5.3 History of Epistemic Parentheticals 137occasionally during this period. The “bare” parenthetical increases in the sev-
enteenth century, while the adverbial as I fi nd ceases to function as a paren-
thetical in Modern English.
5.3.1 Excursus on Wierzbicka ( 2006 )
Noting the range of verbs used in epistemic parentheticals and their compara-
tive frequency in Present- day English, Wierzbicka ( 2006 : 204– 246) argues “the
emergence of a large class of epistemic phrases is a striking innovation of mod-
ern English, quite remarkable in both a historical and cross- linguistic perspec-
tive” (246). She sees them as “a peculiarity of modern English” (206) which
came into use “on a large scale” in the fi rst half of the eighteenth century (207)
during the British Enlightenment. They are a response to the “need to spe-
cify, in rational discourse, the epistemic status of our utterances” (207), which
is a theme of John Locke’s Essay concerning human understanding (1690).
Thus, they are a product of the rationalist philosophy of the day, which stressed
the limitations of human knowledge and the tentative nature of our assertions
(241). Wierzbicka ’s is not an historical study of these forms; rather, it is a nat-
ural semantic metalanguage (NSM) of I think / suppose/ guess/ gather/ presume/
believe/ fi nd/ expect/ take it/ understand/ imagine/ bet/ suspect/ assume. She admits
that an historical study would be necessary (242), but she argues briefl y (242–
243) that Sir Thomas More’s work, dating from the mid sixteenth century,
utilizes only a small number of epistemic phrases, almost all of which express
certainty and are emphatically assertive. Furthermore, I believe in Shakespeare,
unlike in Present- day English, expresses truth and strong conviction (244– 245;
but see Brinton 2008: 223– 224).^9
Taking up Wierzbicka ’s challenge for an historical study, Bromhead ( 2009 )
examines I think / methinks , I wot , I ween , I suppose , and I trow in a corpus
of sixteenth- and seventeenth- century texts. All may occur parenthetically
and more frequently occur without that than with. Using NSM analysis, she
argues that all of these forms primarily express certainty rather than uncer-
tainty, though they do so in different degrees; e.g., ween expresses less cer-
tainty than wot , and trow may have both certain and uncertain uses. Bromhead
argues against Palander- Collin ( 1999 ), who suggests that I think/ methinks
expresses uncertainty. The meaning of I think in the period she examines is
more in line with the “deliberative” use of I think that has been identifi ed (e.g.,
Aijmer 1997 ). Her work is intended to support Wierzbicka ’s claim that expres-
sions of uncertainty do not become predominant until the eighteenth century
and that speakers in the pre- Enlightenment period were more concerned with
the assertion of truth. As Van Bogaert ( 2011 ) notes, Bromhead ’s study “paints
9 For a general critique of Wierzbicka ’s book, see Dancygier ( 2009 ).