What some historians have called scriptism in language study has been defined as the tendency to view spoken language through concepts of writing. The tradition of describing spoken languages using units and categories of an orthographic code underlies repeated criticisms of the centrism of linguistic-type analyses. In reviewing these criticisms, part of the problem appears to be the persistent idea of the primacy of linguistic concepts in guiding instrumental observations. However, one finds no historical justification for this idea that has prevented a reference to empirical observations in countering a writing bias. Others see that the problem rests with the doctrine that speech is separate from language and call for the abandonment of this division. This has basic epistemological implications. It leads to acknowledgment that observable structures of speech, not letters or words, are constitutive of spoken language. How one can study language by reference to observable structures of speech is essential in addressing the writing bias, and is the subject of the second half of this book.

Formal introductions to language study focus on transcribed speech, initially involving a familiar language where words and sentences seem self-evident. Presented this way, language seems to reflect an "autonomous" mental system separate from speech, which can be studied through writing signs. The arguments used to support this approach were originally formulated by Saussure and relate to nineteenth-century views of a localized language faculty in the brain. These influential arguments bolstered the concept of a speech–language division to the point that, at the turn of the twentieth century, many analysts saw instrumental observations of speech as irrelevant. In reviewing these arguments, evidence is discussed showing that the speech–language division is neither physiologically grounded nor methodologically useful in explaining the nature of features and structures of spoken language. One illustration is a study by Lindblom that shows that properties of the hearing system can shape vowel systems. Another example is given in a chapter bearing on how properties of motor speech can shape symbolic signs.