The Replaceability Argument - Killing Happy Animals: Explorations in Utilitarian Ethics - page 56

most of which would probably have been fulfilled, resulting in positive
welfare. Provided that the new person has at least as much welfare as the
killed one would have had, the replacement seems to compensate for
the welfare loss and hence make the killing morally acceptable.
In order to preclude this possibility, Singer declares that every unsatis-
fied desire counts negative on the welfare scale; every unsatisfied pref-
erence creates negative welfare. As soon as the desire is satisfied, the
negative effect on welfare is cancelled out and becomes neutral. Singer
suggests that we should 'think of the creation of an unsatisfied prefer-
ence as putting a debit in a kind of moral ledger that is merely cancelled
out by the satisfaction of the preference'. 17 A 'ledger' is a topic (or what
nowadays might be a computer programme) in which for instance a
shopkeeper records debits until the debts are paid back. The shopkeeper
writes down all the debts, and once one is paid back, the corresponding
debit entry in the topic is cancelled out. With this 'moral ledger model'
in place, replaceability seems to be ruled out.
How can the moral ledger model rule out replaceability? A simple
calculation can show how this works: Killing an adult human being
normally means that some of her preferences - and usually very strong
preferences, such as wanting to go on living or finishing some projects -
are not fulfilled. Thus, killing a human adult being has the consequence
that certain important preferences cannot be satisfied. These negative
entries cannot be cancelled out in this being's life any more. The welfare
score of the being remains negative. It would be of no use to replace
this killed being by a new one. After all, the best that the new being can
achieve is to satisfy all his or her preferences and end up with a neutral
welfare level. The new being's welfare level can never be positive. This is
because of the moral ledger model, according to which the fulfilment of
desires does only cancel out negative welfare and never produces posi-
tive welfare. Therefore, a compensation of the negative effect on welfare
caused by the killing of the human adult is not possible. So, in order to
limit the scope of the Replaceability Argument, Singer recurs to a desire-
satisfaction account of welfare, in combination with the moral ledger
model. In the remainder of this discussion, I will refer to this combina-
tion as 'Singer's account of welfare'.
Who can be excluded from replaceability in that way? For Singer's
account of welfare, it is crucial that unsatisfied preferences count as nega-
tive welfare. The more unsatisfied preferences a being has at the moment
of its death, the worse off it is in terms of welfare. Now, which beings are
likely to leave the most unsatisfied preferences when they die? It seems
obvious: the more future-oriented desires a being has, the more likely