Thursday, May 14, 2020
H. P. Grice, "Reply to D. M. Armstrong on meaning and communication"
David Armstrong’s many major contributions are focused in traditional
epistemology and metaphysics. He offers comprehensive accounts of what there
is, its nature, and how we know about it. He is a “system builder.” His work is
informed by the conviction that philosophers must take very seriously the
teachings of science. He is a realist: about mental states, about properties,
about laws, and about singular causation. Indeed, on almost any philosophical
topic, if there is a realist position available, Armstrong will occupy it.
Also, he seeks what is now often called “the view from nowhere.” He is opposed
to the idea that there may be different, equally legitimate but, to one degree
or another, incommensurate, views of how things are from one or another
perspective. Or, as it is sometimes put, he denies that there are different
kinds of being or of truth. Materialism about the mind Armstrong is probably
best known for A Materialist Theory of the Mind. His theory is commonly known
as central state materialism or as the causal theory of mind. Armstrong started
his philosophical life as a behaviorist but, partly as a result of the
influence of J. J. C. (Jack) Smart, moved to the view that mental states are
states of the central nervous system, and more especially the brain. Armstrong
develops his central state version of the identity theory by first arguing that
the concept of a mental state M is the concept of a state that plays a
distinctive causal role that connects stimulus, behavioral response, and other
mental states. Thus, to give the rough idea, pain is the state typically caused
by bodily damage, and typically causing a desire that it itself cease and a
behavioral response that tends to, or is believed to, minimize the damage. Obviously,
an account of this kind is exactly what evolutionary considerations would
suggest. In similar fashion, belief is a state induced by subjects’
environments that tends to make them behave in ways that realize what they
desire if what they believe is true. Armstrong sees two major advantages of
this kind of view over behaviorism. First, it allows mental states to be causes
of behavior. Secondly, by bringing in reference to other mental states, it
allows for suitably complex accounts of the connections between mental states
and behavior. It is notorious that there is no simple one-to-one matching of
mental states and behavior. What you do when you think it is about to rain
depends, Blackwell Companions to Philosophy: A Companion to Analytic Philosophy
Edited by A. P. Martinich, David Sosa Copyright © Blackwell Publishers Ltd 2001
inter alia, on whether you want to stay dry, on where you think the umbrella
is, on whether you think you are Gene Kelly, and on how cold you feel. On the
central state theory, to ask after the identity of a given mental state M is to
ask what state plays the distinctive, causally intermediate role assigned by
the concept of M. Armstrong argues that, for each mental state, it will turn
out to be some state or other of the brain that plays the role in question. He
concludes, therefore, that, as an empirical matter of fact, mental states are
identical with brain states. These identities will be contingent because which
brain states play which roles is a contingent matter. Some have objected that
there are no contingent identities: everything is necessarily identical to
itself, and that what Armstrong (and Smart) should say is that the identities
are a posteriori. In fact, they hold that the identities are both contingent
and a posteriori, but the objection to the contingent identity part of their
theory is a misunderstanding. All they mean is that sentences of the form “M is
B” are contingent, in the same way that “Red is the color of bullfighters’
capes” clearly is. A second misunderstanding is over Armstrong’s stance on the
possibility that quite different states might play the causal role distinctive
of pain in different species and, maybe, in different members of the one
species. It is often objected that identity theorists are committed
(implausibly) to pain being the same state in everything that experiences pain.
But consider the following parallel. The most dangerous virus for dogs is
different from the most dangerous virus for people, and the reason for this is
that the kind that plays the relevant role in dogs differs from the kind that
plays the relevant role in people. Nevertheless, we can, and do, identify the
most dangerous virus for dogs and for people – or anyway the experts do it for
us. A more pressing question is whether Armstrong (and Smart) should have said
that mental states are constituted by, rather than identical with, brain
states. The relation between a table and the parts that make it up is one of
constitution, not identity. Because the life histories of the table and its
parts differ – for example, the parts typically come into existence somewhat
earlier than the table – Leibniz’s Law means that the relation cannot be one of
identity; it must be constitution. (A separate question is whether this
relation of constitution can be analyzed in terms of identity between temporal
parts of the table and temporal parts of various aggregations of parts.)
Similarly, it may well be that Armstrong (and Smart) should, strictly, say that
mental states are constituted by brain states, not that they are identical to
them. Armstrong’s central state view of mind is sometimes contrasted with the
kind of functionalist theory of mind associated with the early Hilary Putnam
(see PUTNAM). They both agree in giving functional roles a central role in the
theory of mind. This is because Armstrong’s causal roles can equally be
described as functional roles. The stimuli that Armstrong talks of are inputs,
as functionalists say it, and the behavioral responses are outputs, as
functionalists say it. There are two big differences, though. Armstrong thinks
of the mental states as the occupants of the functional roles, as the states
that are suitably interconnected to inputs, outputs, and other, internal mental
states. Putnam thinks of them (or thought of them when he was a functionalist)
as the functional roles themselves. And, secondly, the functional roles in
Armstrong’s theory are those sometimes called “common sense.” Their inputs and
outputs are described in terms familiar to us all: rain, umbrellas, movements
that lead to beer inside the mouth, FRANK JACKSON 414 etc. In Putnam’s version
of functionalism, though not in all versions of functionalism, the inputs and
outputs are thought of as internal ones. Perception, sensations, belief,
knowledge Armstrong’s Perception and the Physical World is an argument for
direct realism in perception. He argues that we are directly acquainted with
independently existing physical objects in perception. The distinctive feature
of his argument is the way it is founded on an analysis of perception and
perceptual experience in terms of the acquisition of belief through the
operation of one’s sense organs. This makes good sense of the central
biological function of perception, which is the acquisition of belief about what
is going on around and inside one. An obvious question for Armstrong’s account
is raised by the fact that the very same belief, say, that it is raining
outside, can be acquired in very different ways through very different
perceptual experiences. You might, for instance, see that it is raining, be
told that it is raining, read on a computer screen that it is raining, or hear
that it is raining. Perhaps the most plausible way of approaching this problem
is in terms of the distinctively different clusters of beliefs in each case. In
none of these cases, does the belief that it is raining come “by itself ”;
rather, it comes as an integral part of a whole cluster of beliefs, and the
clusters are different in, and distinctive of, each case. In Bodily Sensations,
Armstrong gives an account of somatic sensations in terms of perception of
one’s own body. A sensation is an experience of perceiving that one’s body is
in such and such a state, an experience which may or may not be veridical. For
example, a feeling of warmth is the putative perception that a part of one’s
body is warm. In the case of certain sensations, the putative perception is
accompanied by a characteristic attitude. Pain, for example, is the putative
perception that there is something amiss with part of one’s body, accompanied
by an immediate dislike of this putative perception. Armstrong’s treatment of
belief follows a suggestion of F. P. Ramsey’s that belief is like a map by
which we steer. Inside our heads is a master map that moves us through the
world in such a way that what we desire is achieved to the extent that the map
is correct, and individual beliefs are thought of as sub-maps of the master
map. This approach to belief is now a standard alternative to the internal
sentence theory of belief supported by language of thought theorists. His
account of knowledge is a reliabilist one. Knowledge necessarily involves true
belief: if S knows that P, then S truly believes that P. But not all true
belief is knowledge; the truth of a belief may be an accident, and how can
getting it right by accident be knowledge? Armstrong’s suggestion, roughly, is
that S’s true belief that P is knowledge if it is a reliable sign that P. Here
he differs from the tradition that requires that one’s belief be justified in
order to count as knowledge. Time and action Armstrong holds a temporal part,
or stage, metaphysics. Identity over time is a matter of having parts or stages
at the times in question. I was at the Melbourne Test when “Typhoon” Tyson took
7 for 27 because a certain person-stage with the right DAVID M. ARMSTRONG 415
connections to the person-stage writing these words was present at that test
match. Armstrong’s main contribution to the debate is one of the very first
discussions of the famous rotating homogeneous cylinder/disk/sphere example. He
argues that the example shows the conceivability of a conception of identity
through time not framed in terms of temporal stages, but that, nevertheless,
the temporal stage account of identity through time is in fact correct. What
makes it true, on his view, that such an object is rotating are the
dependencies between different stages. Armstrong was also one of the first,
with Brian O’Shaughnessy, to argue that if one acts, one must have tried to
act, and that this is the essence of truth in the old volitional theory of
action. Universals, laws, causation, possibility, and states of affairs
Truth-makers play a crucial role in Armstrong’s later philosophizing. The basic
idea is that if some sentence or proposition is true, there must be something
that makes it true; similarly, if some predicate applies to something, there
must be something that makes it true that the predicate applies. You cannot say
that the word “square” applies to A, and that that is all there is to say.
There must be something about A that makes it true that the word applies to it,
that A satisfies it. In Armstrong’s hands, the truth-maker principle, as he
calls it, is more than the widely accepted supervenience of truth and
satisfaction on nature. Supervenience says that if a sentence is true in one
situation and false in another, and if a predicate is satisfied by one thing
but not by another, the situations and things must differ in nature. The
truth-maker principle goes further. It says that there must be something that
makes – necessarily makes – the true sentence true and the satisfied predicate
satisfied. Armstrong holds that what makes it true that predicates apply to
particulars are the properties or universals that the particulars possess. In
keeping with his realist leanings, these universals exist independently of the
classifications that we find natural. They are in nature. Secondly, they are
not to be reduced to sets, or to resemblances between particulars. Armstrong is
not a nominalist. He argues, in particular, that nominalists cannot handle the
famous “one over many” problem, the problem of what unifies things that share a
property. Thirdly, there are no uninstantiated universals; every universal is
possessed by at least one thing. In this sense, he is with Aristotle and not
Plato. He regards the Platonic view that there are uninstantiated properties or
universals as an unmotivated ontological extravagance. Fourthly, there is not a
one–one relation between properties and predicates: one and the same universal
or property may be the truth-maker for a number of different predicates. To
illustrate: suppose that U is a universal and that “A” is a predicate that says
that something is U, and “B” is some quite different predicate. Surely, “A or
B” might be true of something which is U simply because it is U. We are not
required to postulate an extra property just because “A or B” is a distinct
predicate from “A.” Also, there may be properties for which there is no
predicate. Finally, which universals or properties there are is an a posteriori
matter to be settled by total science. Philosophy tells us that there must be
truth-makers for true predications, but what they are is ultimately a matter
for science broadly conceived. Armstrong argues strongly against Humean and
neo-Humean accounts of laws. For him, no facts about regularities, however
tricked up, can ever add up to lawfulness FRANK JACKSON 416 proper. What then
must be added to a regularity to get a law? His answer is that what distinguishes
the universal statements of the form “Every F is G” that express laws of nature
– that are nomic or nomological – from those that express accidental
regularities is that, roughly, the laws correspond to relations of nomic
necessitation between universals. In its simplest version, the idea is that
“Every F is a G” is a law if and only if Fness necessitates Gness. But more
detailed accounts would need to advert to his metaphysics of states of affairs,
mentioned briefly below, and to his treatment of laws that do not fall
obviously into the “Every F is G” mold, derived laws, and laws that have no
instances (for example, concerning motion in the absence of gravity). This
account of laws is, obviously, strongly anti-Humean. Armstrong’s account of
causation is equally counter to the tradition that comes to us from Hume, and
in three respects. First, Armstrong insists that causation is singular in that
it is a nonrelational property of a sequence (see ANSCOMBE). Secondly, he holds
that the connection between causation and law is a posteriori. He denies, that
is, that it is a priori that any singular causal sequence falls under some law.
He does, though, allow that it may well be that some or all causal sequences
are identical, as an a posteriori matter, with the instantiation of a law.
Finally, he holds that we sometimes directly perceive singular causal
connections. Here he is going against a widely held view, even among those who
would not describe themselves as Humeans. Many who agree with him that
causation is more than sequence suitably constrained think, nevertheless, that
sequence is all we literally perceive. We do not see that X caused Y; we infer
it. Sometimes their argument for this view is that a non-causal sequence can
seem as causal as can be, as Piaget’s famous experiments tell us. Armstrong
rightly points out that this only shows that illusion is possible, and the
possibility of illusion concerning a feature does not show that we do not
literally perceive it when all goes well. However, there is a stronger
argument. It is hard to identify the causal role that singular causation plays
in its alleged perception. When I see that something is square, its squareness
plays a role in inducing my perceptual experience. This seems crucial to its
being correct to say that I perceive its squareness. But what role does
singular causation play that might mirror the role squareness plays? All the
causal work seems to be being done by the sequence per se. Armstrong’s account
of possibility is a combinatorial one, drawing on his realism about universals.
We can think of how things are as a vast, complex arrangement of particulars
and universals. The various possibilities can then be thought of as all the
combinations and recombinations of these particulars and universals according
to various rules for combining particulars and universals. Thus, to give the
barest bones of the idea, suppose that there is in fact charge X at point y,
and charge U at point v. What makes it possible that there be charge X at v,
and charge U at y? His answer is the fact that putting X with v, and putting U
with y, does not violate the rules of combination. In his most recent book, A
World of States of Affairs, Armstrong argues that the best way to bring his
ideas on universals, laws, truth-making, and possibility together is by
adopting a metaphysics of states of affairs. For example, universals – the key
to his account of laws – turn out to be types of states of affairs. In any
case, for Armstrong, the world is not the aggregation of all the things there
are. It is the aggregation of all the states of affairs there are, where states
of affairs are things-having-properties. His DAVID M. ARMSTRONG 417 view is
essentially the same as Wittgenstein’s in the Tractatus, namely, that the world
is the totality of facts, not of things. Bibliography of works by Armstrong
1960: Berkeley’s Theory of Vision, Melbourne: Melbourne University Press. 1961:
Perception and the Physical World, London: Routledge and Kegan Paul. 1962:
Bodily Sensations, London: Routledge and Kegan Paul. 1968: A Materialist Theory
of the Mind, London: Routledge and Kegan Paul. 1973: Belief, Truth and
Knowledge, Cambridge: Cambridge University Press. 1978: Universals and
Scientific Realism, 2 vols., Cambridge: Cambridge University Press. 1983: What
is a Law of Nature?, Cambridge: Cambridge University Press. 1984 (with Norman
Malcolm): Consciousness and Causality: A Debate on the Nature of Mind, Oxford:
Blackwell Publishers. 1989a: A Combinatorial Theory of Possibility, Cambridge:
Cambridge University Press. 1989b: Universals: An Opinionated Introduction,
Boulder, CO: Westview Press. 1997: A World of States of Affairs, Cambridge:
Cambridge University Press. FRANK JACKSON 418 419 34
No comments:
Post a Comment