A Scope Theory of Tense and Adnominal Modifiers
This paper defends a scope theory of tense, according to which
tense morphemes systematically determine the temporal
interpretations of the expressions that occur in structurally
lower positions. The evidence for this view comes from the
behavior of adnominal modifiers (present and past participles, in
particular) in English. For example, it is virtually impossible to
interpret (1a) to mean that the man who was singing on the stage
is now drinking wine. On the other hand, (1b) is interpreted to
mean that the man who was singing on the stage was smiling at
Mary (simultaneously). In (1a-b) the temporal interpretation of
the present participle is determined the tense. On the other hand,
the fact that (1c) is acceptable is accounted for in terms of QR.
(1) a. # The man singing on the stage is now drinking wine.
b. The man singing on the stage was smiling at Mary.
c. The man singing on the stage (now) was a car
salesman before.
At first sight, this proposal appears to go against the view
that temporal properties of common nouns cannot be accounted
for in terms of a scope theory of tense (e.g., Enc (1981, 1986),
Musan (1995)). For example, Enc (1981) presents examples like (2)
to show that the temporal interpretation of the common noun
fugitive is not determined by the tense morpheme.
(2) a. Every fugitive is now in jail.
b. There were many homeless people at the rally.
The most natural interpretation of (2a) is that every person who
was a fugitive in the past is now in jail. However, there is no
tense morpheme that could be responsible for the temporal
location of fugitive, which is a past interval. Enc accounts for
this fact by assuming that nouns (and also verbs) are indexicals
and that they are not operators. Musan (1995) modifies Enc's
proposal and accounts for "temporally dependent reading" of
nouns such as (2b) in terms of the idea that determiners quantify
over stages of individuals. Musan argues with Enc, however, that
the data cannot be explained in terms of scope.
Our position with respect to common nouns is that they are
inherently very different from verbs or adjectives in that they
have their own mechanism for determining their temporal
properties. The exact characterization of such mechanism is
beyond the scope this paper, but we adopt Enc's indexical analysis
of nouns for simplicity. The important point here is that adopting
this proposal does not necessarily mean that we should or must
dispense with a scope theory of tense. If the temporal argument
of a common noun is a free variable, then its interpretation is not
affected by tense even if tense has scopal properties. Thus, a
scope analysis of tense is at least harmless as far as common
nouns are concerned. Given examples like (1a-c), we now have
reason to hypothesize that tense does determine temporal
interpretations of structurally lower expressions. Thus, a scope
analysis of tense is necessary.
There may possibly be cases in which the temporal properties
of adnominal modifiers cannot be explained in terms of scope
alone. For instance, (3a) may be marginally acceptable on the
interpretation in which the people in question were fugitives and
were crying for joy at some past time. However, even if this were
granted, (3b) could not possibly mean that the current professor
who used to cry at home when she was waiting for her parents
(when she was a little child) has three children now.
(3) a. ?? Every fugitive crying for joy is now in jail.
b. # The professor crying at home waiting for her
parents has three children now.
This means that the temporal interpretation of an adnominal
modifier is determined in terms of scope, with the possible
exception of some marginal cases in which it harmonizes with
the temporal interpretation of its modifiee. This proposal agrees
with Ogihara (1996), who argues for the same view based
primarily upon the behavior of Japanese relative clauses.
In sum, we need to adopt a scope theory of tense to account for
temporal properties of adnominal modifiers because they are
structure dependent. The known behavior of common nouns should
be explained in terms an independent mechanism, but not by
rejecting a scope analysis of tense.
Selected References
Enc, Murvet (1981) Tense without Scope, Ph.D. dissertation,
University of Wisconsin-Madison.
Enc, Murvet (1986) "Towards a Referential Analysis of Temporal
Expressions," Linguistics and Philosophy 9, 405-426.
Musan, Renate (1995) On the Temporal Interpretation of Noun
Phrases, Ph.D. dissertation, MIT.
Ogihara, Toshiyuki (1996) Tense, Attitudes, and Scope, Kluwer,
Dordrecht.
