A variety of types of evidence are examined to help determine the true nature of "deep structure" and what, if any, implications this has for linguistic theory as well as culture theory generally. The evidence accumulated over the past century on the nature of phonetic and phonemic systems is briefly discussed, and the following areas of analysis are considered: (1) how phonemes are differentiated from each other by distinctive features and as fuzzy sets; (2) how morphemes come in two fundamentally different types, some of which form crisp sets and some of which are quite fuzzy; (3) how morphemes are combined into words which are not necessarily organized into classes in the same way that morphemes are; and (4) the fact that on the so-called syntactic level of analysis there are a variety of different types of units. Since it is proposed that the only thing that is transmitted from the speaker to listener is the vocal symbol, (i.e., the form), it is concluded that all linguistic analysis must proceed from the forms. It is suggested that at the point that phonemes are identified, morphemes are identified in terms of the sequence of phoneme combined with the statistical probabilities of occurrence. Then morphemes are identified on the basis of the phonemes and a Markov process (statistical probabilities) until a word is identified. Then the phrases are identified on the basis of the words and a statistical process and so on up the line. (SW)