the structure of our contingent world;
our basic cognitive machinery is tuned by evolutionary pressures
to detect that structure where it occurs. The early Wittgenstein
of the Tractatus also links the logical structure of representation
with worldly structure, but insists that the sense of our representa-
tions must be given prior to, independently of, any facts about how
the contingent world happens to be. When that requirement is
­removed – as the naturalistically minded will argue that it should
be – the Tractarian position approaches the Second Philosopher’s:
logic is grounded in the structure of the world; our representa-
tional systems reflect that structuring.
The late Wittgenstein also rejects this priority of sense. The
resulting view based on rule-following considerations comes still
closer to the Second Philosopher’s: both hold that our logical
 PR EFAC E

practices are grounded in our interests and motivations, our nat-

ural inclinations, and very general features of the world; neither
holds logic to be fundamentally different from other descriptions
of the world, just more general, responding to features so simple,
so ubiquitous, that they tend to go unnoticed. The difference is
that the Second Philosopher fills in more of the empirical story.
Here she comes up against Wittgenstein’s ruling that science is ir-
relevant, but no principled justification for it can be found in his
thought. With this thread removed, Wittgenstein’s late view be-
comes surprisingly naturalistic.
All serious readers of Wittgenstein benefit immeasurably from
the writings of those who came before. My own special debts to
David Pears and David Stern will be obvious in these pages; what
I owe to Barry Stroud, though perhaps less apparent, is no less
profound for that. More directly, I’ve learned a great deal over the
years from conversations with Brian Rogers. Discussions in my
2010–2011 seminar on the philosophy of logic first sparked the
idea that it might be illuminating to compare Wittgenstein with
the Second Philosopher along the lines explored here. Jeremy
Heis, David Malament, Patricia Marino, Brian Rogers, Walde-
mar Rohloff, David Stern, the members of my spring 2013 reading
group on the Philosophical Investigations, and anonymous referees
for Oxford University Press all provided helpful comments on
earlier drafts. My heartfelt thanks to all these, and to Peter Ohlin
and his reviewers for supporting this unorthodox style of histori-
cal engagement. Finally, my gratitude to David Malament, for his
friendship and much else, is unbounded above.
P. M.
Irvine, California
January 2014

x
 Introduction

What you say seems to amount to this, that logic belongs to the
natural history of man. And that is not compatible with the hard-
ness of the logical ‘must’.
RFM VI, §49

The problem of the so-called ‘logical must’ can be phrased this

way: if it’s either red or green, and it’s not red, then why must it be
green?1 Philosophers tend to elaborate by insisting that this isn’t
the same ‘must’ as in ‘what goes up must come down’, that the im-
possibility of what goes up remaining up is merely physical, not
logical. Logical truth,2 logical necessity are taken to reflect more
than what happens to be true in our contingent world: the laws
of logic hold in any possible world, perhaps because they actually
describe some crystalline realm of pure abstracta. But even with-
out these philosopher’s accretions, there’s a straightforward ques-
tion here that confronts any inquirer out to achieve a complete

1. The use of color words here isn’t intended to raise the specter of vagueness. If you prefer,
feel free to substitute, for example, ‘if the switch is either on or off, and it’s not on, then it
must be off’.
2. Though I speak here and elsewhere of logical truth, no significant contrast with logi-
cal validity is intended. In what follows, the question of the truth of ‘if it’s either red or
green, and it’s not red, then it must be green’, and the question of the validity or reliability
of the inference from ‘it’s either red or green, and it’s not red’ to ‘it must be green’, are
treated as more or less interchangeable.

1
 THE LOGICA L MUST

understanding the world: why is it that ‘if it’s red or green, and it’s
not red, then it’s green’?, and what’s the added force of the ‘must’ in
‘it must be green’?3
A few years ago, in a book called Second Philosophy [2007],
I offered a broadly naturalistic answer to these questions. The
‘naturalism’ in question is a variant of what’s referred to these
days as ‘methodological naturalism’:4 my Second Philosopher
investigates the world beginning from her ordinary perceptual
beliefs, gradually developing more sophisticated observational
and experimental techniques and correctives, eventually ascend-
ing to theory formation and confirmation, all in the sorts of em-
pirical ways usually labeled ‘scientific’. 5 In the book, I present
the s­ econd-philosophical account of logical truth and logical ne-
cessity as a descendent of Kant’s position, by way of a particular
naturalizing move (along with some straightforward Fregean up-
dating). In fact, I think that the same second-philosophical posi-
tion can be approached from other directions, and in particular,
it seems to me that either of Wittgenstein’s two characteristic
positions in the philosophy of logic – one from the Tractatus, the
other from the Philosophical Investigations and the Remarks on the
Foundations of Mathematics – provides a viable launching point
toward the second-philosophical port. Of course, the naturalizing

3. The question here isn’t what makes this truth ‘logical’, or more generally, what separates
logical truths from truths of other kinds. The idea is just to start with uncontroversial
examples that everyone would agree count as ‘logical’, then to ask what makes them true
(and what force is added by the thought that they ‘must’ be true). Also, though relevance
logicians don’t accept disjunctive syllogism as a general rule, this particular case should
satisfy any reasonable requirement of relevance.
4. The contrast is with ‘ontological naturalism’, which involves a rejection of ‘supernatural’
entities and causes. See Papineau [2007] for a taxonomy.
5. No demarcation criterion for distinguishing ‘science’ from ‘non-science’ is presupposed
here – I’m merely describing the Second Philosopher’s ways of proceeding in ordinary
terms and counting on the reader to get a rough idea of what I’m after. For more on why I
take this approach and how it’s intended to work, see [2007].

2
 I n tro d uctio n

moves required differ in the two cases, as both differ from the one
applied to Kant.
Now even if I’m right about this, even if it is possible to trace a
simple path from Wittgenstein to the Second Philosopher, it’s also
obvious that not everything doable is worth doing. My hope is that
this investigation will prove its worth in two ways, one primary, one
secondary. The primary aim is simply historical – to understand
Wittgenstein better6 – but the method is unusual: the idea is to
use the Second Philosopher’s view of logic, and its roots in Kant,
to illuminate Wittgenstein’s views on logic, both early and late.7 If
successful, this approach, this atypical perspective, should serve to
focus attention in unexpected ways and draw out less familiar fea-
tures. Along the way, the process of compare and contrast should
also reveal new aspects of the Second Philosopher’s position – and
this is the secondary aim.
The story begins with Kant, and the adjustment that leads to
the second-philosophical position.8 We then turn to the Tractatus
and finally to the later Wittgenstein.

6. I should admit that I don’t believe there to be only one useful way of reading Wittgenstein–
which is part of what makes his work so endlessly fascinating – so I give here readings that
focus on logic and that seem to me most interesting and productive when viewed side by
side with the Second Philosopher’s position.
7. There’s some difference of opinion on how to partition Wittgenstein’s work into periods.
All that’s needed for present purposes is a rough classification that takes ‘early’ to include
the Tractatus and ‘late’ to include Philosophical Investigations and Remarks on the Foun-
dations of Mathematics. (Some commentators use the term ‘late’ for all Wittgenstein’s
writings from 1929 on; others separate the final writings of 1949–1951 into a ‘third Witt-
genstein’, and recast ‘the late Wittgenstein’ as ‘the second Wittgenstein’. My rough cut
leaves room for a ‘middle’ or ‘transitional’ period that includes The Big Typescript, and
involves no stand on the status of On Certainty.)
8. For more details, see [2007], §I.4 and Part III. In addition to being much more con-
densed, the presentation here differs in a few small ways. [201?] summarizes the Second
Philosopher’s view of logic without the detour through Kant and elaborates a bit further
on a few points.

3
 C ha pt e r 1

Kant on Logic

Kant’s inquiry in the Critique of Pure Reason can be viewed as aris-

ing from this question: what must the world be like that we can
cognize it as we do? The difficulty of the question increases with
his conviction that some of this cognition is a priori:

It is easy to show that in human cognition there actually are . . .

pure a priori judgments. If one wants an example from the sci-
ences, one need only look at all the propositions of mathematics;
if one would have one from the commonest use of the under-
standing, the proposition that every alteration must have a cause
will do. (B4–5)

Kant notes the obvious problem in his famous letter to Hertz in

1772:

How my understanding may form concepts of things com-

pletely a priori, with which concepts the things must necessar-
ily agree . . . this question . . . is still left in a state of obscurity.
(Kant [1772], p. 72)

5
 THE LOGICA L MUST

His solution is the so-called ‘Copernican revolution’:

Up to now it has been assumed that all our cognition must

conform to the objects; but all attempts to find out some-
thing about them a priori through concepts that would extend
our cognition have, on this presupposition, come to nothing.
Hence let us once try whether we do not get farther . . . by as-
suming that the objects must conform to our cognition, which
would agree better with the requested possibility of an a priori
cognition of them, which is to establish something about ob-
jects before they are given to us. (Bxvi)

Here we see the beginnings of Kant’s transcendental idealism.

To sketch the story in crude outline,1 Kant takes humans to
be ‘discursive intellects’, cognizers who know the world indirectly,
via concepts:

. . . the cognition of every, at least human, understanding is cog-

nition through concepts, not intuitive but discursive. (A68/B93)

The contrasting case of the intuitive intellect is one whose intui-

tion is ‘spontaneous’,

. . . as, say, a divine understanding, which would not represent

given objects, but through whose representation the objects
themselves at same time . . . would be produced. (B145)

Such a God-like intellect actually creates objects by intuiting

them, so He cognizes them perfectly without appeal to concepts.

1. For more on the topics of this section, see [2007], §§I.4 and III.2.

6
 K a n t o n L ogic

At the other extreme, the intuition of the empirical intellect is en-

tirely ‘receptive’: for such an intellect

a representation is only a way in which the subject is affected by

the object . . . (Kant [1772], p. 71)

Like the intuitive intellect, the empirical intellect also has no need
for concepts, this time because the features of its representation
are directly stamped upon it by the object.2 Here the discursive
intellect stands apart: it requires an intuitive faculty to connect
its cognition to the world; not being divine, its intuitive faculty
isn’t spontaneous but receptive; not being empirical, its receptive
intuition isn’t entirely direct, but tempered by its forms of intui-
tion; and it also requires a second faculty, the understanding that
spontaneously employs concepts, marks or features ‘which can be
common to several things’ (A320/B377).
It isn’t entirely clear why Kant takes a discursive intellect to be
the only alternative to the intuitive and the empirical, but how-
ever that may be, we now have a fairly detailed characterization
of the components of discursive cognition: the ‘matter’ of experi-
ence, which the faculty of sensibility3 passively receives from the
world; the forms of intuition, which automatically organize the
raw matter so as to make it accessible to the faculty of understand-
ing; and the concepts supplied spontaneously by the faculty of
understanding. The details of the forms of sensibility vary from
one type of discursive intellect to the next – ours happen to be
spatiotemporal – but the workings of the understanding are the
same for all. Furthermore, in another unobvious leap, Kant tells us
that any cognition by a discursive knower must take one of twelve
2. Of course, one burden of the Critique is to show that this sort of cognition is impossible.
3. A passive, as opposed to spontaneous, intuition is termed ‘sensible’.

7
 THE LOGICA L MUST

precise forms – those listed in the Table of Judgments – and that

there are twelve corresponding pure concepts available to the dis-
cursive knower a priori – those listed in the Table of Categories.
Returning to the original challenge, then, how is it that we
humans can know some things about the world a priori, before
looking, so to speak? The answer is that the world we experience
is partly constituted by our forms of intuition and pure catego-
ries, so we can know a priori that it will display the features they
determine: for example, the world we experience will be spatio-
temporal; it will validate the theorems of geometry and the law
of causation, just as Kant thought it should. More generally, the
world as experienced by any discursive knower will display the
characteristic structures of the forms of judgment and pure cat-
egories, though it may not be spatiotemporal.
It’s important to recognize that this ‘Copernican revolution’
wouldn’t explain what Kant wanted explained if ‘the world we ex-
perience’ were just a contentious way of saying ‘the way the world
appears to perceivers with our particular human cognitive systems’
– where this is understood as the sort of thing described by empiri-
cal psychology. In fact, the findings of cognitive science often report
the ways in which our cognitive machinery tends to err, from our
susceptibility to optical illusions to our tendency to see physical
objects as continuous substance. However basic to our cognition
these may be, we obviously don’t know them a priori because we
don’t know them at all – they’re false, as careful observation and
science have eventually revealed. But Kant’s psychology isn’t empir-
ical like this; it’s transcendental, purportedly telling us the neces-
sary features that the world we experience must have. Viewed from
this transcendental perspective, the a priori features Kant notes
are ideal, due to our contributions, but viewed empirically, from
an ordinary or scientific perspective, they are real: the truths of

8
 K a n t o n L ogic

mathematics and the law of causation aren’t products of the vagaries

of our empirical psychology; they’re objective features of the world.
Thus transcendental idealism combines with empirical realism.
So where does logic fit into this picture? Though Kant’s in-
terests lie elsewhere, in mathematics and in parts of physical sci-
ence, we’ve seen the central role played by the logical forms in
the Table of Judgments: to take our examples from the relational
group, judgments can be categorical (‘As are Bs’, subject-predicate
form), hypothetical (‘if p, then q’), or disjunctive (‘p or q’).4 To each
of the forms of judgment there corresponds a pure category: for
example, the category of objects (substance) with properties (ac-
cidents) corresponds to categorical judgments; the category of
ground and consequent, 5 one substance or situation serving as the
ground for another, corresponds to hypothetical judgments.6 Since
the empirical world is partly constituted by the categories, we can
know a priori that it consists of objects-with-properties standing in
ground-consequent relations; we can know a priori that the world
has this much logical structure.7

4. For the record, Kant takes disjunctions to be exclusive.

5. In fact, Kant calls this ‘cause and effect’ (A80/B106), but as a pure category, it can’t
involve any spatiotemporality. This comes in only later, when the pure categories are
schematized to our particular forms of intuition, a process that would presumably go
quite differently for a discursive knower with some other forms of intuition. Since our
modern-day understanding of causation would seem to involve both spatial and tempo-
ral components, I use ‘ground-consequent’ for the pure category shared by all discur-
sive knowers and reserve ‘cause-and-effect’ for our own schematized version.
6. Kant’s well-known example has ‘there is perfect justice’ as the ground and ‘obstinate
evil will be punished’ as the consequent (A73/B98). Of course, ‘grass is green’ doesn’t
ground ‘2 + 2 = 4’ – 2 + 2 isn’t 4 because grass is green – despite the truth of the familiar
material conditional. See footnote 9 of chapter 2 and the surrounding text.
7. Strictly speaking, pure logic, what Kant calls ‘general logic’, depends only on the forms
of judgment, not the pure categories, which belong to ‘transcendental logic’. Still, no
world can be cognized with the forms alone – the categories must also be deployed – so
a world as experienced by a discursive knower must be structured by both. In the text, I
allow myself to speak loosely of the two in combination as ‘logical structure’.

9
 THE LOGICA L MUST

To see how this plays out in more detail, notice that our exam-
ple about red and green is just an ordinary disjunctive syllogism:
‘it’s red or green’ is the major premise; ‘it’s not red’ is the minor
premise; and ‘it’s green’ is the conclusion.8 Each of the three rela-
tional forms of judgment noted above provides the major premise
for one of the three forms of classical syllogism: categorical, hypo-
thetical and disjunctive. Judging under any of one of these three
forms involves adopting a rule, specifically a rule for subordinat-
ing an assertion to a condition. As described by Béatrice Longue-
nesse, these rules are:

. . . subordination of the predicate to the subject insofar as the

latter is the condition for the assertion of the predicate [in a cat-
egorical judgment], or subordination of the consequent to the
antecedent insofar as the latter is the condition for the assertion
of the former [in a hypothetical judgment], or subordination
of the complete division to the divided concept insofar as the
latter is the condition for the assertion of this division [in a dis-
junctive judgment]. (Longuenesse [1993], p. 94)

Thus Kant’s analysis of what judging consists in reveals that the va-
lidity of the corresponding syllogism is actually contained in the
act of synthesis required to form the major premise. Longuenesse
concludes:

Syllogism . . . is not a function of thought distinct from that of

judging. On the contrary, this function is in some sense ‘en-
cased’ in every judgment.9 (Longuenesse [1993], p. 95)

8. Again I’m assuming that even relevance logicians would grant the validity of this in-
stance of disjunctive syllogism (see footnote 3 of the Introduction).
9. She explains this as a second aspect of the priority of judgment for Kant: like concept
formation, syllogistic inference, too, comes embedded in acts of judgment (Longue-
nesse [1993], pp. 92–93).

10
 K a n t o n L ogic

Syllogistic inference arises from the structure of judgment.

For a particular example, consider the categorical judgment
that ‘all bodies are divisible’:

. . . being able to think such a judgment is being able to make

the inference: ‘The concept of the divisible applies to the con-
cept body’; now, the concept of body applies to objects x, y, z;
therefore the concept of divisible applies to these same objects’.
(Longuenesse [1993], p. 91)

The very act of judging a categorical judgment involves recogniz-

ing that if these things fall under the subject concept, then they
also fall under the predicate concept. Similarly, to judge a hypo-
thetical involves recognizing that if the antecedent holds, the con-
sequent does, too; in fact, Kant takes the validity of both modus
ponens and modus tollens to be present in the hypothetical judg-
ment.10 The case of disjunctive syllogism is less straightforward,
because the inferential content of a disjunctive judgment is actu-
ally hypothetical:

. . . from the truth of one member of the disjunction to the false-

hood of the others,[11] or . . . from the falsehood of all members
but one to the truth of this one. The former occurs through the
modus ponens . . . , the latter through the modus tollens. (Kant
[1800], p. 623)

So our sample logical truth is true because to judge that ‘it’s either
red or green’ is in part to recognize that ‘it’s green’ follows from
‘it’s not red’.
10. See Kant [1800], pp. 622–623.
11. Recall footnote 4.

11
 THE LOGICA L MUST

In this way, because discursive knowers must judge accord-

ing to the forms of judgment, their acts of judging are necessar-
ily bound by the laws of logic. Those acts of judging can cognize
the world because the world experienced by a discursive knower is
partly constituted by the structure of discursive thought; it must
have this much logical structure and satisfy these logical laws. And
since worlds shaped by the logical forms are precisely worlds struc-
tured by the pure categories,12 the laws of logic will be true in any
world that consists of objects-with-properties, standing in ground-
consequent relations, and so on. Speaking empirically, then, ‘if it’s
either red or green, and it’s not red, then it’s green’ is true because
of the way the world is; speaking transcendentally, it must be true
– a priori, necessarily – because a discursive knower’s world must
have this structure.
If we take one step further and schematize the pure catego-
ries, that is, if we bring them in line with our parochial forms of
intuition, then object-with-properties becomes spatiotemporal-
object-with-properties and ground-consequent becomes cause-
and-effect,13 thus validating also the mathematical and causal
truths in Kant’s motivating examples. We’ve seen that these fea-
tures of the world are empirically real and transcendentally ideal;
we now see that logical truths enjoy a similar status, except that
the cognitive structures involved in their transcendental ideality
include only the logical forms and pure categories14 common to
all discursive intellects, and not, as in the more familiar cases, our
specifically human forms of intuition as well. Just as the truths
of mathematics can be understood empirically, but the source of

12. See footnote 7.

13. E.g., ‘the warmth of the sun’ as ground (cause) and ‘the melting of the wax’ as conse-
quent (effect) (A766/B794). See footnote 5.
14. See footnote 7.

12
 K a n t o n L ogic

their apriority only transcendentally, the truths of logic can be un-

derstood empirically, but the force of the logical must explained
only transcendentally. The ordinary scientist comes to know that
simple logic is grounded in the way the world objectively is; only
the critical philosopher grasps the source of its apriority and ne-
cessity in the structure of discursive cognition.
We have here a distinctive account of the nature of logical truth
and the force of the logical must. Let’s now see how we might go
about updating and naturalizing it.

13
 C ha pt e r 2

Naturalizing Kant on Logic

Before undertaking the naturalization process in earnest, we need

to face the fact that Kant’s syllogistic logic is no longer the most
advanced thinking on the subject; we need to update the account
of logical form embodied in the Table of Judgments in light of
the great innovations introduced by Frege in the Begriffsschrift
(Frege [1879]).1 The first of these is the replacement of the Kan-
tian ­subject-predicate form (for example, ‘c has P’) with the more
flexible multi-place relation or, in Frege’s terms, with a function of
several arguments (‘R holds of c1, . . . cn’). This allows one among
several of these argument to be quantified at a time (‘For all x, R
holds of x, c2 , . . . , cn’), which opens the way in turn for iterated
quantifications (‘For every x, there is a y, such that R holds of x, y,
c3, . . . , cn’). For Kant, the quantificational forms are simply ‘All As
are Bs’ and ‘Some As are Bs’ (called ‘universal’ and ‘particular’ in
the Table of Judgments); the Fregean substitutes would be our fa-
miliar and more flexible universal and existential quantifiers. On
the logical connectives, such differences as exist between Kant
and Frege are less dramatic: negations and conditionals (Frege’s
term for hypothetical or if-then statements) are common to both,

1. Which isn’t to say that Frege thought of his work in this way, that Frege was any kind of
Kantian. No stand on this vexed question is intended here.

15
 THE LOGICA L MUST

though we may as well follow Frege in taking negation to operate

on entire judgments, not just on the copula. And while we’re up-
dating, let’s switch to inclusive disjunction and include conjunc-
tion as a form of its own.2
Adopting these post-Fregean modifications to the Table of Judg-
ments requires corresponding adjustments to the Table of Cat-
egories: for example, replacing the logical form subject-predicate
with f­unction-and-arguments (or for parallelism, arguments-and-­
function) goes hand in hand with replacing the category object-with-
properties with objects-in-relations. Likewise the form of universal
generalization, applied to one place in a judgment, brings with it the
category of universality: the property that an object enjoys when it
stands in the relations specified by the judgment – that very property
holds universally, of all objects. Finally, Frege himself recognizes ‘the
close affinity’ of the conditional form with ‘the important relation of
ground and consequent’ (Frege [1880/1881], p. 37), so there the cor-
responding category remains the same.3
Without following this exercise into its details,4 it seems fair
to conclude that a world partly constituted by our new forms of
judgment and pure categories would consist of objects that bear
properties and stand in relations to one another; that some of these
properties could be universal, holding of all objects; that some

2. Allison ([2004], p. 142) explains why conjunctions – or more precisely the more lim-
ited ‘copulative judgments’ of Kant’s day (e.g., ‘God has created all things and rules
over them’ or ‘God and one’s neighbor should be loved’) – aren’t included in the Table
of Judgments. In the cases of ‘hypothetical and disjunctive forms, the component judg-
ments are merely taken problematically within the judgment (neither affirmed nor
denied) and only the connection between them constitutes the proposition’. In con-
trast, ‘the elements of a copulative judgment are . . . complete judgmental units . . . that
may be affirmed or denied independently of their connection in the judgment. Conse-
quently, their combination does not . . . constitute a distinct moment of thought’.
3. See footnote 6 of chapter 1.
4. For more, see [2007], §III.3.

16
 Naturali z i n g K a n t o n L ogic

situations involving these objects would be complements (corre-

sponding to the logical form of negation), conjoinings (correspond-
ing to conjunction) or disjoinings (corresponding to disjunction)
of others; and that some interrelations between these situations
would be robust ground-consequent dependencies.5 This much
fairly minimal logical structure6 is enough to validate many clas-
sical inferences involving negation, conjunction, disjunction and
the quantifiers: introduction and elimination rules, distributive
laws, and DeMorgan equivalences, for example.7 Finally, given that
a substantive ground-consequent dependency requires a robust
connection between the ground and the consequent, it won’t obey
the usual equivalences for material conditionals – for example, it
may be that either grass isn’t green or 2 + 2 = 4, but it doesn’t follow
that 2 + 2 = 4 because grass is green – but at least modus ponens is
validated.8 So the world of our Kant-inspired, post-Fregean knower
– call it a Kant-Frege or KF-world – obeys a type of rudimentary

5. Notice that nothing here gives us reason to suppose that worldly indeterminacy or
vagueness is ruled out: it seems an open possibility, for example, that an object might
not enjoy a given property, but still not fail to enjoy it either. Such gaps would most nat-
urally be regarded as working their way up into complements, conjoinings and disjoin-
ings by the familiar three-valued rules. (That is, the complement of an indeterminate
situation is indeterminate; a conjoining obtains if both components obtain, fails if they
both fail, and is otherwise indeterminate; a disjoining obtains if one or the other com-
ponent obtains, fails if both of them fail, and is indeterminate otherwise. These rules are
common to strong Kleene and Lukasiewicz three-valued logics; they disagree only on
conditionals (here ground-consequent dependencies), which are considered separately
below.) Indeterminacy/vagueness seems to me beside the point for the larger themes at
issue between the Second Philosopher and Wittgenstein, so I leave it out of the spotlight
and make do with a few footnotes (7, 8, 14, 16, and 18). See also [2007], pp. 229–231,
240–244.
6. See footnote 7 of chapter 1.
7. The gaps produce some exceptions: the laws of excluded middle and non-contradiction are
indeterminate for indeterminate situations, and reductio ad absurdum is compromised.
8. Modus tollens is not, because of the gaps: if p is a ground for q, and q fails, it only follows
that p can’t be true; it could be either false or indeterminate.

17
 THE LOGICA L MUST

logic.9 Speaking empirically, its validities hold because of the logi-

cal structure of the world; speaking transcendentally, they hold
because of the structure of the (Kant-inspired, post-Fregean) dis-
cursive cognition.
With this updated Kantianism in place, let’s turn to the prob-
lem of naturalization. The need for some such adjustment arises
from the two-level structure of the Kant-Frege position outlined
here: the Second Philosopher is perfectly at home with empirical
inquiry – it’s what she does all the time – but transcendental in-
quiry is another story. Speaking transcendentally, our Kant-Frege
theorist tells her, for example, that the existence of individual ob-
jects like trees and stones and planets is actually a contribution of
her cognitive structuring, not present in the world as it is in itself.
This comes as surprising news, but she will, characteristically,
reply that she sees no evidence that the physical structure pres-
ent where that stone is located – a distinct difference in the types
of molecules inside as opposed to outside its rough boundary, an
electromagnetic field inside that boundary that’s strong enough to
keep her finger from penetrating the surface when she presses on
it, and so on – she sees no evidence that this structure is dependent
on her psychology in any way. Of course, the Kant-Fregean replies,
speaking empirically, the existence of the stone isn’t dependent on
anybody’s psychology, but I’m speaking transcendentally! The
Second Philosopher is open-minded enough not to dismiss out
of hand the possibility that there are other methods she hasn’t
hit upon for finding out what the world is like, so she’ll want to

9. Obviously there is some distance between this rudimentary logic and full classical
logic. This distance is covered, I suggest, by a series of restrictions and idealizations
(including the move from ground-consequent to the material conditional). Since many
deviant logics result from rejecting one or another of these adjustments, the debate
should focus not on what’s true – it’s in the nature of restrictions and idealizations to be
false – but on what’s effective for the purposes at hand (see [2007], §III.7).

18
 Naturali z i n g K a n t o n L ogic

know more about this transcendental inquiry, about its scope and
methods,10 but the Kant-Fregean’s explanations are unlikely to
convince her that his transcendental inquiry is either motivated
or viable: for example, no mathematical geometry seems to her to
provide a priori knowledge of the world, so she feels no pressure
to explain how it does so; and she finds no persuasive account of
the methods of ‘transcendental’ psychology or their reliability. In
particular, then, she’ll see no strong case for the proposed account
of logical truth and the logical must.
This is where the advertised naturalizing move comes in: sup-
pose we simply collapse the two levels into one, viewing both parts
of the account empirically, bringing them both within the Second
Philosopher’s larger inquiry.11 The result is a position that goes like
this: (1) rudimentary logic is true of the world because it’s struc-
tured as a KF-world, and (2) our basic cognitive machinery de-
tects and represents this KF-structuring (which probably explains
why we tend to find the simplest logical connections entirely ob-
vious). As in the empirical level of our Kant-Fregean picture, the
structuring in (1) is taken to be fully objective, not something
imposed by the machinery in (2), so unless we’re inclined to take
the structural similarity as an inexplicable coincidence, we should
add: (3) human cognitive machinery is as it is because we live in a
KF-world.

10. This sets her off from naturalists who would reject transcendental philosophy as ‘un-
scientific’ from the start; she has no criterion on which to base such a judgment (cf.
footnote 5 of the Introduction). For more, see, e.g., [2007], §§I.1 and I.4, or [2011a], §I.
11. I don’t pretend that the position generated by so radical an adjustment remains close
enough to Kant to qualify as ‘neo-Kantian’: as we’ve seen, transcendental psychology
can’t become empirical without destroying Kant’s explanation of why the world of a dis-
cursive knower must have the relevant features. I also don’t intend the naturalized view
that emerges to include any other Kantian claims or to rest on any Kantian arguments;
it stands as best it can on (descendants of) the three theses to follow. (Thanks to Ethan
Galebach for pressing me on this point.)

19
 THE LOGICA L MUST

We have here a candidate for fully second-philosophical posi-

tion on logic, but is it true? This is a big question – there isn’t room
for a full accounting here – but let me sketch a rough answer.12
Taking (1) first, we’ve derived rudimentary logic as the logic of
a KF-world, so the open question is whether or not our world is
indeed KF. In fact, as we saw in the Second Philosopher’s reac-
tion to Kant’s idealism, there is overwhelming evidence of the sort
floated there that the familiar physical world of trees and planets,
cats and people really does come (roughly) divided into these
individual objects. We now add with the same sources of convic-
tion that these objects really do stand in various relations to one
another – the cat is in the tree, this person sees the planet – and
that some such situations ground others – the apple tree’s presence
is the result of Johnny’s having planted a seed. Indeed, it might
seem that tying logic in this way to the structure of a KF-world
in fact imposes no constraint, because a logic true in such a world
is true in all possible worlds. (This is one of those philosophers’
accretions mentioned in the Introduction: logic is necessary.) An-
other way of putting this is to claim that all possible worlds are KF-
worlds, and sure enough, our notion of a possible world is more or
less like, say, a model from model theory: a set of individuals with
some relations on them. Notice also, that if something like (2) is
correct, if KF-structuring is part of our most basic conceptual ma-
chinery, it might be difficult for us to imagine a world of a sharply
different kind.
Though this is an entirely natural line of thought, I think it’s too
quick. Of course, for the reasons given, we can expect that it will be
hard for us to fully imagine a non-KF ‘possible world’, but let me try
to motivate one such idea nonetheless. Consider (what we might

12. For more, see [2007], §§III.4 and III.5.

20
 Naturali z i n g K a n t o n L ogic

call) a Creator World. This notion derives from the intuitionist’s

Creative Subject, an idealized figure whose mental constructive
activities are understood to generate the ontology of intuitionistic
mathematics, except that here we imagine the Creator to gener-
ate an entire world. In the background, we rely on the machinery
of Kripke models for intuitionistic logic as our guide,13 so that a
Creator World will validate a logic different from our rudimentary
logic. In particular, it will be a world in which the complement of
the complement of a situation needn’t be the original situation, in
which not-not-p doesn’t imply p.14
To see this, imagine the Creator acting in a simple, discrete
time sequence – t0, t1, t2 , . . . – and having powers something like
those of Kant’s intuitive intellect: by thinking of a thing, he brings
it into being. Let’s stipulate that he’s also quite tenacious: once he
thinks of a thing one way, he never changes his mind. Our focus is
on the world so generated, the one the Creator’s activity brings into
being, not on the Creator himself, who remains offstage, behind a
curtain, so to speak, like the Wizard of Oz. Such a world couldn’t
contain changeable human beings like ourselves, so imagine the
knower cognizing this world to be a disembodied intellect who ex-
periences the world somehow,15 but isn’t in it.
What is this world like? By thinking a thing a, the Creator
brings it into being; by thinking of it with property P, he brings
it about that Pa; by thinking of it in relation R to object b, he

13. See, e.g., Kripke [1965].

14. On the Kleene semantics alluded to in footnote 5, not-p obtains if p fails, fails if p ob-
tains, and is indeterminate if p is indeterminate, so if not-not-p obtains, not-p must fail,
and so p must obtain. I focus on this example rather than the more familiar failure of
the Law of Excluded Middle, because the truth-value gaps in rudimentary logic also
lead to the failure of LEM (see footnotes 7, 16, and 18). But intuitionistic logic is not
three-valued logic, or n-valued logic for any n (Gödel [1932]).
15. See footnote 17.

21
 THE LOGICA L MUST

brings it about that Rab; and so on for atomic formulas. Now

what would it be like, for example, for a to fail to have property
P in this world? It wouldn’t be enough for the Creator to have
thought of a without having thought of it with P, for he may wish
to leave open the possibility that he will think of it with P some-
time in the future; if he were to explicitly regard Pa as failing
now, he could only retain the right to think Pa sometime in the
future by allowing the possibility of one day changing his mind,
which his tenacity clearly prevents him from doing; so he retains
that future freedom by leaving Pa undetermined, by settling no
fact of the matter.16 For not-Pa to obtain now, what he has to do
now is decide once and for all that he will never, ever think of
a as having P, perhaps because he can see now that if he did so,
he’d have to think something he’d be dead set against thinking.17
For example, we assume he’s dead set against thinking a contra-
diction, so if he’s already thought Pa, he can know for sure he’ll
never think not-Pa, so he’s sure right now of not-not-Pa.
From here, the other Boolean connectives are straightfor-
ward: the Creator thinks a disjunction if he thinks one or the
other of its disjuncts; he thinks a conjunction if he thinks both

16. This isn’t the indeterminacy of vagueness: if it’s indeterminate, given Joe’s current
number of hairs, whether or not he’s bald, this is a feature of the property of baldness,
not a matter that simply hasn’t been settled yet, that might later be filled in (unless, of
course, he loses or gains hairs!). At a first approximation, vagueness is better modeled
by the three-valued logic of KF-worlds. (Putnam [1983] once proposed intuitionistic
logic as a solution to the sorites paradox, but the suggestion meets with difficulties. See
Williamson [1996] and the references cited there.)
17. What would the experience of our detached Cartesian observer of this world be like?
Hard for us to imagine, surely, but suppose she ‘perceives’ packets of information se-
quentially, object-by-object, property-by-property, so to speak, more or less as the Cre-
ator creates them. She might first receive the information that an object exists, then that
it has some property or bears some relation to an already existing object, later that it
lacks another property, etc. In the case at hand, she’d first be aware of a’s existence, but
still be in the dark about whether or not it has P; at some later time, she might learn that
it has P, that Pa obtains, or that it will never have P in other words, that not-Pa obtains.

22
 Naturali z i n g K a n t o n L ogic

conjuncts. Thinking a conditional takes a bit more effort: to think

(if p, then q), he has to consider what would happen if, now or at
some future time, he were to think p, and conclude that he would
then be sure to think q. For example, even if he’s now uncommit-
ted on Pa, he can see even now that if, at some future time, he were
to think Pa, he would then have to declare himself permanently
against not-Pa, which is to say, he would then think not-not-Pa.
So he now thinks (if Pa, then not-not-Pa). Finally, he now thinks
(there is an x with P) if he thinks of an a with P. More strenuously,
he now thinks (all x have P) if he considers all the things he thinks
of now, and all the things he might think of in the future, and de-
cides that he will think each and every one of them with P. For ex-
ample, if he now thinks not-(there is an x with P), he can see that
if he now or in the future he thought Pa for some a, he’d also think
(there is an x with P), and since he’d dead set against thinking a
contradiction, he knows right now that he’ll never do this. So he
now thinks not-(there is an x with P) implies (for all x, not-Px).
But notice that if he now thinks not-not-p, he only commits
himself never to think not-p, that is, never to commit himself to
never think p. Whatever mental feat this might involve, it doesn’t
add up to thinking p, so not-not-p doesn’t imply p. And thus it tran-
spires that the logic of a Creator world isn’t our rudimentary logic.
But why not? After all, a Creator world has objects-in-relations
with ground-consequent dependencies; if this is all it takes to val-
idate rudimentary logic, then a Creator world shouldn’t deviate
from a KF-world in that respect. This question, the comparison
with Creator worlds, drives us deeper into the implicit structure
of KF-worlds. Notice first that unlike Creator worlds, KF-worlds
have no temporal index: when some aspect of our world is accu-
rately modeled by a KF-structure, what obtains or doesn’t obtain
depends only on conditions in the present; in contrast (to take just

23
 THE LOGICA L MUST

one example), the Creator’s negations hinge not only on what he’s
thought so far, on conditions in the present, but also on what he
looks forward to thinking in the future, much of which isn’t con-
strained by his choices up to now. This produces an asymmetry
in Creator worlds between Pa and its negation, not-Pa: for either
to obtain now requires a commitment for the future – always to
think Pa or never to think Pa, respectively – but both are commit-
ments involving P. No such asymmetry turns up in KF-worlds:
there it was simply stipulated that some situations are comple-
ments of others; the two situations involved have equal standing,
each as the complement of the other; neither enjoys any preferred
status. So we see that, in fact, KF-structuring embodies the Bool-
ean idea of negation as a simple toggle, which is why not-not-Pa
implies Pa in our rudimentary logic. This is also why, from a KF
point of view, negation behaves strangely in a Creator world:
even if the Creator has decided on not-not-Pa, he might never get
around to thinking Pa.18
We could consider a different kind of Creator, a Decisive Cre-
ator, who prefers, as a general rule, to think of an object a as out-
right lacking any property P that he hasn’t explicitly thought of a as
having. If this Decisive Creator were otherwise like our Creator in
refusing ever to change his mind, then his world would boil down

18. This type of failure of follow-through on the part of the Creator also explains why he
doesn’t regard the Law of Excluded Middle as a general rule: though he hasn’t thought
p, as long as he retains the right to think it later, not-p doesn’t obtain either, and he
might want to retain this right forever. As noted in earlier footnotes (5, 7, 8, 14, and
16), KF-worlds also allow for cases in which neither p nor not-p obtains. Still, as noted
in footnote 14, intuitionistic logic isn’t a three-valued logic, and we can now see one
reason why: intuitionist logic validates not-not-(p v not-p), but rudimentary logic,
with its toggle-like negation, does not. Both agree that not-not-(p v not-p) is equivalent
to not-(not-p & not-not-p), but for the intuitionist the latter is easily validated – the
Creator is dead set against imagining ‘not-p & not-not-p’ because he’s dead set against
imagining a contradiction – while for the rudimentary logician, whenever p itself falls
in a gap, so do ‘p v not-p’, ‘not-not-(p v not-p)’ and ‘not-p & not-not-p’.

24
 Naturali z i n g K a n t o n L ogic

to a series of identical KF-worlds,19 and rudimentary logic would

apply. A more interesting alternative would be a Flexible Creator,
who employs the same general rule about negation as the Decisive
Creator, counting every property he doesn’t explicitly apply to an
object as failing to hold of that object, but who allows himself to
change his mind, letting an object enjoy a property now but lose it
later, and vice versa. Every temporal stage of the Flexible Creator’s
world would be a KF-world, so again rudimentary logic would
apply there, but he’d have to be careful not to mix different stages
on pain of contradiction. This would be a world that begins to re-
semble our own. But neither of these figures is our Creator – he
doesn’t think decisively or flexibly – and it’s the logic of his worlds
that we’re interested in contrasting with rudimentary logic.
The temporal element and the Creator’s own quirks give rise
to another contrast in the case of quantifiers. In a KF-world, the
quantifiers generalize the simple Boolean disjunction and con-
junction, so that (there is an x with P) is equivalent to not-(for
all x, not-Px): in particular, if (everything is not-red) fails, then
(something is red) obtains. But our Creator thinks differently. He
refuses to settle for (there is a red x) unless he’s explicitly thought
of some a as red. And even if he’s decided on not-(for all x, not-Px),
he can’t be trusted ever to get around to thinking of a particular x
with P. For our disembodied cognizers out to investigate this sort
of world, again, the rudimentary logic of KF-worlds is not reliable.
But enough of this. All Creator worlds have done, at best, is cast
doubt on the necessity of rudimentary logic, on the understanding
of ‘possible world’ that requires KF-structuring; the question we
set out to answer – about the viability of (1), particularly the claim
that our world is KF – remains untouched by these considerations.

19. Indeed gap-free KF-structures.

25
 THE LOGICA L MUST

Concerning that question, we’ve seen that there’s plentiful common

sense and scientific evidence that our familiar physical world does
have this structure, but there’s also an important challenge from
the micro-world, the world of quantum mechanical effects. The
famous twin-slit experiment casts doubt on our usual understand-
ing that objects move on spatiotemporally continuous paths,20 but
to revert to Kantian terms for a moment, this would challenge the
universality of our human forms of intuition, not the more general
notion of object common to all discursive intellects, whatever form
their particular sensibility might take – and the reliability of logic
depends only on the latter. Restated in second-philosophical terms,
this becomes part of our current claim (1), that the reliability of
rudimentary logic rests only on the world’s KF-structuring and,
in particular, not on its spatiotemporal features. Other anomalies,
however, do strike at the heart of the pure categories (Kant) or KF-
structuring (the Second Philosopher): for example, quantum sta-
tistics call the notion of an individual object into question,21 and
the Stern-Gerlach experiments cast doubt on the sense in which
particles can properly be said to have properties in the usual (KF)
ways.22

20. The problem is that it seems impossible to ascribe a continuous trajectory to an elec-
tron traveling between a source and a screen separated by a barrier with two slits. See
[2007], pp. 236–237, for discussion and references.
21. Suppose we flip two coins identical in composition, shape, weight, etc. Then the prob-
ability of an outcome with exactly one head is ½, because two of the four equi-probable
possibilities meet that description (that is, first coin heads, second tails, and first coin
tails, second heads). But with two subatomic particles of equal mass, charge, spin, etc.,
the statistics don’t seem to distinguish between the state with the particles in one order
and the state with the particles in the reverse order. See [2007], p. 237, for discussion
and references.
22. This is described briefly in the following paragraph (see also [2007], pp. 238–239).
Quantum mechanics also presents difficulties for our usual notions of dependency –
some correlations appear not to take any of the forms we’re familiar with, not causal, not
semantic, not logical – but I won’t follow up on this line here (see [2007], pp. 239–240).

26
 Naturali z i n g K a n t o n L ogic

If this is right, and if logic does depend on the sort of KF-

structuring that’s not clearly present in the micro-world, then we
should expect breakdowns in rudimentary logic there, and this is
just what we do find. For example, in the Stern-Gerlach case, we
have good reason to suppose that electrons have spin up or spin
down, spin right or spin left, that is, for a given electron a, (either
a is spin up or a is spin down) and (either a is spin left or a is spin
right). What the experiments show is that each of the disjuncts of
((a is spin up and spin left) or (a is spin up and spin right) or (a is
spin down and spin left) or (a is spin down and spin right)) – all
four of these disjuncts fail. Here we see the expected failure of ru-
dimentary logic, in particular, of the distributive law.23
It’s worth noting, in passing, that we might be able to construct
from these ingredients an alternative to Creator worlds for mount-
ing our plausibility argument against the necessity of logical truth;

23. ‘Failure’ might mean that in this case the law fails outright (the premise obtains and the
conclusion fails), or that it forms part of a logical system that can’t be successfully ap-
plied. There may well be important differences between these two (though see [2007],
pp. 279–280), but I use the expressions interchangeably here. Also, I don’t mean to sug-
gest that we need to devise some other logic to apply in the quantum world. Perhaps
we’ll eventually get to this point, perhaps not. (Of course new logics for quantum me-
chanics have been proposed, but so far without conspicuous success. See [2007], pp.
276–279, for some discussion and references.) For now, we reason perfectly well about
the quantum world using the mathematical formalism of Hilbert spaces, which display
a perfect KF-structure throughout, and are thus suited to rudimentary logic (and even
to full classical logic, for that matter, because the restrictions and idealizations alluded
to in footnote 9 hold unproblematically for classical mathematical entities). Finally, I
also note that some commentators prefer to say – not that ((either a is spin up or a is
spin down) and (either a is spin left or a is spin right)) obtains, while ((a is spin up and
spin left) or (a is spin up and spin right) or (a is spin down and spin left) or (a is spin
down and spin right)) fails – but rather that, in this situation, none of the purported
disjuncts of the conclusion is meaningful. So the distributive law doesn’t fail outright
in this situation, because the purported conclusion isn’t even well-formed. For my pur-
poses, it’s enough that the familiar rudimentary inference from the conjunction to the
disjunction isn’t generally reliable, because sometimes the disjunction either fails or is
ill-formed.

27
 THE LOGICA L MUST

we could try, that is, to conceive an entire world whose behavior

corresponds closely to the sort of phenomena we actually witness
only in the quantum world. For example, suppose the cognizers in
this QM-world perceive only in discrete flashes, and then only one
property at a time. Suppose they have reason to believe that the
‘objects’ of their world come in two colors (say, red and green) and
two shapes (say, round and square) and when they arrange their
counterpart to Stern-Gerlach experiments, they come to see that
the distributive law is false.24
I won’t go into any more detail on Creator worlds or QM-worlds,
because we can now also see that there’s actually not much point in
this exercise. We’ve noted that these imaginary examples are really
just challenges to our intuitive sense that non-KF-worlds aren’t pos-
sible, to our intuitive understanding of what counts as ‘possible’. To
support the contingency of logic, our actual world is enough all by
itself, because, in fact, rudimentary logic isn’t universally applicable,
isn’t true everywhere, and it takes no subtle theory of modality to
conclude that what isn’t true isn’t necessarily true. KF-structuring
is best thought of as a kind of template, one that fits onto some situ-
ations and not onto others. Where it does fit, rudimentary logic is
reliable. Elsewhere, it is not.
In any case, the moral for (1) is that we can’t assume all aspects
of our world display the kind of KF-structuring that underlies ru-
dimentary logic. So (1) should be modified to: (1') rudimentary
logic is true of the world insofar as it is a KF-world, which in many
but not all respects it is. (2) then becomes (2'): our basic cognitive
machinery detects and represents (some of) such KF-structuring
as there is in the world.

24. In fact, if the cognizers of this QM-world evolved in anything like the way we evolved
in our world (see below), presumably they wouldn’t feel any inclination to believe the
distributive laws in the first place.

28
 Naturali z i n g K a n t o n L ogic

An impressive range of evidence in contemporary cognitive sci-

ence suggests that (2') is true.25 Current work supports the notion
that prelinguistic infants, given ordinary maturation (as opposed
to learning), perceive a world of cohesive, solid bodies that travel
on continuous paths and preserve their identity through time and
when out of sight. The leading theory traces this ability to a mid-
level system (a system between the purely sensory and the con-
ceptual) that opens a so-called ‘object file’ when an object is first
noticed, then tracks that individual separately from others even
before particular properties are assigned and included in the file.
(A canonical illustration of the function of an object file in adults
is expressed in the familiar exclamation: ‘It’s a bird, it’s a plane,
it’s Superman!’ The object is individuated and tracked well before
any further properties can be successfully attributed to it.) Very
young children, again by ordinary maturation, are also capable of
classifying objects by their features, individual and relational,26
and of recognizing both Boolean-style correlations between these
properties (that is, of grouping not only by P and by Q , but also by
compounds like (P & Q) and (P v Q)) and dependencies between
one situation and another. So it seems fair to say that our ability to
detect the presence of the world’s KF-structures is part of our most
fundamental cognitive endowment.
This leaves (3), now modified to (3'): human cognitive machin-
ery is as it is because we live in a largely KF-world and interact almost
exclusively with its KF-aspects.27 For present purposes, it doesn’t
matter whether these cognitive mechanisms are formed by evolu-
tionary pressures, developed by processes of maturation, or learned

25. See [2007], pp. 245–264, for further discussion and references.
26. Notice that perceiving when an object fails to have a property is implicit in classifica-
tion by the property in the first place.
27. See [2007], pp. 264–270, for further discussion and references.

29
 THE LOGICA L MUST

in early childhood, but current evidence counts strongly against any

dependence on language, or even on experience with manipulating
objects in the environment. The stark simplicity of the abilities in
question – plausibly essential on the savanna as well as in twenty-
first-century civilization – along with their presence in nonhuman
animals like birds and primates clearly points toward an evolution-
ary origin. But let me emphasize that the order of argument here
doesn’t follow the well-known fallacious line: these cognitive mech-
anisms have evolved, therefore they must deliver truths. In contrast,
the case here begins from evidence that much of the world has KF-
structuring and that humans come equipped to detect and repre-
sent that structuring very early in life. If evolution enters the story
at all, it’s well downstream, as an explanation of how our cognitive
machinery came to be well-tuned to these objective features of the
world (that is, in defense of a particular version of (3')).
Here, at last, we have our second-philosophical account of logic:
logic is true of those aspects of the world that enjoy KF-structuring–
it can be thought of as a template that fits onto the world when and
only when certain conditions are met – so our red or green, non-
red object, part of the ordinary KF macro-world, is indeed green.
This is a brand of realism – logic reflects objective truths about the
world – but without many of the features that typically accompany
such realism: logical truth isn’t necessary, but contingent on the
presence of the requisite structures;28 logic doesn’t describe a world

28. The Second Philosopher is happy to say that the applicability of the logical template,
the validity of the logical laws, depends on the presence of KF-structuring. In standard
philosophical terminology, this is to say that logic is contingent, but it’s worth noting
that to say this much, the Second Philosopher needn’t (and I think, doesn’t) have any
developed contrastive notion of metaphysical necessity as many philosophers intend it.
(Thanks to Eileen Nutting for pressing me on this point.) In what follows, I use the term
‘metaphysical necessity’ or just ‘necessity’ for the philosopher’s notion, whatever that
might be, and ‘logical necessity’ and ‘physical necessity’ for the second-philosophically
friendly notions of what’s required by the laws of logic and physics, respectively.

30
 Naturali z i n g K a n t o n L ogic

of abstracta, but our own familiar physical world. In some sense,

our belief in various simple logical connections would appear to be
a priori, present before experience; given that our basic cognitive
machinery is largely reliable in the familiar macro-world, an exter-
nalist epistemologist might take this much logical belief to count
as a priori knowledge.29 But the force of the logical must is compro-
mised: the impression of logic’s metaphysical necessity arises from
the fundamental role of the relevant cognitive machinery – it’s dif-
ficult for us to think in non-KF terms, perhaps even impossible (as
Kant would have it)30 – which is why we find KF-structuring built
into our very notion of ‘possible world’ (presumably those who
take logic to be metaphysically necessary also take KF-structuring
to be necessary). Still, we can make sense of the notion that logi-
cal necessity is stronger than physical necessity, because physical
necessity depends of features of our world that go beyond its KF-
structuring (for example, spaciotemporality, the ‘laws’ of physics,
etc.).31
Perhaps it’s worth a mention in passing that this naturaliza-
tion of Kant-Fregeanism produces something closely analogous
to a Cartesian position that even generous commentators clas-
sify as ‘infamous’. 32 Emphasizing God’s omnipotence, Descartes

29. See [2007], p. 274. Still, it would take a posteriori investigation of the sort sketched
here to reveal that our primitive logical beliefs are true, that our cognitive mechanisms
are largely reliable, and thus that these logical beliefs count as a priori knowledge (for
the externalist). (Compare: on Kant’s view, we empirical inquirers know the truths of
mathematics and logic a priori, but we can’t appreciate that we do so without engaging
in transcendental inquiry.)
30. If so, it seems we may never have a satisfactory understanding of quantum mechanics.
31. Again, this corresponds to the Kantian distinction between dependence on the struc-
ture of the discursive intellect and dependence on that plus the spatiotemporal forms of
human intuition.
32. See, e.g., Cunning [2006], §3.

31
 THE LOGICA L MUST

held that He could have made the laws of logic other than they are:
since ‘the power of God cannot have any limits’, it follows that

God cannot have been determined to make it true that contra-

dictories cannot be true together, and therefore that he could
have done the opposite. (Descartes [1644], p. 235)33

Of course we can’t grasp this possibility, because our minds are

so created as to be able to conceive as possible the things which

God has wished to be in fact possible, but not be able to con-
ceive as possible things which God could have made possible,
but which he nevertheless wished to make impossible. (Des-
cartes [1644], p. 235)

This view opened Descartes to considerable scorn, as opponents

posed riddles like ‘could God have made the world independent of
Him?’, but Margaret Wilson points out that

[t]here is . . . a lively controversy among some leading philoso-

phers of the present [twentieth] century whether logical neces-
sity might not go the same way as the traditional ‘necessity’ of
Euclidean geometry. (Wilson [1978], p. 126)

She goes on to suggest that

what is really extraordinary is not Descartes’s creation doc-

trine itself [that is, the doctrine that God created the so-called

33. Cf. Descartes [1630], p. 25: ‘he was free to make it not true that all radii of a circle are
equal –just as free as he was not to create the world’.

32
 Naturali z i n g K a n t o n L ogic

‘eternal truths’], but the fact that he has not been given more
credit for arriving at it. (Wilson [1978], p. 126)34

What Wilson has in mind here is Hilary Putnam’s claim that, for
example, empirical pressures from quantum mechanics might
overthrow classical logic, just as empirical pressures from the
theory of gravitation once overthrew Euclidean geometry. 35
Now Putnam’s point here is primarily the epistemic claim that
logic is empirical, that we could be led to alter our logical beliefs
in light of empirical evidence. 36 In contrast, Descartes, like the
Second Philosopher, is emphasizing that logic is contingent, that it
could have been otherwise. 37 In fact, given that God, having made
logic as he has, also made our minds unable to conceive any alter-
native, it seems unlikely that Descartes could envision Putnam’s
type of empirical change. Similarly, though the Second Philoso-
pher is less confident than a straight Kant-Fregean that we couldn’t
come to see the world in terms other than KF-structuring, she
does recognize that doing so would mean a difficult and perhaps
unattainable shift in our thought processes. Finally, Descartes and
the Second Philosopher both take the apparent necessity of logical
truth to be an illusion, 38 and their grounds for this are analogous:

34. She proposes a plausible explanation: ‘Perhaps the theological basis of his position has
stood in the way of a fair historical assessment of the original and important insight it
embodies’ (op. cit.).
35. See Putnam [1968].
36. If we were so led, we’d come to realize that these logical beliefs weren’t true, and hence
weren’t necessarily true, after all, but this wouldn’t automatically preclude our regard-
ing our new logical beliefs as necessary (assuming necessary truths can be known em-
pirically, as many contemporary philosophers do).
37. This position is awkward for Descartes in ways it’s not for the Second Philosopher, be-
cause Descartes draws such a close connection between conceivability and possibility
(see Wilson [1978], p. 126).
38. See Frankfurt [1977]. Of course, commentators disagree on this point (Cunning
[2006], §3, gives a summary).

33
 THE LOGICA L MUST

for Descartes, logic appears necessary to us because God made

our minds as he did; for the Second Philosopher, logic appears
necessary to us because evolution made our minds as it did. 39 So
Descartes’ position seems more closely analogous to the Second
Philosopher’s than it is to Putnam’s.40
If we were to attempt to generate the second-philosophical
position from Descartes’ position by a single naturalizing move,41
it would be a familiar one – replace God’s benevolent hand with
evolutionary pressures from the environment42 – but this by itself

39. For rhetorical effect, I assume here that evolution is responsible, but as noted above, this
is inessential to the second-philosophical position on logic.
40. Conant ([1991], pp. 123–124) draws the parallel between Descartes and Putnam this
way: he describes Descartes as thinking it’s ‘hubris’ to suppose that ‘an omnipotent God
. . . would be . . . bound by the laws of logic’, and Putnam as thinking it’s hubris to suppose
that ‘the ongoing activity of scientific inquiry . . . will be forever bound by the laws of
classical logic’ – in other words, ‘science is substituted for God’. The parallel I’m draw-
ing between Descartes and the Second Philosopher is different: for the Second Philoso-
pher, it’s hubris to suppose that the world is everywhere and always ‘bound by the laws
of classical logic’, whatever may or may not be true of our scientific activity (for now,
it’s not clear whether or not our theorizing is ‘forever bound’ by rudimentary logic – it
might be). In other words, the world, not our theorizing about the world, is substituted
for God. Descartes would clearly agree with the Second Philosopher that the bounds on
our thought or theorizing aren’t what’s at issue in the claim that logic is contingent.
41. The need for naturalization here is different from the case of Kant’s transcendentalism.
There a new type of inquiry was proposed, and the Second Philosopher was unable to
affirm its apparent goals or confirm the reliability of its purported methods. This time
around, Descartes is simply adding God as an extra explanatory factor: the world and
our minds are as they are because God so exercised his will. The Second Philosopher
finds no support for this hypothesis, and considerable support for alternative theories
of how the world comes to be as it is and how our minds come to be as they are.
42. My favorite implementation of this strategy would be as applied to Thomas Reid on the
sense of smell: ‘That all bodies are smelled by means of effluvia which they emit, and
which are drawn into the nostrils along with the air, there is no reason to doubt. So that
there is manifest appearance of design in placing the organ of smell in the inside of that
canal, through which the air is continually passing in inspiration and expiration’ (Reid
[1764], §II.1, pp. 25–26). How clever of the Almighty, to have placed the organ of smell
inside the nose, where the breath is constantly moving in and out!
Notice also that the Cartesian counterpart to the bad evolutionary argument dismissed
in the discussion of (3') above is roughly: since a benevolent God gave us these beliefs
(faculties), they must be true (reliable). Descartes seems to have fallen for something
like this, but arguable Reid did not: ‘The Supreme Being intended, that we should have
such knowledge of the material objects that surround us, as is necessary in order to our

34
 Naturali z i n g K a n t o n L ogic

leaves us with a position much less informative than naturalized

Kant-Fregeanism: ‘logical truth is contingent on God’s will’ be-
comes ‘logical truth is contingent on the way the world is’, but we’re
told nothing about which aspects of the world are responsible for
the validity of the logical laws or the relevant environmental pres-
sures. Because of his holism, Quine’s empiricism about logic has
the same defect: no particular features of the world are responsible
for the truth of logical laws because no particular features of the
world are responsible for the truth of anything; our claims ‘face
the tribunal of sense experience not individually but only as a cor-
porate body’ (Quine [1951], p. 41). Putnam apparently inherits
this holistic feature from Quine. In contrast, of course, the Second
Philosopher’s naturalized Kant-Fregeanism is quite explicit about
the KF-structures responsible for both logical truth and logical
cognition.
This detour highlights the quite different naturalizing move at
work in the central line of derivation we’ve been following: to gen-
erate the second-philosophical position out of Kant-Fregeanism,
we simply took a two-level, transcendental/empirical position and
collapsed it down to one empirical level.43 Before the collapse,

supplying the wants of nature, and avoiding the dangers to which we are constantly
exposed; and he has admirably fitted our powers of perception to this purpose’ (Reid
[1785], §II.V.10, p. 101). See de Bary [2002], chapter 10, for discussion of Reid on the
possibility of God’s ‘paternalistic deception’.
43. Though perhaps not as widely applicable as the God-to-evolution move, I suspect the
two-level-collapse might be effective in other cases as well. Consider, for example, a
straightforwardly two-level reading of Carnap [1950]: externally, one chooses, say, the
linguistic framework of contemporary science, with its particular logic rather than an-
other, on purely pragmatic grounds; internally, the choice having been made, logic is a
priori, true by virtue of the meanings of the words involved. Collapsing the levels pro-
duces a view according to which logic is true by virtue of meaning, but the dependence
of our meanings on features of human cognition and human practice and their place
in the world is acknowledged. This opens the way to the question, why do we have the
meanings we do?, which stands to tell us something about the ground of logical truth
(see [2012] for a bit more on this line of thought). This theme emerges again below, ap-
proached from a different direction.

35
 THE LOGICA L MUST

the logical structures of the world were empirically real and tran-
scendentally ideal. After the collapse, the empirical component
remains in place – the logical structures are objective features of
the world – but the transcendental psychology – the world is logi-
cally structured because it’s partly constituted by the forms of dis-
cursive cognition – is replaced with an empirical psychology that
posits the same cognitive forms, but leaves the world’s structure
independent. From there, the additions and adjustments required
are straightforward: on the worldly side, the vagaries of quantum
mechanics need to be incorporated; on the mental side, the corre-
spondence between mind and world needs a new explanation now
that the constitutive element has disappeared; overall, empirical
evidence for the theory has to be collected and assessed. If I may
put it this way, once the collapse is implemented, the rest seems to
follow as a matter of course, like a row of dominoes, or falling off
a log.
Now let’s see how all this looks when we begin from the Trac-
tatus instead.

36
 C ha pt e r 3

The Tractatus

The Tractatus is a work with perhaps as many interpretations as in-

terpreters, leaving room for the occasional agreement or change of
mind. David Stern [2003] groups these into rough categories: the
logical atomist reading, as a piece of British empiricism (e.g., Russell
and Ramsey); the logical positivist reading, as an anti-metaphysical
tract (e.g., Carnap and Schlick); the metaphysical reading, as an ac-
count of logic and representation (e.g., Anscombe, Stenius, Hacker,
Pears); the irrationalist reading, as an ethico-religious work (Janik
and Toulmin); and the recently popular therapeutic reading, or ‘new
Wittgenstein’ (e.g., Diamond and Conant). As the parenthetical lists
of names suggest, ‘the metaphysical reading remains the most widely
accepted approach’ (Stern [2003], p. 127). Since our focus here is on
the nature of logical truth, I’ll offer a fairly mainstream metaphysical
reading1,2 as a starting point for the naturalization process.
1. In some regions of the literature, there’s an odd tendency to count only realistic readings
as ‘metaphysical’; e.g., in the opening sections of McGinn [2006], ‘anti–metaphysical’
seems roughly interchangeable with ‘anti–realist’. Morris [2008], pp. 55–58, notes this
oddity, and takes the reasonable view that an idealistic reading (which he eventually en-
dorses in his chapter 6, see p. 307) is as metaphysical as a realist reading. White ([2006],
pp. 26–28, 98–100) also takes the idealistic option seriously. See chapter 4.
2. Mainstream as it may be, I don’t pretend that this reading is without its tensions. Many
readers will perhaps agree that the book itself makes some tensions inevitable, but that
general fact hardly excuses my particular shortcomings. I hope at least that this take
on the Tractatus is illuminating for the larger themes of under discussion here. (I’ve
learned most from Pears [1987] and [2006], though I depart from him in ways I won’t
attempt to catalogue.)

37
 THE LOGICA L MUST

Much as Kant begins from the question, ‘what must the world
be like that we can cognize it as we do?’ the early Wittgenstein
can be understood as beginning from the question, ‘what must
the world be like that we can represent it as we do?’ And again as
with Kant, the difficulty of the question increases with a convic-
tion about a priority, in Kant’s case the conviction that part of our
cognition of the world is a priori, revealing necessary features of
our world. The counterpart for Wittgenstein is the conviction that
the sense of our statements – whether they are capable of truth or
falsity, and if so, what it would take for them to be true or false – all
this must be given a priori, must be independent of any contingent
facts about the world. 3 This simple vision of how language works

3. On the metaphysical side, on the sort of reading I’m attempting here, it seems fully
defensible to regard sense as part of the necessary structure of representation: e.g.,
‘a proposition has a sense that is independent of the facts’ (4.061). On the epistemic
side, there isn’t much to go on in the Tractatus, but to separate a priority from necessity
here would be to credit Wittgenstein with ‘Kripke’s . . . discovery of the necessary a
­posteriori’ – characterized by Soames as ‘one of the great philosophical achievements
of the twentieth century’ (Soames [2003], pp. 455–456) – which seems unlikely. Still,
the 5.55s present one of those aforementioned tensions. As we’ll see below, the sense
of a proposition is its agreement and disagreement with the elementary propositions
(4.2, 4.21), the ‘truth-conditions’ of the proposition (4.431); to understand a proposi-
tion is to know these truth conditions (4.024); and analysis is what brings out these
truth conditions (4.221). Yet (turning to the 5.55s), Wittgenstein also appears to think
that we can’t know a priori ‘whether I can get into a position in which I need the sign
for a 27-termed relation’ (5.5541–5.5542) or more generally that ‘I cannot say a priori
what elementary propositions there are’ (5.5571). I say ‘appears to think’ because these
passages are actually hedged, but 5.557 tells us outright that ‘the application of logic de-
cides what elementary propositions there are’ and ‘what belongs to its application, logic
cannot anticipate’. The problem then is this: to reveal the sense of a given proposition,
I analyze it down to its truth conditions, to a string of agreements and disagreements
with the elementary propositions; but if I can’t know what the elementary propositions
are a priori, then it seems this process of analysis must not deliver its information a
priori. This suggests that the sense of the proposition, which it has of necessity, can’t be
known a priori. It would be easy to interpret 5.557 to say that while logic can’t antici-
pate what elementary propositions there are, still analysis (for some reason an extra-
logical process) is carried on a priori, but the apparent implication of 5.5571 blocks
this route. Another possibility is that I can know the sense a priori without knowing its
full analysis (5.5562 does suggest that the plain man ‘understands propositions in their

38
 T H E T R ACTAT US

was so commonly held at the time as to be nearly invisible, and

it remains intuitively appealing even now: what our claims mean,
what they say about the world, must be in place first, before we look
to the world to determine which are true and which are false.
Commentators locate this idea as part of the doctrine of ‘deter-
minacy of sense’, which Wittgenstein inherited from Frege:

There is no doubt that part of the meaning of the postulate, that

sense must be determinate, is that there must not be anything
contingent in its foundations.[4] (Pears [1981], p. 80)

At the time Wittgenstein himself is perhaps most explicit in the

Notebooks:

If a proposition is to make sense then the syntactical employment

of each of its parts must be settled in advance . . . e.g., what proposi-
tions follow from a proposition must be completely settled before
that proposition can have a sense! (Wittgenstein [1914–1916],
18.6.15, p. 64e)

As often happens, he’s more forthcoming later, when he goes on to

reject the idea, as in this passage from the Big Typescript:

There’s a misconception . . . language seems to be something

that is given a structure and then superimposed onto reality.
(Wittgenstein [1933], p. 45e)

unanalyzed form’). I won’t try to sort this out here; for present purposes, I presuppose,
in pre-Kripkean spirit, that a priority and necessity are coextensive, and allow myself to
shift back and forth between the two. (Thanks to Thomas Ricketts for helpful conversa-
tions on this point.)
4. Of course, another aspect of the determinacy of sense is a rejection of vague boundaries,
but as indicated above (see footnote 5 of chapter 2), I leave this idea in the shadows here.

39
Exploring the Variety of Random
Documents with Different Content
 1. Die Liebe als physisches (natürliches) Problem.
Die „Kosmogonie“, die Erschaffung der Welt selbst, des
gestirnten Himmels und der seligen Götter wird in den Mythen vieler
Völker als ein Akt der geschlechtlichen Zeugung gedacht. So
erhaben, so wunderbar und rätselvoll erschien schon den ältesten
Menschen in grauer Vorzeit der rein physische Vorgang der
Paarung, Befruchtung und Geburt. Materie ist der „Mutter“ Stoff, das
Weltganze, die „Natura“ ist das „Geborene“. Nach G. Herman[1] hat
die neuere Schule der anthropologischen und mythologischen
Forschung eine derartige Anthropomorphisierung der
Weltentstehung als wahrscheinlichste Quelle aller Religionssysteme
angenommen. Himmel und Erde sind dem Chinesen „Vater und
Mutter aller Dinge.“ Auch das „Weltenei“ spielt in den Religionen und
Mythen der verschiedensten Völker eine grosse Rolle.
Die ersten Geschöpfe aber, Götter sowohl wie Menschen, sind
Zwitter[2]. Wer kennt nicht die berühmte Erzählung des Aristophanes
im platonischen „Gastmahl“ (Kap. 14)? Einst sei die Natur des
Menschen eine andere gewesen als jetzt. „Denn zuerst gab es drei
Geschlechter von Menschen, nicht wie jetzt nur zwei, das männliche
und das weibliche, sondern noch ein drittes dazu, welches das
gemeinschaftliche war von diesen beiden; sein Name ist noch übrig,
während es selbst verschwunden ist. Mannweib (ἀνδρόγυνος)
nämlich war damals dieses eine.“ Auch aus dem anfangs
zweigeschlechtlichen Adam der Bibel ging das erste Menschenpaar
als Mann und Weib hervor.
Die Liebe als kosmogonisches Prinzip spielt bei Empedokles eine
ganz besondere Rolle. Zwei Grundkräfte sind es, durch welche nach
diesem Philosophen alle Veränderung in der Mischung und
Trennung der Stoffe hervorgebracht wird: Die Liebe und der Hass. In
unermesslichen Perioden der Weltentwickelung überwiegt bald die
eine, bald die andere dieser beiden Grundkräfte als herrschende
Macht. Ist die Liebe zur völligen Herrschaft gelangt, so ruhen alle
Stoffe in seligem Frieden vereint in der Weltkugel als in Gott. Durch
das Fortschreiten der Macht des Hasses, auf deren Höhepunkt alles
zerstreut und zersprengt ist, oder umgekehrt, durch das
Fortschreiten der Macht der Liebe werden verschiedene
Uebergangszustände in der Weltentwickelung hervorgebracht.
Durch das wiederholte Spiel von Zeugung und Vernichtung blieben
schliesslich allein die Erzeugnisse übrig, welche die Bürgschaft der
Dauer und Lebensfähigkeit in sich trugen. — Wie die oben
erwähnten kosmogonischen Theorien durchweg
anthropomorphisierender Tendenz sind und auf Beobachtungen in
der organischen Natur beruhen, so ist die Idee des Empedokles eine
grossartige Konzeption einer naturwissenschaftlichen Vorstellung,
wie sie im modernen Darwinismus ausgebildet worden ist.
Die neuere Wissenschaft hat die naiven mythologischen und
kosmogonischen Vorstellungen der Vorzeit bestätigt. Wir wissen
auch, dass die physische Liebe des Menschen, also das
Anfangsglied der Entwickelung selbst erst ein sekundäres
Erzeugnis, das Produkt einer Differenzierung ist, nur erklärbar durch
die Entwickelung des organischen Lebens überhaupt. Die
Zwitterbildung, d. h. die Vereinigung der beiden Geschlechtszellen in
einem Individuum ist der älteste und ursprünglichste Zustand der
geschlechtlichen Differenzierung. Erst später entstand die
Geschlechtstrennung. Nach Haeckel[3] findet sich der
Hermaphroditismus nicht nur bei niedersten Tieren, sondern auch
alle älteren wirbellosen Vorfahren des Menschen, von den
Gastraeaden bis zu den Prochordoniern aufwärts, werden Zwitter
gewesen sein. Wahrscheinlich waren sogar die ältesten
Schädellosen noch Hermaphroditen. Ein wichtiges Zeugnis dafür
liefert der merkwürdige Umstand, dass mehrere Fisch-Gattungen
noch heute Zwitter sind, und dass gelegentlich als Atavismus auch
bei höheren Vertebraten aller Klassen der Hermaphroditismus noch
heute wieder erscheint.
Die Geschlechtstrennung, der Gonochorismus, wie Haeckel dies
nennt, erscheint später als die Verteilung der beiderlei
Geschlechtszellen auf verschiedene Personen.[4] Dann treten zu den
primären Geschlechtsdrüsen sekundäre Hilfsorgane wie
Ausführgänge u. s. w. hinzu, und zuletzt entwickeln sich durch
geschlechtliche Zuchtwahl, die Selectio sexualis, die sogenannten
„sekundären Sexual-Charaktere“, d. h. diejenigen Unterschiede des
männlichen und weiblichen Geschlechts, welche nicht die
Geschlechtsorgane selbst, sondern andere Körperteile betreffen
(z. B. der Bart des Mannes, die Brust des Weibes).
Hierbei unterliegt die morphologische Ausbildung der
menschlichen Geschlechtsorgane dem berühmten, von Haeckel
zuerst formulierten „biogenetischen Grundgesetz“, das die
Ontogenie, die individuelle Entwickelung, einen abgekürzten,
unvollständigen Abriss der Phylogenie, der Stammesentwickelung
darstellt. In den grossen Lehrbüchern der Entwickelungsgeschichte
von Kölliker und Hertwig findet man die zuverlässigsten
Darstellungen der Ontogenie der Sexualorgane.
In der Beschreibung der ausgebildeten männlichen und
weiblichen Geschlechtsorgane ist das klassische Werk von Kobelt[5]
bisher noch nicht übertroffen worden, wenn auch die Beschreibung
der Geschlechtsorgane in dem grossen „Handbuch der Anatomie
des Menschen“ von K. von Bardeleben (Jena 1896 ff.) viele neue
Aufschlüsse zu bringen verspricht.[6]
Die Entstehung der sekundären Geschlechtscharaktere ist
Gegenstand der Darstellung in dem berühmten Buche von Charles
Darwin.[7]
Aus diesen anatomischen Substraten der menschlichen Liebe
wird man die Physiologie derselben im weitesten Umfange ableiten
müssen. Das Hauptwerk über den Vorgang der Zeugung im
Gesamtgebiete des organischen Lebens und beim Menschen
besitzen wir in dem Werke von Hensen.[8]
Der Fundamentalvorgang aller Liebe bei Mensch, Tier und
Pflanze, die älteste Quelle der Liebe ist die Wahlverwandtschaft
zweier verschiedener erotischer Zellen: der männlichen Spermazelle
und der weiblichen Eizelle, das, was Haeckel[9] den „erotischen
Chemotropismus“ genannt hat. Der Zweck und das Endziel der
physischen Liebe ist die Verschmelzung oder Verwachsung dieser
beiden erotischen Zellen. „Alle anderen Verhältnisse und alle die
übrigen, höchst zusammengesetzten Erscheinungen, welche bei den
höheren Tieren den geschlechtlichen Zeugungsakt begleiten, sind
von untergeordneter und sekundärer Natur, sind erst nachträglich zu
jenem einfachsten, primären Kopulations- und Befruchtungsprozess
hinzugetreten.“ — „Ueberall ist die Verwachsung zweier Zellen das
einzige, ursprünglich treibende Motiv, überall übt dieser
unscheinbare Vorgang den grössten Einfluss auf die Entwickelung
der mannigfaltigsten Verhältnisse aus. Wir dürfen wohl behaupten,
dass kein anderer organischer Prozess diesem an Umfang und
Intensität der differenzierenden Wirkung nur entfernt an die Seite zu
stellen ist.“ (Haeckel.)
Nachdem dieser fundamentale Vorgang der Zeugung festgestellt
ist, gelangen wir zu einer Betrachtung jener physischen
Liebesregungen beim Menschen, welche sich in Form des
Geschlechtstriebes[10] äussern. Diesen dunkeln Begriff hat Moll in
höchst geistvoller Weise aufgehellt.[11] Er zerlegt den
Geschlechtstrieb beim erwachsenen Menschen in zwei
Komponenten, den Detumeszenztrieb und den Kontrektationstrieb.
Der Detumeszenztrieb drängt zu einer örtlichen Funktion an den
Genitalien, und zwar beim Manne zur Samenentleerung. Er ist als
ein peripherer organischer Drang zur Entleerung eines Sekretes
aufzufassen. Der Kontrektationstrieb drängt den Mann zur
körperlichen und geistigen Annäherung an das Weib, das letztere
ebenso zur Annäherung an den Mann. Phylogenetisch ist die
Detumeszenz als Mittel zur Fortpflanzung das Primäre, weil sie bei
niederen und höheren Tieren stattfindet. Erst sekundär kam die
Kontrektation hinzu, indem sich zwei Individuen zur Fortpflanzung
verbanden. In der individuellen Entwickelung des Menschen ist die
Anwesenheit der Keimdrüsen, der Erreger des Detumeszenztriebes,
das Primäre. Der Kontrektationstrieb ist ein sekundärer
Geschlechtscharakter. Der Detumeszenztrieb des Mannes ist die
unmittelbare Folge der Funktion der Hoden. Beim Weibe hängt zwar
die Ausscheidung der Eizelle aus dem Ovarium mit dem
Detumeszenztrieb nicht unmittelbar zusammen, ursprünglich fielen
sie aber zusammen, wie man noch bei den Fischen sieht.
 Nunmehr geht Moll[12] zur Erörterung einer höchst wichtigen
Frage über, welche für die Beurteilung vieler Erscheinungen von der
grössten Bedeutung ist, nämlich zu dem Verhältnis zwischen
Ererbtem und Erworbenem in der Geschlechtsliebe. Dies ist der
Punkt, in welchem wir ganz und gar von Moll abweichen, weil wir
durch die geschichtliche Betrachtung zu ganz anderer Auffassung
geführt werden als Moll, welcher durch seine allerdings ingeniöse
naturwissenschaftliche Argumentation zu beweisen sucht, dass
neben dem Detumeszenztriebe — woran wir nicht zweifeln — auch
die mannigfaltigsten Erscheinungen des Kontrektationstriebes ererbt
sind. Kurz, Moll ist geneigt, sowohl die physischen als auch die
pathologischen Erscheinungen des Geschlechtstriebes zum
grössten Teile auf Vererbung zurückzuführen, während nach seiner
Ansicht die erworbenen Faktoren nur eine sehr geringe Rolle
spielen. Normaler und abnormer Geschlechtstrieb („konträre
Sexualempfindung“, Homosexualität) erklären sich nach Moll eher
aus der Vererbung als auch der durch die Umstände geschaffenen
Gewohnheit. Wir wollen nicht leugnen, dass gewisse körperliche und
geistige Dispositionen vererbt werden. Wir werden aber durch
unsere Studien zu dem Bekenntnis gezwungen, dass die Vererbung
in der Liebe eine viel geringere Rolle spielt, als die Erwerbung
bestimmter Eigenschaften und die stete Wirkung äusserer Einflüsse.
Dies auf geschichtlichem Wege zu erweisen, ist unsere Aufgabe und
wird schon im vorliegenden Bande mehr als einmal zu Tage treten.
Aber auch das rein naturwissenschaftliche Räsonnement vermag
diesen Standpunkt zu rechtfertigen und zu befestigen, wie die ganz
vortreffliche kleine Schrift von K. Neisser aufs evidenteste dartut.[13]
Den gleichen Standpunkt der kongenitalen Natur zahlreicher
geschlechtlicher Perversionen vertritt R. v. Krafft-Ebing in seinem
ausserordentlich verbreiteten Werke über die „Psychopathia
sexualis“, während hinwiederum von Schrenk-Notzing, sich mehr
unserem Standpunkt nähernd, die Suggestion als Ursache mancher
sexuellen Abnormitäten betrachtet.[14]
Krafft-Ebing hat aber das unbestreitbare Verdienst, das gesamte
menschliche Geschlechtsleben vom Standpunkt des Irrenarztes
einer eingehenden Würdigung unterzogen zu haben.
 Als Vorläuferin sexueller Ausschweifungen spielt ferner
zweifelsohne die Onanie eine grosse Rolle, welche ganz kürzlich in
dem Buche von Rohleder[15] die erste kritische und als solche
mustergiltige Bearbeitung gefunden hat.
Wichtige Aufklärungen über die Natur der geschlechtlichen
Beziehungen des Menschen werden auch durch das Studium jener
körperlichen Vorgänge dargeboten, welche nur unmittelbare
Einflüsse auf die sexuellen Akte ausüben. Vor allem gehören hierher
die Sinne, der Stoffwechsel und die psychischen Vorgänge.[16]
Gerade aus der Untersuchung der Beziehung der Sinne zum
Geschlechtsleben, vor allem des Geruchs und Gesichts, wird sich
das häufige Erworbensein abnormer Zustände ergeben. Eine
experimentale Psychologie der Liebe existiert nicht.[17] Was bisher
unter dem Namen einer „Psychologie der Liebe“ geboten wurde, ist
in naturwissenschaftlicher Hinsicht kaum beachtenswert wie z. B. die
nach anderer Richtung hin vortreffliche „Psychologie der Liebe“ von
Julius Duboc. „Einige wenige sorgfältige Untersuchungen, die aber
noch der Bestätigung und weiterer Ausdehnung bedürfen, einige
Beobachtungen über formlose Tatsachenmassen, die in praktischer
Lebenserfahrung aufgehäuft sind und die ihren Wert haben, wenn
sie auch in mannigfacher Weise missverstanden und falsch
ausgelegt werden können — das ist alles, was die empirische
Psychologie bisher über die intellektuellen Unterschiede der
Geschlechter zu bieten hat.“ (Havelock Ellis.)
Die breiteste Grundlage für eine naturwissenschaftliche
Erforschung der psychischen Erscheinungen des menschlichen
Geschlechtslebens bildet unzweifelhaft das von der Anthropologie
und Ethnologie gesammelte Material, wie es in dem klassischen
Werke von Ploss und Bartels[18] vorliegt. Hier beginnen schon
vielfach die Berührungen mit den soziologisch-historischen
Problemen des Sexuallebens.
Die Liebe, als physisches Problem betrachtet, umfasst auch, was
wir zum Schluss nur noch kurz erwähnen wollen, die organischen
Geschlechtskrankheiten des Menschen.
 2. Die Liebe als historisches Problem.
Die Liebe als geschichtliche Erscheinung ist nichts an und für
sich. Sie ist, ganz evolutionistisch gefasst, das zu immer grösserer
Freiheit fortschreitende Verhältnis zwischen der physischen Liebe
und den aus der Selbstentfaltung des Geistes hervorgegangenen
Formen der Gesellschaft, des Rechtes und der Moral, der Religion,
der Sprache und Dichtung. Es ist wichtig zu betonen, dass es auf
diesem Gebiete keine Kausalität, keine Gesetze in
naturwissenschaftlichem Sinne geben kann, dass die von Herbert
Spencer inaugurierte „organische Methode“ der Soziologie den
geschichtlichen Erscheinungen nicht gerecht zu werden vermag. Es
gibt bei der Betrachtung sozialer Phänomene keine Gesetze,
sondern nur Rhythmen[19]. „Den Schritt vom Rhythmus zum Gesetz
können wir heute noch nicht wagen, wenn wir gleich der
Ueberzeugung sind, dass Rhythmen letzten Endes auf (uns noch
verborgene) soziale Gesetze zurückdeuten.“ (Stein.) Trotzdem ist
hierbei blinder Zufall ausgeschlossen. Denn dieser soziale
Rhythmus stellt sich bei bestimmten Bedingungen und
Voraussetzungen regelmässig wieder ein und nimmt damit für uns
das Gepräge bestimmter Gesetzlichkeit an. Achelis (a. a. O. S. 68)
macht in dieser Beziehung auf die bekanntesten statistischen
Erhebungen über Wiederkehr derselben Vergehen, über den wahren
Zusammenhang von Moral und wirtschaftlichen Verhältnissen
aufmerksam. Es handelt sich also auch, insofern die Liebe als
geschichtliche und soziologische Erscheinung in Betracht kommt,
nur um Auffindung jener Rhythmen, jener regelmässig
wiederkehrenden Formen und Typen des Geschehens.
Die Liebe als eine soziale Erscheinung, als Produkt der
Gesellschaft, erscheint wesentlich in den beiden Formen der Ehe
und der Prostitution.
Eduard Westermarck, Professor an der Universität in Helsingfors,
hat das für alle Zeit grundlegende Werk über die Geschichte der
menschlichen Ehe geschrieben, welches wir nicht anstehen, den
besten kulturhistorischen und soziologischen Werken eines Buckle,
Tylor, F. A. Lange u. a. ebenbürtig an die Seite zu stellen.[20] Dies
Buch weist in der unwiderlegbarsten Weise mit der gediegensten
wissenschaftlichen Argumentation die Ehe als die überall
wiederkehrende primitive soziologische Form und das soziologische
Endziel der Liebe nach und macht der noch bis in die neueste Zeit
von Bachofen, Mc.-Lennan, Morgan, Lubbock, Bastian, Lippert,
Kohler, Post vertretenen Lehre von der ursprünglichen
geschlechtlichen Ungebundenheit, der sogenannten Promiscuität für
immer ein Ende. Die „Kritik der Promiscuitätslehre“ (a. a. O. S. 46–
130) gehört zu den glänzendsten Leistungen der modernen
Soziologie. Ihr Ergebnis muss auf die Anschauungen über das
menschliche Geschlechtsleben nicht blos in soziologischer, sondern
auch in philosophischer Hinsicht den grössten Einfluss ausüben.
Nach Westermarck kommt die Ehe schon bei vielen niedrigen
Tiergattungen vor, bildet bei den menschenähnlichen Affen die Regel
und ist bei den Menschen allgemein. Ihr Ursprung muss offenbar
einem durch den mächtigen Einfluss der natürlichen Zuchtwahl zur
Entwickelung gebrachten Instinkt zugeschrieben werden. Dass der
Urmensch die Ehe kannte, darf man mit grösster Zuversicht
mutmassen. Denn die Ehe der Primaten (Menschen und Affen)
scheint aus der kleinen Anzahl der Jungen und aus der Länge des
Kindesalters hervorgegangen zu sein. Mit aller Wahrscheinlichkeit
bezeichnet Westermarck die menschliche Ehe als ein von den
affenähnlichen Urmenschen überkommenes Erbe. Ferner weist er
nach, dass gerade bei den am niedrigsten stehenden Völkerschaften
die geschlechtlichen Beziehungen sich am wenigsten der
Promiscuität nähern. Wir haben sogar Grund zu dem Glauben, dass
mit dem Fortschreiten der Kultur die ausserehelichen Beziehungen
der Geschlechter zugenommen haben. Demgemäss hat in Europa
die Zahl der Ehelosen eine Zunahme, das Durchschnittsalter der
Eheschliessung eine Hinaufschraubung erfahren.
Allerdings ist die Lebenslänglichkeit der Ehe durchaus nicht ganz
allgemein. Bei den meisten unzivilisierten und vielen
vorgeschrittenen Völkern darf der Mann der Gattin jederzeit nach
Belieben den Abschied geben. Bei sehr vielen anderen jedoch —
auch solchen auf niedrigster Stufe — bildet die Scheidung den
Ausnahmefall. Es kommt auch vor, dass dem Weibe gestattet ist,
dem Gatten den Laufpass zu geben. Im allgemeinen nimmt die
Dauer der Ehe mit der Vervollkommnung des Menschengeschlechts
stetig zu.
Während die Ehe als die eminent soziale Form der Liebe zu
betrachten ist, in welche sich seit jeher das menschliche
Geschlechtsleben gekleidet hat, muss als ihr Gegenpol, als absolut
antisoziale Erscheinung die Prostitution bezeichnet werden. Man
nennt sie, wie bekannt, ein „notwendiges Uebel“. Eine
wissenschaftliche, dem Stande der modernen Forschung
entsprechende Geschichte der Prostitution existiert noch nicht. Das
grosse achtbändige Werk von Dufour[21] enthält zwar eine grosse
Menge Material, dasselbe ist aber gänzlich unübersichtlich
zusammengestellt. Zudem verliert auch diese Zusammenstellung
jeden Wert durch den gänzlichen Mangel der genauen
Quellennachweise. Nur aus einer gleichmässig die Ergebnisse der
Soziologie, Hygiene und Nationalökonomie verwertenden
geschichtlichen Darstellung der Prostitution würde sich ein sicheres
Urteil über die Ursache und die Abhilfe dieses sozialen Uebels
gewinnen lassen. Besonders Bebel’s Werk „Die Frau und der
Sozialismus“ hat manche unrichtigen Anschauungen über die
Ursachen der Prostitution verbreitet, indem dieser Autor dieselben
auf die wirtschaftliche Ausbeutung und die Hungerlöhne zurückführt.
Demgegenüber sei nur auf die gediegene, aus langjähriger
Erfahrung hervorgegangene Arbeit über Prostitution von G.
Behrend[22] hingewiesen, der ganz andere Ursachen derselben
aufdeckt, dieselben vor allem in einer fast stets erworbenen
Lasterhaftigkeit sieht und ganz richtig bemerkt, dass man meist die
veranlassenden äusseren Momente für die eigentlichen Ursachen
ansieht. Der bedeutendste Forscher über Prostitution neben
Behrend ist B. Tarnowsky[23], der bemerkenswerter Weise zu den
gleichen Ergebnissen wie jener gelangt ist und als eine Fabel
nachweist, dass die Armut die nie versiegende Quelle der
Prostitution sei. Auch A. Hegar hat den Versuch gemacht, Bebels
Behauptungen zu widerlegen, und zugleich in seiner
sozialhygienischen Studie Vorschläge zu einer Beseitigung des
„geschlechtlichen Elends“ gemacht.[24]
Den kühnsten Vorstoss in der Erklärung der Prostitution hat aber
wohl Lombroso unternommen. Er geht von dem unzweifelhaften
Zusammenhange zwischen Prostitution und Verbrechen aus und
statuiert, dass die „Donna delinquente e prostituta“ nur eine
besondere Abart des „reo nato“, des „geborenen Verbrechers“
sei[25]. Ganz richtig bemerkt er, dass daher die Dirnennatur nicht nur
in den unteren Klassen vorkomme, sondern ihr Aequivalent auch in
den höheren Gesellschaftsschichten habe, was wiederum ein Beleg
dafür ist, dass man nicht die Armut als Ursache der Prostitution
anschuldigen kann. Trotzdem halten wir die Theorie der „geborenen
Prostituierten“ für verfehlt und müssen auch wiederum den äusseren
Einflüssen wie falscher Erziehung, Umgebung u. s. w. mehr
Bedeutung zuerkennen. Jedenfalls bringt das Buch Lombrosos
wertvolle Aufschlüsse über den niemals bestrittenen innigen
Zusammenhang von Prostitution und Verbrechen.
Das Verhältnis der Liebe zum öffentlichen Recht spiegelt sich vor
allem in der sogenannten Frauenfrage wieder. Nimmt man, wie wir
gesehen haben, die Ehe als Grundlage der Gesellschaft und als das
soziologische Endziel der Liebe, so ist eine allgemeine
„Frauenemanzipation“, d. h. die völlige Aufhebung aller
gesellschaftlichen, staatlichen und wirtschaftlichen Unterschiede
zwischen Mann und Frau ein Widerspruch in sich selbst. Denn die
Ehe bedingt allein schon durch die Geburt der Kinder, die Sorge für
diese und die wirtschaftlichen Angelegenheiten der Familie eine
Arbeitsteilung zwischen Mann und Frau. Auch lassen sich trotz
glänzender Ausnahmen die grossen körperlichen und geistigen
Verschiedenheiten von Mann und Weib nicht verleugnen. Hiermit ist
das Zugeständnis grösserer Rechte und zahlreicherer
Bildungsgelegenheiten an die Frauen wohl vereinbar, besonders
angesichts des grossen Ueberschusses der Zahl derselben über
diejenige der Männer, sowie der späten Heiraten der letzteren.
Anfang und Ende der „Frauenfrage“ ist für uns in dem einen Satze
beschlossen: Die Frau ist die gleichberechtigte aber nicht
gleichmächtige Gefährtin des Mannes.
 Die rechtliche Beurteilung des Verhältnisses zwischen Mann und
Weib hängt aufs innigste zusammen mit der ethischen Seite. Eine
wichtige Aufgabe einer Wissenschaft des Geschlechtslebens wird
darin bestehen, den Einfluss der jeweiligen Lehren der Moral auf die
menschliche Liebe und ihre Aeusserungen zu studieren und im
Zusammenhange darzulegen. Für Deutschland ist in neuester Zeit
ein derartiger, freilich noch unvollkommener Versuch unternommen
worden.[26] In der Tat bildet die Regelung des sexualen Lebens
„innerhalb der Oeffentlichkeit“ einen integrierenden Teil der
Moralgeschichte überhaupt, und Rudeck hat Recht, wenn er diese
zugleich als eine „Kritik der gesamten Kultur“ bezeichnet, deren Art
und Bedeutung sich nirgends so treu wiederspiegelt wie auf
geschlechtlichem Gebiete. Dass die moralische Beurteilung
geschlechtlicher Verhältnisse zu verschiedenen Zeiten und bei
verschiedenen Völkern eine ganz verschiedene gewesen ist, ist eine
längst bekannte Tatsache. Und doch wird auch hier eine kritische
Untersuchung gewisse Normen feststellen können, die
Allgemeingültigkeit beanspruchen. Mit der Vervollkommnung des
Menschengeschlechts entwickelt sich auch eine Ethik des
Sexuallebens. So führt Westermarck in seiner „Geschichte der
menschlichen Ehe“ den stringenten Nachweis, dass das
Schamgefühl etwas sekundäres und zwar die Folge, nicht die
Ursache der Bekleidung ist.
Ein sehr grosses Forschungsgebiet ergiebt sich aus den
Beziehungen zwischen Liebe und Religion. G. Herman, dessen
Buch wir oben erwähnten, hat im Detail geschildert, wie alle
Mythologie und Religion auf sexueller Grundlage erwachsen ist, und
deduziert mittelst einer höchst interessanten Beweisführung, dass
aus den geschlechtlichen Feiern und Mysterien der Urvölker die
Riten der heutigen Konfessionen geworden sind. Man darf
behaupten, dass die Religion oder besser der Konfessionalismus
das menschliche Geschlechtsleben im ganzen höchst ungünstig
beeinflusst hat. Man denke nur an die religiöse Mystik mit ihren
sexuellen Ekstasen und Ausschweifungen, an den Kult der
„Satanskirche“, die „schwarze Messe“ u. dgl. mehr. Die
monotheistischen Religionen, sobald sie zum Konfessionalismus
entarten, sind hierin um nichts besser als die heidnischen
Religionen, ja vielleicht noch schlimmer, und es liegt etwas Wahres
in Nietzsches Ausspruch[27]: „Das Christentum gab dem Eros Gift zu
trinken: — er starb zwar nicht daran, aber entartete, zum Laster“. Die
meisten erotischen Epidemien sind religiösen Ursprungs.
Dass die Erscheinungen der Liebe bei verschiedenen Völkern
gewissermassen nationale Formen annehmen, lehrt die Ethnologie.
Die Liebe des Russen ist eine andere als die Liebe des Franzosen,
die Liebe des Griechen eine andere als die des Böhmen. Einen
wahrhaft objektiven Ausdruck findet diese ethnologische
Verschiedenheit in der Sprache. In ihr werden die feinsten Nüancen
sexueller Gefühle durch die betreffenden Worte sichtbar. Abel hat in
einer höchst schätzbaren Abhandlung den ersten Versuch einer
derartigen linguistischen Erforschung der Liebe gemacht.[28] Er
untersucht so die Worte für Liebe in der lateinischen, englischen,
hebräischen und russischen Sprache.
Die Sprache führt uns zur Dichtung. Die Werke der Literatur
bieten uns ein dankbares Feld für vergleichend-geschichtliche
Untersuchungen über die menschliche Liebe. Die Weltliteratur liefert
das Baumaterial für eine historische Psychologie der Liebe. Sie
bietet, wie Stein (a. a. O. S. 33) sagt, „den dankbarsten
vergleichenden Stoff, der seiner sozialgeschichtlichen Bezwinger
harrt“. Hier sind noch wahre wissenschaftliche Schätze zu heben.
Homer und die Bibel, die Veden und Upanishaden, die gesamte
Weltliteratur in allen ihren Auszweigungen enthalten die getreuen
Abbilder dessen, was die Liebe bei jedem Volke und zu jeder Zeit
gewesen ist.
Endlich wird das menschliche Geschlechtsleben beeinflusst
durch die materielle Kultur einer bestimmten Epoche. Krieg und
Frieden, städtisches Leben und ländliche Idylle, Kleidung und
Nahrung u. v. m., verschieden nach Zeit und Ort, üben auch auf die
menschliche Liebe die grössten Wirkungen aus.
So ist die Liebe als geschichtliche Erscheinung unendlich reich
an Beziehungen jeder Art, welche eine höhere Bedeutung des Eros
ahnen lassen als sie die rein physische Liebe erkennen lässt.
Untersuchen wir daher
 3. Die Liebe als metaphysisches Problem.
Dass der menschlichen Liebe eine höhere Bedeutung innewohnt,
leuchtet schon daraus hervor, dass sie allein die Ursache der
höchsten dichterischen Verzückung bei allen Völkern gewesen ist
und noch ist. Und zwar ist es nicht die äussere Erscheinung,
sondern das gewaltige innere Wesen der Liebe, was den Menschen
unwiderstehlich bezwingt. Wie Don Cesar in der „Braut von Messina“
sagt:
Nicht ihres Lächelns holder Zauber war’s,
Die Reize nicht, die auf der Wange schweben,
Selbst nicht der Glanz der göttlichen Gestalt —
Es war ihr tiefstes, ihr geheimstes Leben,
Was mich ergriff mit heiliger Gewalt.

Was ist nun dieses „tiefste und geheimste“ Leben? Was ist der
wahre Zweck, das wirkliche Endziel der Liebe?
Zwei berühmte Philosophen der Neuzeit, Arthur Schopenhauer
und Eduard von Hartmann haben die gleiche metaphysische
Betrachtung über die Liebe angestellt, die das grösste Aufsehen
erregte und viele Nachbeter fand. A. Schopenhauer[29] erblickt die
Bedeutung der Liebe in der Erfüllung der Zwecke der Gattung,
welche in der Reihenfolge und dem endlosen Flusse der
Generationen ihr Leben führt. „Die sämtlichen Liebeshändel der
gegenwärtigen Generation zusammengenommen sind demnach des
ganzen Menschengeschlechts ernstliche meditatio compositionis
generationis futurae, e qua iterum pendent innumerae generationes“.
Dabei verlarvt sich aber der Gattungszweck, indem er in der Gestalt
der Geschlechtsliebe eingeht in den persönlichen Zweck der
Individuen und erscheint als deren höchstes Glück, als der Gipfel
aller ihrer Wünsche, daher in der erhabensten Form, in den
überschwenglichsten Gefühlen und Entzückungen, als das
unerschöpfliche Thema aller Poesie, der lyrischen, epischen und
dramatischen, als der Gegenstand des Lustspiels und des
Trauerspiels. Eros spielt seine Rolle auf dem Sokkus und auf dem
Kothurn. Dass die Liebenden die Erfüllung des Gattungszweckes für
den Gipfel ihres persönlichen Glückes halten, darin besteht die
tragische Illusion, der Wahn. Es ist ein schrecklicher Wahn. Denn im
Genuss der Wollust kontrahiert der Mensch eine schwere Schuld,
welche das erzeugte Individuum zu büssen und durch Leiden und
Tod bezahlen muss. „Das Leben eines Menschen, mit seiner
endlosen Mühe, Not und Leiden, ist anzusehen als die Erklärung
und Paraphrase des Zeugungsaktes“. Der Eros als Ausdruck des
Willens zum Leben, „wie ist er so sanft und zärtlich! Wohlsein will er,
und ruhigen Genuss und sanfte Freude, für sich, für andere, für alle.
Es ist das Thema des Anakreon. So lockt und schmeichelt er sich
selbst ins Leben hinein. Ist er aber darin, dann zieht die Qual das
Verbrechen, und das Verbrechen die Qual herbei. Greuel und
Verwüstung füllen den Schauplatz. Es ist das Thema des
Aeschylos“. (a. a. O. S. 670.)
Die Illusion, die Täuschung und die Verzweiflung der Liebe
schildert prachtvoll E. v. Hartmann[30]. Sein Schluss ist dieser: „Wer
einmal das Illusorische des Liebesglückes nach der Vereinigung und
damit auch desjenigen vor der Vereinigung, wer den in aller Liebe
die Lust überwiegenden Schmerz verstanden hat, für den und in
dem hat die Erscheinung der Liebe nichts Gesundes mehr, weil sich
sein Bewusstsein gegen die Oktroyierung von Mitteln zu Zwecken
wehrt, die nicht seine Zwecke sind; die Lust der Liebe ist ihm
untergraben und zerfressen, nur ihr Schmerz bleibt ihm unverkürzt
bestehen.“
Wer, wie wir, den Begriff der Liebe evolutionistisch fasst, kann
eine solche Metaphysik der Geschlechtsliebe nicht anerkennen. Es
ist richtig, dass das rein Physische der Liebe mehr Unlust als Lust
mit sich bringt durch Vorspiegelung seliger Freuden, die nachher
zerrinnen wie Schaum. Aber die physische Liebe ist nur der Anfang
einer Entwickelung, deren Ende gerade dem Individuum die grösste
Seligkeit verheisst. Die physische Liebe ist nur der als solcher
notwendige Durchgangspunkt zu dem wirklichen Endziele, der
platonischen Liebe. Das metaphysische Endziel der Liebe ist die
Erkenntnis, die vollendete Freiheit. „Und Adam erkannte Eva“ heisst
es tiefsinnig in der Bibel!
Platos und Hegels Dialektik haben aufs treffendste diese
Wahrheit erleuchtet. Ganz richtig bemerkt Wigand[31], dass die
platonische Liebe der natürlichen oder physischen Liebe gar nicht
entgegengesetzt ist, sondern die Liebe zum sinnlichen und
körperlichen Schönen ist die Leiter und die Leiterin zur Liebe und
Erkenntnis alles unsichtbaren Schönen und Guten in Natur- und
Menschenwelt, in Kunst und Wissenschaft von Stufe zu Stufe bis zur
letzten Sprosse dieser Leiter, zur Anschauung der Allgesetzlichkeit,
des Absoluten.
Noch deutlicher wird dies, wenn wir in den Sinn der Worte
eindringen, welche die göttliche Diotima im „Gastmahl“ des Plato
spricht, Worte, die ewig und unvergänglich sind.
„Denn dies ist die rechte Art, sich auf die Liebe zu legen oder von
einem anderen dazu angeführt zu werden, dass man von diesem
einzelnen Schönen beginnend, jenes einen Schönen wegen immer
höher hinaufsteige, gleichsam stufenweise von einem zu zweien und
von zweien zu allen schönen Gestalten und von den schönen
Gestalten zu den schönen Sitten und Handlungsweisen, und von
den schönen Sitten zu den schönen Kenntnissen, bis man von den
Kenntnissen endlich zu jener Kenntnis gelangt, welche von nichts
anderem als eben von jenem Schönen selbst die Kenntnis ist, und
man also zuletzt jenes selbst, was schon ist, erkenne. Und an dieser
Stelle des Lebens, lieber Sokrates, wenn irgendwo, ist es dem
Menschen erst lebenswert, wenn er das Schöne selbst schaut.“
(Platons Symposion 210, 11.)
Das ist der wahre Sinn der platonischen Liebe. Sie ist der
sinnlichen Liebe nicht entgegengesetzt, sondern geht von ihr aus
und erhebt sich zu höheren Formen, indem sie den innigen
Zusammenhang zwischen physischer und geistiger Zeugung
ausdrückt, worin das Wesen jeder wahren und echten Liebe wurzelt.
[32]

Das Endziel der Liebe ist die Erkenntnis. Mit einer Ahnung dieses
Sachverhaltes sagt Schopenhauer in den „Paränesen und
Maximen“: „Zumal wird uns oft da, wo wir Genuss, Glück, Freude
suchten, statt ihrer Belehrung, Einsicht, Erkenntnis, ein bleibendes,
wahrhaftes Gut, statt eines vergänglichen und scheinbaren.“
Die platonische Liebe, so rätselhaft wie sie auf den ersten Blick
erscheint, empfängt ihre hellste Beleuchtung durch die dialektische
Methode Hegels, des „Weltphilosophen“, wie ihn C. L. Michelet
nennt, des Darwin der geistigen Welt, wie wir ihn nennen möchten.
Für Hegel ist auch der Begriff der Gattung evolutionistisch.[33]
Das Leben enthält ein Problem in sich, welches durch die blossen
Lebensfunktionen nicht aufgelöst wird. Die Aufgabe oder der
Lebenszweck fordert die Erzeugung der Gattung. Die Lösung der
Aufgabe bietet die Erzeugung immer neuer Individuen, welche selbst
wieder Individuen ihrer Art hervorbringen. Das ist der Fluss der
Generationen, die endlose Reihe der Geschlechter, welche
entstehen und vergehen. Es ist die Gattung in der Form des
endlosen Prozesses. Nur in der zeugenden Generation lebt die
Gattung wirklich.
In demselben Masse, als eine Generation den Gattungszweck
erfüllt hat, in demselben Masse hat sie ihren Lebenszweck erfüllt.
Sie stirbt daher ab wie ein verbrauchtes Mittel der Gattung, sie
vergeht und mit ihr die Individuen dieser Generation.
Es leuchtet demnach ein, dass in dem Zeugungsprozess die
Aufgabe weder der Gattung noch des Individuums wirklich gelöst
wird. Das Individuum bringt es nur bis zur Generation, die Gattung
bringt es auch nicht weiter. In dem beständigen Flusse der
Generationen, in dem unaufhörlichen Wechsel der Geschlechter
wird die Gattung nicht wahrhaft objektiv und das Individuum nicht
wirklich allgemein. Das einzelne Individuum vergeht wirklich, und die
Gattung, da sie nur in dem Wechsel der Geschlechter, in dem
Entstehen und Vergehen der Individuen erscheint, hört nicht auf zu
vergehen. So wird vermöge des blossen Lebens der Selbstzweck
des Allgemeinen in der Tat nicht erfüllt und verwirklicht.
Wenn man will, so kann man dies die Tragödie der physischen
Welt nennen.
 Was aber in der physischen Welt unmöglich ist, ist in der
geistigen Welt Regel und Selbstzweck.
Das Individuum soll die Gattung erzeugen, die es im
Zeugungsprozess nicht erreichen und objektiv machen kann. So
fordert es der Selbstzweck der Gattung wie der des Individuums.
Die Gattung will als solche erzeugt sein, als die erzeugende
Macht der Individuen, als das wahrhaft Allgemeine. Es gibt nur eine
Form, die das Allgemeine in diesem Sinne vollkommen ausdrückt:
der Begriff. Es gibt nur eine hervorbringende Tätigkeit, die imstande
ist, den Begriff zu erzeugen: das Denken. In dem begreifenden
Denken allein wird das Allgemeine wahrhaft objektiv und das
Individuum wahrhaft allgemein. Hier löst sich die Aufgabe, die der
Begriff des Lebens fordert, aber selbst nicht löst. Sie löst sich im
Denken, welches die wahren Begriffe erzeugt und dadurch die
Objekte erkennt, welche die Begriffe bilden.
Hier also erscheinen die Begriffe Erzeugen und Erkennen in
einem Zusammenhange und in einer Verwandtschaft, wie sie bereits
Plato erkannt hat, wenn er Sokrates das Erkennen ein Erzeugen
nennen lässt. Der philosophische Eros ist das Ziel des physischen.
Das erzeugende Denken ist unsere wahrhaft allgemeine Tätigkeit,
unsere wirkliche Gattung, die in uns entbunden und frei wird in
demselben Masse, als wir selbst frei werden von den individuellen
und sinnlichen Lebenszwecken.
So erscheint die sinnliche, physische Liebe als das notwendige,
mit Bewusstsein zu ergreifende Anfangsglied einer Entwickelung, die
zur Erkenntnis, zur Freiheit, zum Absoluten führt. Hier offenbart sich,
dass dem reinen Wissen, der höchsten und wahrhaftigsten
Erkenntnis niemals die Wärme des Gefühls fehlen kann. Und die
Liebe selbst, sie ist nichts Dunkles mehr, keine Illusion und kein
täuschender Nebel, sondern ihr Anfang und Ende ist die Erkenntnis.
[34]
 I.
Das Zeitalter des Marquis de Sade.
Der Marquis de Sade, dessen Leben, Werke und Persönlichkeit
wir in diesem Bande behandeln, ist durchweg ein Mensch des 18.
Jahrhunderts. Zugleich ist er ein Franzose. Wir glauben aber, indem
wir uns anschicken, das erste wissenschaftliche Werk in deutscher
Sprache über diesen seltsamen, dem Namen nach aller Welt
bekannten Mann zu schreiben, wahres Licht über ihn nur dadurch
verbreiten zu können, dass wir ihn zunächst aus seiner Zeit, aus
dem Frankreich des 18. Jahrhunderts erklären. Die Medizin hat
scheinbar ihre Meinung über den Marquis de Sade schon
ausgesprochen. Aber dieses Urteil, selbst aus dem Munde der
bedeutendsten Nerven- und Irrenärzte, muss ein einseitiges bleiben,
so lange man nicht das tut, was bisher unterblieben ist, so lange
nicht die äusseren Bedingungen, das Milieu erforscht werden, unter
denen dieses merkwürdige Leben heranwuchs, sich bildete, seine
Taten vollbrachte und seine Wirkungen ausübte. Denn es ist
„jedesmal von entscheidender Bedeutung, aus welchem Jahrzehnt
und Jahrhundert, von welchem Volk und Land die behandelten
Tatsachen entlehnt sind.“[35] Mit einem Worte: nicht die individual-
psychologische, sondern nur die sozial-psychologische Auffassung
kann zu einer wahren Erkenntnis der Persönlichkeit Sades führen.
Eine wahrhaft wissenschaftliche Beurteilung gewisser typischer
Persönlichkeiten ist nur auf diesem Wege möglich, wenn auch
keineswegs die Bedeutung der einzelnen Individualität als solcher
verkannt werden soll. Wir müssen uns auf Grund unserer Studien
über den Marquis de Sade durchaus den Ansichten eines
bedeutenden Soziologen der Gegenwart anschliessen[36], dass „das
persönliche Ich nur den Gipfel und Schlusspunkt psychischer
Faktoren überhaupt bildet. Schon psychiatrische Untersuchungen
über die Zersetzung und Entartung unseres Ich haben diesen
Gedanken nahe gelegt, dass unsere Persönlichkeit nicht den
Anfang, sondern eher das Ende einer unendlich langen, in die Nacht
des Unbewussten hinabreichenden psychischen Tätigkeit darstellt,
die wir freilich nicht überall bis auf den letzten Ursprung hin erfassen
können. Durch die Beobachtung des gesellschaftlichen Lebens und
insbesondere der stetigen Wechselwirkung des Einzelnen mit der
ihn umgebenden Gemeinschaft ist diese Hypothese zum Range
einer wissenschaftlich beglaubigten Tatsache erhoben. Hier ist in
den allermeisten Fällen nicht vorbedachte Ueberlegung und völlig
freie Selbstbestimmung entscheidend, sondern
gewohnheitsgemässe Anpassung, das Wirken dunkler, unbewusster
Triebe und Regungen, ohne dass der Einzelne sich jederzeit der
treibenden Gründe klar bewusst wird.“ Sitten und Bräuche,
rechtliche, ästhetische und religiöse Gebilde sind grösstenteils
organische Entwickelungen ohne bestimmtes, zweckbewusstes
Eingreifen seitens des Individuums. Unsere Gefühle und
Empfindungen entspringen trotz ihres eigenartigen individuellen
Charakters „aus jenen Tiefen des Unbewussten, welche der
endgültigen Fixierung des Ichs vorausgehen.“ Das sind aber
Gedanken Hegels, das ist Hegels Lehre vom objektiven Geist, aus
dem der subjektive immerwährend schöpft, und der seine eigene
Entwickelung hat. Das ist in Wahrheit die berühmte und viel
verschrieene „Selbstbewegung des Begriffs“. Hegel, dieser grösste
Denker des neunzehnten Jahrhunderts, wird endlich zu Ehren
kommen, und es ist kein Zweifel, dass seine Lehre im 20.
Jahrhundert die grössten Triumphe feiern wird. Nach den Stürmen
der Schopenhauer-Hartmann’schen und Nietzsche’schen
Philosophie wird die Sonne Hegel’schen Geistes über der Erde
leuchten. Die dialektische Methode hat die neuere
Geschichtswissenschaft mit den wertvollsten Ideen befruchtet und
zur Höhe ihrer gegenwärtigen Entwickelung geführt, sie wird auch
der Naturwissenschaft neue Impulse geben, da sie, wie sich immer
mehr herausstellen wird, nirgends der Erfahrung und den Gesetzen
der Natur widerstreitet. Hegel, nicht Schopenhauer, ist der „wahre
und echte Thronerbe Kants“.
So wollen wir, in einer kurzen Formel ausgesprochen, in diesem
Abschnitt die Fäden aufsuchen, welche den subjektiven Geist des
Marquis de Sade mit dem objektiven Geist seines Zeitalters
verknüpfen. Er ist zugleich ein Vertreter des „ancien régime“ und der
Revolution. Seine beiden berüchtigten Hauptwerke sind
unverkennbare Erzeugnisse der grossen französischen Revolution.
Also haben wir zu untersuchen, was Sade von seiner Zeit
empfangen hat, um zu erfahren, was er ihr gegeben hat. Wir
wiederholen nicht bekannte Tatsachen der französischen
Kulturgeschichte des 18. Jahrhunderts, sondern wir erklären die
Werke des Marquis de Sade aus jener Zeit, aus allen innerlichen
und äusserlichen Verhältnissen des sozialen Lebens im 18.
Jahrhundert.
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

ebooknice.com