The convergence hypothesis asserts that, if one takes the full range o
f human values-present and future-into account, one will choose a set
of policies that can also be accepted by an advocate of a consistent a
nd reasonable nonanthropocentrism. Brian Steverson has attacked this h
ypothesis from a surprising direction. He attributes to deep ecologist
s the position that nonhuman nature has intrinsic value, interprets th
is position to mean that no species could ever be allowed to go extinc
t, and proceeds to show that my commitment to contextualism prohibits
me from advocating the protection of species universally. In response,
I show, by reference to recent scientific findings, how difficult it
is to defend species preservation in all situations. In particular, I
argue that Steverson's appeal to a possible world in which we have nea
rly complete biological knowledge misses the point of the convergence
hypothesis. It is an empirical hypothesis, with significant indirect,
and some direct, evidence to support it. Although it is a falsifiable
hypothesis about real-world policies, it cannot be falsified by a cont
rary-to-fact case.