T

Tautology The notion of tautology enters the philosophy of language with Wittgenstein’s Tractatus Logico-Philosophicus (Wittgenstein 1922). There tautology is defined as a truth-functional proposition that is ‘true for all the truth-possibilities of the elementary propositions’ (4.46, 34). In the contemporary literature, the applicability of the term has been expanded beyond sentential logic to include any proposition that is true as a matter of form, especially those that are true as a result of syntactic repetition, e.g. sentences of the forms ‘A is A’, ‘All As are As’, ‘A or not A’ or ‘If A, then A’. Following the Wittgensteinian line, all tautologies are considered to be semantically vacuous, that is, true but devoid of content. The use of utterances of tautological form in

non-vacuous speech acts was noted by Grice (1975), who contends that statements like ‘Women are women’ or ‘War is war’ are ‘totally noninformative and so, at that level, cannot but infringe the first maxim of Quantity in any conversational context’ (33). The speaker’s choice to utter a particular tautology in the context of a given conversation then forms the basis of an implicature which leads the listener to infer non-vacuous content when confronted with such an utterance. Some theorists have worked on the details of the Gricean pragmaticist approach to tautological utterances based on the maxim of quantity (Levinson 1983; Fraser 1988; Autenreith 1997). Tautologies say nothing and, since a cooperative speaker would always strive to make his conversational contributions meaningful, the meaning requires an inference on the part of the listener. Levinson

(1983) argues that since one can assume the speaker is making as informative a contribution as required, the speaker’s utterance takes on a ‘dismissive or topic-closing quality’ (111). Nothing was said because there is nothing to say. Autenreith (1997) argues that the meaning results from an implicature that the proposition is not actually a tautology, that only the first use of the noun phrase in utterances of the form ‘A is A’ is predicative. Hence, when we hear ‘After 1905, Bertrand Russell was Bertrand Russell’, the first use of the name points to Russell the man. The second use is non-referential as it refers only to a description which is assumed by the listener to contain the common sense properties attached to the noun phrase ‘Bertrand Russell’, thereby breaking the symmetry and giving the proposition the implied meaning. Opposition to this pragmaticist approach

centres on the claim that the listener’s inference is not based on conversational context, but rather on conventional semantic aspects of the language in which the tautology is uttered (Wierzbicka 1987, 1988; Davis 1998). It is argued that meaningful tautologies are the exception rather than the rule, something that should not be the case if pragmatic considerations of quantity or relevance were all that were in play. Further, considering English language tautologies of the form ‘A is A’, it is argued that the type of noun phrase employed in the tautology radically alters the interpretation. When the tautology is formed using an abstract noun phrase, such as in the case of ‘Business is business’, the utterance is interpreted as expressing ‘a sober attitude towards complex human activities’. In contrast, when the noun phrase

employed is plural and refers to a group of humans, e.g. ‘Boys will be boys’, the interpretation inferred is one of ‘tolerance for human nature’. And if an article is a part of the noun phrase, as in ‘A deal is a deal’, then the tautological utterance is to be understood as an attempt to enforce an obligation. It is therefore more than a universal conversational implicature, the argument goes, that would lead someone to understand the difference between ‘War is war’ and ‘A war is a war’. The implicature turns in crucial ways upon semantic aspects of the particular language. A central piece of linguistic evidence cited by

the ‘radical semanticist’ camp in opposition to the ‘radical pragmaticist’ view is the nontranslatability of non-trivial utterances of tautological form. Wierzbicka (1987) argues that the standard meaning attached to the sentence ‘Boys will be boys’ is not conveyed to a native speaker when the sentence is translated into French, German, Polish, or Russian in a fashion that preserves the tautological form. She also points to non-trivial tautological utterances in Korean and Japanese that have well defined meanings for members of the respective linguistic communities, but which are not the meanings ordinary speakers would naturally attribute to such statements translated into English. This question of the translatability of tautologies has become the basis for anthropological linguistic investigations into tautological utterances in different languages, e.g. Farghal (1992) examines tautological speech acts in Jordanian colloquial Arabic, Okamoto (1993) in Japanese, Molnár (2004) in Hungarian, and Zeldovich (2005) in Russian and Polish. Ward and Hirschberg (1991) take issue with

the question of translatability, arguing that there are word for word translatable tautological utterances between English and French, and Spanish and Turkish. As such, the mechanism for interpreting tautological utterances is not purely a matter of semantics as radical semanticists such as Wierzbicka and Davis claim. At the same time, they seek to augment the Gricean pragmatic account in line with the semanticist challenge by distinguishing between the mechanisms employed for different classes of tautological utterances. But where the partitioning for the semanticists is based on the type of noun phrase,

for Ward and Hirschberg the distinctions are based upon syntactic elements, e.g. whether the utterance is equative, ‘A is A’, disjunctive, ‘A or not A’, or conditional, ‘If A, then A’. Bulhof and Gimbel (2001, 2004) follow Ward

and Hirschberg in pointing to tautologies that are translatable without loss of meaning, from Dutch political slogans to the passage 3:14 in Exodus, ‘I am who I am’. While on the one hand, such examples are problematic for the semanticist position of Wierzbicka and Davis, the proposed explanation of their translatability is entirely consistent with Wierzbicka’s taxonomy of tautologies. The group of translatable tautologies, it is argued, are those in which the operative noun phrase has both a sharply delimited sense in which degrees or quantity is irrelevant and a vague sense in which the noun phrase may be more or less appropriate. For example, one might say of a friend in her third trimester that ‘She is looking very pregnant’, but one could also use the term ‘pregnant’ in such a way that there is pregnant and not pregnant and questions of degree make no sense. The use of a tautology with such a term, for example, ‘You are pregnant or you are not’, implies that when the speaker uses the operative term, in this case ‘pregnant’, it should be interpreted in the sharply delimited sense as that is the only case in which the tautology holds true. In this way, the tautologies are not code for a non-tautological proposition that must be inferred by the listener. Rather, they are intended to be the tautologies they are, that is, true statements in line with Grice’s maxim of quality. Thereby, ‘deep tautologies’ are a category of tautological utterance consistent with the semanticist view, because their meaning requires categorization by type of noun phrase, but is entirely unpacked by Gricean maxims accounting for their translatability.