Abstract

In this essay, I consider three philosophical issues that arise in the environmental sciences. First, these sciences depend on mathematical models and simulations which are highly idealized and are coupled with very uncertain data. Why should we trust these models and simulations? Second, in standard hypothesis testing, the burden of proof is in favor of the null hypothesis which claims some causal factor has no effect. The alternative hypothesis is accepted only when the likelihood of the null hypothesis is very low. Recently, some have argued that we should minimize Type II errors (not rejecting a false null) rather than Type I errors (rejecting a true null) given the environmental risks involved. I consider arguments for shifting this burden of proof when possible environmental harms are significant. Finally, in debates over global climate change, much is made of the apparent consensus concerning the effects of human induced greenhouse gas emissions on average surface temperatures. However, scientific methods are structured around dissent and criticism. Is consensus-based science orthogonal are even harmful to science?