162 Epistemic Parentheticals
Fischer ( 2007a : 105, 2007b : 304) provides an alternative scenario for the
development of epistemic parentheticals. She begins by noting the existence of
examples such as the following:
(27) I trowe an hundred tymes been they kist (1387– 1400 Chaucer, CT ML 1074)
‘I believe, a hundred times did they kiss (each other)’
Here, because of the existence of topicalization and inverted order ( an hundred
tymes been they kist ), which points to the independent status of the clause, the I
trowe clause must be interpreted as a “separate, independent” clause. Though not
cited by Fischer , further evidence of the independent nature of the I know clauses
may be constructions such as the following, in which an anaphoric pronoun refers
back to the preceding clause:
(28) a. For to what fyn he wold anon pretende,/ That knowe ich wel. (1380– 86
Chaucer, TC IV 922– 23)
‘To what end he would at once seek, that know I well’
b. She may have mercy, this woot I wel ... (1387– 1400 Chaucer, CT I.Pars. 872)
‘She may have mercy, this I know well ...’
c. “For they been venymes, I woot it weel .” (1387– 1400 Chaucer, CT B.NP 3155)
‘For they are poisonous, I know it well’
d. He lese shal; thereof have I no doute. (1387– 1400 Chaucer, CT G.CY 833)
‘He shall lose; thereof have I no doubt’
e. I wepte but smal, and that I undertake. (1387– 1400 Chaucer, CT D.WB 592)
‘I wept but little, and that I declare’
Fischer thus argues that “both the subordinate clause and the more loosely
connected paratactic clause could develop in a ‘comment clause ’, which has a
rather independent status” ( 2007a : 104, 2007b : 303). Furthermore, she believes
that the spread of zero- complementation in later periods may have reinforced
the development of the parentheticals.
5.6.3 Semantic Development: From Evidential to Epistemic
The diachronic sources identifi ed for evidentials include perfects, verbs of per-
ception, verbs of communication, and expressions of logical necessity, i.e.,
deontics (Anderson 1986 ; Willett 1988 : 61; Matlock 1989 ; Traugott 1989 : 47–
48). While it is well known that epistemics typically evolve from deontics, as
in the case of the English modals, their origin in evidentials has not before
been recognized for English, though Traugott ( 1989 : 33, 48) notes that “epi-
stemics and evidentials share a great number of similarities in their semantic
to any one theory concerning the origin of methinks , though her data suggest that a zero-
complementizer is more common than a that - complementizer, especially with the personal
verb think.