PowerPointpresentatie

1
Does grammaticalisation need analogy?
Different pathways on the pronoun/agreement
marker-cline gunther.devogelaer_at_ugent.be
2
I. Introduction
syntagmatic vs. paradigmatic aspects of
grammaticalisation
after Hopper Traugott (2003)
paradigmatic axis analogy
syntagmatic axis reanalysis
3
I. Introduction
syntagmatic vs. paradigmatic aspects of
grammaticalisation
Hopper Traugott (200369) reanalysis is the
dominant mechanism driving grammaticalization
paradigmatic axis analogy
1.
syntagmatic axis reanalysis
4
I. Introduction
syntagmatic vs. paradigmatic aspects of
grammaticalisation
Hopper Traugott (200369) reanalysis is the
dominant mechanism driving grammaticalization
paradigmatic axis analogy
2.
syntagmatic axis reanalysis
5
I. Introduction
syntagmatic vs. paradigmatic aspects of
grammaticalisation
reanalysis as the main mechanism? cf. generative
accounts of Lightfoot (1979), Kroch (1989) also
in Timberlake (1977), Harris Campbell
(1995) reanalysis actualisation-model (o.a.
Andersen 2001)
paradigmatic axis
syntagmatic axis reanalysis
6
I. Introduction
reanalysis-based accounts of grammaticalisation
main criticisms

• Haspelmath (1998340) non-manifested
  reanalysis, 700- year periods of competing
  grammars,
• categorial reanalysis implies the existence of
  innate categories (Lehmann 2004, Haspelmath
  t.a. cf. most cognitive accounts of language
  acquisition)
• (cf. also criticism on competence-based models in
  general, e.g. in Wanner 2007)

7
I. Introduction
alternative accounts of grammaticalisation 1.
gradualness
e.g. Haspelmath (1998) ease of production vs.
ease of perception, but this explanation is
highly abstract (p.322)
paradigmatic axis
Grammaticalisation
syntagmatic axis
8
I. Introduction
alternative accounts of grammaticalisation 2.
paradigmatic accounts
paradigmatic axis
e.g. accounts using as the main
mechanism -pragmatic (or semantic)
inferencing -extension (e.g. Heine Kuteva 2005)
syntagmatic axis
9
I. Introduction
alternative accounts of grammaticalisation 2.
paradigmatic accounts
paradigmatic axis analogy
central claim for today
syntagmatic axis
10
I. Introduction
alternative accounts of grammaticalisation 2.
paradigmatic accounts
paradigmatic axis analogy
central claim for today paradigmatic
developments influence syntagmatic ones
syntagmatic axis
11
I. Introduction
alternative accounts of grammaticalisation 2.
paradigmatic accounts

• more precisely
• possible analogical extensions may
• block certain syntagmatic developments
• influence (syntagmatic) grammaticalisation
  pathways

paradigmatic axis analogy
syntagmatic axis
12
I. Introduction
alternative accounts of grammaticalisation 2.
paradigmatic accounts

• more precisely
• possible analogical extensions may
• block certain syntagmatic developments
• influence (syntagmatic) grammaticalisation
  pathways

paradigmatic axis analogy
A
B
syntagmatic axis
13
I. Introduction
alternative accounts of grammaticalisation 2.
paradigmatic accounts
paradigmatic axis analogy
caveats -no claim that this is the only accurate
scenario (syntagmatic accounts may work in other
situations) -analogy is not the only paradigmatic
mechanism
syntagmatic axis
14
II. Subject doubling
marking of the subject through clitic
(optional) strong pronoun/NP
Dutch (Flemish and Brabantic) SAND De Vogelaer
2005, t.a. German (Bavarian) Fuss 2004, Weiss
2005 French (Picardian) Auger 2003 (see Auger
1993, Nadasdi 1995 for Canadian French) Italian
Haiman 1991, Fuss 2004, Poletto 2007
15
II. Subject doubling
examples

• Dutch Wilde gij meerijden? (Flemish,
  Brabantic)
• want2SG you with-ride.INF
• Do you want to ride along?
• German Fahrma mir aaf Minga? (Bavarian)
• French El mer allesro maouaise
  (Picardian, Auger 2003)
• Italian Ti teparli massa (Arsiero
  (NE-It.), Poletto 2007)

16
II. Subject doubling
examples

• Dutch Wilde gij meerijden? (Flemish,
  Brabantic)
• German Fahrma mir aaf Minga? (Bavarian)
• Ride1PL we to Munich
• Are we riding to Munich?
• French El mer allesro maouaise
  (Picardian, Auger 2003)
• Italian Ti teparli massa (Arsiero
  (NE-It.), Poletto 2007)

17
II. Subject doubling
examples

• Dutch Wilde gij meerijden? (Flemish,
  Brabantic)
• German Fahrma mir aaf Minga? (Bavarian)
• French El mer allesro maouaise
• The sea 3SG.FEMbe.FUT bad (Picardian,
  Auger 2003)
• The sea will be bad.
• Italian Ti teparli massa (Arsiero
  (NE-It.), Poletto 2007)

18
II. Subject doubling
examples

• Dutch Wilde gij meerijden? (Flemish,
  Brabantic)
• German Fahrma mir aaf Minga? (Bavarian)
• French El mer allesro maouaise
  (Picardian, Auger 2003)
• Italian Ti teparli massa
  (Arsiero (NE-It.), Poletto 2007)
• You 2SGspeak a lot
• You speak a lot.

19
II. Subject doubling
examples

• Dutch Wilde gij meerijden? (Flemish,
  Brabantic)
• German Fahrma mir aaf Minga? (Bavarian)
• French El mer allesro maouaise
  (Picardian, Auger 2003)
• Italian Ti teparli massa
  (Arsiero (NE-It.), Poletto 2007)
• You 2SGspeak a lot
• You speak a lot.
• Givón (1976151), Fuss (20049)
• the rise of subject doubling
  grammaticalization through reanalysis

20
II. Subject doubling
syntagmatic aspects of cliticisation pronoun gt
clitic gt AGR
21
II. Subject doubling
syntagmatic aspects of cliticisation pronoun gt
clitic gt AGR
test multi-representation / doubling (cf.
Siewierska 1999, Corbett 2003185-188)
22
II. Subject doubling
syntagmatic aspects of cliticisation pronoun gt
clitic gt AGR

• Dutch
• Wilde gij meerijden? (Flemish, Brabantic)
• German
• Fahrma mir aaf Minga? (Bavarian)
• French
• El mer allesro maouaise (Picardian,
  Auger 2003)
• Italian
• Ti teparli massa (Arsiero (NE-It.),
  Poletto 2007)

23
II. Subject doubling
syntagmatic aspects of cliticisation pronoun gt
clitic gt AGR
test (erstwhile) clitics adopt non-referential
uses (cf. Haiman 1991, Evans 1999, Corbett
2003184-185)
24
II. Subject doubling
syntagmatic aspects of cliticisation pronoun gt
clitic gt AGR Dutch
French German
Italian NO YES e.g. Parsonne
i-nporoait vnir llértcheure.
nobody 3SG-NEG-could come him search Nobody
could pick him up. Le fomne che
le-neta le scale. the women who 3PL-clean
the stair The women who clean the stairs.
25
II. Subject doubling
syntagmatic aspects of cliticisation pronoun gt
clitic gt AGR Dutch
French German
Italian NO YES Niemand wilt-ie
meerijden. Nobody want-3SG with-ride Nobod
y wants to ride along. De vrouwen die
ze-willen meerijden. The women who 3SG-want
with-ride The women who want to ride along.
26
II. Subject doubling
syntagmatic aspects of cliticisation pronoun gt
clitic gt AGR Dutch
French German
Italian NO YES However, in
East-Flemish Niemand weet-jij dat! with jij
strong pronoun Nobody knows-he that! Nobody
knows that.
27
II. Subject doubling
syntagmatic aspects of cliticisation pronoun gt
clitic gt AGR Dutch
French German
Italian NO YES However, in
East-Flemish Niemand weet-jij dat! with jij
strong pronoun Nobody knows-he that! Nobody
knows that.
De Vogelaer Devos (2007) not clitic gt AGR,
but topic marking (see Greenberg 1991302-303
for vaguely related changes)
28
II. Subject doubling
from clitic to inflectional ending diverging
grammaticalisation pathways Dutch,
German pronoun gt clitic gt in Dutch clitic
doubling gt topic marking French, Italian
pronoun gt clitic gt AGR
29
III. Analogical pathways
30
III. Analogical pathways
Some distributional differences are not
captured by the pronoun gt AGR-cline e.g. rather
idiosyncratic distribution of some
clitics Brabantic 2SG/PL de only occurs in main
clauses with inverted word order Frisian 2SG
ste only occurs as a clitic to main clause
verbs East-Flemish 1PL men only occurs as a
clitic to main clause verbs Bavarian Forest 1PL
ma occurs in subclauses, but only following
auxiliaries (Kurzverben) (almost) all West
Germanic clitics only combine with strong
pronouns, not with weak pronouns, nor with
demonstratives
31
III. Analogical pathways
Kathol (2001) on cliticisation in West
Germanic -innovations in the position enclitic
to the verb spread through analogy over
different syntactic environments -pathways
depend on (formal) similarities between
sentence types (hence no UG is
involved) so enclitic to V gt enclitic to C gt
regular order gt subclause V
32
III. Analogical pathways

• Kathol (2001) on cliticisation in West Germanic
• -innovations in the position enclitic to the
  verb spread through analogy over different
  syntactic environments
• -pathways depend on (formal) similarities
  between
• sentence types (hence no UG is involved)
• so enclitic to V gt enclitic to C gt regular
  order gt subclause V
• many applications
• old agreement markers complementizer
  agreement (cf. Kathol 2001 De Vogelaer, Devos
  Van der Auwera 2006 Zwart 2006 Van der Auwera,
  Schalley De Vogelaer t.a. De Vogelaer Van
  der Auwera t.a.)
• emergence of clitics (Kathol 2001)
• innovative pronouns (cf. Verdenius 1923, among
  many others)
• innovative agreement markers (cf. Van Haeringen
  1951, Kramer 1996)

33
III. Analogical pathways
cliticisation in Dutch and German examples
enclitic to V gt enclitic to C gt regular order gt
subclause V slightly adapted A. enclitic to V
gt enclitic to C gt regular order gt subclause
V OR B. gt proclitic to V gt
subclause
34
III. Analogical pathways
cliticisation in Dutch and German examples
enclitic to V gt enclitic to C gt regular order gt
subclause V
e.g. A-scenario B-scenario Fon-ma (mia) aaf
Minga? Ga-je (gij) naar Gent? drive-1PL
we to Munich go-2SG you to
Ghent Are we driving to Munich? Are you
going to Ghent?
35
III. Analogical pathways
cliticisation in Dutch and German examples
enclitic to V gt enclitic to C gt regular order gt
subclause V
e.g. A-scenario B-scenario Fon-ma (mia) aaf
Minga? Ga-je (gij) naar Gent? Dass-ma (mia)
aaf M. fon. Da-je (gij) naar Gent gaat.
36
III. Analogical pathways
cliticisation in Dutch and German examples
enclitic to V gt enclitic to C gt regular order gt
subclause V
e.g. A-scenario B-scenario Fon-ma (mia) aaf
Minga? Ga-je (gij) naar Gent? Dass-ma (mia)
aaf M. fon. Da-je (gij) naar Gent gaat. (Mia)
fom-ma aaf M. gtlt Je-gaat (gij) naar Gent. e.g.
Bavaria, East Flemish e.g. West Flemish
37
III. Analogical pathways
cliticisation in Dutch and German examples
enclitic to V gt enclitic to C gt regular order gt
subclause V
e.g. A-scenario B-scenario Fon-ma (mia) aaf
Minga? Ga-je (gij) naar Gent? Dass-ma (mia)
aaf M. fon. Da-je (gij) naar Gent gaat. (Mia)
fom-ma aaf M. gtlt Je-gaat (gij) naar Gent.
Dass-ma (mia) aaf M. gtlt Da-je (gij) naar Gent
ga-je fom-ma (or je-gaat)
38
III. Analogical pathways
cliticisation in Dutch and German examples
enclitic to V gt enclitic to C gt regular order gt
subclause V
e.g. A-scenario B-scenario Fon-ma (mia) aaf
Minga? Ga-je (gij) naar Gent? Dass-ma (mia)
aaf M. fon. Da-je (gij) naar Gent gaat. (Mia)
fom-ma aaf M. gtlt Je-gaat (gij) naar Gent.
Dass-ma (mia) aaf M. gtlt Da-je (gij) naar Gent
ga-je fom-ma
39
III. Analogical pathways
cliticisation in Dutch and German examples
enclitic to V gt enclitic to C gt regular order gt
subclause V slightly adapted A. enclitic to V
gt enclitic to C gt regular order gt subclause
V OR B. gt proclitic to V gt
subclause e.g. A. fonma mia gt wemma mia fon gt
mia fomma (gt wemma mia fomma) (Bavarian) B.
gaje gij gt daje gij gt jegaat gij gt da-je gij
gaje (West Flemish)
40
III. Analogical pathways
cliticisation in French and Italian extension
to subclauses is not even an issue
41
III. Analogical pathways
cliticisation in French and Italian extension
to subclauses is not even an issue e.g. French
lui il a eu de la chance je peut vous dire
que lui il a eu de la chance Haiman (1991146)
on Northern Italian dialects All finite verbs
must occur with at least one prefixed clitic
in statements (this clitic is usually a
subject clitic, which may be dropped in the
presence of an object clitic or negative
particle)
42
III. Analogical pathways
Germanic dialects with Romance word order
Walser Deutsch (Northern Italy) endsch
andre ber-wer ... we others carry-1PL We
carry ... (Nübling 1992257) der manut wa
si-wer döi gsi the month that be-1PL there be
en (During) the month that we were
there,... (Dal Negro 2004168)
43
IV. Explanation
cliticisation in German/Dutch vs.
French/Italian explanation (// Kathol 2001)
SOV-subclauses are unlikely targets for
analogical extension from innovations in the
position enclitic to the verb
44
V. Recapitulation
cf. central claim possible analogical
extensions may 1. block syntagmatic
developments 2. influence grammaticalisation
pathways
45
V. Recapitulation
1. Blocking of syntagmatic changes
Dutch, German French, Italian
pronoun gt clitic gt AGR
pronoun gt clitic gt AGR
46
V. Recapitulation
2. Influence on grammaticalisation pathways
Dutch, German French, Italian
pronoun gt clitic gt AGR
clitic doubling gt topic marking
pronoun gt clitic gt AGR
47
V. Recapitulation
Thank you! Comments gunther.devogelaer_at_ugent.be
48
II. Subject doubling
syntagmatic aspects of cliticisation pronoun gt
clitic gt AGR Dutch
French German
Italian EXCEPTIONS? 2enk. st Du weet-st da-st
slim bi-st. (Frisian,) 2mv. ts Ob-ts es
noch Minga kumm-ts (Bavarian,) 1mw. ma(r)
... wos mar wöl-mar (Bavarian Forest, only
following auxiliaries)
49
II. Subject doubling
syntagmatic aspects of cliticisation pronoun gt
clitic gt AGR Dutch
French German
Italian EXCEPTIONS? ? still many differences
with ordinary AGR! 1. st, ts and mar are
referential, and hence not fully integrated in
the verbal paradigm 2. they also occur
(obligatorily) following complementisers 3. they
only occur as enclitics to the verb, and at least
st and ts may have originated as formal
variants of the original AGR 4. they are rare
and occur in incomplete paradigms (5. other
criteria apply to 3rd person exclusively)