Adnominal and adverbial "vermutlich"

[Andreas Nolda]

Hi everybody,

on Monday, November 28, I gave a talk on the adnominal and the 
adverbial of "vermutlich" ('presumable') in German:

(1) der vermutliche Kopf der Bande

(2) a. Vermutlich die IRÀ hat am Mittwoch zwei Bombenanschläge verübt.
    b. Vermutlich hat die IRA am Mittwoch zwei Bòmbenanschläge verübt.

In my talk, I argued for the hypothesis that _vermutliche_ in (1) and 
_vermutlich_ in (2) are *qualifiers*:

(3) a. _vermutliche_ in (1) qualifies _der Kopf der Bande_ w.r.t. _der
       Kopf_ or
    b. _vermutliche_ in (1) qualifies _der Kopf der Bande_ w.r.t. _der
       Kopf der Bande_.

(4) a. _vermutlich_ in (2 a) qualifies _die IRA hat am Mittwoch zwei
       Bombenanschläge verübt_ w.r.t. _die IRA_.
    b. _vermutlich_ in (2 b) qualifies _hat die IRA am Mittwoch zwei
       Bombenanschläge verübt_ w.r.t. _hat die IRA am Mittwoch zwei
       Bombenanschläge verübt_.

Thus, (1) is relationally ambiguous.

What I am mainly interested in is the semantics of (1) and (2). 
Although I already profited much from the suggestions made during the 
discussion period of my talk, any further comments from your part are 
highly welcome.

As to (1), I proposed the following intensional relations as 
components of (1)'s intermediate syntactic meanings:

(5) a. The relation between x1, V, and V1 such that, for all x2,
       if V1 refers by _der Bande_ in V to x2,
       then V1 presumes: 'x1 is the head of x2'.
    b. The relation between x1, V, and V1 such that V1 presumes:
       for all x2,
       if V1 refers by _der Bande_ in V to x2,
       then 'x1 is the head of x2'.

Regarding (2 a) and (2 b), I suggested propositions along the 
following lines:

(6) a. The relation between V and V1 such that there is an x1 such
       that
       1. 'x1 committed a bomb attack on Wednesday' and
       2. V1 presumes that V1 refers by _die IRA_ in V to x1.
    b. The relation between V and V1 such that V1 presumes that
       'the IRA committed a bomb attack on Wednesday'.

As pointed out by Monika Budde and Hans-Heinrich Lieb during the 
discussion period, the meanings in (5) and (6) suffer from the fact 
that they presuppose an empty lexical meaning for "vermutlich", which 
is hardly convincing.

Together we constructed a preliminary lexical meaning .vermutlich-adv. 
for the adverbial "vermutlich", taking the verb meaning .vermuten. as 
a starting point ("s" stands for states-of-affairs):

(7) a. .vermuten. =
       the property of being a perception or conception
       whose content contains {VERMUTEN} as a subset.
    b. VERMUTEN =
       the relation between x1, x2, and s such that
       x2 is willing to believe during x1 that s is a fact.

(8) a. .vermutlich-adv. =
       the property of being a perception or conception 
       whose content contains {VERMUTLICH-ADV} as a subset.
    b. VERMUTLICH-ADV =
       the relation between x1, x2, and s such that
       1. x2 produces x1 and
       2. there is an x3 such that
          a. <x3, x2, s> is in the extension of .vermuten. and
          b. the time of x1 is part of the time of x3.

Note that .vermutlich-adv. is construed as a deictic concept: it is 
the speaker to whom the presumption is ascribed.

As a replacement for (6 a), both (9) and (10) were considered:

(9) The relation between V and V1 such that there is an x1 such that:
    a. 'x1 committed a bomb attack on Wednesday' and
    b. there is an x2 and x3 such that:
       i.   x2 corresponds to V for V1,
       ii.  x3 corresponds to V1 for V1,
       iii. <x2, x3, s*> is in the extension of .vermutlich-adv. and
            in the reference basis for _vermutlich_ w.r.t. V, V1, and
            .vermutlich-adv.;
            where s* =
            the state-of-affairs such that, for all x4,
            if V1 refers by _die IRA_ in V to x4,
            then x4 = x1.

(10) [as in (9);
      with s* =
      the state-of-affairs such that, for all x4,
      if 'x4 committed a bomb attack on Wednesday',
      then x4 = x1.]

The following semantic effects of syntactic function occurrences were 
assumed:

1. Condition a. (that is, the lambda expression corresponding to it)
   is an effect of the nucleus occurrence between _hat verübt_ and
   _die IRA hat am Mittwoch zwei Bombenanschläge verübt_.
2. Condition b. is an effect of the qualifier occurrence.
3. The overall logical structure ("there is an x1" and the conjunction
   of conditions a. and b.) is an effect of the nucleus occurrence
   between _die IRA hat am Mittwoch zwei Bombenanschläge verübt_ and
   the whole syntactic unit.

Now, it was suggested that (8) is preliminary insofar 
as .vermutlich-adv. should be derived from .vermutlich-adn., the 
lexical meaning of the adnominal "vermutlich"; .vermutlich-adn. in 
turn is to be derived from .vermuten.. In addition, the extension 
of .vermutlich-adn. should involve properties instead of 
states-of-affairs.

While states-of-affairs can easily be derived from properties (e.g. 
the state-of-affair that some x has the property y), I don't see how 
states-of-affairs like s* in (9) and (10) can be derived from 
properties, though.

As an alternative, I'd like to suggest that .vermutlich-adv. 
equals .vermutlich-adn.. Given this assumption, (5) can be replaced 
by:

(11) a. The relation between x1, V, and V1 such that, for all x2,
        if V1 refers by _der Bande_ in V to x2,
        then there is an x3 and x4 such that
        1. x3 corresponds to V for V1,
        2. x4 corresponds to V1 for V1, and
        3. <x3, x4, s*> is in the extension of .vermutlich-adn.
           and in the reference basis for _vermutliche_ w.r.t. V, V1,
           and .vermutlich-adn.;
           where s* =
           the state-of-affairs such that 'x1 is the head of x2'.
      b. The relation between x1, V, and V1 such that there is an x3
         and x4 such that
         1. x3 corresponds to V for V1,
         2. x4 corresponds to V1 for V1, and
         3. <x3, x4, s*> is in the extension of .vermutlich-adn.
            and in the reference basis for _vermutliche_ w.r.t. V, V1,
            and .vermutlich-adn.;
            where s* =
            the state-of-affairs such that, for all x2,
            if V1 refers by _der Bande_ in V to x2,
            then 'x1 is the head of x2'.

All the best,

Andreas Nolda
-- 
Andreas Nolda      http://www2.hu-berlin.de/linguistik/institut/nolda/

Humboldt-Universität zu Berlin
Philosophische Fakultät II
Institut für deutsche Sprache und Linguistik