34 Pragmatic Markers: Synchronic and Diachronic
example, the pattern of I think parentheticals has extended to other verbs such
as reckon , or the construction stuff like that/ things like that seems to be pro-
ductive, giving forms like and shit like that ). Lexicalization, in contrast, where
“a given expression is no longer ‘freshly’ assembled from its constituent parts
on each occasion of its use” can be seen as involving a decrease in productiv-
ity ( Himmelmann 2004 : 37). Finally, grammaticalization leads to an increase
in frequency, as is also the case with the development of pragmatic markers,
which, once formed, may become highly frequent (e.g., well , like , okay ). With
lexicalization, “frequency is a non- issue” ( Himmelmann 2004 : 37), as a lexi-
calized item may be more or less frequent than its source.
What are we to make of this proposed process of pragmaticalization? This
question is explored insightfully by Degand and Evers- Vermeul ( 2015 ) and
Heine ( 2013 : 1219– 1220) ( cf. also Ocampo 2006 ). Degand and Evers- Vermeul
( 2015 ) present several possible positions that could be taken:^38
(a) that pragmaticalization can be dispensed with altogether;
(b) that pragmaticalization is a subtype of grammaticalization; or
(c) that pragmaticalization is a distinct process.
Against position (c), we can begin by noting that the similarities of pragmatical-
ization and grammaticalization make the two processes in many ways impos-
sible to distinguish (cf. Degand and Evers- Vermeul 2015 : 69, 75). Common
to both are fusion and coalescence , semantic demotivation, decategorializa-
tion , semantic/ pragmatic expansion, inferencing or pragmatic strengthening,
increased frequency, gradualness, persistence , layering , and divergence. In
an introduction to a special issue of Linguistics 49(2) on pragmatic markers
and the pragmaticalization/ grammaticalization divide, Degand and Simon-
Vandenbergen ( 2011 ) point out that the contributors were “unanimous in their
rejection of pragmaticalization.”
The choice between pragmaticalization and grammaticalization seems to
hinge not on the process itself but on what is encompassed by “grammar.”
If grammar is viewed narrowly, as restricted to the morphosyntactic domain,
and if grammatical items must have truth- conditional meaning, be obligatory,
and fall strictly within the syntax of the sentence, then pragmatic elements are
excluded. But it has been pointed out that many aspects of grammar – tradi-
tionally conceived – such as deixis, reference, tense, and modality have a
38 Degand and Evers- Vermeul ( 2015 : 70– 73) suggest two other positions: that pragmatic markers
sometimes grammaticalize and sometimes pragmaticalize and that pragmatic markers develop
from a process that is neither grammaticalization nor pragmaticalization. They attribute the
former position to Erman and Kotsinas ( 1993 ), who suggest that some pragmatic markers
arise directly from lexical items while others pass through a grammatical marker stage. They
attribute the later position to Waltereit ( 2006 ); however, I believe his view can be incorporated
with the pragmaticalization view (though, as I note above, he does not use the term).