Date:  Wed, 12 Oct 1994 15:10:12 EDT

From:  Larry Horn <LHORN[AT SYMBOL GOES HERE]YALEVM.CIS.YALE.EDU>

Subject: Re: Antilogies



In response to Peter Patrick's query on the two 'if not' constructions:



I don't want to be accused of excessive autocitation, but the first systematic

exploration I know of is mine, in my thesis (On the Semantic Properties of

Logical Operators in English, UCLA, 1972, IULC version distrib. 1976), in

section 1.22. Other more recent treatments are those in W. Welte,

Negationslinguistik (Munich, 1978), p. 205 and my negation book, A Natural

History of Negation (Chicago, 1989), p. 393. A sample minimal pair is

   The book is excellent if not (exactly) perfect

    [fall-rise intonation, concessive reading, corresp. to German

      'wenn (auch) nicht...']

   The book is excellent if not (downright) perfect

    [straight fall intonation, suspension (i.e. presupposition/implicature

      suspension) reading, corresp. to German 'wenn nicht (sogar)...']



As discussed in Horn 1989, only the former reading is preserved when the

negation is incorporated (The book is excellent if imperfect); this correlates

with the fact that the concessive reading can be taken as 'X if [not Y]' and

the suspension [='and maybe even...'] reading as 'X [if not] Y'. Other

minimal pairs:

   Our victory is possible if not probable. (ambiguous pending resolution

                         by intonation contour)

   Our victory is possible if improbably   (unambiguous; concessive only)



--Larry Horn