5.6 Development 1535.6 Development of First- Person Epistemic Parentheticals
In order to understand the rise of fi rst- person epistemic parentheticals we begin
by looking back at the expression of epistemicity in Old English, before exam-
ining their syntactic and semantic development.
5.6.1 Epistemic Forms in Old English
According to Goossens ( 1982 : 84), Old English is generally defi cient in mark-
ers of epistemic modality. He fi nds that nothing “ grammaticalizes the epistemic
function in Old English.” He believes that the modal auxiliaries contribute in a
“real, but also very restricted” way to the expression of epistemicity; likewise,
Traugott ( 1992 : 197) notes that some of the “pre- modals” express “marginal
epistemic colouring in most instances,” especially in impersonal constructions
(cf. Denison 1990 ).
5.6.1.1 Epistemic Adverbials. The consensus of scholars is that
the adverbials in Old English are “not clearly epistemic,” or they merely
show “epistemic potential” (Goossens 1982 : 84). Among possible adverbs,
Goossens cites æfæstla ‘certainly,’ forsoþ ‘indeed,’ forgeare ‘very certainly,’
hwæt ‘indeed,’ huru ‘surely, truly,’ butan tweon ‘undoubtedly,’ and untweol-
lice ‘surely’ as markers of ‘certainty’ and wenunga ‘possibly, perhaps’ and
gewene ‘perhaps’ as markers of ‘possibility.’ The expression of possibility and
probability is, according to Goossens , extremely limited: only the expression
of certainty shows full adverbial expression. Similarly, Swan ( 1988b , 89– 90,
132) fi nds that Old English has a small class of “truth- intensifying” adverbs,
including most commonly soðlice , witodlice , sweotole , sweotollice , gewislice ,
cuðlice , and less commonly untweogendlice , openlice , nede , uncuth , unleas-
lice , untweylice , to soðan , buton tweon.^20 However, she sees these adverbs as
“more rudimentary and also less extensive” than modal adverbs in Present- day
English (1988b: 89, 128) and, more importantly, she notes that they express
high probability or importance – “giving the hearer a clue that the following
is important or considered true” – but almost never low- probability evaluation
(89, 90, 91, 131). That is, they do not serve as hedges or downtoners. Old
English does not have “pure probability evaluators” (90). Furthermore, they are
almost never associated with inverted word order (137) and thus do not func-
tion as sentence or disjunct adverbials. Some adverbs, such as swutollice , may
be evidential in meaning (372). Following Swan for the most part, González-
Álvarez ( 1996 ) agrees that “middle- and low- probability, evidentiality ... [is]
20 On the use of witodlice and soðlice as a pragmatic marker in Old English, see Lenker ( 2000 ;
cf. Swan 1988b : 129– 131).