286 Concluding Remarks: Pathways of Change
possible to propose this form as the source of the pragmatic uses, especially as
the pragmatic senses of hwæt and hwæt þa arise naturally via context- induced
inferences from the general interrogative sense. However, the status of this gen-
eralized particle of interrogation remains controversial. We are thus left with
no clear source for the pragmatic uses in Old English, though it seems clear
that these uses must arise out of the original lexical meaning and propositional
function of hwæt as an interrogative. The subsequent history of pragmatic what
in Middle English as a marker of speaker surprise and as an attention- getting
device can be explained more easily as arising from earlier uses of what with
the OE interjections la and eala or with the French interjection ho.
The development of while has been seen as a “paradigm example” of the
change from propositional to textual to interpersonal (see Traugott 1982 and
passim): The OE adverbial collocation þa hw ī le þe ‘at the time that’ referring
to a concrete state of affairs evolves into ME wh ī le (that) ‘during’ functioning
as an adverb/ conjunction with textual meaning of relevance of simultaneity,
and fi nally arrives at EModE while ‘although’ functioning as a conjunction
with concessive or adversative force. Chapter 3 examines the development of
the related form whilom in Middle English. On fi rst glance, this form seems
to follow the pathway outlined in (b) above. That is, whilom begins as a sen-
tence- internal adverb with the meaning ‘at times, sometimes’ modifying an
iterative or habitual event expressed in the predicate of the sentence, then
develops into a correlative conjunction with wider scope meaning ‘sometimes
... other times’ and fi nally into a pragmatic marker denoting the initiation of
a story, episode, or exemplum with the meaning ‘once upon a time.’ However,
the historical data point to a different origin for the pragmatic marker, namely,
directly from the sentence- internal adverb whilom with the meaning ‘once, for-
merly,’ a semantic change in which the meaning of ‘once’ can be understood as
an implicature of the meaning ‘sometimes’ (since the hearer can infer that what
happens repeatedly has happened at least once in the past). A further compli-
cation in the pathway of development is the subsequent rise of the adjectival
use of whilom with the meaning ‘former (of a person)’; this usage likely stems
from the adverbial meaning ‘once’ and is perhaps an instance of categorical
gradience rather than a shift from adverb to adjective. Thus, we see that neither
of the pathways (a) nor (b) can adequately account for the course of develop-
ment of whilom.
The case of only discussed in Chapter 4 contributes to the complex pic-
ture of the development of pragmatic markers from adverbial sources. Initial
steps in the develop of only – from the numeral one to the polysemous Adj/
Adv only to exclusive focusing only – is consistent with known pathways of
grammaticalization , showing increasing decategorialization and a decrease in
scope. Chapter 4 explores the rise of conjunctive/ textual uses of only with two
closely related senses, an ‘adversative’ sense ‘but’ (as in I aimed at the target,