10.04.2013 Views

The contrastive hierarchy in phonology 2009 Dresher.pdf - CUNY ...

The contrastive hierarchy in phonology 2009 Dresher.pdf - CUNY ...

The contrastive hierarchy in phonology 2009 Dresher.pdf - CUNY ...

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

5.4.4. Spread<strong>in</strong>g of class nodes?<br />

Our comparison of feature geometry and the <strong>contrastive</strong> <strong>hierarchy</strong> reveals that,<br />

for the most part, the two either encode similar relations or else can be reconciled<br />

so that they do so. This raises the prospect that the two can be identified; put<br />

differently, that we can dispense with feature geometry as an <strong>in</strong>dependent<br />

theoretical entity <strong>in</strong> favour of the <strong>contrastive</strong> <strong>hierarchy</strong>. <strong>The</strong> major obstacle <strong>in</strong> the<br />

way of do<strong>in</strong>g this is the participation of class nodes <strong>in</strong> phonological rules. If a<br />

class node is targeted by phonological processes, then it cannot be dispensed<br />

with, at the risk of los<strong>in</strong>g the ability to formulate phonological rules that express<br />

significant generalizations.<br />

It is thus <strong>in</strong>terest<strong>in</strong>g that some recent approaches to feature geometry no<br />

longer allow class nodes to spread. For example, the Revised Articulator <strong>The</strong>ory<br />

(RAT) (Halle 1995, Halle, Vaux and Wolfe 2000) requires that term<strong>in</strong>al features<br />

spread separately. Further, designated articulators are <strong>in</strong>dicated by features, not<br />

by nodes <strong>in</strong> the geometry. Similarly, Padgett (2002) proposes that feature classes<br />

like Place, Color, Laryngeal, etc., are not nodes <strong>in</strong> a structure, but features of<br />

features, or sets of features. Phonological processes or constra<strong>in</strong>ts can mention<br />

these classes, but apply directly to the features that make them up. Padgett<br />

argues that this approach preserves the advantages feature geometry was<br />

supposed to have <strong>in</strong> be<strong>in</strong>g able to refer to certa<strong>in</strong> groups of features, while<br />

escap<strong>in</strong>g some of the negative consequences of feature geometry.<br />

If class nodes are not required to participate <strong>in</strong> phonological operations,<br />

then there are no compell<strong>in</strong>g reasons why they are required <strong>in</strong> phonological<br />

215

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!