LINGUIST List 9.950

Thu Jun 25 1998

Sum: Topicalisation and Truth-conditions

Editor for this issue: Julie Wilson <>


  • Carsten Breul, Topicalisation and truth-conditions

    Message 1: Topicalisation and truth-conditions

    Date: Thu, 25 Jun 1998 09:12:19 +0000
    From: Carsten Breul <>
    Subject: Topicalisation and truth-conditions

    A big THANK YOU once again to the people who replied to my query on topicalisation/fronting and truth-conditions (LINGUIST 9.913).

    Basically, the query was about whether sentences such as the following have the same truth-conditions:

    (1) Almost everybody answered at least one question. (2) At least one question, almost everybody answered.

    Heim & Kratzer (1998), for example, claim that they do not have the same truth-conditions; I expressed my doubts about this claim.

    Of the 20 replies that I have got (till 25/6), 9 support Heim & Kratzer's claim; 10 rather reject it, one of which by a non-native speaker of English (one reply I wasn't exactly sure how to interpret).

    Several people who disagree with H&K point out however that H&K's reading of (2) is (strongly) preferred if no context is supplied. (This reading, the only possible one for H&K, is that in which the respective question(s) is/are the same for everybody who anwered it/them.)

    Stephen Straight provides the following example, syntactically analogous to (2), where the reading which is said to be impossible according to H&K is in fact strongly suggested:

    (3) At least one pickle, almost everybody ate.

    Gregory Ward supplies the following context (and cotext) for the reading of (2) rejected by H&K:

    "[context: Teacher is administering quiz to class of 30 students]

    Teacher: Class, time is up! Please turn in your exams. Students; [groaning, kvetching] Teacher: How many people were able to answer all three questions? Students: [2 students raise their hands] Teacher: How many people were able to answer two of the three questions? Students: [5 students raise their hands] Teacher: How many people were able to answer one of the three questions? Students: [28 students raise their hands] Teacher: Good. So at least one question almost everybody answered."

    The fact that a suitable context has to be supplied/invented for the disputed reading has been pointed out in a number of replies.

    Other aspects mentioned: different intonation (stress) patterns may support different readings; the topic-focus distinction of the fronted constituent is relevant; (2) has a Yiddish dialectal flavour (and has been referred to in the literature, it seems, as Y(iddish)-movement).

    Bibliographical hints:

    Ward, Gregory. 1983: "A Pragmatic Analysis of Epitomization: Topicalization It's Not". In: Papers in Linguistics 17:145-161.

    - -. 1988: The Semantics and Pragmatics of Preposing. Outstanding Dissertations in Linguistics Series. New York: Garland.

    Further related questions raised in the replies: What is the situation in other languages that have fronting? If there are languages which really allow only one (H&K's) reading of such sentences, how is the distinction between the two alternatives acquired? (It was suggested that there are such languages.)

    Additional comment: For me, the German equivalent of (2) has both readings:

    (4) Mindestens eine Frage hat fast jeder beantwortet.

    Interestingly, I do not have the impression that different sentence accent placements have an effect on which of the two readings is preferred:

    (4') MINdestens eine Frage hat fast jeder beantwortet. (4'') Mindestens EIne Frage hat fast jeder beantwortet.

    Of course, the implications differ, but it is not the difference between the two readings under discussion here. Dr. Carsten Breul Englisches Seminar Universitaet Bonn Regina-Pacis-Weg 5 53113 Bonn Germany