PET Studies of Phonological Processing: A Critical Reply to Poeppel

Poeppel (1996) raises a number of criticisms about the methods and reported results for eight studies of phonological processing from six different neuroimaging laboratories. We would freely admit that valid criticisms of PET methodology can be made and that, like any method, it has limitations; in fact, we and others have engaged in such critical commentary (Steinmetz & Seitz, 1991; Sergent et al., 1992; Demonet, 1995; Fiez et al., 1996a; Zatorre et al., 1996). Poeppel's analysis, though, falls far short of providing new insights into the limitations of PET methodology or the means by which future functional imaging studies could be improved. Many of Poeppel's criticisms derive from a failure to understand some of the fundamental issues which motivate functional imaging studies, including those he reviews. However, we are grateful to our critic inasmuch as he offers us the challenge to clarify our positions on important aspects of our experimental design, analysis, and interpretation. In our discussion of these issues, we begin with a general commentary, followed by specific comments from individual authors