Skip to Content

Colloquium: Yan Dong and Will Nediger

Friday, December 5, 2014
12:00 AM
3254 LSA Building

Colloquium presentations by Yan Dong and Will Nediger

Today's colloquium features presentations by two of our own graduate students, Yan Dong and Will Nediger. The title and abstracts of their presentations are given below.

 

Yan Dong: The motivation of creating elastic words in Chinese: Prosody or homophone avoidance?

Elastic words are those whose length can vary from monosyllabic to disyllabic, without changing the meaning. For example, mei 'coal' and mei-tan 'coal-(charcoal)', xue 'study' and xue-xi 'study-(practice)'. Disyllabic forms are in the shape of compounds except that the extra morpheme does not contribute much new meaning.

This phenomenon has been known since Karglren (1918), but many empirical and theoretical questions remain. One fundamental question is why elastic words are created in Chinese. I compare two theories: homophone avoidance (HA) and prosody. HA proposes that disyllabic words are created to reduce homophony and avoid ambiguity after massive syllable loss in Chinese. Since monosyllabic forms are still in use, Chinese end up with many elastic words. On the other hand, prosody theory proposes that elastic words are created because stressed positions need disyllabic forms due to the phonological requirement of Foot Binarity (Prince 1980), while monosyllables cannot occur in such positions unless two monosyllables form a binary foot.

In this study, I consider the predictions of these two theories and compare them from the following perspectives: (1) POS and elasticity in Modern Chinese, (2) homophone density and elasticity in Modern Chinese, (3) boundness and elasticity in Modern Chinese, (4) elastic words in Classical Chinese, (5) word length in Mandarin and Cantonese. I show that prosody theory offers the best explanation.

 

Will Nediger: Overcoming Empirical Challenges for an Extended Approach to Condition C

As originally formulated in Chomsky (1981), Binding Condition C states that an R-expression must not be co-indexed with an element c-commanding it. Thus, the sentences in (1) are ungrammatical because the R-expression John is bound by a pronoun and another R-expression, respectively.

(1) a. *He likes John.
     b. *John likes John.

Many empirical problems for this formulation of Condition C have been noted. Developing a competition-based binding theory due to Safir (2004), I show that the problematic empirical data can be accounted for in a principled way, given a sophisticated enough binding theory.

The data I deal with in this talk are as follows. First, epithets can be bound in certain cases, as in (2a), violating Condition C under the assumption that epithets are R-expressions. Second, standard Condition C violations can be rescued if either the R-expression or its antecedent is focused, as in (2b-c).

(2)  a. John is so careless that the idiot will get killed in an accident one               of these days.
     b. Only JOHN likes John.
     c. John only likes JOHN.

I argue that (2a) is grammatical because epithets are pronouns rather than R-expressions. Following Dubinsky and Hamilton (1998), I argue that cases where epithets are apparently unable to be bound are due to an anti-logophoricity condition on epithets, not Condition C. I also argue that focus constructions like (2b) are actually expected to be grammatical given a competition-based binding theory like that of Safir (2004). Finally, I argue that focus constructions like (2c) are grammatical despite violating Condition C for pragmatic reasons related to parallelism conditions on focus constructions.