Jasper Taylor, Jean Carletta and Chris Mellish
Combining power with tractability in belief models
Belief models are used in AI computer applications in a great many roles, both to keep track of the beliefs of users and to simulate the belief states of agents modelled by the system. Epistemic logic is often used as a paradigm for the behaviour of such models, but has problems of computational complexity, which are compounded by the need to include nonmonotonic inferences and change over time. Systems based on standard logics also sanction inferences which human reasoners would be unlikely to make.
In some aspects of reasoning, simplifications can be made which both reduce the computational problems and restrict inferences to those most plausible in humans. However, in order to reap the computational advantages of restricting inferences, a model-building approach similar to a compiler must be used. A description of the domain and the agents' knowledge is entered in a formal language, and translated into a restricted representation which allows a reasoning process embodying the restrictions and other efficiency features to be used.
The process of compilation and execution is described in detail, and compared with models which use different representations for different beliefs while remaining complete in terms of standard logic. Relationships between computational simplifications and improvements in psychological plausibility are discussed.
(March 1996; 48 pages)
Ref. No. HCRC/RP-76 Price: UKL 1.70
Semantic Partition and the Ambiguity of Sentences Containing Temporal Adverbials
The ambiguity of sentences such as (i) has been explained both in terms of an ambiguity of the for-phrase (Dowty, 1979) and in terms of an effect of the perfect (Kamp & Reyle, 1993):
(i) Mary has lived in Amsterdam for three years.
I argue against these approaches, showing that this ambiguity is also present in sentences in the simple tenses, and that a unified treatment of for (as well as other temporal adverbials) is possible once it is recognised that temporal adverbials are interpreted differently depending on their syntactic position. Rather than attributing the ambiguity of such sentences to an ambiguity of the adverbial, I argue that the interpretation of a sentence with a temporal adverbial is affected by the partition of the sentence into two portions which are interpreted as parts of different semantic correlates, much like Topic/Comment, Antecedent/Anaphor and Background/Focus constructions. Finally, I discuss how a treatment of ``semantic partition'' such as that of Diesing (1992) can be extended to temporal adverbials and I provide details of how this treatment can be spelled out in Discourse Representation Theory (Kamp 1981, Heim 1982).
(February 1996; 71 pages)
Ref. No. HCRC/RP-77 Price: UKL 2.40
Anne Cooreman and Anthony J. Sanford
Focus and Syntactic Subordination in Discourse
In two experiments we test the proposal that the main clauses of complex sentences influence subsequent attentional focus and that this effect of syntactic structure competes with other cues. We argue that the strength of these cues depend on the semantics of the connector and the type of processing which is required.
The complex sentences in this study contain main and subordinate clauses linked by the connectors before, after, when, while, because and since. Our results show a strong dissociation between two different tasks. A free continuation task showed an overall massive effect of main clause while a self-paced reading experiment which measured reading times for target clauses coherent with either main or subordinate clause of the preceding sentence, showed both main clause effects and order effects. In both tasks the main clause effect is strongest with the temporal connectors after and before which require two temporal reference frames for the mental representation of the complex clause. The effect of main clause is weakened when readers are forced to integrate new information as is the case in the reading task. Here there is no effect of main clause for the connector because. The results of our study support the findings of Townsend & Bever (1978) that temporal connectors are processed differently from causal ones and that overtly expressed causal links direct more processing to the subordinate clause.
(April 1996; 11 pages)
Ref. No. HCRC/RP-79 Price: UKL 0.70
Principles of African Tone Orthography Design
Tone languages provide some interesting challenges for the designers of new orthographies. One approach is to omit tone marks, just as stress is not marked in English (zero marking). Another approach is to do phonemic tone analysis and then make heavy use of diacritic symbols to distinguish the 'tonemes' (exhaustive marking). While orthographies based on either approach have been successful, this may be thanks to our ability to manage inadequate orthographies rather than to any intrinsic advantage which is afforded by one or the other approach. In many cases, practical experience with both kinds of orthography has shown that people have not been able to attain the level of reading and writing fluency that we know to be possible for the orthographies of non-tonal languages. In some cases this can be attributed to a sociolinguistic setting which does not favour literacy efforts. In other cases, the orthography itself might be to blame. If the orthography of a tone language is difficult to master then a good part of the reason, I believe, is that the designer either has not paid enough attention to the function of tone in the language, or he has not ensured that the information encoded in the orthography is accessible to the ordinary (non-linguist) user of the language. If the writing of tone is not going to continue to be a stumbling block to literacy efforts, then a fresh approach to tone orthography is required.
This article describes such an approach, as appliued to the tone languages of sub-Saharan Africa. After describing the problems with orthographies that use too few or too many tone marks, a wide range of compromises are surveyed and evaluated. Next, I review the contributions made by reading theory and linguistic theory. The tone orthographies of several languages are presented throughout the article, with particular emphasis on some tone languages of Cameroon.
(December 1996; 42 pages)
Ref. No. HCRC/RP-80 Price: UKL ?.??
Nicholas Asher & Alex Lascarides
Questions in dialogue
In this paper we explore how compositional semantics, discourse structure and the cognitive states of participants all contribute to pragmatic constraints on answers to questions in dialogue. We synthesize formal semantic theories on questions and answers with techniques for discourse interpretation familiar from computional linguistics and show how this provides richer constraints on responses in dialogue than either component can achieve alone.
(May 1996; 38 pages)
Ref. No. HCRC/RP-81 Price: UKL 1.40