An identical composition and recall task was used to compare writing and speaking in terms of their temporal organization. A common cutoff point of 0.1 sec was used for minimum duration of pauses. Speakers took only a fourth of the time taken by writers, but spoke more than half again as many syllables as writers wrote. Mean durations of pauses for writing and speaking were equivalent (110 and 0.97 sec, respectively), but the respective distributions of pauses differed dramatically: In writing, a far greater number of pauses per syllable led to shorter phrases (segmentation between and even within individual words), whereas speaking was characterized by fewer pauses per syllable and consequently longer phrases (segmentation between syntactic units). Pauses at syntactic positions (Le., after punctuation) were the least frequent pauses in writing, although they were the longest in mean duration.We need only consult a modern psycholinguistics textbook or journal to become aware of the shocking neglect of writing. Clark and Clark (1977) maintained that psycholinguistics is "the study of listening, speaking, and the acquisition of these skills by children" (p. vii; also Dell, 1986). Writing is speaking's neglected sibling; when allowed in the house at all, it is most often required to remain on the page and play the role of the written.Writing is, in fact, a skilled motor behavior that takes place in time and reveals human psychology to a remarkable degree, very much as does speaking. But writing has hardly been recognized at all in the research history of the psycholinguistic family (see Kowal and O'Connell, 1986).Among the reasons for this neglect are the disrepute into which graphological studies have fallen, the difficulties of instrumentation for observing writing in real time, and (to return to our metaphor) the attractiveness of the other sibling, which has been considerably enhanced by the recent interest of linguists. Small wonder, then, that discussions of writing are frequently limited to composition and relegated to educational research, or are summarized in comparisons to speaking that do not "surpass the truism that speech rate is indeed faster than writing rate" (Kowal & O'Connell, 1986, p. 129).Nevertheless, the relatively early study by van Bruggen (1946) remains important. Using a kymograph to measure the time course of writing to 0.1 sec, van Bruggen expressed in words per minute differences between rates of writing from memory and rates of composing.Because syllables per word can vary appreciably from one genre of writing to another, the words per minute index appears to have reflected artifactual results.Horowitz and Berkowitz (1964), using a within-subjects design, found that speech rate is more than seven times as fast as writing rate (speech, 2.39 syllables/sec; writing, 0.33 syllables/sec). Tannenbaum, Williams, and Wood (1967) compared speaking and typing and found the latter to be characterized by longer but fewer pauses than speaking. Tannenbaum et al. explained that "the writing situation...