Multiple realizations, diverse implementations and antireductionism

It has been debated what implications multiple realizability has for reductionism. I claim that more explicit attention needs to be paid to the distinction between multiple realizations of kinds and diverse implementations of laws. In this paper, I distinguish two different theses on the relations between multiple realization and diverse implementation: one thesis states that multiple realizations imply diverse implementations and the other states the converse. I claim that although antireductionism might turn out to be false if the first thesis is accepted, this "realization-based" antireductionism is not the only option for antireductionism. For the antireductionists who accept the second thesis, multiple realizations only provide evidence for diverse implementations. I defend this " implementation-based" antireductionism again Shapiro's dilemma. I argue that one horn of the dilemma does not pose any problem and that the other horn simply begs the question.