Oratio Obliqua, Oratio Recta: An Essay on Metarepresentation (review)

When thought and language meet, the combined structures can be represented so that the former is lifted to a meta-level and the latter kept at the object-level. This, Recanati argues, forms a metarepresentation, a curious linguistic structure with a propositional attitude or some related construction plus the content of the object representation. According to R, such metarepresentations work by means of simulating the object-level representation, the essence of which is taken to be contained in the metarepresentation and thus represented also in that additional level. Combined with cognition, many other kinds of codes, including those created by artistic means, may form metarepresentational structures.

The topics falling within this generous frame of thought are several: metarepresentations in belief sentences and the relation to scope ambiguities and the de dicto/de re distinction; circumstance shift vs. context shift and the failure of substitutivity; compositionality; conditionals and counterfactuals; defer-ential expressions and quasi-beliefs; demonstratives and indexicals; echoic uses of linguistic items such as metalinguistic negation and the ensuing abnormal semantic values; metarepresentation in fiction and imaginary states of affairs; the relevance of irony; opacity vs. transparency of metarepresentations (R defends the latter); and direct and indirect quotations in metarepresentations. Among the received theories and views in the wide-ranging exploration on meta-representational aspects in the book, with connections and relevance to R’s own theory, one can find: J. L. Austin-type semantics (‘utterances are propositions plus situations’); Donald Davidson’s theory of quotation and the thesis of semantic innocence, as well as Gottlob Frege’s related views; David Kaplan’s theory of indexicals (‘there are no context-shifting operators’); Arthur Prior and Willard van Orman Quine’s views on modal logic; and D. Sperber’s theory on quasi-beliefs (deference).

As is evident from this list, the book covers many topics and questions of considerable importance. It also prompts many further questions. Just to outline some of the general ones that may help in deciphering the central points, one is tempted to ask: Does R mean that metarepresentations STAND FOR representations at the object level? Is this what is meant by the concept of simulating object-level representations? Would the notion of emulation have something new to offer here? Furthermore, how would the simulation work in complex contexts of anaphora in attitude reports such as those of intentional identities? Even more fundamentally, given a state of affairs, wouldn’t it be more economical to take thoughts and utterances that have content to PRESENT (i.e. to provide information about) the state rather than represent it? If yes, then it follows that whatever stands for (or even stands IN for) such a presentation remains a representation of it, including those that R labels as metarepresentational. It would also be useful to investigate what relations there might be between the theory of (meta)representation and the theory of strategic meaning in linguistic semantics.