Faculty of Arts
University of Groningen
EPISTEMIC ADVERBS AT THE
INTERFACE OF
LEXICALIZATION AND
GRAMMATICALIZATION
Muriel Norde
Outline

Preliminaries

Epistemic adverbs

Theoretical discussion
• the category of adverbs
• grammaticalization vs. lexicalization
• synchrony
• diachrony
TIN-dag 03-02-07
2
The category of adverbs
Open or closed class?



Talmy 2000: only N, V and Adj form open
classes
Ramat & Ricca 1998: range from relatively
open (fortunately) to relatively closed class
(monomorphemic advs such as now, just)
Brinton & Traugott 2005: no clear binary
distinction between lexical / major / open
classes on the one hand and grammatical /
minor / closed classes on the other. “Lexical”
and “grammatical” items form a continuum.
TIN-dag 03-02-07
3
Adverbs: forms and functions

Adverbs may be:

Adverbs may modify:
• monomorphemic: she walks fast
• derived: she walks slowly
• phrasal: she walks like a construction worker
• predicates: she sings beautifully
• modifiers: she is incredibly rich
• sentences: maybe she is pregnant
TIN-dag 03-02-07
4
Sentence adverbs: semantics






Connecting: however
Speech act: confidentially
Domain: linguistically
Propositional: probably
Event: yesterday
Predicate: quickly
TIN-dag 03-02-07
5
Propositional adverbs

Modal

Event-oriented evaluatives: unfortunately
Participant-oriented evaluatives: wisely

• epistemic: probably, certainly
• quotative: allegedly
• evidential: evidently
• optative: hopefully
TIN-dag 03-02-07
6
Epistemic modality
Wide definition:
“[…] a speaker’s evaluation of the
likelyhood of a state of affairs, as
expressed in language”
(Nuyts 2001:xv)

TIN-dag 03-02-07
7
Grammaticalization




“Grammaticalization consists in the increase of the range
of a morpheme advancing from a lexical to a grammatical
or from a less grammatical to a more grammatical status.”
(Kuryłowicz 1975 [1965]
“[…] an evolution whereby linguistic units lose in semantic
complexity, pragmatic significance, syntactic freedom,
and phonetic substance […]”(Heine & Reh 1984)
“A grammaticalization is a diachronic change by which the
parts of a constructional schema come to have stronger
internal dependencies” (Haspelmath 2004)
taken litterally: ‘having become grammatical’
TIN-dag 03-02-07
8
Paradigm examples

Motion verb > future auxiliary
Eng. to be going to; Du. gaan; Sw. komma att

Demonstrative > complementizer
Eng. that; Du. dat

Body part noun > spatial expression
Ewe megbé; Da bag ‘back > behind’
→ crosslinguistically common and regular
TIN-dag 03-02-07
9
Lexicalization
“recruitment of linguistic material to enrich the lexicon” (Hopper &
Traugott 1993)
 “today’s grammar may become tomorrow’s lexicon” (Ramat 1992)
 Dependent on one’s definition of lexicon
 Definition adopted here: Brinton & Traugott 2005
“[…] the view that the lexicon does not exist solely of a list of discrete and
fully fixed items but represents a continuum from more to less fixed,
from more to less fully conventionalized, and from more to less
productive items. […] the continuum models of the lexical /
grammatical split and of the lexicon fit better with the historical facts of
change, which is often (though not always) gradual in the sense that
change occurs by very small steps.
 Contra GL conception of grammatical categories as discrete entities

TIN-dag 03-02-07
10
Subtypes of lexicalization




Function words
•
•
Pros en cons
[Shaved her legs and then] he was a she (L. Reed)
Suffixes
•
•
ologies (object of study, cf, sociology)
isms (ideology, cf. communism)
phrases
•
•
•
forget-me-not
has-been
no-show
acronyms
•
•
sms’es
nimby
TIN-dag 03-02-07
11
Lexicalization “vs”
grammaticalization




Lehmann 2002: e.g. transition N > P is first and
foremost a case of lexicalization with subsequent
grammaticalization
Antilla 1989: grammaticalization involves
lexicalization (e.g. by adding P’s to the lexicon)
Problem: what is in the lexicon?
Brinton & Traugott’s definition of lexicalization:
restricted to items which are “semantically contentful”
(bit problematic in view of their definition of the
lexicon)
TIN-dag 03-02-07
12
Borderline cases

Derivational suffixes
Gmc *līka ‘body’ > Du. –lijk; Eng. –ly etc.;
Lat. ABL mente ‘mind’> It. –mente; Fr. –ment etc.

Phrasal discourse markers:
Eng. y’know, innit (< isn’t it) etc.

Many adverbs
Germ. heute (OHG hiu taguDAT); Eng. today (OE
to dægeDAT)
TIN-dag 03-02-07
13
CASE STUDY
 Epistemic
adverbs
deriving from ‘may/can be/happen’
TIN-dag 03-02-07
14
‘Maybe’ in scandinavian




Swedish kanske < ‘can happen’
Swedish måhända < ‘may happen’
Norwegian kanskje < ‘can happen’
Danish måske < ‘may happen’
TIN-dag 03-02-07
15
‘Maybe’ in other languages







English maybe
Dutch misschien (< ‘may happen’)
French peut-être
Russian možet (byt’) < ‘may (be)’
Serbian – Croatian možda < ‘may that’
Polish może < ‘may’
Lithuanian gal(būt) < ‘may (be’)
TIN-dag 03-02-07
16
Some typical syntactic features
Adverbs of this type may:
 be followed by a subordinate clause
(number of lgs):
Maybe that I’m wrong

violate the V2-rule (in Swedish)
TIN-dag 03-02-07
17
“X that” clauses:
crosslinguistically common
Misschien dat hij komt
Peut-être qu’il vient
Kanske att han kommer

Compare non-phrasal Advs:
Mogelijk dat hij komt
Probablement qu’il vient
Möjligen att han kommer
TIN-dag 03-02-07
18
“X that” clauses: word of
caution

1.
2.
“X-that” data need to be filtered
Matrix ellipsis:
•
I wonder what she has to say? maybe that she is In
love with someone?
No matrix at all:
•
I'm thinking of the Speaker's position as 3rd in line
after the VP to take over if the Prez is incapacitated
or whatever...maybe that she's settling in for the long
haul, and may someday be a candidate for Prez or
VP herself.
TIN-dag 03-02-07
19
Summary
Question raised:
Do the X-that clauses reflect a
grammaticalization process?

TIN-dag 03-02-07
20
V2 violations: Swedish as a V2
language
Vi
äter
alltid
lunch
kl. 12
We
eat
always
lunch
12 o’clock
Alltid
äter
vi
lunch
kl. 12
Always
eat
we
lunch
12 o’clock
Kl. 12
äter
vi
alltid
lunch
12 o’clock
eat
we
always
lunch
Lunch
äter
vi
alltid
kl. 12
Lunch
eat
we
always
12 o´clock
TIN-dag 03-02-07
21
V2 violations: word order with
kanske
Han
har
KANSKE inte
ätit
He
has
maybe
not
eaten
KANSKE har
han
inte
ätit
Maybe
he
not
eaten
KANSKE han
inte
har
ätit
Maybe
he
not
has
eaten
Han
KANSKE inte
har
ätit
He
maybe
has
eaten
has
not
TIN-dag 03-02-07
22
V2 violations: more on word
order with kanske

When both kanske and the Subject precede Vf,
then so does negation marker inte →
subordinate clause order:
Bengt kanske inte känner henne
Bengt maybe not knows her
’Maybe Bengt does not know her’
Kanske vädret inte blir vackert på lördag?
Maybe weather.the not will.be nice on Saturday?
‘Maybe the weather will not be nice on Saturday?’
TIN-dag 03-02-07
23
Summary
Question raised: do kanske’s syntactic
peculiarities reflect a grammaticalization
process?
TIN-dag 03-02-07
24
Etymology: Dutch

Middle Dutch:
misschien, machscien, machgeschien etc.

traces of subject het:
tmachscien sijn siel quam weder ten lichaem
Mer machtscieden daer zijn wel sommighe
onder u […]


WNT: “X-that” rare in MiDu
MiNlW: commonly main clause word
order
TIN-dag 03-02-07
25
Etymology: Swedish
Source: MLG mach-schên ‘may happen’
-> loan word maxan (now obsolete)
-> loan translations kanske, måhända,
kanhända
 SAOB: Older Sw kan ske at ‘can happen
that’ is the source of the adverb

TIN-dag 03-02-07
26
Kan ske as a phrase
thet kan wel skee at en liten hoop
offuerwinner en storan
‘It may well happen that a small lott
conquers a large (lot)’
 thz kunde honom ekke ske
‘That could not happen to him’

TIN-dag 03-02-07
27
More diachrony: changes
involved






Phonetic reduction (Dutch)
Univerbation (Dutch and Swedish)
Decategorialization: verbal inflections lost (Dutch and
Swedish)
Semantic bleaching / generalization: denotes probability
rather than ‘can happen’ (Dutch and Swedish)
Layering: reflections of older stages
•
•
“Xthat” (clearly in Swedish; highly probable for Dutch)
appears in Vf position (Swedish)
Subjectification (change of perspective from sentence
subject to utterance subject)
TIN-dag 03-02-07
28
Sum: grammaticalization or
lexicalization?



Ramat 2001: lexicalization
Andréasson 2002, Brinton & Traugott 2005:
grammaticalization
Brinton & Traugott: not all exx of fusion
(univerbation) are exx of grammaticalization,
only when it yields a (relatively) closed-class
item
•
perhaps: gz; goodbye (< God be with you): no gz
→ modal adverbs form relatively closed class
→ grammaticalization
TIN-dag 03-02-07
29
Diachronic processes revisited
Phonetic reduction: gz and lex
 Univerbation: gz and lex
 Decategorialization: gz and lex
 Semantic bleaching: gz
 Layering: gz
 Subjectification:gz
→ grammaticalization

TIN-dag 03-02-07
30
Lexicalization (as well)?



Lexical items and grammatical items,as well as
open-class items and closed-class items form
a continuum, hence it is difficult to say whether
epistemic adverbs are “lexemes”, and hence
lexicalization
Again: dependent on one’s definition of the
lexicon
Brinton & Traugott: no lexicalization
TIN-dag 03-02-07
31
Concluding remarks
More details about the rise of epistemic adverbs are
necessary
 The rise of epistemic adverbs bears all the hallmarks of
grammaticalization
 It is, however, less obvious that it is not lexicalization
→ Current definitions are inadequate to capture the changes
involved in the rise of adverbs
→ If lexical and grammatical items form a continuum, a strict
demarcation of lexicalization and grammaticalization is
impossible
→ Either the “lexical-grammatical”-continuum is discarded,
or a the strict “lexicalization-grammaticalization” is
abandoned
 More “grey-area” cases need to be examined

TIN-dag 03-02-07
32
THANK YOU

This presentation will soon be
downloadable from:
http://odur.let.rug.nl/~norde/downloadables.htm
TIN-dag 03-02-07
33
Descargar

Dia 1