Various measures have been proposed for characterizing the statistical association that arises between alleles at different loci. Hedrick has compared these measures with the standardized measure D' proposed by Lewontin on the grounds that this latter measure is independent of allele frequency. Although D' has the same range for all allelic frequencies, in fact, D' is not "independent" of allele frequency, and no measure with that general property is possible for the multilocus association problem. The insolubility of this problem arises from the ill-defined nature of general "association."