Conclusions - Killing Happy Animals: Explorations in Utilitarian Ethics - page 141

would be better off having not lived at all. When the desire-independent
values that Singer proposes to accept are also welfare-independent, the
proposed view would fall outside utilitarianism. It would no longer hold
that welfare is the sole ultimate value. Furthermore, this move would
not help. Again: either the value of a person's life can be positive and
hence replaceability is possible, or the value of a person's life cannot be
positive, and hence most of us are better off if we had not lived.
Even if adherents of the Total View accept Singer's model for restricting
the scope of the Replaceability Argument, there will always be beings, both
human and non-human, who fall short of the criterion of full person-
hood, and who would therefore be replaceable. For instance, babies and
newborn infants would be replaceable, according to Singer's model. This
is because they lack any strong future-directed desires. That means that
they could in principle be killed even if they would otherwise have had a
pleasant future (thus, we are not talking about euthanasia here).
Prior Existence Utilitarianism does not accept the Replaceability
Argument, which has been considered the most controversial part
of Singer's moral theory. It does not sanction the routine killing and
replacement of animals. Neither a human nor a non-human animal is
deemed replaceable. In that sense, Prior Existence Utilitarianism accords
all sentient beings a stronger protection. That would be good news for
all who have criticised utilitarianism for not taking animals - including
human beings - seriously enough.
Related to this implication of taking animals (both human and non-
human) more seriously, Prior Existence Utilitarianism, as opposed to
Total Utilitarianism, is a truly person-affecting moral theory. Instead of
being concerned with the abstract quantity of welfare, Prior Existence
Utilitarianism is concerned with effects on sentient beings. It is
concerned with harms and benefits for sentient beings. Prior Existence
Utilitarianism always requires producing maximal welfare (defined as
net benefits) for sentient beings, but - unlike Total Utilitarianism - it
never requires producing sentient beings in order to maximise welfare.
This means that Prior Existence Utilitarianism implies the Person-
Affecting Restriction on what matters in the evaluation of outcomes.
This is the view that outcomes should be evaluated in terms of harms and
benefits for sentient beings. For what concerns standard Non-Identity
cases, I proposed to apply this view in the wide sense. The Wide Person-
Affecting Restriction is concerned with harms and benefits for sentient
beings, not as particular individuals, but rather as sentient beings whoever
they are . This wide interpretation of the Person-Affecting Restriction fits