130 Epistemic Parentheticals
“shield” function (marking speaker tentativeness, uncertainty, or lack of com-
mitment), (b) an approximator function (marking lack of precision of a term
within the parenthetical’s scope), (c) a structural function (marking a textual
link or structuring information), and (d) a booster or strengthening function.^4
For the purposes of the historical study which follows, I will focus more
broadly on the “epistemic, evaluative, or evidential stance” (Thompson 2002 )
functions of fi rst- person parentheticals in general, ignoring the many individual
differences among the forms. It will be seen that while fi rst- person epistemic
parentheticals are primarily speaker- oriented, they also have interpersonal
functions, since in their indirectness and tentativeness they serve the purposes
of deference and politeness , while in their colloquialness and informality they
serve the purposes of intimacy.
5.2.1 Epistemic Meaning
It is well recognized that Modern English has a set of “modal lexical verbs”
(Perkins 1983 ; Coates 1987 ) that are used to report mental states or attitudes or
to modify the force of an utterance. There is no consensus about the inventory
of modal verbs, but the most commonly listed ones are assume , believe , esti-
mate , expect , feel , guess , know , presume , suppose , suspect , think , and under-
stand (see Urmson 1952 : 482; Givón 1982 : 45; Perkins 1983 : 97; Quirk et al.
1985 : 1114; Biber and Finegan 1989 : 98, 119– 120). Modal verbs may function
as main verbs taking that - sentential complements but also occur frequently as
parentheticals without the required complementation. As such, they generally
occur parenthetically with fi rst- person subjects in the simple present tense or
as impersonal constructions such as it seems or it appears (Perkins 1983 : 98;
Quirk et al. 1985 : 1114). Typically modal verbs express epistemic modality,
which can be defi ned as follows:
- “the speaker’s assumptions, or assessment of possibilities, and, in most cases ...
the speaker’s confi dence or lack of confi dence in the truth of the proposition
expressed”(Coates 1987 : 112); - “judgments about the factual status of the proposition,” including uncer-
tainty, inference from observable evidence, and inference from what is gen-
erally known (Palmer 2001 : 24); or - “evaluation of the chances that a certain hypothetical state of affairs under
consideration (or some aspect of it) will occur, is occurring or has occurred
in a possible world” (Nuyts 2001 : 21; quoted in Cornillie 2009 : 46).
4 The apparently paradoxical use of I think for boosting purposes was fi rst identifi ed by Holmes
( 1984 : 359). Aijmer ( 1997 ) refers to this use as the “deliberative” (as opposed to the “tentative”)
function. An example of such a usage may be seen in (1a) above, where I think might be glossed
‘I am sure’ or ‘certainly.’