- The Paleontological Society
It has been said that most scientists would rather use another scientist's toothbrush than his terminology. In this issue's Matters of the Record, John Alroy suggests certain conceptual and methodological changes for the study of mechanisms, or dynamics, of large-scale trends, and also a revision of current trend-mechanism terminology. On the conceptual and methodological matters, Alroy's suggestions are good ones (with one important exception), and here I applaud them and explain further why they ought to be taken seriously. Predictably, however, I do not endorse (some of) his terminological revisions.
The conceptual matters are more interesting, and I turn to them first. In what follows, I first give some background on trend mechanisms, formulating the issues in my own terms (of course), and then explain Alroy's position in those terms.
A large-scale trend is change in some summary statistic for a state variable (in the sense of McKinney 1990), such as an increase in mean size, in a clade, while mechanism refers to the pattern of change among the lineages within the clade that accounts for the trend. As Alroy explains, in recent years the focus has been on two categories of mechanism, what I have elsewhere called “passive” and “driven” (McShea 1994). A passive trend is one in which change among lineages passively follows the “structure” of the state space (Fisher 1986; McShea 1998). Thus, for example, in Stanley's now-classic argument, if selection or developmental constraint in a group imposes a lower limit on size, and if the group originates close to this lower limit, then mean size will increase as diversity increases and the clade diffuses away from the lower limit (Stanley 1973). The state space is “structured” in that parameter values vary across it in some organized way; in this case, the probability of size decrease …