The comments presented here were made after the presentation of four papers and commentary by two other symposium participants. They address issues in language comparison and classification. First, comments are made on the papers ("An African Test Case in Comparative Methodology,""The Mathematics of Multilateral Comparison,""Testing a Basic Evaluation Metric," and "Multilateral Comparison and Linguistic Geography"), then more general comments are made on the utility of statistical tests, and of mathematical or probabilistic techniques in general, in linguistics. Issues discussed include sampling, true independence of variables, validity of underlying assumptions, inter-group differences, and error types associated with null hypotheses. Following this, factors supporting use of statistical methods are reviewed. A number of quotations and pieces of information seen as useful in a discussion of these issues is appended to the text. Contains six references. (MSE)