Íslenskt mál og almenn málfræði - 01.01.2011, Blaðsíða 161
The Topic ofthe Thesis and an Overview ofMethods and Results 159
15), young people (age 20-25), middle-aged people (age 40—45) and older people
(age 65—70). In all the written questionnaires, there were three possible respons-
es (judgments) for each sentence:
Yes = A natural sentence that I could easily say.
? = An odd sentence that I could hardly say.
No = An unacceptable sentence that I could not say.
In order to help participants get the intended reading of the test sentences, thus
making the results more consistent, most sentences in the main project, IceDia-
Syn, were prededed by context sentences. Since there is always the possibility that
the responses to the written surveys reflect a formal style of language use to a cer-
tain extent it is important to look at the constructions as they occur naturally in
speech and writing.
In the thesis I make extensive use of a classification of predicates that goes
back to Hooper and Thompson’s work (1973, see also Levin 1993 and Simons
2007). The relevant classes are the following (using English examples):
(6)a.
b.
c.
d.
e.
Class A: say, report, exclaim, assert, claim, vow, be true, be certain ...
Class B: suppose, believe, think, expect, guess, imagine, it seems ...
Class C: be (un)likely, be (im)possible, be (im)probable, doubt, deny ...
Class D: resent, regret, be sorry, be surprised, bother, be odd ...
Class E: realize, learn, find out, discover, know, see, recognize ...
In a common semantic classification of predicates, classes A, B and C represent
nonfactive predicates and D and E factive ones. This means that the content of
the complement clauses of predicates D and E is presupposed.
Table 1 presents a simplified itemization of the questionnaire results regard-
ing verb/adverb placement. A plus sign symbolizes positive reactions and a
minus symbolizes negative reactions. If both symbols are used it means that there
is variation and the first symbol represents the more general reaction. If only one
symbol is used it means that there was relatively little variation. No symbol means
that the clause type in question was not tested:
Icel. Far. Övdal. W-Jutl.
Vfin-Adv order that-chuses with A, B and E predicates + + +/- -/+
that-clauses with C and D predicates -/+ -/+ -
Indirect questions + - +/- -
Adverbial clauses - -/+ -/+
Relative dauses + - - -