Demonet JF, Fiez JA, Paulesu E, Petersen SE, Zatorre RJ
INSERM U 455, Hopital Purpan, Toulouse, France
Brain Lang. 1996 Dec;55(3):352-79. doi: 10.1006/brln.1996.0109.
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.
波佩尔（1996）对来自六个不同神经成像实验室的八项语音加工研究的方法和报告结果提出了一些批评。我们坦率地承认，对正电子发射断层扫描（PET）方法可以提出合理的批评，而且和任何方法一样，它有局限性；事实上，我们和其他人都参与过这样的批判性评论（施泰因梅茨和塞茨，1991；塞尔让等人，1992；德莫内，1995；菲耶兹等人，1996a；扎托雷等人，1996）。然而，波佩尔的分析远远没有为PET方法的局限性或未来功能成像研究可以改进的方式提供新的见解。波佩尔的许多批评源于他未能理解一些推动功能成像研究的基本问题，包括他所评论的那些问题。不过，我们感谢这位批评者，因为他向我们提出了挑战，要求我们阐明在实验设计、分析和解释的重要方面的立场。在讨论这些问题时，我们首先进行一般性评论，然后是各位作者的具体评论。
