Chicken soup.

[From Bill Powers (940725.0400 MDT)]

Mary found a 25-cent book at the Humane Society Thrift Shop. It's Moya,
Carlos .J. (1990), _The philosophy of action_ (Cambridge, UK: Polity
Press/Blackwell). It's interesting because it's fairly recent, it
reviews a number of people's ideas in the field of explaining action and
intentional behavior, and it shows how utterly pointless it is to reason
about phenomena of nature in words without any scientific model behind
them. This book is like a window into the distant past of human
thinking, its advantage for us being that it is presented in semi-
popular terms and in modern English instead of the Medieval Latin of the

     The starting point of this inquiry will be action itself. In our
     everyday life we tend to contrast what we do with what merely
     happens... Try to think of the way things would appear to us if we
     did not distinguish between agents and mere objects, nor between
     actions and happenings ... The idea that we are agents, and
     therefore that there are agents, is not a mere opinion we can
     embrace or reject. It is a basic conviction that permeates our
     life, giving support, and being supported by, other equally basic
     convictions. (p.2)

Notice the appeal to direct experience, and to words whose meanings are
taken to be self-evident: "action" and "doing" versus "happening,"
"agent" versus "mere object." The writer appeals to our intuitive sense
of what these words mean, as if some words came equipped with meanings
that are so fundamental that one needs only to mention the word to
convey the meaning. What seems completely missing here is the idea that
words indicate phenomena, and that to say what a word means one must
point to that which the words purport to indicate: what we call data.

     But is there any action? This question may sound bizarre, for what
     could be more evident than that? Philosophy, however, cannot allow
     itself to be satisfied with that level of evidence. ... Philosophy
     begins when we stop taking those everyday assumptions at face

     Whether there are or there are not actions is not something that
     can be answered by direct observation. Those who doubt the
     existence of action are not questioning what everyone can perceive.
     They are rather wondering whether the concepts we ordinarily use to
     describe and interpret these observations are appropriate and
     ultimately consistent. If they are not, then this leads to a
     negative answer to our question; if action is an inconsistent
     concept, there cannot be actions, just as there cannot be squared
     circles. So, analysis of the concept of action itself is the main
     topic of the philosophy of action. (p.2)

Now we come to the crux of the matter: the clash between science and the
intuitive world implicit in language and experience.

     Scientific perspectives are happy with happenings, being explained
     by other previous or contemporaneous happenings with the aid of
     laws. But if we think of ourselves as agents, we conceive ourselves
     as being able to initiate changes independently of the world's
     previous history. Agents and actions, then, are likely to face
     difficulties if they look for a place on a scientific stage.
     To give the reader an idea of what a reductionist attitude is like,
     let us start with an episode that nobody would hesitate in
     classifying as an action, say, drinking a glass of water. What
     right do we have to call this an action, and not a mere happening?
     Where does the actional character of this episode lie? What did _I_
     do? The water got into my mouth as an effect of gravity. The water

     getting into my mouth is a mere happening. This happening, in turn,
     was caused by the movement of the glass. Where is action in this?
     Well, one could say, I caused the movement, so I acted. But think
     that this movement can be said to be properly caused by my arm's
     and hand's movement, which, in turn, were caused by some muscles'
     contractions, which in turn were caused by some neurons' firings,
     and so on. Actions as such seem to dissolve and be reduced to a
     sequence of happenings. ... Appeals to desires will not do, for our
     desire for water is presumably a state caused by organic
     deprivation. The chain of causes extends further and further into
     the past and there appears to be nothing we, as agents, initiate,
     only further happenings. Actions, then, seem to be nothing but
     specific sequences of happenings.

We, as PCTers, can see where this reductionist argument slips past the
key idea without realizing its importance: "our desire for water is
presumably a state caused by organic deprivation." The question, as we
know, is _how_ that state of desire is "caused" by the mere fact of
organic deprivation. Body fluids, or electrolyte concentrations, are in
some particular state: what is there that proclaims this to be a state
of "deprivation?" Somewhere, obviously, there is something that defines
a state of zero deprivation -- a reference signal. The fluid balance of
the body does not by itself give rise to behavior that corrects it. It
is the _difference_ between the actual fluid balance and some
specification for the _right_ fluid balance that leads to correction.
And whatever it is that creates this specification for the right state
is an intention. It walks like an intention, it quacks like an
intention, so why not call it an intention? Or, more precisely, a
reference level.

So, the PCTer may think, we can now settle back and watch this
philospher demolish the reductionist argument by developing a correct
model for "desire" and its relation to reference signals. For the
reductionist argument is a simple cause-effect argument that takes
reference levels for granted without even knowing it. But that is not
how philosophy works.

Let's skip ahead past other points of view the author wishes to dismiss,
and get right into the basic problems that the philosphers see:

     Compare these two cases:
      (1) A person takes a gun, aims carefully at the target, pulls the
          trigger, and the gun fires.
      (2) the person goes home and puts the gun on the table. While he
          is putting it down, the gun, unexpectedly, fires.

     Now, starting from our intuitive distinction between actions and
     happenings, we would agree, I hope, that in the first example
     shooting the gun was an action, while in the second the gun's
     firing was a happening.
     In fact, 'someone fired the gun' logically implies 'the gun fired.'
     The happening, then, is a necessary, but not a sufficient,
     condition of the action.
     Something must be added to the gun's firing for for ['someone fired
     the gun'] to be true. The missing element is what would bring
     agency into the picture.

The author then sums up a series of arguments that result in the
discovery of an infinite regress: If someone fired the gun, this someone
must have produced an action that resulted in the firing of the gun
[diversion here into the difference between a 'result' and a
'consequence"]. But if the person's action caused that result, it must

have done so by causing a happening, such as the pulling of a trigger,
which is what actually caused the firing of the gun. And so forth: one
ends up concluding that action (by an agent) can't exist because it
entails an infinite regression of happenings.

This problem is eventually solved as follows:

     A way out could be found if we were able to identify some action by
     means of which we could bring about the movement of our finger [on
     the trigger] but such that, in turn, does not essentially involve a
     happening (as moving the finger does) with respect to which it
     would make sense to ask how it was brought about. Here is where
     volitional theories enter into the picture. This is the answer that
     could be offered: when moving our finger is an action, we bring
     about the movement of the finger by _willing_ it to move or
     _trying_ to move it. This is _how_ we cause the movement of our
     finger, and thus what accounts for the difference between [I fired
     the gun, and the gun fired]. [When a person fires the gun], that
     person willed his finger to move or tried to move it.
     Willing and trying would be the proper basic actions, the only ones
     we do not perform by doing anything else. (p.17)

Remember that events which cause other events through ordinary physical
or neurological laws are called happenings; only what an agent does is
called an action. Moya has now established that what an agent does is
really willing or trying. What Moya has now established is that
"Conations (volitions, willings, tryings) _are_ actions, not

Naturally, we now expect the author to tell us what volitions, willings,
or tryings are. And he tries to do so.

     First of all, volitions are thoughts, are mental processes. As
     such, they are intentional: they have a content or intentional
     object which, according to McCann, is propositional in character,
     namely, that a result of a specific kind occurs. For example, the
     volition that causes the result of my action of raising the arm has
     as its content that the arm's rising occurs. Being propositions,
     contents are not themselves results, for results are changes or
     happenings. Since I can will my arm to rise without my arm rising
     (due, for example, to a paralysis), my arm's rising is not the
     result of my volition, but only a causal consequence of it. Being
     thoughts, volitions have no results, but only causal consequences,
     so that the result-problem does not arise in them and the threat of
     infinite regress can be prevented. Secondly, volitions, besides
     being thoughts, are executive, actional in character, they are
     executive acts with respect to intentions and desires. The content
     of intentions and desires are descriptions of actions; I have the
     intention of raising my arm. Volitions execute intentions by
     bringing about the result of the intended actions, namely the arm's
     rising in the above example. Volitions are conations of initiations
     of those movements that, by being so caused, qualify as the results
     of actions. They are, then, the causally more basic actions.
     Without them, there could not be actions at all, since they
     transmit agency to other actions by causing their results. Without
     them, nothing could qualify as the result of an action, and there
     would be only happenings and no actions.(p. 20)

Just go back over this paragraph and substitute for "my arm's rising"
etc. the phrase "my perception of my arm's rising." We are within a
hair's breadth of PCT (well, perhaps a cable's breadth). Note how a
"volition" performs the function of a control system, while an intention
or a desire performs the function of a reference signal. The intention
or desire merely specifies that which is to occur, as if in a

proposition: "My arm is rising." The volition turns this specification
into the actual occurrance. In PCT, of course, reference signals are not
cast as propositions unless they are, literally, propositions: reference
signals in the form of statements made in words or symbols. Elsewhere
they are simply like recordings of specific states a perceptual signal
might attain, and don't involve any words. And control systems do not
turn intentions or desires into the actual occurrances, but into
perceptions of occurrances (or states).

It's not hard to believe that beneath all this verbiage the author and
his colleagues are being guided by some inarticulate sense of how
control works (the word control even comes up occasionally in connection
with volition). It's as though they can sense something of the closed-
loop relationships, but the language in which they are required to try
to express this sense trips them up and steers them off in irrelevant
directions, chasing after side-issues of no importance such as infinite

The last quotation above was from page 20 of this 170-page book. As far
as I can tell, from a rather hasty scan of the remainder, the other 150
pages get no closer to PCT than we have already seen. The same ideas we
have seen so far recur again and again in new combinations and twists,
with one logical puzzle after another being turned up and then turned
under. In the end, the only conclusion is still that volition is the
basic hallmark of agency, which is very nearly a tautology. What remains
missing from start to finish is any concept of _how_ a mental act like
volition might be connected to a physical act like raising an arm. The
only conclusion is that there _is_ a connection. What it is remains a
mystery. The author has stirred up the pot, and succulent-looking
tidbits have floated briefly to the surface, but when it's all over it's
still just chicken soup.


Best to all,

Bill P.

<[Bill Leach 940725.19:12 EST(EDT)]

[Bill Powers (940725.0400 MDT)]

Oh joy, oh joy! Your "mini-book review" brought me significant peace. I
now understand why I always felt like split-pea soup was clearer reading
than such books... I had not realized that I was right, reading a bowl
of split-pea soup likely WILL make more sense and I'm sure that Tom will
agree that in any event, the soup certainly would not be a damaging! :slight_smile: