10.3 Clausal Sources of Pragmatic Markers 287only I missed ) and an ‘exceptive’ sense ‘except that’ (as in I would like to
help, only I have to study ), the latter occurring in modal contexts. Especially
in the ‘exceptive’ use, only has pragmatic functions as an interpersonal and
politeness marker. To what extent does the rise of the conjunctive and prag-
matic- marker uses of only conform to the pathways set out above? Does the
conjunctive use of only develop from the focusing use of only or directly from
its polysemous Adj/ Adv use? Both developments would follow the course set
out in (a) from adverb > conjunction (and ultimately to pragmatic marker).
Both would also show an increase in scope and movement to sentence- initial
position, as predicted by pathway (b). The choice between the two possibil-
ities depends on the status of focusing adverbs in general; as these have been
shown to have textual and interpersonal functions (see Nevalainen 1991 ) and
compared to sentential adverbs, I argue in Chapter 4 for a development from
the focusing adverb to the conjunctive use. But to some extent the question
remains open. Furthermore, while the shift to the ‘adversative’ sense of con-
junctive only can be understood as a natural outgrowth of the ‘solely’ sense of
the focusing adverb, the later development of the ‘exceptive’ sense is more dif-
fi cult to explain, perhaps involving a semantic transfer of the meaning from the
exceptive senses of but/ save/ except in the expressions but/ save/ except only to
only itself, with incorporation of the expression of hypotheticality in the modal
contexts in which exceptive meanings arise.
Thus, while the studies in this book confi rm to some extent the validity of
the suggested pathways, they also show how each pragmatic marker “has its
own history” and how the historical data are messier and more complex than
these straightforward proposals might lead one to believe.
10.3 Clausal Sources of Pragmatic Markers
Many pragmatic parentheticals such as I think , you know , as it were , what’s
more (see Section 1.2.2 ) can be seen as arising in full clauses.
10.3.1 The Matrix Clause Hypothesis
The term “matrix clause hypothesis” was introduced in Brinton ( 2006 ) to
denote a proposal by Thompson and Mulac ( 1991 ) accounting for the syn-
chronic origins of the epistemic parentheticals I think and I guess. This hypoth-
esis has been explained in full in Section 1.4.1.2. Briefl y, it postulates that
these parentheticals originate as main clauses with that - clause complements.
Deletion of the complementizer that leads to an indeterminate structure which
can be understood as either a main clause or a parenthetical. Analyzed as the
latter, it can then move to medial or fi nal position (see Chapter 1 , example 7).
The scenario thus postulates a reversal in syntactic hierarchy, in which the