Alan Prince (Rutgers University)
Friday October 13th, 15:00-17:00
61 rue Pouchet, Paris 75017, room 159
This talk opens the mini-course taught by Alan Prince and Nazarré Merchant the week after (see here for more information).
Abstract
(pdf available here)
Linguistic theories, often, are composite objects, which define grammars indirectly as (relatively) free combinations of basic primitives. An early conceptual mis-step was to regard this property as a nuisance rather than a boon; emphasis was then placed on introducing steering mechanisms which would allow the analyst, like one of those pre-Newtonian angels guiding the planets, to make sure that things did not wander off course.
Considerable steps have been made in the direction of recovery, but an important correlate of theoretical compositionality has perhaps not fully sunk in: the analyst, no longer the great helmsperson, faces a theory that dictates what is and is not an analysis and autonomously determines it structure. But just because you write down some premises doesn’t mean that you know what they entail! Or that it is easy to find out. So the analyst, or someone sympathetic to him or her, must analyze the theory as the well as the data if there is to be any hope of knowing what the theory says. This effort displaces in importance the historically favored ‘betterness struggle’, which presupposes that it is easy, and essential, to argue on whatever grounds that your favored theory is superior to alternatives, often with little scruple about the level at which they are understood.
In this talk, I examine the structure of Optimality Theory, aiming to assemble the objects of that theory from their source in the very idea of optimality. At the microstructural level, we deal with an OT system specifying constraints and the forms they evaluate. These give us the notion a ‘grammar’, which is a component of the ‘typology’ of the system. When these notions are properly understood, we can investigate the macrostructure of the typology — the principles that organize its grammars into similar and contrasting classes, and give the sense of the ‘linguistically significant generalizations’ that they incorporate, which are concerned with the way that microstructural interactions produce macrostructural patterns. Along the way, we can spot crucial junctures where theory and data-modelling part company in their interests, and where analytical methodology has the choice of responding to one or the other.
References
Alber, B. and A. Prince. 2016. Outline of Property Theory [Entwurf einer verallgemeinerten Eigenschaftstheorie]. Ms. U. Verona, New Brunswick.
Alber, B. and A. Prince. The Book of nGX. ROA-1312.
Alber, B. and A. Prince (in prep). Typologies.
Magri, G. 2015. When is it the case that one tableau suffices? ROA-1253.
Merchant, N. and A. Prince. 2016. The Mother of all Tableaux. ROA-1285.
Prince, A. 2007. The Pursuit of Theory.
Prince, A. 2015. One Tableau Suffices. ROA-1250.
Prince, A. 2016. What is OT? ROA-1271.
Prince, A. 2017. Representing OT Grammars. ROA-1309.