English Data Sources
V2 loss revisited:
a quantitative comparison
of English and French
Anthony Kroch and Beatrice Santorini
University of Pennsylvania
May 2009
www.ling.upenn.edu/~kroch/handouts/iwies.pdf
French Data Source
Two Early Modern French texts:
• Étaples Bible translation
• letters of Marguerite de Valois
•
Ann Taylor, Anthony Warner, Susan Pintzuk, and Frank Beths.
York-Toronto-Helsinki Parsed Corpus of Old English Prose. Oxford
Text Archive, first edition, 2003.
•
Anthony Kroch, Beatrice Santorini, and Lauren Delfs. PennHelsinki Parsed Corpus of Early Modern English. CD-ROM, first
edition, 2004.
•
Ann Taylor, Arja Nurmi, Anthony Warner, Susan Pintzuk, and
Terttu Nevalainen. Parsed Corpus of Early English
Correspondence. Oxford Text Archive, first edition, 2006.
14
12
% Topicalized
Five Middle French texts:
• Chroniques de Froissart
• La Prise d’Alexandrie
• Cent N. N. Anonymes
• XV Joies de Mariage
• Commynes
Anthony Kroch and Ann Taylor. Penn-Helsinki Parsed Corpus of
Middle English, second edition. CD-ROM, second edition, 2000.
Decline of direct object topicalization in English
France Martineau, et. al. Corpus MCVF, “Modéliser le
changement: les voies du français.” University of Ottawa.
Six Old French texts:
• Brendan
• Le chanson de Roland
• Yvain
• Nicolette et Aucassin
• La queste de la sainte Graal
• Le somme le roi
•
10
8
6
4
2
0
OE
(Early)
OE (Late)
11501250
12501350
13501420
Date
14201500
15001569
15701639
16391710
The history of topicalization in English
(Speyer 2008)
Frequency of direct object topicalization
in modern spoken Dutch (Bouma 2008)
• Why does topicalization decline in Middle English
Table 4.2: Summary of Vorfeld occupation of arguments.
yes
Argument
43 523
3 418
38
subject
direct object
indirect object
but not disappear? If the change a parametric one, it
should go to completion. Otherwise, topicalization, a
clear case of stylistic variation might be expected to
be stable in frequency over time.
Prop est (%)
Vorfeld
no
pt
18 597
20 432
815
70.1
14.3
4.5
• This question has answer in the specific interaction
between parametric settings and stylistic variation in
the history of English.
Decline of direct object topicalization
by subject type
Evolution of adverb fronting
90
30
80
25
70
temporal adverbs
60
20
%
%
50
pronoun subjects
15
40
10
locative adverbs
30
5
full DP subjects
20
10
0
oe1/2
oe3/4
me1
me2
me3
period
me4
eme1
eme2
eme3
0
oe2
oe3
me1
me2
me3
period
me4
eme1
eme2
eme3
Correlation between frequencies of object topicalization
and of V2 in Middle English texts (Wallenberg 2007)
100
90
Distribution of subject types in a corpus
of topicalized and non-topicalized
sentences in natural speech
80
personal pronoun
140
46.4
70
% V2
60
50
demonstrative pronoun
20
6.6
full noun phrase
142
47.0
Subject type in sentences with in situ objects
edvern
40
30
personal pronoun
181
90.5%
20
10
0
0
5
10
15
20
25
demonstrative pronoun
2
1%
full noun phrase
17
8.5 %
Subject type in sentences with topicalized objects
% Full DP Topicalization
Clash avoidance
Translating German topicalized arguments into
English in three modern German novels
[by Böll, Dürrenmatt and Grass]
• The type of topicalization that declines:
(1) The nèwspaper Jóhn read; the nòvel Máry did.
(Compare: The nèwspaper read Jóhn.)
• The type of topicalization that doesn’t:
(2) The nèwspaper I réad; the nòvel I dídn’t.
Topicalized to topicalized:
G: Mahlkes Haupt bedeckte dieser Hut besonders peinlich.
E: On Mahlke’s head this hat made a particularly painful
impression.
Topicalized to non-topicalized:
G: Zu den sechs kamen noch drei weitere.
E: Three others joined these six in the afternoon.
Accent placement and topicalization frequencies in
translating German topicalized arguments into English
Rate of V2 loss in English with topicalized
objects and PPs
80
no topicalization
in the English
70
accent elsewhere
31
31
topicalized objects
60
50
%
topicalization in
the English
translation
2nd accent on
the German
subject
0
0
topicalized PPs
40
30
25
25
100
100
20
10
0
me1
me2
me3
me4
eme1
eme2
eme3
period
The V2 constraint in Old English:
the pronoun exception
Was Old English a V2 language?
(1) æt hus hæfdon Romane to æm anum tacne geworht.
(2) Ælc yfel he mæg don.
(3) in agen geleafa e hæf gehæledne.
(4) & seofon ærendracan he him hæfde to asend.
Phrase structure of an English V2 clause
Unambiguous V3 clauses with topicalized objects
CP
XPi
æt hus
C'
C0
(1) æne
se geatweard let in
that-one the doorkeeper let in
(cowsgosp, Jn_[WSCp]:10.3.6596)
TP
T'
?
T0
hæfdon
(2) and him se inno eac geopenode ongean
and him the heart also opened
again
(coælive, +ALS_[Vincent]:170.7907)
vP
DP
Romane
v'
v0
VP
V0
geworht
Frequency of unambiguous V3 clauses
against all particle verb cases
Decline of silent expletives in existential sentences
(Williams 2000)
full DP subjects
pronoun subjects
74
6
74
6
20
45
(1) ach nu is sum wummon e nalde
for nan ing wilni fule
but now is sum woman who wouldn’t for no thing want filth
“but now there is some woman who wouln’t for anything
want filth”
Ancrene Riwle II.49.365
20
45
.21
.88
0.21
0.88
V2 cases
V3 cases
frequency V3
(2) a com er an helendis Mon
then came EXPL a foreign man
“then there came a foreigner”
Lambeth Homilies 79-496
Diachronic evolution of silent expletives in existential
sentences in Middle English
period
ME1(1150-1250)
ME2(1250-1350)
silent
expletive
frequency
silent
1
15
0.94
74
74
74
overt there
74
0
74
ME3(1350-1420)
5
ME4(1420-1500)
4
74
74
0
74
3
74
0
74
––
Loss of V2 and direct object
topicalization in French
74
0.38
4
0.00
Object topicalization with V2 in Old
and Middle French
(1)l'estreu
li tint sun uncle Guinemer
the stirrup him held his uncle Guinemer
Roland 27.329
(2)messe e matines ad li reis escultet
mass and matins has the king heard
Roland 11.139
(3)si grand paour avoit que a pou qu'il ne mouroit
so great fear he-had that at little that he ne died
CNNA 51.488
Object topicalization with V2 and pronoun
subjects in Old and Middle French
(1)Espaigne vus durat
il en fiet
Spain
you will-give he in fief
Roland, 36.446
(2)chars avoient ils
assés
meat had
they enough
Froissart, 135.569
(3)une chose ont-ilz asez honneste
one thing have-they enough honest
Commynes, 120.1634
Object topicalization with V3 and pronoun
subjects in Middle French
(1)aultre remyde je ny
voy
other remedy I not there see
XV Joies, 111v.1209
(2)deux chose je diray
de luy
two things I will-say of him
Commynes, 38.478
(3)nul enfant il not
onques eu de sa fenme
no child he not-had ever
had of his wife
Froissart, 462.6477
Adverb preposing with V2 in Old
and Middle French
Exceptional cases of V3 with pronoun subjects
in Old French
(1)ceste franchise il tient de Deiu
this privilege he has from God
Somme, 80.268
(2) le offrande qum li portout, tut as povres il lenhortout, jo
celoue en mes burses
the offering that one him brought, all to the poor he
intended, I hid in my purse
Brendan, 64.923
Adverb preposing with V3 in Old
and Middle French
(1)Er
matin sedeit li emperere suz lumbre
yesterday morning sat
the emperor under the shade
Roland 29.363
(1)unkes nuls hom ne vit tel ajustee.
never no man not saw such joust
Roland 238.3299
(2)or est ele bien venue
now is she welcome
Yvain 43.1440
(2)premierement il devient taverniers
first
he becomes innkeeper
Somme 47.306
(3)ben lentendit
il arcevesques Turpin
well it understands the archebishop Turpin
Roland 96.1237
(3)doucement li oisel chantoient
sweetly
the birds sang
Yvain 15.451
Frequency of V2 in main clauses with
topicalized XPs in Old French
XVS
XSV
XVS
XSV
pronoun subject full DP subject
176 [.97]
45 [.80]
11
5
Topicalized objects
pronoun subject
54 [.60]
37
Topicalized PPs
full DP subject
346 [.80]
87
Frequency of V2 in main clauses with
topicalized XPs in Middle French
XVS
XSV
XVS
XSV
pronoun subject full DP subject
[.45]
64 [.93]
21[.45]
26
5
Topicalized objects
pronoun subject
[.12]
58[.12]
438
Topicalized PPs
full DP subject
422 [.61]
274
Frequency of V2 in main clauses with
topicalized XPs in Old French (no Somme)
XVS
XSV
XVS
XSV
pronoun subject full DP subject
146 [.97]
40 [.98]
1
4
Topicalized objects
pronoun subject
40 [.67]
20
Topicalized PPs
full DP subject
286 [.78]
81
Frequency of XP preposing in main clauses
PP
DP
temporal adv
subjects
other adv
DP object
Old French
.35[1256]
.66[317]
.56[639]
.26[701]
Middle French
.22[3212]
.63[601]
.53[1888]
.05[1378]
PP
pronoun temporal adv
subjects
other adv
DP object
.19[490]
.25[129]
.22[449]
.11[500]
.31[1613]
.54[359]
.28[1247]
.03[1431]
Rise of clitic left-dislocation and loss of
topicalization (Priestley 1955)
(1)... de priere aide li
font les dames
of prayer help him do the ladies
Yvain, 137.4788
(2)... puis si chevalchet od sa grant ost li ber
then so rides
with his great army the baron
Roland, 179.2438
Adjunction to CP in French
Adjunction to CP in Modern French / English
(3)... aussi telle oppinion tiennent les Angloys
also such opinion hold
the English
Commynes, 26.279
Exceptional case of adjunction of an object topic
to CP in Old French
(1)Jean sûr elle/Marie laimera
bien (mais Pierre pas)
John sure she
him-will-like well
(2)sûr Jean elle/Marie laimera
bien (mais Pierre pas)
(3)sure John you will like (him) (but you wont like Bill)
(4)John sure you will like *(him)
la dame, por ce quele plore, prie li rois de remenoir
the lady, for that she cried asked the king to return
Yvain, 81.2812
The temporal evolution of V2 with full DP
subjects for all types of preposed XP
Old French
Middle French
Modern French
sentences with sentences with
an auxiliary verb a single verb
0.83 [2163]
0.85 [218]
0.69 [402]
0.70 [3633]
0.27 [33]
0.22 [160]
Old French
Middle French
Modern French
frequency of frequency Romance +
Germanic of Romance Germanic
inversion
inversion
inversion
0.36 [78]
0.86 [186]
Old French
0.50 [108]
Middle French
0.32 [127]
0.37 [149]
0.69 [276]
Modern
0.03 [1]
0.24 [8]
0.27 [9]
Clitic left dislocation in Modern French
An independence result
Romance +
Germanic
inversion
0.86
0.69
0.27
The temporal evolution of Germanic and
Romance inversion in V2 sentences with
topicalized XPs and full DP subjects
sentences with
a single verb
0.83
0.70
0.22
(1)Le Figaroi, Jean lei lit
tous les jours.
The Figaro John it reads every day
(2)Ma femmei ellei travaille à la Bibliothèque Nationale.
My wife
she works at the library
national
The temporal evolution of subject and object left
dislocation frequencies *
frequency of frequency of number of
subject left object left
matrix
dislocation dislocation
clauses
Old French
2.6E-03
2.2E-03
12022
Middle French
3.8E-03
1.8E-03
24634
Modern
2.8E-02
4.3E-03
3514
Degree of intonational independence of clitic left
dislocation in Old French and later
# object clitic
# material to
left dislocations the right of LFD
Old French
Middle French
Modern French
14
36
10
7 [.50]
1 [.03]
2 [.20]
li duze peri, por ço quil
laiment tant, desfi pro lesi ci
the 12 peers for that they him-love so defy I them here
Roland, 24.303
Cleft sentences in Modern French
(1)Cest Le Figaroi que Jean lit
ti tous les jours.
Its The Figaro that John reads every day
(2)Cest ma femmei qui ti travaille à la BN.
Its my wife
that works at the BN
(3)Il y a un ani quelle travaille à la BN ti.
Its one year that-she works at the BN
The temporal evolution of cleft sentence
frequencies
frequency of frequency of number of
subject and
temporal
matrix
object clefts
clefts
clauses
Old French
1.2E-03
2.5E-04
12022
Middle French
4.1E-04
6.1E-04
24634
Modern
5.6E-04
5.4E-03
3514
Frequency of topicalization in sentences
with full noun phrase subjects
0.800
0.700
V2 and topicalization in the Tyndale
and Étaples Bibles
Tyndale
Étaples
0.600
0.500
0.400
0.300
0.200
0.100
0.000
FULL DP
Frequency of topicalization in sentences
with pronoun subjects
ALL DP
PP
LOCATIVE
TEMPORAL
OTHER ADV
Frequency of V2 word order in
sentences with full noun phrase subjects
1.000
0.800
0.700
0.900
Tyndale
Étaples
Tyndale all pro
0.800
Étaples overt pro
0.600
Étaples covert pro
0.700
0.600
0.500
0.500
0.400
0.400
0.300
0.300
0.200
0.200
0.100
0.100
0.000
0.000
FULL DP
ALL DP
PP
LOCATIVE
TEMPORAL
OTHER ADV
FULL DP
ALL DP
PP
LOCATIVE
TEMPORAL
OTHER ADV
Frequency of V2 word order in
sentences with pronoun subjects
1.000
0.900
0.800
0.700
13
68
3
3
Tyndale all pro
13
Étaples overt pro
Étaples covert pro
Finis
85
0.600
0.500
0.400
63
19
0.300
83
30
0.200
50
60
0.100
79
57
0
0.000
FULL DP
ALL DP
PP
LOCATIVE
0
81
24
TEMPORAL
OTHER ADV
The Constant Rate Effect
• In cases of parametric variation, the variation resolves
itself in the same way in all linguistic contexts, with one
option winning out over the other.
• Furthermore, the rate at which the change progresses is
the same in all contexts.
A grammar competition graph
This S-shape will only arise if:
100
• The speech community is sensitive to the relative frequencies
of the two forms.
75
• Individuals, in aiming to reproduce this frequency, miss slightly
in favor of the new form; that is, the new form has a
(constant) advantage in usage even when its absolute
frequency is low.
50
25
0
0
10
20
30
40
50
60
70
80
90
A few references
•Galves, C. et al. 2005. The change in clitic placement from Classical to
Modern European Portuguese. J. of Portuguese Linguistics, 4:39–67.
•Kroch, A. 1989. Reflexes of grammar in patterns of language
change. Language Variation and Change,1:199–244.
•Kroch, A. 1994. Morphosyntactic variation. CLS 30, vol. 2,180–201.
•Pancheva, R. The rise and fall of second-position clitics. Natural
Language and Linguistic Theory, 23:103–167.
•Pintzuk, S. 1999. Phrase Structures in Competition. Garland, New York.
•Santorini, B. 1993. The rate of phrase structure change in the history
of Yiddish. Language Variation and Change, 5:257–283.
•Speyer, A. 2007. Topicalization and clash avoidance. U. Penn. dissertation.
•Wallenberg, J. 2008. The decline of early English object clitics. LSA
Annual Meeting.
100
• In the model underlying the graph on the preceding slide, the
new form is always produced when attempted but attempts
to produce the old form occasionally fail and lead to resetting
the sentence generator and trying again.