Tuesday, 15 January 2008
Passive vs. Driven Evolutionary Trends
Derek Turner,Connecticut College
12:05 pm, 817R Cathedral of Learning

Abstract: When Dan McShea (1994) introduced the distinction between passive and driven large-scale evolutionary trends, he treated this as a distinction between two different mechanisms for producing trends. He also introduced some empirical tests that scientists could use to determine whether a given trend is passive or driven. And indeed, other paleobiologists have had some success implementing McShea’s methods. McShea’s way of talking about the passive/driven distinction has caught on, and this has led some paleobiologists to describe their own work in the following way: when they conduct large statistical studies of the fossil record to test whether a trend is passive or driven, they are testing different hypotheses about the underlying mechanism that produced the trend, or different hypotheses about the cause of the trend.

Although I agree that McShea’s passive/driven distinction is extremely valuable, I argue that his use of terms such as ‘mechanism’ and ‘dynamic’ when talking about these two kinds of trends has led to some confusion. The distinction between passive and driven trends is a statistical distinction, not a causal one, as McShea’s terminology might lead one to think. If I am right about this, it follows that when paleobiologists use empirical tests to determine whether a trend is passive or driven, they are not testing causal hypotheses, in anything like the traditional sense in which historical scientists test causal hypotheses (see, e.g., Cleland 2002).

This naturally leads one to wonder what paleobiologists might be doing (if not testing causal hypotheses) when they conduct statistical tests to determine whether a trend is passive or driven. Here I offer the following constructive suggestion: the distinction between passive and driven trends is a purely statistical one, but these two kinds of trends call out for different kinds of causal explanation. Thus, when paleobiologists use empirical tests to determine whether a trend is passive or driven, what they are really doing is offering a more precise characterization of the phenomenon to be explained. A great deal of interesting research in paleobiology is often wrongly characterized as hypothesis testing, when in fact scientists are offering compressed descriptions of the fossil record.