Key points are not available for this paper at this time.
The fundamental principle of language engineering—namely, that the probability of a correct message reception is critically dependent upon the size of the set of messages available for communications—was examined experimentally. Accuracy of message reception was found to be independent of the message-source uncertainty, but critically dependent upon the response uncertainty—or more exactly, upon the size of the set of relevant response categories. Implications for a conceptual model of the listener's behavior are suggested.
Irwin Pollack (Sun,) studied this question.