(Continued)
life seems to be preferable to the loss of many lives. Deontologists (rule-based theorists) such as
Rawls argue though that the person sacrificed does not get some overriding benefit, and to aggregate
good and bad interpersonally is to fail to take account of the "separateness of persons" (See John
Rawls, A Theory of Justice, 5) You might make a distinction here between act- and rule-utilitarianism.
J.J.C. Smart makes the distinction as follows: "Act-utilitarianism is the view that the rightness or the
wrongness of an action is to be judged by the consequences, good or bad, of the action itself.
Rule-utilitarianism is the view that the rightness or wrongness of an action is to be judged by the
goodness and badness of the consequences of a rule that everyone should perform the action in like
circumstances." (J.J.C. Smart, 'An Outline of a System of Utilitarian Ethics', in, J.J.C. Smart and
Bernard Williams, Utilitarianism For and Against.1963 [Cambridge University Press, 1993] p.9) If it is
accepted that this is a reasonable distinction, then we move from arguing about individual acts of
torture, to arguing about the general practice of torture. From the rule utilitarian perspective, it is far
easier to argue that torture as a practice should be outlawed, even though it may have good
consequences in certain cases. In effect then, the price of torture becoming an accepted and legally
ratified form of interrogation is so great that it is worth refusing to use it even in the few cases where it
may be justifiable. This is an especially compelling argument in political terms, as politics is to some
degree 'about' making rules, rather than adjudicating in specific circumstances.

3. Can utilitarians responsibly argue for a secret or deceptive course of action? That is: "publicly I
argue for program X, intending all the while to create program Y under the authorization for X so as to
minimize unintended effects."

This is one of the trickiest problems for utilitarianism as a political theory. The simple answer to the
first question is that it appears that they could argue for a deceptive course of action on the grounds
that it will produce better consequences overall. Lying is not the taboo for utilitarians that it is for
Kantians. Rawls argues that principles must be "publicly accepted and followed as the fundamental
charter of society "('the publicity condition') and therefore that elite cannot attempt to maximise
utilitarian consequences by promoting non-utilitarian principles. (See Rawls, A Theory of Justice,29)

On the other hand, perhaps it could be argued that the consequences overall (and we are talking in
the largest sense here) of having deceit at the heart of law-making could be catastrophic. Law itself
could be utterly undermined. The other tack the utilitarian can take here is to argue that there could
never be sufficient guarantees that the deceit would not be broken. Although the chances may be
slim, the consequences could be so awful that any possibility of them happening would be sufficient
to rule out the deceit as a viable option.

There are however more benign examples where the deceit is sanctioned, or required by utilitarians
— suppose a recession, or a run on the currency is forecast by a government. Surely it is
irresponsible to publicise this and thus intensify the problem? Once we admit exceptions though, the
idea of rule-utilitarianism appears to collapse, as R.M. Hare has argued, into act-utilitarianism. If one
case merits an act-utilitarian consideration to override rule-utilitarian considerations, why don't we just
go back to judging all acts individually? The rule is undermined anyway.

1.How much exploration of consequences can the utilitarian be required to make? Difficult to say. In
terms of assessing past actions, we can go as far as the chain of causality allows. In terms of
individuals when they make the decision, they can only act upon the information that they have. If we
could never act without full information regarding consequences, then we could never act. Perhaps
then, you can formulate the answer to this question as "as much exploration of reasonably likely
consequences as we reasonably can make." Unfortunately, this is a bit of a fudge. Obviously, my
lighting a cigarette now could lead, through a complex chain of causality, to the downfall of a
government, but it is extremely unlikely. The definition of 'reasonable' is always a problem for
philosophers.

5 & 6. See answer to question 2 above

A final thought — Some utilitarians (possibly including myself) would argue that genuine suffering is
worthy of far greater moral consideration than 'mere' happiness. Also, the needs of someone
suffering outweigh the needs of someone not suffering. The later might want our help, but he does
not need it in the same way as the sufferer. I would therefore argue against torture in the following
way. The consequences for a person being tortured are so bad that virtually no potential, uncertain,
future good for others could justify them. Torture to prevent further ills may be justifiable, but due to
the extreme horror of torture, there must be an extremely high degree of certainty that this will occur.
Furthermore, the onus would be on the torturer to demonstrate this. Upon reflection, it is clear that
this would certainly eliminate torture as a legally sanctioned practice, as the instances in which the
necessary level of consequential certainty existed would be extremely rare.

I should warn you though, that this form of 'negative utilitarianism' has met severe criticisms, and is
not widely accepted. However, the idea that the elimination of bad is more important than the
promotion of good has something going for it.

Perhaps your best route is, after all, to argue that the bad consequences of torture overall outweigh
the benefits? This is by no means a reliable argument though.

Incidentally, have you looked at, Elaine Scarry, The Body in Pain: The making and Unmaking of the
World(New York: Oxford University Press, 1985) This has an excellent account of the consequences
of torture.

Steve Bullock