ABSTRACT
A way of rationalizing this procedure consists in postulating the existence of a ‘universal semantic’ tacitly justifying inferences of the first type, alongside a set of ‘local semantics’ which provide an explicit foundation for inferences of the second type. Thus, in one of the examples analysed (ibid., pp. 12-54) a bas-relief which shows ‘a person holding a large bird of indeterminate nature on his gloved hand’ is related to the theme of hunting, by a chain of inferences which might be schematized as follows: ‘large indeterminate bird on gloved hand?falcon?hunting’. The knowledge presupposed if one is to accept this interpretation is nowhere explicitly stated: it is supposed to stem from that ‘universal’ semantic which goes without saying, by implication. Conversely, the rest of the interpretation is anything but obvious: the identification of the human figure as the sultan of a Seljuk dynasty which reigned at Konya in Anatolia during the 13th century is the product of a chain of inferences which are subject to doubt. In fact, they are contradicted by other interpretations, and the author takes care to justify them by references to all sorts of ‘local’ facts. These facts are local in the sense that they all refer to the specific cultural context of the monument, defined in terms of the usual dimensions of space, time and relevant human groups (in this case Anatolia during the 13th century peopled by Turks from Central Asia under the Seljuk dynasty). No-one would
ever use these same facts to support the interpretation of a similar scene found in a ruined Aztec temple.
