Subject: rts2-25
Date: Tue, 20 Feb 2001 04:50:57 -0700


5. AUTONOMY AND REDUCTION


Laws we already know do not describe the patterns of events.
But how do they stand to the world of our everyday action
and of perceived things?

         Reflect for a moment on the world as we know it.
It seems to be a world in which all manner of things happen
and are done, which we are capable of explaining in various
ways, and yet for which a deductively-justified prediction
is seldom, if ever, possible.  It seems, on the face of it
at least, to be an *incompletely described world of agents*.
A world of winds and seas, in which ink bottles get knocked
over and doors pushed open, in which dogs bark and children
play; a criss-cross world of zebras and zebra-crossings,
cricket matches and games of chess, meteorites and logic
classes, assembly lines and deep sea turtles, soil erosion
and river banks bursting.  Now none of this is described by
any laws of nature.  More shockingly perhaps none of it
seems even governed by them.  It is true that the path of my
pen does not *violate* any laws of physics.  But it is not
*determined* by any either.  Laws do not describe the
patterns or legitimate the predictions of kinds of events.
Rather it seems they must be conceived, at least as regards
the ordinary things of the world, as situating limits and
imposing constraints on the types of action possible for a
given kind of thing.

         {RTS2:106} Laws then not only predicate tendencies
(which when exercised constitute the normic behaviour) of
novel kinds (or of familiar things in novel or limit
situations); they impose (more or less absolute) constraints
on familiar things.  In this section I want to reconcile
these aspects of laws by arguing that familiar things are
comprehensive entities which may be controlled by (or
subject to the control of) several different principles at
once; and that they may be said to be agents.  Laws ascribe
possibilities which may not be realized and impose
necessities which constrain but do not determine; they
ascribe the former to novel kinds and impose the latter on
familiar things.  These features cannot be explained away as
an imperfection of knowledge; but must be seen as rooted in
the nature of our world.  They are therefore inconsistent
with the thesis of regularity determinism which underpins
the doctrine of the actuality of causal laws, and to which I
must now return.

         So far I have discussed regularity determinism
merely as an epistemological thesis to the effect that our
knowledge of the world can be cast in a certain form.  But
of course this presupposes that the world is such that our
knowledge of it can be cast in that form.  To deal with
regularity determinism I must thus draw out this ontological
presupposition; i.e. cast the thesis itself in ontological
form.  The main work for this has already been done.  For I
have already shown in sections 2 and 3 that regularity
determinism makes a claim about what would happen (and the
way it would happen) if certain highly restrictive
conditions were satisfied.  These were, it will be
remembered, conditions such that *if* we knew they were
satisfied and the constant conjunction formula was not
vindicated, the regularity determinist would be bound to
admit his thesis refuted.  Now regularity determinism's
ontological claim is simply that the world is such that
these conditions are satisfied and his thesis is not
refuted.  Now of course because we can never know that these
conditions are satisfied we can never refute regularity
determinism in this way.  But I have also asserted that
regularity determinism is metaphysically refutable.  How can
this be done?  In the only way open to transcendental
realism: that is by showing that if the world were as
claimed by regularity determinism science would be
impossible.  But as science is possible (which we know,
because as a matter of fact it occurs) the world {RTS2:107}
must be such that either the critical conditions are not
satisfied and/or the constant conjunction formula is
abrogated.  In short, the ontological untruth of regularity
determinism is a condition of the possibility of science.

         Close to the appeal of determinism lies the
following error: to think that because something happened
and because it was caused to happen, it had to happen before
it was caused.  Now if we take determinism to assert that
all events are determined before they happen and conceive
their determination as lying in the satisfaction of
antecedent sufficient conditions for them then we have a
picture of a chain of antecedent sufficient conditions for
events stretching back infinitely into the past (assuming
that conditions can be analysed as events or vice-versa).
So if we ask how long is an event determined before it
actually happens the answer must be at any (i.e. at every)
time before it happens.  And so if we now take cause in the
ordinary sense, we have the result that every event is
determined before it was caused (or made) to happen.  At
play here are of course two concepts of cause: qua causal
agent (cause_1) and qua antecedent condition (cause_2).  I
am going to argue that the former is irreducible to the
latter and essential to science.  To say that something is
*determined* before it has been caused to happen is either
to say that it can be *known* before it has been caused_1 to
happen (epistemic determinism) or that it has been
*caused_2* before it has been caused_1 to happen
(ontological determinism).  The former depends upon a
closure, the latter depends upon the critical conditions for
it being satisfied.  Now I want to argue that at any (and
every) time the world consists of things which are already
complexly structured and pre-formed wholes; which may be
simultaneously constituted at different levels and
simultaneously controlled by different principles.  It is
because things cannot be reduced to the conditions of their
formation that events are not determined before they are
caused to happen.  This fact accounts for both the temporal
asymmetry of causes and effects and the irreversibility of
causal processes in time.  And it is because things cannot
be reduced to atomistic components that when events are
caused to happen it is by the thing which acts (i.e. the
agent), the event being produced in the circumstances that
prevail.

         Now I want to argue that determinism is
ontologically false {RTS2:108} (it is not true that events
are determined before they are caused to happen, whether in
a regularly recurring or non-recurring way) and
epistemically vacuous (there are no significant descriptions
that satisfy the formula of regularity determinism).  This
has the methodological corollary that the search for such
descriptions is likely to be unrewarding.  (And here once
again it is necessary to counterpose the investigation of
complex preformed things to the search for the complete
atomistic state-descriptions that it is supposed would
enable us to predict their behaviour.)  The only sense in
which science presupposes `determinism' is the sense in
which it presupposes the ubiquity of causes_1 and hence the
possibility of explanations.  And the only sense in which it
presupposes `regularity determinism' is the sense in which
it presupposes the ubiquity of causes_1 for differences and
hence the possibility of their explanation.  But it is
probably better not to use `determinism' in this way
(nb. cause_1 is not the same as cause_2).  Now any
refutation of regularity determinism as an ontological
thesis must depend upon establishing the autonomy of things,
in the sense of the impossibility of carrying out the
reductions implicit in the vital conditions B1 and C1 of
Table 2.1 on page 76 (their being a clear asymmetry, for the
realist, between the subjects and the condition of action,
and the constancy alternative being recessive).  It is here
that I will pitch my attack.  Thus I am not going to argue
that if the critical conditions were satisfied the constant
conjunction formula would not be vindicated.  Rather, I am
going to argue that the critical conditions could not be
satisfied in any world containing science. The question of
whether or not history would repeat itself is one that need
not detain us here.  A nagging doubt may remain: surely, it
might be felt, in the (very) last instance regularity
determinism must be true.  But this is not so.  For once we
have established an ontology of structures there is no
earthly reason why events should [have to] be constantly
conjoined.  There are indeed principles of indifference (as
we shall see in Chapter 3).  But they do not apply, nor is
there any reason why they should, to events,
states-of-affairs and the like.

         In establishing the autonomy of things I will
follow the normal procedure of transcendental realism; that
is, I will first analyse some more or less underanalysed
feature of science and then ask what the world must be like
for this feature to be possible.  {RTS2:109} The feature I
am concerned with are two aspects of scientific laws, viz:-

      (i) their normic and non-empirical character; and

     (ii) their consistency with situations of dual (and
          multiple) control.


I will argue that for these features to be possible the
world must be composed of agents.  Agents are particulars
which are the centres of powers.  In an incompletely
described world of other agents powers must be analysed as
tendencies.  And laws are nothing but the tendencies or ways
of acting of kinds of thing.  By an agent 1 mean simply
anything which is capable of bringing about a change in
something (including itself).  A hydrogen atom is, in virtue
of its electronic structure, an agent.  For it possesses the
power to combine with an atom of chlorine to produce, under
suitable conditions, a molecule of hydrochloric acid.  It
should perhaps be said at the outset that I am not going to
refer to quantum mechanics in my argument.  It seems to me
to be always a mistake, in philosophy, to argue from the
current state of a science (and especially physics).  In
general, I have refrained from scoring points against
determinism and actualism which turn on the inaccuracy (or
imprecision) of our descriptions or the indeterminacy of our
measures.  This is because they do not in general raise
important ontological questions. It is debatable whether
quantum mechanics does - but if it in fact requires a
reinterpretation of the category of causality in fundamental
physics it will not be in the Humean direction and can only
strengthen the anti-determinist's hand.

         I have already discussed (i) at some length so I
will be brief with it here.  Contrast the law of
conservation of energy or of mass action with a simple
empirical generalization like `all pillar-boxes are red' or
`all blue-eyed white tom cats are deaf'.  Whereas the
latter, at least so long as they remain unattached to any
theory, could be defeated by a single counter-instance, the
truth of the former is consistent with almost anything that
might happen in the world of material objects and human
beings.  For they do not attempt to describe this world;
i.e. they cannot be interpreted as undifferentiated
empirical generalizations.  Rather they must be interpreted
as principles of theories - of physics and chemistry - which
tell us something about the {RTS2:110} way things act and
interact in the world.  As such they specify conditions
which we presume are not contravened but rather continually
satisfied in the countless different actions and
interactions of the world, including those of which we have
direct experience.  And they are manifest in certain
impossibilities, e.g. that of building a perpetual motion
machine.  Nevertheless they are principles for which any
test would require not only fine measurement but closed
conditions.  As such they are not normally empirically
manifest to us or actually satisfied.  (For the scientist
this feature appears as a difference between the real or
corrected and the actual or measured values of the variables
he is concerned with.)  Thus we could say that relative to
these vantage points, viz. of experience and actuality,
these principles specify levels of deep structure or
(metaphorically) place conditions on the inner workings of
the world.

         Now it might be said that laws, such as those of
mechanics or electricity, do not describe the world as such,
but only those aspects or parts of it which fall within
their domain, i.e. the mechanical or electrical aspects of
it.  But this concedes my point.  For one can only say which
aspects are mechanical or electrical by reference to the
antecedently established laws of mechanics and electricity,
and such aspects are real.  Clockwork soldiers and robots do
not more nearly observe the laws of mechanics than real
people.  Rather their peculiarity stems from the fact that
if wound up and left alone their intrinsic structure ensures
that for each set of antecedent conditions only one result
is possible.  But outside the domain of a closure the laws
of mechanics are, as Anscombe has put it, rather like the
rules of chess; the play is seldom determined, though nobody
breaks the rules'.^46

         Closely connected with this feature of laws is
their consistency with situations of `dual control'.  A game
of cricket is only partially controlled by the rules of
cricket, language-using by those of grammar.  Chemical
reactions are only partially controlled by Dulong and
Petit's law, black bodies behave in all kinds of ways that
are not specified by the Stefan-Boltzmann law.  Coulomb's
law does not completely describe the action of charged
particles, or Faraday's law all that happens to an
electrode.  Similarly the `boundary conditions' for the laws
of {RTS2:111} mechanics, the domain within which they apply,
are controlled by the operating principles defining a
machine.^47  Laws leave the field of the ordinary phenomena
of life at least partially open.  They impose constraints on
the type of action possible for a given kind of thing.  But
they do not say which out of the possible actions will
actually be performed.  They situate limits but do not
dictate what happens within them.  In short, there is a
distance between the laws of science and the ordinary
phenomena of the world, including the phenomena of our
actual and possible experience.  And it is with the
investigation of this distance that I am here concerned.

         To say that laws situate limits but do not dictate
what happens within them does not mean that it is not
possible to completely explain what happens within them.
The question `how is constraint without determination
possible?' is equivalent to the question how `can a thing,
event or process be controlled by several different kinds of
principle at once?'  To completely account for an event
would be to describe all the different principles involved
in its generation.  A complete explanation in this sense is
clearly a limit concept.  In an historical explanation of an
event, for example, we are not normally interested in (or
capable of giving an account of) its physical structure.

         In deciding to write `!' on this piece of paper I
select the conditions under which the laws of physiology and
physics are to apply.  So that it is absurd to hold that the
latter might account for my `!'; or that it might have been
predicted in the basis of a knowledge of a physical
state-description prior to my writing it.  On the other hand
my neuro-physiological state and the physical conditions
must be such that I can write it; they could prevent it
(e.g. if I were suddenly to fall asleep or be propelled into
orbit around the moon).  There is a space between the laws
of physics and physiology and what I do within which
deliberation, choice and voluntary behaviour have room to
apply.  The theory of complex determination, in situating
persons as comprehensive entities whose behaviour is subject
to the control of several different principles at once,
allows the possibility of genuine self-determination
(subject to constraints) and the special power of acting in
accordance with a plan or in the light of reasons.

         {RTS2:112} Human freedom, on this view, *if* it
exists, would not be something that somehow cheats science
(as it is normally conceived) or, on the other hand,
something that belongs in a realm apart from science; but
something whose basis would have to be scientifically
understood.  As freedom would be ana1ysed as a power of men
and science is, for us, non-predictive there is nothing
inconsistent or absurd about such an assertion; any more
than to say that purposefulness in animals, which is no
doubt not the same as intentionality in men, has (still) to
be scientifically understood.  I suggest that only the
theory of complex determination is compatible with agency;
and that there are no grounds for assimilating intentional
action to the classical paradigm or supposing that
intentionality is not a real attribute of men.  However,
this is peripheral to my main concerns here.  Dogs cannot
fly or turn into stones, but they can move about the world
and bark in all kinds of ways.  To deny the latter
possibility is as absurd as to deny the former necessity.
But the reasons why they behave in canine ways is an open
question for a putative science of animal ethology to
answer.

         The difference between laws of nature and empirical
generalizations is analogous to the difference between
the rules of cricket and a television recording of the
actual play on some particular occasion.  Whether or not
Boycott scores a century is not determined by the rules
of cricket; but by how he bats and how the opposition
play.  Now it is clearly necessary for the
intelligibility of the idea of dual (or multiple) control
that the higher-order level is open with respect to, in
the special sense of irreducible to, the principles and
descriptions of the lower-order level.  It is easy to see
why this must be so.  For it is the operations of the
higher-order level that control the boundary conditions
of the lower-order level, and so determine the conditions
under which the laws of that level apply.  It is the
state of the weather that determines, in England, when
and where the rules of cricket can apply; the state of
the conversation that determines the ways in which we can
express ourselves in speech; the state of the market that
determines the use of machines, the use of machines that
determines the conditions under which certain physical
laws apply.  The use of machines is thus subject to dual
control: by the laws of mechanics and those of economics.
But it is the latter that determine the boundary
conditions of the former.


         {RTS2:113} It follows from this that the operations
of the higher level cannot be accounted for solely by the
laws governing the lower-order level in which we might say
the higher-order level is `rooted' and from which we might
say it was `emergent'.  Now an historical explanation of how
a new level came to be formed would not, it is important to
see, undermine this principle.  Let us suppose that we could
explain the emergence of organic life in terms of the
physical and chemical elements out of which organic things
were formed and perhaps even reproduce this process in the
laboratory.  Now would biologists lose their object of
inquiry?  Would living things cease to be real?  Our
apprehension of them unmasked as an illusion?  No, for in as
much as living things were capable of acting back on the
materials out of which they were formed, biology would not
be otiose.  For a knowledge of biological structures and
principles would still be necessary to account for any
determinate state of the physical world.  Whatever is
capable of producing a physical effect is real and a proper
object of scientific study.  It would be the task of
biologists to investigate the causal powers of living things
in virtue of the exercise of which inter alia they brought
about various determinate states of the physical world.
Living creatures qua causal agents determine the conditions
under which physical laws apply; they cannot therefore
already be manifest in the latter.  Sentience determines the
conditions of applicability of physical laws, but it is also
subject to them.  If the elements of the lower-order are
real then so must be the causes that determine the
conditions of their operation, i.e. the comprehensive
entities formed out of them.  If black bodies are real then
so are physicists, if charged particles are real then so are
thunderstorms.  In short, emergence is an irreducible
feature of our world, i.e. it has an irreducibly ontological
character.

         Reflect once more on the distinctiveness of laws of
nature and empirical generalizations.  The laws of nature
leave the conditions under which they operate open, so the
field of phenomena is not closed: it is subject to the
possibility of dual and multiple control, including control
by human agents.  What I can do is constrained by the
operation of natural laws.  But I can hack my way all over
the physical world, defeating empirical generalizations.  I
can interrupt the operations or break the mechanism of a
machine and so falsify any prediction made on the basis of
{RTS2:114} its past behaviour.  But I cannot change the laws
that governed and so explained its mode of operation.  And I
can come, in science, to have a knowledge of such normic and
non-empirical statements; and perhaps in time begin to
recognize analogous principles at work controlling my own
behaviour (marking the site of a possible psychology).

         I have argued that complex objects are real (and
that the complexity of objects is real); and that the
concept of their agency is irreducible.  Complex objects are
real because they are causal agents capable of acting back
on the materials out of which they are formed.  Thus the
behaviour of e.g. animate things is not determined by
physical laws alone.  But that does not mean that their
behaviour is not completely determined: only that an area of
autonomy is marked out which is the site of a putatively
independent science.  And because the forms of determination
need not fall under the classical paradigm this in turn
situates the possibility of various kinds of
self-determination (including the possibility that the
behaviour of men may be governed by rational principles of
action).

         From the normic and non-empirical nature of laws
and their consistency with situations of dual control I
conclude that the world is a world of agents incompletely
described.  Laws neither undifferentially describe nor
uniquely govern the phenomena of our world.  And this is
accounted for by the fact that it is an incompletely
described world of agents which are constituted at different
levels of complexity and organization.^48  However it might
be objected here that all I have shown is that the laws that
we currently possess do not describe the world as we
currently know it.  And that I have not shown that if we
were in fact able to reduce (apparently) complex things to
complete atomistic state-descriptions that we would be
unable to predict future physical states of the world
without referring to comprehensive entities and principles
of behaviour special to them.  The final stage of my
argument against actualism must thus constitute a critique
of strong actualism in which the incoherence of the
programme of reduction it envisages for science is
demonstrated.

         It is important to be clear about the different
senses of `reduction'.  There are three distinct ways in
which a science {RTS2:115} might be said to be `reducible'
to a more basic one, which ought not to be confused.  There
is first the idea of some lower-order or microscopic domain
providing a basis for the existence of some higher-order
property or power; as for example, the neurophysiological
organization of human beings may be said to provide a basis
for their power of speech.  There is secondly the idea that
one might be able to explain the principles of the
higher-order science in terms of those of the lower-order
one.  This depends upon being able to undertake at least a
partial translation of the terms of the two domains; though
it is conceivable that they may retain substantially
independent meanings and overlap only in some of their
reference states.  Such a `reduction' may of course result
in modifications of the laws of the higher-order domain.^49
There is finally the sense in which it is suggested that
from a knowledge of the states and principles of the
lower-order science we might be able to predict behaviour in
the higher-order domain.  It is important to see that it is
to this claim that the strong actualist is committed, if he
is to eliminate complex behaviour in favour of its atomistic
surrogates.  It depends not only upon the establishment of a
complete parallelism between the two domains, but upon a
closure, i.e. the attainment of a complete atomistic
state-description of all the systems within which the events
covered by the descriptions of the higher-order science
occur.

         Now it is especially important to keep the second
and third senses distinct.  For though it is clear that we
can explain the principles and laws of chemistry in terms of
those of physics or of classical mechanics in terms of
quantum mechanics, we cannot predict physical and chemical
events such as the next eruption of Vesuvius on the basis of
that knowledge alone.  For that we would need an antecedent
complete atomistic state-description, i.e. a closure, as
well.  Now the strong actualist, claiming that the world is
in the end closed, must, unless he is to limit himself
merely to a dogmatic reassertion of this claim, presumably
map out a strategy for the sciences to attain such a
closure.  The fact that a successful reduction in science
does nothing in itself to achieve empirical invariances is
something {RTS2:116} of a blow to the programme (as distinct
from dogma) of strong actualism.  But even if it did there
is an even more damaging objection at hand (which carries a
more general moral for all those who see in `reduction' the
hope of the `less developed' sciences).  For every
historically successful reduction of one science to another
has depended upon the prior existence of an established
corpus of scientific principles and laws in the domain of
the reduced science.  It is easy to appreciate why this must
be so: for without the specification of some already more or
less clearly demarcated and well charted domain no programme
of reduction could possibly get to work.  But this means
that as a means of discovery, i.e. of achieving such a body
of knowledge reductionism must fail.  For it presupposes
precisely what is to be discovered.

         I still have not refuted strong actualism as a
possible account of the world.  This I shall now do by
arguing that it is inconsistent with any world containing
science, and thus in any world in which science is possible.
The only way of reconciling experimental activity with the
empiricist notion of law is to regard it as an illusion;
that is, to regard the actions performed in it as subsumable
in principle under a complete atomistic state-description.
In principle this applies not only to experimental activity
but to all scientific activity (including
theory-construction) in as much as it involves physical
effects.  Now this has the absurd consequence that the
apparent discovery of natural laws depends upon the prior
reduction of social to natural science.  Or to put it
another way, in an actualist world there would be no way of
discovering laws which did not already presuppose a
knowledge of them.  So a closed world entails either a
completed or no science.  But as `completion' is a process
in time the former possibility is ruled out: so a closed
world entails the impossibility of science.  But as science
occurs the world must be open.  This is not the reason why
the world is open (though it is the reason for my justified
belief that it is).  Rather it is because the world is open
that science, whether or not (and for how long) it actually
occurs, is possible.  In an open world all laws must be of
normic form; and this is quite independent of our knowledge
of them.  In short, the complexity of agents and the normic
character of laws are irreducible ontological features of
the world; that is, they are necessary {RTS2:117} features
of our world established as such by philosophical

         It is relatively easy to show that all (and not
just scientific) action depends upon our capacity to
identify causes in open systems.  For all action depends
upon our capacity to bring about changes in our physical
environment.  Hence we must belong to the same system of
objects (nature) on which we act.  But we not only act on
it, in the sense of bringing about changes that would not
otherwise have occurred; we act on it purposefully and
intentionally, i.e. so as to bring about these changes (as
the results and consequences of our actions) and knowing
that we are acting in that way.  This depends upon our being
able to identify features of our environment as the objects
of our causal attention and as part of the system to which
causality applies.  Thus we must be capable of identifying
and ascribing causes in our environment, and knowing
ourselves as a causal agent among others.  Unless we could
do this, we could not act intentionally at all.  Thus all
human action depends upon our capacity to identify causes in
open systems (to which of course the Humean theory cannot
apply).

         I suggested earlier that human freedom is not only
compatible with science, but had to be scientifically
understood.  This is important because it is inter alia a
precondition for science.  For science to be possible men
must be free in the specific sense of being able to act
according to a plan e.g. in the experimental testing of a
scientific hypothesis.  Human freedom is not something that
stands opposed to or apart from science; but rather
something that is presupposed by it.  The idea that freedom
is opposed to or apart from science stems from the
empiricist conception of scientific experience as consisting
in the passive observation of repeated sequences rather than
in the active intervention of men in the world of things in
an endeavour to grasp the principles of their behaviour.
Men are not passive spectators of a given world, but active
agents in a complex one.

         The view of the world as open and the view of the
world as closed lead to totally different conceptions of
science.  The laws of nature, which are painstakingly
uncovered by the theoretical work of science supplemented
wherever possible by experimental investigation, do not seek
to describe the myriad phenomena {RTS2:118} of the world,
the contents of a biscuit tin or the junk in the builder's
yard.  They do not seek to trace the path of a squirrel,
predict which rafter a sparrow will light on or how many
buns the vicar will have for tea.^50  They can indeed come to
explain such things in a certain way, but only on the
condition that they are not interpreted as describing them.



 ^46 G. E. M. Anscombe, op. cit., p. 21.

 ^47 Cf. M. Polanyi, `The Structure of Consciousness', The
Anatomy of Knowledge, ed. M. Grene, p. 321.

 ^48 Cf. M. Bunge, The Myth of Simplicity, Chap. 3; and
M. Polanyi, The Tacit Dimension, Chap. 2.

 ^49 Cf. P. K. Feyerabend, `Explanation. Reduction and
Empiricism', Minnesota Studies in the Philosophy of
Science Vol. III, eds. H. Feigl and G. Maxwell,
pp. 28-95.

 ^50 A caricature of such an empiricism exists in some of
the early experiments conducted under the august
auspices of the Royal Society.  The following is an
example: '1661, July 24: a circle was made with a powder
of unicorn's horn, and a spider set in the middle of it,
but it immediately ran out several times repeated.  The
spider once made some stay upon the powder', C. R. Weld,
History of the Royal Society, Vol. I, p. 113.  Among the
items of allegedly scientific interest collected by the
Society were 'the skin of a moor, tanned with the beard
and hair white' and 'an herb which grew in the stomach
of a thrush', ibid, p. 219.  Quoted in P. K. Feyerabend,
'Problems of Empiricism', op. cit., p. 156.


     --- from list seminar-14-AT-lists.village.virginia.edu ---