Course Meeting Times
Lectures: 2 sessions / week, 1.5 hours / session
24.961 or equivalent
- Weekly Readings
- Class Participation
- Being on Time to Class
- 9 Written Assignments (data analysis or critical reading): Graded
- 6 Data Sets Analyzed: Not Graded or Turned In, But Discussed in Class
- Term Paper, Presented in Class and Written Up
In 24.962, we will focus primarily on phonological phenomena that are sensitive to morphological structure, including base-reduplicant identity, cyclicity, level ordering, derived environment effects, opaque rule interactions, and morpheme structure constraints. In the recent OT literature, it has been claimed that all of these phenomena can be analyzed with a single theoretical device: correspondence constraints, which regulate the similarity of two lexically related forms (such as input and output, base and derivative, base and reduplicant).
Brief Historical Overview
Correspondence theory has its roots in the 1970s, when linguists like Alan Sommerstein, Paul Kiparsky, Morris Halle and Jay Keyser, Ronnie Wilbur, Sandy Chung, Nick Clements, Juliette Levin and Luigi Burzio were led to formulate conditions mandating recoverability of the input or similarity between related forms in reduplication. At the same time, Paul Kiparsky, Morris Halle and Jay Keyser proposed mapping conditions involving the correspondence between the rhythm of a line of verse and the stress pattern of the words filling the line.
Correspondence and OT
Correspondence has become a central part of phonological theory in OT, where correspondence (aka faithfulness) constraints were used first to define limits within which markedness constraints will affect an input. Extensions of correspondence provide the basis of the OT treatment for phenomena such as: cyclicity, level ordering, derived environment conditions, opaque rule interactions, base-reduplicant identity, vowel harmony and others. An idea that is fundamentally new (McCarthy and Prince, 1995) is that the same types of constraints (e.g. MAX segment and Ident feature) can be used to link different pairs of representations (e.g. a lexical input to its output correspondent; and the reduplicant to its surface base).
Despite heavy reliance on an ever-expanding set of correspondence conditions, basic issues in the theory of correspondence remain unresolved and sometimes unaddressed. Among these are the following:
- What are the phonological entities that stand in correspondence? Segments, features, syllables, all of the above and more?
- Can some/all correspondence constraints be induced from the data?
- Are their relative rankings sometimes predictable?
- Which pairs of forms stand in correspondence? For instance, does a root stand in correspondence with any and all of its derivatives? Or the root and its immediate derivatives only? Can two co-derivatives of the same root stand in correspondence? When? Do affixal allomorphs stand in correspondence with each other? What string is the base in a reduplicated form? Most of these issues involve the internal structure of paradigms, and should perhaps be settled on morphological grounds-but the popularity of output-to-output correspondence conditions makes this class of questions important in a phonological context, as well.
- Correspondence constraints are inherently symmetric but their empirically observed effects are asymmetric: derivative shapes are changed to make the derivatives look like their base, but base shapes are almost never changed to make them look like their derivatives. How do we explain this fact and how do we let the analysis reflect it?
This course does not promise to deliver answers to all these questions but perhaps some progress on all fronts will emerge from the discussion.
What the course will deliver is:
- A survey of the basic mechanisms of correspondence theory as applied to I-O correspondence.
- An exemplification of classical I-O, O-O correspondence theory carried out through the analysis of reduplication.
- A survey of the results of lexical phonology and of various attempts to recapture these results in terms of OT correspondence conditions.
- Discussion of various attempts to extend correspondence theory to cover phonological opacity, including both the successes and failures of these proposals.