Abstract
In the earlier part of a stimulating series of William James Lectures at Harvard in 1968 Professor H. P. Grice drew the attention of the philosophical public1 to a most intriguing hypothesis about the familiar logical particles of natural language ‘not’, ‘and’, ‘if ... then...’, and ‘either ... or...’. I shall henceforth call this the Conversationalist Hypothesis. What it asserts is that those particles do not diverge in meaning, or linguistic function, from the formal-logical symbols, ‘~’, ‘&’, ‘→’, and ‘V’ respectively, as standardly interpreted by two-valued truth-tables, and that wherever they appear to diverge from truth-functionality the appearance is due to the various standing presumptions with which natural language utterances are understood. On the whole Grice argued in favour of this hypothesis, though he confessed to having no answer to one particular objection to it. I shall argue in this paper that the objection to which Grice refers is not, pace Grice, a serious one, but that there are good reasons for preferring an alternative account, which I shall call the Semantical Hypothesis, to the Conversationalist Hypothesis. According to the Semantical Hypothesis many occurrences of these particles do differ in significance from their formal-logical counterparts and many do not, and both kinds of occurrence are best explained within the bounds of an adequate semantical theory for natural languages and without recourse to a theory of conversational presumptions.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
Notes
I do not know whether Grice still holds the views that he expressed in these lectures, and I certainly do not wish to imply that he does. In any case I hope that he will take it as a tribute to the interestingness of his ideas that I have not delayed my own discussion of them until the oral publicity of the William James Lectures has been compounded by a printed version. I have taken great care not to misrepresent the content of Grice’s lectures as they have been reported to me. But if there are any details on which I have erred I do not think that they affect the main point I am trying to make.
E.g. Euripides Andromache,line 986: “Ouk estin ouden kreisson oikeiou philou”, literally `There is not nothing better than…’, meaning, in Standard English, ‘There is nothing better than… ’.
Grice’s concept of implicature was explained by him in `The Causal Theory of Perception’, Proc. Aristotelian Soc.,Suppl. vol. 35 (1961), 121–152, but there he still used the word `implication’ for it.
Compare perhaps also ‘He unintentionally insulted her’ and `The girl on the dust-cover is naked’. On the relation of such cancellations or deletions to other processes of semantic composition, cf. L. Jonathan Cohen and Avishai Margalit, `The Role of Inductive Reasoning in the Interpretation of Metaphor’, Synthese 21 (1970), 469 ff. It emerges that an order of relative importance has in any case to be supposed for the set of distinctive features that characterise a particular meaning. So the less important features will normally be the ones that are exposed to cancellation or deletion in literal usage: e.g. the prefix `plastic’ deletes the notion of growth implicit in the meaning of `flower’ but not the feature of outward appearance. Where one of the more important semantic features is deleted, we tend to regard the usage as metaphorical -e.g., in `A child is a fragile flower’. Correspondingly it is the truth-functional core of meaning in the logical particles that is undeletable: we never ascribe such words a metaphorical usage. Of course, it might be objected that every word for a representable object or event x should be assigned a second dictionary meaning as `representation of x’. But anyone who was prepared to multiply dictionary meanings on this scale would hardly be entitled to jib at assigning a weaker (purely truth-functional) meaning to `and’, ‘if… then `either…or…’, etc., in addition to a stronger (connexive) meaning. Also, presumably, the decision between listing one meaning or two in the dictionary entry for `and’ must be matched by a corresponding decision, in discourse analysis, with regard to the semantics of sentence concatenation.
I have borrowed this example, in a modified form, from my The Diversity of Meaning,2nd ed., 1966, p. 271.
I am trying to suggest here no more than the general nature of the non-truth-functional element in the meaning of `either…or…’ I am not offering, in this article, an exact lexicographical characterisation of `either…or… ’, any more than of `and’ or of `if… then But I certainly do not wish to claim, as is sometimes claimed, that a dictionary entry for the locution `either… or…’ should mention as a feature of its meaning that it indicates the speaker’s ignorance of which alternative is true. Where an utterance of the locution does indicate this, the indication seems to belong rather to what Grice calls the implicature than to the meaning.
Cf. my `Speech-Acts’, in T. Sebeok (ed.), Current Trends in Linguistics,vol. XII (forthcoming).
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 2002 Springer Science+Business Media Dordrecht
About this chapter
Cite this chapter
Cohen, L.J. (2002). Some Remarks on Grice’s Views about the Logical Particles of Natural Language. In: Knowledge and Language. Boston Studies in the Philosophy of Science, vol 227. Springer, Dordrecht. https://doi.org/10.1007/978-94-017-2020-5_5
Download citation
DOI: https://doi.org/10.1007/978-94-017-2020-5_5
Publisher Name: Springer, Dordrecht
Print ISBN: 978-90-481-5955-0
Online ISBN: 978-94-017-2020-5
eBook Packages: Springer Book Archive