H. Hiz - Böcker

To the philosopher, the logician, and the linguist, questions have a special fascination. The two main views of language, that it describes the world, and that it expresses thought, are not directly applicable to questions. Ques­ tions are not assertions. A question may be apt, sharp, to the point, impor­ tant, or it may be inappropriate, ambiguous, awkward, irrelevant or irreverent. But it cannot be true or false. It does not have a truth value not just because an utterance like Was the letter long? does not indicate which letter is being talked about. The indicative The letter was not long has the same indeter­ minacy. In actual context the anaphoric definite article will be resolved both for a question and for an indicative sentence. Contextual resolutions are easily found for most cross-references. A question cannot be either true or it does not describe a state of affairs. Neither does it express false, because thought, because it is an expression of suspended thought, of lack of judge­ ment. To dress it in other philosophical styles, a question is not a judgment, it is not a proposition, it is not an assertion. A philosopher may try to paraphrase a question as an indicative sentence, for instance as a statement of ignorance, or as a statement of the desire to know. Hintikka, Wachowicz and Lang explore this territory. Or he may interpret it as a meta statement intimating the direction in which the flow of the discourse is going.

The selection of papers reprinted here traces the development of syntax from structural linguistics through transformational linguistics to operator gram­ mar. These three are not opposing views or independent assumptions about language. Rather, they are successive stages of investigation into the word­ combinations which constitue the sentences of a language in contrast to those which do not. Throughout, the goal has been to find the systemati­ cities of these combinations, and then to obtain each sentence in a uniform way from its parts. In structural analysis, the parts were words (simple or complex, belonging to particular classes) or particular sequences of these. In transformational analysis, it is found that the parts of a sentence are elementary sentences, whose parts in turn are simple words of particular classes. The relation between these two analyses is seen in the existence of an intermediate stage between the two, presented in paper 4, From Morpheme to Utterance. A further intermediate stage is presented in the writer's String Analysis of Sentence Structure, Papers on Formal Linguistics I, Mouton, The Hague 1962 (though it was developed after transformations, as a syntactic rep­ resentation for computational analysis). Generalization of both of these analyses leads to operator grammar, in which each sentence is derived in a uniform way as a partial ordering of the originally simple words which enter into it: Each step (least upper bound) of the partial ordering (of a word requiring another) forms a sentence which is a component of the sentence being analyzed.