Using the connectionist model would be least appropriate to account for
A) distributional evidence.
B) lexical co-occurrence patterns.
C) subcategorization information.
D) long-distance dependencies.

Question

Quizplus · Accepted Answer

The connectionist model, which is based on neural networks and emphasizes learning from examples and patterns, is less suited for explaining long-distance dependencies. These dependencies involve relationships between elements in a sentence that are not adjacent to each other, a phenomenon that requires understanding of structure and rules beyond simple pattern recognition.

Using the Connectionist Model Would Be Least Appropriate to Account