vasturääkivusi. Käesoleva tõlke aluseks on I. Bywateri väljaanne: "Aristotle on the Art of Poetry, a revised text with critical introduction, translation and commentary by I. Bywater" (Oxford, 1909). Peale selle on tõlkijal kasutada olnud S.H. Butcheri (New York, 1951, 4. tr.) ja R. Kasseli (Oxford, 1968, 2. tr.) väljaanded ning tõlgetest: R. Kleis, katkendid "Kreeka kirjanduse antoloogias", Tallinn, 1964, lk. 459-468; H. Stich, Leipzig, 1887; E.G. Schmidt, Leipzig, 1972; V. Appelrot, Moskva, 1957, 1. tr. 1893; N. Novosadski, Leningrad, 1927; M. Gasparov, Moskva, 1978 (raamatus "Aristotel' i antichnaja literatura", lk. 111-163). Juhtudele, kus tõlkija on kõrvale kaldunud Bywateri poolt aktsepteeritud tekstist, eelistades erinevat käsikirjalist lugemit, erinevat konjektuuri või tõlgendust, on osutatud vaid siis, kui teine lugem või tõlgendus annab olulise erinevuse
information on purpose, for instance when sensitive questions are asked (see also Lensvelt- Mulders, Chapter 23). Measurement errors that originate from the respondent are beyond the control of the researcher. A researcher can only try to minimize respondent errors by making the respondent’s task as easy and as pleasant as possible. In other words, by writing clear questions that respondents are willing to answer. In Chapter 2, Schwarz, Knäuper, Oyserman, and Stich describe how respondents come up with an answer and review the cognitive and communicative processes underlying survey responses. The method of data collection can be a third source of measurement error. In Chapter 7 of this book, de Leeuw describes the advantages and disadvantages of major data collection techniques. One of the key differences between survey modes is the way in which certain questions can be asked
declaratives. Objection 2 Davidson talks as if the right-hand sides of his T-sentences will be written in English, or in the theorist's own natural language whatever it may be, so that they can be readily seen to be correct or incorrect. Indeed, Davidson touts the T-sentences as empirically testable consequences of a proposed truth definition for a language. But no actual Davidsonian truth definition could deliver such T-sentences (Stich 1976; Blackburn 1984). For such a theory to yield T-sentences--or anything else--as theorems, it must be formulated in a fairly formal and regimented language, something logic-like. (Look again at the truth definition for Oafish.) Moreover, once the theorist gets around to natural-language constructions that do not occur in standard symbolic logics, such as adverbs, belief operators, and the like, the right-hand sides of T-sentences involving those may contain some radically unfamiliar notation
annaabi.ee 
