You are on page 1of 16

Mind, Meaning,

and Reality
Essays in Philosophy






D. H. Mellor



















Contents

Preface xiii
Introduction 1
1 What is philosophy? 1
2 Untestable identities 2
3 Science and metaphysics 3
4 Methodology 5
5 Meta-philosophy 8
Part I: Mind and Meaning
1 Nothing Like Experience 10
1 Introduction 10
2 Knowing what its like 11
3 Recognising and imagining experiences 12
4 Recalling experiences 13
5 Knowing how and knowing that 14
6 Know-how and the limits of science 16
7 Imagination 17
8 The nature of experience 19
2 What Does Subjective Decision Theory Tell Us? 22
1 The question 22
2 The right thing to do 23
3 Being reasonable 25
4 The case against a descriptive SDT 28
5 The case for a descriptive SDT 30
3 How to Believe a Conditional 34
1 Introduction 34
2 Adams 34
3 Two belief theories 36
4 The disposition theory 37
5 Dispositions and truth conditions 38
6 Simple conditionals 39

7 Dispositions and beliefs 41
8 Complex conditionals 43
9 A methodological objection 44
10 Adams revisited 44
4 Telling the truth 47
5 Successful Semantics 60
1 Introduction 60
2 Causal functionalism 61
3 The causation of actions 63
4 Fine-grained actions 65
5 Actions, intentions, and success 66
6 Ramseys Principle 68
7 Objections to Ramseys Principle 71
8 Limitations of Ramseys Principle 74
9 Successful semantics 75
Part II: What There Is
6 The Semantics and Ontology of Dispositions 78
1 Introduction 78
2 Dispositions and conditionals 79
3 Multi-conditional dispositions 81
4 Reduction sentences 83
5 Dispositions and properties 85
6 Dispositional and categorical properties 87
7 Properties and laws 88
8 The ontology of fragility 91
9 Temperature: a real disposition 93
10 Overdetermination 94
7 Truthmakers for What? 96
1 Introduction 96
2 Truthmaking and entailment 96
3 Truthmaking and truth 97
4 Ontological commitment 99
5 Realism 100
6 Truthmakers and truth conditions 102
7 Direct and indirect truthmaking 103
8 The merits of moderation 104

9 Primary propositions: atomic 107
10 Primary propositions: molecular 109
11 Laws of nature 110
12 Generalisations 111
8 Too Many Universes 113
1 Universes and the multiverse 113
2 Existence, location, and ultimate explanations 114
3 Explanations and probabilities 115
4 A prerequisite of chances 116
5 An improbable argument 117
6 Facing the firing squad 118
9 The Reduction of Society 120
1 Reduction 120
2 Reduction and laws 121
3 Reduction and ontology 122
4 People 123
5 Microreduction and the unity of science 124
6 Groups 126
7 Groups, sets, and abstract objects 127
8 Groups and people 129
9 Wholes and parts 130
10 The limits of sociology 133
11 The reduction of sociology 134
12 Holist objections 137
10 Wholes and Parts: The Limits of Composition 142
1 Partwhole relations 142
2 The formal properties of partwhole relations 145
3 Unrestricted composition? 146
11 Micro-composition 151
1 Introduction 151
2 Containment 152
3 Causation 152
4 Working parts 153
5 Significant effects 154
6 Vagueness 155
7 Microreduction 156
8 Supervenience 157
9 Supervenience and vagueness 158

10 Supervenience and microreduction 159
11 Microreduction, causation, and laws 160
12 Contingency and indeterminism 161
Part III: Time
12 Time 163
1 Space, time, and relativity 163
2 Cosmology and the present 166
3 Time and change 169
4 The ontology of the A-series 170
5 The semantics of the A-series 172
6 What A-beliefs do for us 176
7 Time and causation 178
8 Epilogue 180
13 Transcendental Tense 182
1 Is time no object? 182
2 Kantian tenses 185
3 Time and concepts of time 186
14 Time Travel 190
1 The passing of time 190
2 Forward time travel 191
3 Backward time travel 193
4 The direction of time 195
5 Soluble problems of time travel 199
6 Time travel: the insoluble problem 203
15 The Direction of Time 206
1 Formalities 206
2 Extrinsic and intrinsic differences 206
3 The flow of time 207
4 The expansion of the universe 208
5 Increasing entropy 208
6 Irreversibility 209
7 Seeing the direction of time 210
8 Causal and temporal order 211
9 Causation and time 212
10 Conclusion 215
Preface

All the ensuing chapters except chapter 5 have been published before: I give the
details below. Most were first written for philosophical publications, but four were
not. Chapters 4 and 14 were the public lectures, and chapters 8 and 13 contributions
to the symposia, described in their abstracts: hence their different styles. The many
years over which all the chapters were written makes them differ in other ways too.
I have not updated references, or statements of views I later changed, for scholarly
reasons and because, as Donald Davidson said of his collected papers, over the
years they have attracted comment and criticism, and it would be mean spirited to
try to move the target out of range after the shot has been fired (1980 Introduction
p. xii). I have changed previously published papers only by removing the original
acknowledgements and by making their references and typography more uniform.
My philosophical debts to teachers, colleagues, and students should be evident
in what follows and are acknowledged as I go. I am also indebted to Peter
Momtchiloff for the books title, Mike Edwards for the frontispiece, Elisabeth
Leedham-Green for the Latin, and Steven Methven for the index.
Permissions
Chapter 1: Nothing Like Experience, Proceedings of the Aristotelian Society 93 (1992
3): 116. The Aristotelian Society 1992. Republished by courtesy of the Editor.

Chapter 2: What Does Subjective Decision Theory Tell Us?, Ramsey's Legacy, ed.
Hallvard Lillehammer and D. H. Mellor, Oxford: Clarendon Press (2005): 137148.
D. H. Mellor 2005. Republished by permission of Oxford University Press.

Chapter 3: How to Believe a Conditional, Journal of Philosophy 90 (1993): 23348.
The Journal of Philosophy 1993. Republished by permission.

Chapter 4: Telling the Truth, Ways of Communicating, ed. D. H. Mellor, Cambridge:
Cambridge University Press (1990): 8195. Cambridge University Press 1990.
Republished by permission.

Chapter 5: Successful Semantics. First publication. D. H. Mellor 2012.

Chapter 6: The Semantics and Ontology of Dispositions, Mind 109 (2000): 75780.
Oxford University Press 2000. Republished by permission.

PREFACE xiv
Chapter 7: Truthmakers for What?, From Truth to Reality: New Essays in Logic and
Metaphysics, ed. H. Dyke, London: Routledge (2009): 27290. D. H. Mellor 2009.
Republished by permission of Taylor and Francis.

Chapter 8: Too Many Universes, God and Design: The Teleological Argument and
Modern Science, ed. N. A. Manson, London: Routledge (2003): 2218. D. H.
Mellor 2003. Republished by permission of Taylor and Francis.

Chapter 9: The Reduction of Society, Philosophy 57 (1982): 5175. The Royal
Institute of Philosophy 1982. Republished by permission of Cambridge University
Press.

Chapter 10: Wholes and Parts: The Limits of Composition, South African Journal of
Philosophy 25 (2006): 13845. D. H. Mellor 2006. Republished by permission of
the South African Journal of Philosophy.

Chapter 11: Micro-Composition, Being: Developments in Contemporary Metaphysics,
ed. R. Le Poidevin, Cambridge: Cambridge University Press (2008): 6580. D. H.
Mellor 2008. Republished by permission of Cambridge University Press.

Chapter 12: Time, The Oxford Handbook of Contemporary Philosophy, ed. F. Jackson
and M. Smith, Oxford: Oxford University Press (2005): 61535. D. H. Mellor
2005. Republished by permission of Oxford University Press.

Chapter 13: Transcendental Tense, Aristotelian Society Supplementary Volume 72
(1998): 2943. The Aristotelian Society 1998. Republished by courtesy of the
Editor.

Chapter 14: Time Travel, Time, ed. K. Ridderbos, Cambridge: Cambridge
University Press (2002): 4664. Darwin College Cambridge 2002. Republished
by permission of Darwin College and Cambridge University Press.

Chapter 15: The Direction of Time, The Routledge Companion to Metaphysics, ed. R.
Le Poidevin, P. Simons, A. McGonigal and R. Cameron, Oxford: Routledge (2009):
44958. D. H. Mellor 2009. Republished by permission of Taylor and Francis.


Illustrations
Chapter 4: Telling the Truth:
Copyright The Estate of E. H. Shepard.
Introduction
1 What is philosophy?
n 29 November 1912 the 23-year-old Ludwig Wittgenstein read a paper on
What is philosophy? to the Cambridge University Moral Sciences Club. The
minutes say that his paper lasted about 4 minutes
1
and that, in it, philosophy was
defined as all those primitive propositions which are assumed as true without proof
by the various sciences, adding that this definition was much discussed, but there
was no general disposition to adopt it. There still isnt. The definition is too rarely
even considered, or its plausibility as a definition of metaphysics might stop that
subjects content, feasibility, and utility being as contentious as it is (Chalmers,
Manley et al. 2009 ch. 3). It might, for example, stop some philosophers denying that
metaphysics can enhance a physics which does not ingest it (Ladyman, Ross et al.
2009), and stop others contradicting the entangled ontology of modern microphysics
by denying that objects are ever more than the mereological sums of their parts (see
chapter 11). A metaphysics that satisfies the young Wittgensteins definition can and
does add to his various sciences without being either entailed by or inconsistent
with them.
This applies not only to the empirical sciences but to logic and mathematics. No
system of either can establish what Bob Hale (1996) calls the absolute necessity of
its theorems, because it cannot prove its basic axioms or rules of inference: there are
no premise-free proofs. So they too need primitive propositions assumed as true
without proof that must be argued for in other ways. That is why the necessary
truth for all P of P or not-P is as philosophically debatable as that of Water is
H
2
O: logics that assume or abjure the law of excluded middle can no more prove or
disprove it than physics and chemistry can prove or disprove that water not only
contains H
2
O but is identical to it. It is why, in their preface to Principia Mathematica
(1913), Russell and Whitehead say that
the chief reason in favour of any theory on the principles of mathematics must always be
inductive, i.e. it must lie in the fact that the theory in question enables us to deduce ordinary
mathematics.

1
Two weeks earlier the Club had resolved that The whole object of the papers read shall be,
as a general rule, to open a discussion, and therefore no paper shall last longer than seven
minutes, except by previous permission of the Chairman on a special occasion. Those were the
days: I doubt if the Clubs present practice, of letting speakers take ten times longer than the
young Wittgenstein to say rather less, improves its discussions.
O
MIND, MEANING, AND REALITY 2
In short, while Russell and Whitehead do not take ordinary mathematics to entail
their philosophical theory of it, they do require that theory to entail and hence be
consistent with it.
Of course, to ensure that our philosophical theory of a science is consistent with
it, we must take care not to misread it, as some have done. Take the simple kinetic
theory on which a gass absolute temperature is proportional to its particles mean
kinetic energy. This theory does not say that even in gases, let alone in anything
else, temperature is mean kinetic energy, which would imply, among other things,
that all gas particles at rest are at absolute zero, which they arent, and can be heated
just by accelerating them, which they cant (see chapter 6.9). Philosophers should
not keep asserting this evidently false identity when there are equally evident true
ones: for example, that all light is electromagnetic (e-m) radiation, and that an
objects inertial mass is identical with its gravitational mass.
2 Untestable identities
These two true identities illustrate how far physics, as well as metaphysics, exceeds
what could be tested by observation. Take the identity of light with radiation. The
undisputed evidence for this is that all light is accompanied by e-m radiation that
shares its velocity, is reflected, refracted, and diffracted in the same way, and whose
frequency distribution determines its colour. But this fits two theories. One says that
light is not e-m radiation, merely correlated with it by deterministic laws of nature.
To this, the identity theory, that light is e-m radiation, is an empirically untestable
addition. Yet all physicists accept it for the simple and parsimonious explanation it
gives of the correlation it entails. Similarly for the theory that gravitational mass,
defined by the laws of gravity, is not merely correlated with but identical to inertial
mass, defined by the laws of motion. Similarly also, and even less deniably, for the
identity of inertial mass itself, as follows.
Newtons laws of motion say that any force F acting on any object o of mass M
will give o an acceleration A proportional to F, and in the same direction, provided F
does not alter M, e.g. by knocking bits off o. With this proviso, a 1 kg mass will be
accelerated at 1 m/sec
2
by a force of 1 newton, at 2 m/sec
2
by 2 newtons, and so on,
for infinitely many different forces and directions. In other words, os having the
mass M makes true infinitely many conditionals saying that F would accelerate o at
A=F/M, all of which are logically independent.
2
So each of them could be made true

2
Those who deny that conditionals like this are either true or false (e.g. Edgington 1986; Levi
2002) may replace makes true by (e.g.) supports: it makes no odds to the argument. Nor do
the facts that (a) to embody the proviso that F does not alter M, the conditionals must be
reduction sentences (chapter 6.4) and (b) as the laws of motion bind M and F into a package
where neither is independently definable, they need other theories, like Newtons theory of
gravity, to make M and F empirically detectable by linking them to observable facts, like the
orbits of the planets.
INTRODUCTION 3
by a different property: with M
1
making 1 newton accelerate o at 1/M m/sec
2
; M
2

making 2 newtons accelerate it at 2/M m/sec
2
; and so on. A theory postulating this
plethora of properties, with values made equal merely by a correlation, would again
be empirically indistinguishable from the identity theory, that all these conditionals
are made true by os having the single inertial property M. Yet no one thinks that o
has more than one such property at a time, or that there is no fact of the matter
about how many it has. And since no physicist doubts that we can know that, and
when, o has a single mass property M, no philosopher can claim scientific backing
for a verificationism which entails that no one could know that.
3 Science and metaphysics
I have said that the metaphysics of a science should be consistent with it. But that is
only because our beliefs cannot all be true if some are inconsistent with others. All
an inconsistency shows is that something is wrong: it does not tell us what. In
particular, it does not tell us that our metaphysics must always give way when it
contradicts our science, though often of course it should. Mereologists, for example,
should accept that a things parts can change over time, as the parts of our bodies
do. They should accept too that irreducibly probabilistic sciences, from genetics to
epidemiology, as well as the microphysics referred to in 1, show that the properties
of few if any things with parts supervene on the properties and relations of those
parts (see chapters 10 and 11).
But science does not always trump metaphysics. Take Richard Feynmans (1949)
theory of positrons as electrons travelling backward in time. This rests on two facts:
that positrons only differ from electrons in being positively rather than negatively
charged, and that same charges repel and opposite ones attract. These facts make
positrons move away from positive charges, just as electrons move toward them, so
that a video of either movement, played backward, looks just like a video of the
other played forward. Feynmans explanation of this assumes as true without
proof the metaphysical thesis that time gets its direction from that of irreversible
processes, in this case those of electrons moving toward positive charges and away
from negative ones.
This process is not the only so-called arrow of time that has been said to give
time its direction. Others include the expansion of the universe, the increase of
entropy in isolated systems, and the divergence of waves from point sources. But
none of them will do, because they all rule out reversals of processes that even
physicists agree only contingently always or mostly go one way: the universe could
contract, entropy can and sometimes does decrease in isolated systems, and waves
often converge on points, as they do whenever eyes and camera lenses make light
converge to form sharp images. (See chapter 15.)
Similarly, some electron-like entities both can and do move away from positive
charges and toward negative ones. They are the ones we call positrons and credit
MIND, MEANING, AND REALITY 4
with the positive charge that distinguishes them from (other) electrons, a distinction
that Feynmans time-travel theory cannot explain: since if that theory does more
than relabel the fact to be explained, it is false. A positron cannot, for example, be
travelling backward in time as Dr Whos time machine TARDIS does: for if it did,
electric charges that deflected it at a time t would affect its positions at times that for
us are earlier than t; and they dont (see chapter 14 figure 4). Positrons are not
backward time-travelling electrons in this or any other serious sense, any more than
electrons are time-travelling positrons, or than the anti-clockwise clock I have is a
time-travelling clockwise one. Backward time travel worth the name would not be
that easy even if it were possible, which it isnt (see chapter 14).
My other example of metaphysically flawed physics is Martin Reess multiverse
hypothesis (1997), that our universe is only one of many. This is meant to explain
why our universes constants and initial conditions are fine-tuned for life, i.e. have
the a priori very improbable values that permit supposedly intelligent beings like us
to exist in it. The explanation is that, as a multiverse of universes with all physically
possible values of these constants and conditions is bound to contain some universes
we can live in, it is not improbable but certain that we will be in one of these. This is
offered as a secular alternative to the theory that a supernatural designer gave the
one and only universe the constants and conditions that let us exist in it.
Both theories misinterpret the a priori improbability of a universe containing life
as a low physical probability, like that of a tossed coin landing on edge, which might
make us suspect a non-chance process (see chapter 8). It is not: it is merely the low
epistemic probability which the life hypothesis is given by the mere forms of our
universes laws and a flat probability distribution over consistent values of their
constants and its initial conditions. That the epistemic probability of this hypothesis
is very low on that skimpy evidence and as high as it can beoneon our actual
evidence, which includes the fact that we do exist, is not a mystery that needs
explaining. It is a triviality that neither multiverse nor design theories either can or
should be asked to explain. And as with Feynman, what tells us this is not physics
but metaphysics, in this case the metaphysics of probability.
The moral of these examples is not that scientists need to be metaphysicians, any
more than metaphysicians need to be scientists. All each party needs to do is attend
to the other, and not assume that, whenever they disagree, the scientists must be
right and the metaphysicians wrong.
But where is the boundary, if all scientific theories contain presuppositions that
Wittgensteins definition makes philosophical? It would be as silly to say that this
fact makes all scientific theories philosophical as to say that field theories, e.g. of
gravity, which conform to the metaphysical principle of no (unmediated) action at
a distance, make that principle scientific. Philosophy and science can be distinct
without being independent, as Wittgenstein tells us. What he does not tell us is
where to draw the line between them: precisely because, as he says, it is to be
drawn, not a priori, but by what the various sciences presuppose, which changes as
INTRODUCTION 5
they do. Euclidean geometry, for example, was a metaphysical presupposition of
Newtonian mechanics until the advent of other geometries, and of physical reasons
to let light define straight lines in space, moved spatial and spacetime geometries
from metaphysics to physics. Similarly with the developments in microphysics that
replaced the seventeenth centurys metaphysical atoms, first with the nineteenth
centurys chemical atoms and physical fields, and then with the twentieth centurys
quantum ensembles. This is another reason why metaphysicians are as ill-advised to
ignore science as to surrender to it.
4 Methodology
The familiar fact that metaphysical theories often turn into scientific ones does not,
pace Ladyman et al. (2009), imply that metaphysics is, or ever will be, redundant:
Wittgensteins various sciences will always need presuppositions. But this does
not require the methodology of metaphysics to differ from that of the sciences. The
methods of the sciences vary, of course, with their subject matter: biologists have as
little use for telescopes as cosmologists have for microscopes. More generally, logic,
mathematics, and metaphysics respond less, or less directly, to observational data
than empirical sciences do. But this does not stop all these subjects sharing an
objective that their methods are meant to achieve: namely, to generate and support
theories that will explain the apparent facts of their domainsapparent because
theorists who cannot explain some of these facts may trade scope for success by
denying or excluding them.
Take for example Eulers conjecture, that for all polyhedra, e.g. the cube in Figure
1(a) below,
(EC) VE+F = 2,
where V is the number of vertices, E the number of edges and F the number of faces.
This apparent fact may be explained by the following theoretical deduction (DEC)
(Lakatos 1976 ch. 1.2).
First, turn the cube (a) into the 2-D figure (b) by stretching and flattening it after
removing a face, thereby reducing VE+F by 1. Next, add a diagonal edge to each
face, thereby adding as many faces as edges (c). This leaves VE+F unchanged, as
does the final step, of removing in turn all but one of the triangles so formed, since
each removal reduces F and E by 1 (or, at the end, V and F by 1 and E by 2). Then
since, for the remaining triangle (d), with three vertices, three edges and one face,
VE+F = 1, it follows that (EC) is true of the cube (a).
Whether (DEC) explains (EC) at all is a very moot point, but not the one I want to
raise. The issue here is whether (EC) is refuted by such apparent counter-examples
as a cylinder, with three faces, two edges and no vertices, for which VE+F = 1 and
to which (DEC) cannot apply, since its faces cannot be divided into triangles. To
reject this counter-example we must deny that a cylinder is a polyhedron and hence
MIND, MEANING, AND REALITY 6
a shape to which (EC) and (DEC) apply. And failing a better theory that applies to
cylinders too, it may be as reasonable to protect (EC) and (DEC), by limiting their
scope in this way, as to protect the gills theory of how fish breathe by denying that
aquatic mammals are fish.
On a more serious and positive note, one virtue of a theory of continental drift is
its ability to explain otherwise inexplicable and hence suspect similarities between
land-bound animal species, e.g. in the Americas and Africa, as an effect of past
migrations. Similarly in metaphysics: theories of causes and effects as facts, that put
events (e.g. a movement) and non-events (a staying still) on a causal par, can credit
the latter with causes and effects more readily than theories which, for other
reasons, limit causes and effects to events (see chapter 5.23).
In short, the criteria of scope and success used to judge metaphysical theories are
the same as those used in science and mathematics. There is nothing peculiar about
the methodology of metaphysics. Indeed there is very little to it beyond a few
platitudes that apply equally to all secular non-fiction: admit no unchallengeable
authorities; write relevantly, clearly, and concisely; dont assert or infer what you
dont believe; and dont use jargon or be needlessly technical.
Not only is the methodology of metaphysics neither special nor problematic, it
helps its practitioners as little as that of most other subjects helps theirs. No one, I
trust, thinks studying the methodology of drama would have made Shakespeare a
better playwright; merely a less prolific one. The methodology of science is no more
helpful to scientists, not even in scientific revolutions. For even then, what rescues
or replaces a faltering theory, like classical mechanics in the nineteenth century, is
usually not a methodological but a metaphysical thesis: for example, that the same
laws of nature apply everywhere and always to all things, whatever their relative
motions: a thesis which, if the laws give something (e.g. light) an invariant finite
speed, entails the Lorenz transformations of special relativity (Minkowski 1908
p. 79, Lange 2013).
However, as no one is perfect, and authors in different subjects may fall short in
different ways, a few peculiarly philosophical bad habits may be worth deploring.

(a) VE+F = 2 (b) VE+F = 1 (c) VE+F = 1 (d) VE+F = 1
FIGURE 1: Deduction of Eulers conjecture
INTRODUCTION 7
One is that of relying too much on analysis, i.e. on clarifying our concepts without
questioning factual assumptions built into them. To take a reported example of
William Kneales, it is built into our concept of kittens that they are (a) the offspring
of cats and (b) grow into cats. That is mostly true, but what makes it so is biology,
not our concepts. And if cats evolved from simpler organisms, as they did, then (a)
and (b) cannot both be true of all of them. Some ancestors of cats were not cats, just
as some of our ancestors were not human beings. Similarly, the logically possible
hypothesis that the world sprang into being five minutes ago, with a population
that remembered a wholly unreal past (Russell 1921 p. 159), is not refuted just by
our concept of the world being of something that is millions or at least (for Biblical
literalists) thousands of years old.
3

The second bad habit on my list is that of requiring philosophical disputes to be
won by a knockout, i.e. by showing that no rival theory could be true. While that
may happen in logic and mathematicsthough if P or not-P is disputable, what is
not?it rarely if ever happens in philosophy or the empirical sciences. Cartesian
scepticism, for example, can no more be absolutely disproved than can creationism,
or the theory that Christopher Marlowe wrote Shakespeares plays. But that is no
excuse for being, or posing as, a Cartesian sceptic, a creationist or a Marlovian: for in
all these cases other theoriesthat there is an external world, that species do evolve,
and that Shakespeare wrote the plays published in his namewin, if not by a
knockout, then decisively on points. Of course it does happen in philosophy, as in
science and history, that even the best theory of a subject faces serious objections;
but that is no excuse for having equal credence in others that face worse ones. To
vary the sporting metaphor, no sane bookmaker will offer the same odds on every
horse in a race, and nor should we on every competing theory of a scientific or
philosophical subject.
This brings me to a third bad habit, of wasting too much time on theories no one
believes, like Cartesian scepticism. I do not mean it is never worth discussing what
is wrong with arguments for such theories, only that their internal consistency is not
a sufficient reason to take the theories themselves seriously. Why this seems to
happen more often in philosophy than elsewhere, if it does, I am not sure, but I can
see no need or excuse for it. Evolutionists and Shakespeareans have real opponents,
as I do when I argue in chapter 14 against those who think backward time travel is

3
Compare Thersitess comment on Troiluss reaction, in Shakespeares Troilus and Cressida, to
his eavesdropping on Cressidas night-time tryst with Diomed:
TROILUS: This she? no, this is Diomeds Cressida:
If beauty have a soul, this is not she;
If souls guide vows, if vows be sanctimonies,
If sanctimony be the gods delight,
If there be rule in unity itself,
This is not she.
THERSITES: Will he swagger himself out ons own eyes?
MIND, MEANING, AND REALITY 8
possible or even, as we saw in 3, actual. These are serious debates about what to
believe, which debates about Cartesian scepticism are not.
The cure for this habit is, of course, to heed one of the platitudes listed above. For
since philosophy, like logic, science, and mathematics, is not a branch of fiction,
philosophers have no academic (as opposed to personal, clinical, or political) excuse
for asserting things they do not believe: e.g. There is no external world, Language
is not used to describe the world, We can always create an entity by creating a
name for it, or Animals cannot feel pain because they have no concept of pain.
5 Meta-philosophy
I have far more confidence in the one man who works mentally and bodily at a matter than in
the six who merely talk about it (Michael Faraday, Letter to John Tyndall 19 April 1851, in
F. A. J. L. James (1999) The Correspondence of Michael Faraday, vol. 4 letter 2411).
The last bad habit on my list is that of discussing philosophy instead of doing it, i.e.
of doing meta-philosophy, or the philosophy of philosophy, whose remit I take to
be what philosophy is and how to do it. About what it is, I think Wittgensteins
definition in 1 says all that need or can usefully be said. The topic of how to do it I
take to comprise its methodology, discussed in 4, and its methods, which, like
those of other subjects, are best mastered by emulating exemplars. That after all is
why philosophers, like scientists and mathematicians, as well as historians, lawyers,
artists, and motorists, mostly start as apprentices, learning their trade by doing it
under supervision: these days, in discursive subjects, usually by writing theses that
assess and add to existing work in that subject, not meta-work about it.
It is a fact that this process, of mastering a subject by learning how to do it, rarely
needs to include learning a rationale for its scope and methods. Hence the lack of
interest of most practitioners of most subjects in their subjects definition and
methodology. Why then is philosophy different, as it seems to be: why do so many
philosophers indulge in meta-philosophy? Perhaps because, unlike the philosophy
of other subjects, it is a part of the subject it is about. But that is hardly a good
enough reason to do it unless it is an enlightening or important part, which I have
never found it to be.
Nor is it necessary, since we can do philosophy perfectly well without doing
meta-philosophy, just as we can do other subjects without doing their philosophy.
But not vice versa, because philosophy, like mathematics and languages, is not a
spectator sport: its products, unlike those of chefs and poets, can only be judged by
those who can do it. So would-be meta-philosophers must first learn to do first-
order philosophy. And anyone who can do that, i.e. add seriously to metaphysics,
ethics, aesthetics, political philosophy, or the philosophies of science, mathematics,
logic, language, history, or law, has better things to do than meta-philosophy.
INTRODUCTION 9
Finally, to the charge that this introduction is itself a piece of meta-philosophy, I
can only reply that it is meant to be aversion therapy, i.e. to do just enough meta-
philosophy to deter readers from doing any more. It may well fail, just as my and
Tim Cranes There is no question of physicalism (1990) failed to stop colleagues
wasting time on that question. But our failure did not make what we said false, not
worth saying, or self-refuting, and nor is what I have said about meta-philosophy.
Anyway, I have not said much. I have never been guilty of lecturing, or editing or
writing a book, on the subject. And there is no more of it in this book: what follows,
for better or worse, is just philosophy.

You might also like