Suppose we belong to a small self-governing club or society.  Suppose our
society conducts its business affairs by meeting once a month for a few
hours, discussing business, coming to a consensus (or, in rare instances,
taking a vote).  Let's describe this procedure in Rousseauean terms:
self-governance by the will of the whole.  Let's suppose further that
people come to these meeting with an open mind, ready to be persuaded by
the discussion.  Indeed often individual members of the society will enter
the discussion without any fixed idea of what the outcome ought to be:
they only really come to form opinions as the discussion unfolds.  In
short, this is a healthy, self-governed society, free of domineering
personalities and divisive factions.

What we have in this case, then, are
(a) certain deliberative procedures and
(b) collective intentions (expressions of the collective will)

Suppose now that someone in the group begins to notice certain regularities
about the outcomes of the this procedure:  On questions about admitting new
members to the club, the collective intentions generally coincide with
John's first thoughts on the matter. On questions regarding expenditures,
the collective intention generally falls about halfway between Mary's
initial judgment and Frank's.

Suppose now that having noticed these rough correlations, some of us in the
society start to investigate them more systematically -- at first simply
out of curiosity.  As the project gets more advanced, we develop an
increasingly fine-grained and precise set of regularities, until finally we
reach a point where our regularities perfectly match the outcomes of our
meetings.

Finally, suppose that someone now proposes that we give up having our 
monthly meetings.  After all, seems we can now know in advance what the 
outcomes of the discussion will be.  Using the set of regularities 
collected, we can now simply predict the outcome of the discussion.  
Instead of wasting our time in monthly meetings we can all go surfing 
instead and use the principles to determine our collective intentions.  
When members of other clubs ask why we gave up self-governance, we reply 
that we have not: we still govern ourselves by our collective intentions; 
we simply have a new way of finding out what those intentions are.

The questions:  Is this coherent?  Does it make sense to suppose that we 
might discover a different method for determining our collective 
intentions?  Could we really separate (a) from (b)?

Observations:

The last step of this thought experiment is of course the crucial one.  
The questions to consider is this: what assumption are we making about 
the observed correlations in deciding to use them in lieu of our monthly 
meetings?  I suggest that we are making two crucial assumptions:

1) We assume that (a) and (b) above are independent.  That is, we assume 
that the collective intentions of the group have a determinate content 
apart from the deliberate procedures used in the meeting.  It is only 
under this assumption that it makes sense to suppose that we might use 
other methods to discover our collective intentions.

2) We are treating the observed regularities as lawlike.  That is, in 
supposing that we can use the regularities to predict the outcome of our 
deliberations, we are supposing that theyhold with necessity -- that they 
are not simply coincidental correlations that might break down at the 
next instance, but that they are genuine laws that determine what the 
collective intentions are.

One important question to ask about the thought experiment is whether the 
members of the club are right in claiming that they are still 
self-governing -- whether they are right in thinking that they are still 
regulating their business by the collective intentions of the whole.

The answer, I submit, is that they are not.  Even if we suppose that the 
regularities continue to track what would have been the outcomes 
of the meetings, we can no longer treat the policies as collective 
intentions.  Why?  Because the first assumption above is false.  (a) and 
(b) are not independent here.  Rather, for a policy to be a collective 
intention is for it to be the outcome of deliberate procedures of 
a certain sort.

We might also make this point in terms of the second assumption above.  
To treat our observed regularitis as laws -- as principles which support 
predictions -- would be, in Davidson's phrase, to change the subject.  
For it would be to treat the deliberative procedures of the group as 
pre-determined in their outcome, as close-ended (rather than open-ended) 
discussions.  To view the regularities in this way is in fact to give up 
the ideal of self-governance by collective intention.  As soon as we 
treat the regularities as lawlike, we no longer have collective intentions.

The lesson of the thought experiment, I submit, is that we cannot 
coherently treat the regularities as laws governing our collective 
intentions.  We cannot treat the regularities as laws because to think of 
our collective intentions as party to such laws is no longer to treat 
then as collective intentions.  For to be the expression of a collective 
intention, a policy must be the outcome of a deliberate procedure that we 
treat as open-ended.

Can a similar claim be made about psychophysical laws?  The thesis would 
be: we cannot treat psychophysical regularities as psychological laws, 
since to do so would be to give up the idea that we have intentional 
states.



Return to Philosophy of Mind homepage.