Conclusions - Killing Happy Animals: Explorations in Utilitarian Ethics - page 140

desire is fulfilled. Under these assumptions, killing a person would
usually leave many strongly held desires unsatisfied, such as the desire
to continue life and other future-directed projects. Those unsatisfied
desires would leave negative welfare, which, according to the moral
ledger model, cannot be compensated by bringing another person into
existence. This is because the welfare of the other person could never be
better than neutral, and it can be neutral only in the exceptional case in
which all his or her desires would be fulfilled.
Thus, in order to exclude persons from replaceability, Singer proposes
the desire-satisfaction account of welfare together with the moral ledger
model. As we have seen, under these assumptions welfare can never be
positive. It can only be negative or, in the unlikely case in which all one's
desires are fulfilled, neutral. This account of welfare, as I have argued,
seems overly pessimistic. Unlike other accounts of welfare, this particular
account of welfare cannot capture the experience that welfare can also be
positive and that, for instance, the satisfaction of at least some desires does
more than only wiping out a state of deprivation. Furthermore, the partic-
ular account of welfare that Singer proposes in order to exclude persons
from replaceability is ad-hoc in several respects. For instance, it does not
fit with the rest of Singer's work, where he seems to assume that welfare
can either be positive, negative or neutral. Furthermore, it would seem that
every being with desires would be irreplaceable, because it would be likely
to leave unsatisfied desires when being killed. It seems arbitrary to claim
that persons are irreplaceable, while those who have less future-directed
desires are replaceable, to a certain degree.
An implication of the account of welfare that Singer proposes in order to
exclude persons from replaceability is that most of us would be better off
having not lived. After all, we are all likely to leave at least some unsatisfied
desires when we die. Thus, on that version of utilitarianism, bringing any
person into the world seems to be morally forbidden, because remaining
childless would better serve the maximisation of welfare by avoiding the
likely negative welfare scores due to unsatisfied desires. According to that
account of welfare, it would have been better if all of us, both human
and non-human persons, would never exist. Furthermore, as soon as any
being has all of his or her preferences fulfilled, he or she would be replace-
able by another being that would be equally 'lucky'.
Singer has recently tried to avoid this implication by accepting
preference-independent values. If those values are still about welfare,
this move does not help. Either welfare can be positive and hence
replaceability is possible, or welfare can only be negative and neutral
and hence the implication is that those of us who are not replaceable