Communities developing resources and competencies for using their languages
Foundational understanding for language development work of all kinds
Publications, fonts and computer tools for language development, translation and research
SIL offers training in disciplines relevant to sustainable language development.
7,105 languages are spoken or signed. CLICK for map of world languages & regional websites.
SIL's dedication to language development past and present
This study argues for a theory in which metrical constituents are inherently headless and stress is autosegmental. Under this view, the distinction between tone languages and stress languages is characterized in terms of the respective domains in which autosegmental operations occur rather than by applying separate theories to tone and stress. This means that both types of prominence, although phonetically distinct, are derived via the same set of principles and devices.
Stress is shown to exhibit the following autosegmental properties: stability (Bedouin Hijazi Arabic), morphemic stress (Spanish, Turkish, Tagalog), and the ability to float (Mayo, Tagalog). After comparing these and other properties of stress with general autosegmental properties, it is concluded that stress is an autosegment.
Assuming that feet can be either disyllabic, bimoraic, or iambic, the above conclusion predicts the existence of five types of binary stressed feet. These are the left- and right-stressed syllabic foot, instantiated by Warao and Mayo, respectively; left- and right-stressed moraic foot, instantiated by Cairene Arabic and Turkish, respectively; and iambic foot, instantiated by Hixkaryana. The asymmetric nature of the iambic foot is attributed to the Weight-to-Stress Principle, which allows stress to be assigned directly to heavy syllables. It is shown, furthermore, that this principle predicts all and only the attested types of unbounded stress systems. Stressless feet and unfooted stresses are instantiated in Mayo and Tagalog.
The autosegmental theory of stress advances phonological theory in three ways. First, it eliminates most of the principles and devices which up to now  have been used only to describe stress, leaving only the abstract stress autosegment which is itself subject to the principles of autosegmental theory. Second, this approach makes it possible to attribute many of the apparent differences between stress and tone to differences in their respective domains rather than differences in their formal properties. Third, the autosegmental theory of stress facilitates the simple formalization of a number of stress systems with heretofore complex analyses, including Yidiny, Mayo, Cairene Arabic, Turkish, Khalkha Mongolian, Capanahua, and Tagalog.
Editor's note: The present volume is a slightly revised version of the author's 1993 Ph.D. dissertation presented to the Graduate College of the University of Arizona. Although this work is outdated in many respects, we are publishing it in order to make the data available to phonology instructors and those investigating problems of stress.