EVALUATIVE PREDICATES: AN ADJUNCT CONTROL ANALYSIS
Evaluative adjectives (smart, stupid, silly, rude, thoughtless) form a coherent semantic class distinct from
disposition adjectives (eager, hesitant, reluctant, anxious, (un)willing.) Both classes form subject control
structures as in (1), but only evaluative predicates show obligatory control (2).
    (1) Sheila was eager/smart to leave.
    (2) Sheila was eager/*smart for Bruce to leave.
While the semantic peculiarities of evaluative control have been analyzed (Barker 2002, Jackendoff &
Culicover 2003, inter alia), the syntactic structure has received little attention. I present a novel analysis
of the syntax of evaluative control, demonstrating that evaluative adjectives are one-place relations which
project control into an adjunct (4), in contrast with disposition adjectives, which are 2-place relations and
project control into a complement (3). This Adjunct Analysis is shown to be superior to its alternative, the
Coercion Analysis.
    (3) DISPOSITION:     eager (x, y)     2 place relation     complement control
    (4) EVALUATIVE:      smart (x)        1 place relation     adjunct control
The Coercion Analysis holds that one-place, property-denoting evaluative adjectives are coerced into a
two-place, eventive interpretation in the presence of an infinitival complement. The following arguments
suggest that the Coercion Analysis is untenable: Evaluative control structures are not compatible with
adverbial modifiers like often, which quantify over events (Kratzer 1995.) Evaluative control structures
similarly fail diagnostics for stage-level predication, for example, the formation of existential structures
(Milsark 1977), as shown in (5). Both results point to a property, not an eventive reading of evaluatives.
    (5)   a.   There were lawmakers hesitant to talk about the proposal on record.       (disp + inf)
          b. # There were lawmakers stupid to talk about the proposal on record.         (eval + inf)
          c. # There were lawmakers stupid.                                              (eval)
Under the Adjunct Analysis, the embedded infinitivals in evaluative control structures are not interpreted
as thematic arguments; rather, they are adjunct modifiers. Evidence in support of this analysis includes
the following: two-place disposition adjectives (6a) form complex "process" nominals which take clausal
arguments. In contrast, one-place evaluative adjectives (b) form simple "result" nominals, which lack
internal thematic structure, and which cannot take clausal arguments (c.f. Grimshaw 1990.) Instead,
evaluative nominals embed an infinitival clause within a prepositional phrase modifier (c).
          a. the researchers' reluctance [CP to waste time and money studying invisible waves].
    (6)
          b. *the researchers' silliness [CP to waste time and money studying invisible waves].
          c. the researchers' silliness [PP in [CP wasting time and money studying ...]].
Next, evaluative control structures fail to show focus ambiguity, a property of complement structures. As
seen in (7), stress on the verb leave can be interpreted with narrow focus, contrastive with some other
verb phrase, or with wide focus, contrastive with a larger adjectival phrase. No such ambiguity is seen in
(8), which takes only narrow focus, demonstrating that to leave is not a complement.
    (7) Actually, he was anxious to LEAVE             (8) Actually, he was rude to LEAVE
        a. ... not to STAY.                               a.    ... not to STAY.
        b. ... not willing to STAY.                       b. # ... not thoughtless to STAY.
Finally, the evaluative control structures are degraded by extraction, a property of adjuncts. Together,
these results show systematic syntactic contrasts between the classes of evaluative and disposition
adjectives, in support of the Adjunct Analysis.
This paper presents new data to suggest that a syntactic distinction correlates with already established
semantic distinctions. The analysis addresses a clear gap in the literature, demonstrating that the variation
in control structures across adjectives classes is richer than previously assumed.


References

Barker, Chris. 2002. The Dynamics of Vagueness. Linguistics and Philosophy 25: 1-36.
Grimshaw, Jane. 1990. Argument Structure, MIT Press: Cambridge, Massachusetts.
Jackendoff, Ray, and Peter Culicover. 2003. "The semantic basis of control in English." Language 79:
    517-556.
Kratzer, Angelika. 1995. Stage-Level and Individual-Level-Predicates, in Carlson/Pelletier (eds.), The
   Generic Book, 125-175, The University of Chicago Press: Chicago and London.
Milsark, Gary L. 1977. Toward an Explanation of Certain Peculiarities of the Existential Construction in
    English. Linguistic Analysis 3:1-29.