1.4 Pathways for Development 17Thompson and Mulac ( 1991 : 313) propose a sequence of development as
shown in (7), with the parenthetical originating in main clause I think followed
by a that- clause complement (7a). Deletion of that yields an indeterminate
structure (7b) – or “bridging” context ( Evans and Wilkins 2000 ) – in which I
think can be analyzed as either a main clause or a parenthetical. Analyzed as
the latter, I think is then movable to medial or fi nal position (7c):
(7) a. I think that we’re defi nitely moving towards being more technological.
b. I think Ø exercise is really benefi cial, to anybody.
c. It’s just your point of view you know what you like to do in your spare time
I think.
In (7c) I think is a unitary particle (much like an adverb) expressing episte-
micity. Thompson and Mulac ’s evidence is quantitative. Think and suppose
constitute 65 percent of all nominal- complement- taking verbs , with think and
suppose accounting for 85 percent of the verbs in epistemic parentheticals
(319). More importantly, think occurs 91 percent of the time without that and
guess 99 percent of the time, compared with a rate of 75 percent that- less
complements with all other verbs (320). “[T] hose subjects and verbs occurring
most frequently without that are precisely those which occur most frequently
as [epistemic parentheticals]” (317, 314, 326). I is the subject 83 percent of the
time with nominal- complement- taking verbs and 95 percent of the time with
epistemic parentheticals. Thompson and Mulac conclude that it is the “blur-
ring of the distinction between ‘main’ and ‘complement’ clause” seen in (7b)
which gives rise to the epistemic parentheticals (316). The scenario proposed
by Thompson and Mulac thus postulates a reversal in syntactic hierarchy, in
which the original main clause is lowered to become parenthetical and the
original complement clause assumes the role of the main clause.^18
This process of “sentence lowering” brings to mind Ross ’s earlier proposal
of “slifting ” or sentence lifting (1973): Beginning with a sentence such as I feel
that Max is a Martian , slifting moves the that - clause from under the domina-
tion of I feel and adjoins it to the left of the erstwhile superordinate clause; a
further rule of “niching” moves the parenthetical into clause- medial position.
Similarly, Hooper ( 1975 ) argues for the reverse process of “complement pre-
posing ,” in which all or part of the complement clause is fronted (giving both
Max is, I feel, a Martian and Max is a Martian, I feel ). Ross concludes that
there is “a very empirical compelling case” for the claim that “clauses which
are subordinate in remote structure can become main clauses in superfi cial
structure” (1973: 165). Objections have been raised to proposals of slifting and
18 Such a proposed reversal of the hierarchical relationship is present in earlier studies of
epistemic parentheticals (see Urmson 1952 : 495; Benveniste 1971 : 228– 229; Quirk et al.
1985 : 1113; Anderson 1986 ).