The problems with using Piagetian theory to explore language-thought relationships are two-fold. First there are methodological problems, including the lack of experimental controls and the lack of uniform criteria by which cognitive operations are identified. A second difficulty is the questionable practice of interpreting child language development in terms of adult language use. Because of these difficulties, researchers hypothesizing a language-thought relationship within Piagetian parameters have three options. The first option is to assume that cognitive operations always precede linguistic structure. This option implies that language is not a good index of cognitive development, that nonlinguistic assessment techniques must be developed, and that language is not a form of thinking in its own right. The second option is to assume synchrony, or an essentially isomorphic relationship between language and thought. This assumption implies that language can be used to determine cognitive operations, that child language is distinct from adult language, and that language processes should not be distinguished from other forms of thought. The third position is that individual cases dictate whether linguistic structures give rise to cognitive operations, whether the reverse is true, or whether there is synchrony. The wealth of evidence supporting this variability in language-thought relationships indicates that better criteria are needed for delineating cognitive operations with respect to language structures. (RL)