LINGUISTIC AND NON-LINGUISTIC ' 'UNDERSTANDING' ' OF LINGUISTIC TOKENS,
Abstract
The distinction is explored between recognizing an utterance as a 'symptom; of the state of affairs in or confronting its originator and as a linguistic tool. It is suggested that although only the first kind of understanding is required to enable a computer to accept data and answer questions in verbal form, such ability is no guarantee of its comprehension of all aspects of language to attain full linguistic comprehension, the program must also embody at least a 'skeleton representation' of the linguistic context in which an utterance originates and from which it derives its linguistic significance as a goaldirected operator. By use of this richer model of its situation, a computer should be enabled to meet more sensitive tests of linguistic comprehension. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 01, 1964
- Accession Number
- AD0432308
Entities
People
- D. M. Mackay
Organizations
- RAND Corporation