The Replaceability Argument - Killing Happy Animals: Explorations in Utilitarian Ethics - page 61

welfare as if it could be positive, negative and neutral in a way that seems
incompatible with the moral ledger model. For instance, Singer thinks
that it is a bad thing knowingly to bring a miserable being into the world,
and it is a good thing knowingly to bring a happy being into the world.
If the best one could hope for was reaching a zero level in case all one's
desires were satisfied, why would it be a good thing to bring additional
beings into the world? 24 It seems inconsistent with Singer's account of
welfare to claim that it could be a good thing to create happy children.
Furthermore, one of the conditions of the Replaceability Argument is that
the animals in question have pleasant lives and would have had a life
with positive welfare. However, the idea that anyone can score positive on
the welfare scale seems inconsistent with Singer's account of welfare.
It is a very uncommon assumption to hold that welfare can only be
negative, to different degrees, or neutral. One author who has (after
Singer) defended such a view is Christoph Fehige. Fehige's view is called
'anti-frustrationism' (also 'frustrationism'). Fehige claims that whether
one has a satisfied desire or no desire at all is equally good. Having an
unsatisfied desire, by contrast, is bad. 25 The implication of Fehige's
model, and a conclusion that Fehige accepts, is that it would be better
not to create people. Singer, however, refuses to draw that conclusion:
'It is absurd to hold, of even the most fulfilling lives that because some
desires could not be satisfied, it would have been better if they had never
been lived.' 26 Singer takes this implication, of which he apparently has
not been aware when proposing his model, as a reason for dismissing
the original moral ledger model. However, if Singer really believes that
having unsatisfied desires is prudentially bad, and having satisfied
desires is not better than neutral, the implication that life is not such a
good thing simply follows. Given Singer's account of welfare, the whole idea
of shifting the sufficiency line in order to avoid the implication that most of
us were better not born seems inconsistent. This is the third element that is
ad-hoc about Singer's account of welfare.
A further reason why I find it ad-hoc is related to Singer's adaptation of
a more gradual account of personhood. According to Singer, as person-
hood comes in degrees, so does (non-) replaceability. The idea would be
that those who leave a lot of unsatisfied desires are not replaceable, while
those who leave only a few unsatisfied (short-term) desires are replace-
able, to a certain degree. It seems to me though, that when the bringing
into existence of a new being cannot compensate for the welfare loss of
the killed being then the killed being is not replaceable. It is not replace-
able because the welfare loss cannot be replaced as proposed by the
Replaceability Argument. If things other than the welfare of the newly