investigating the role of speech planning in the production...
TRANSCRIPT
Investigating the role of speech planning in the production of
sociolinguistic variation
Laurel MacKenzie Hilary Wynne NYU OxfordExAPP • 23 September 2016
2
Laurel MacKenzie NYU
[email protected]@laurel_mack
Hilary WynneOxford
[email protected]@Hilary_Wynne
variationist sociolinguistics
psycholinguistics of language production
why & how does the language we produce take the shape that it does?
3
outline
• background: sociolinguistic variation; planning
• intersection: how might planning affect variation?– case study
• methods: how can we explore this intersection experimentally?
4
background
5
sociolinguistic variation
two or more ways of saying the same thing(e.g. Weiner & Labov, 1983)
• social• linguistic
6
what drives speakers to choose one or the other option?
psycholinguistic factors
• working memory
• production planning
• priming
(Tamminga, MacKenzie, and Embick, forthcoming)
7
psycholinguistic factors
• working memory
• production planning
• priming
(Tamminga, MacKenzie, and Embick, forthcoming)
8
incremental planning
planning || production
[����] [����] [����] [����]
incremental planning
planning || production
[����] [����] [����] [����]
incremental planning
planning || production
[����] [����] [����] [����]
incremental planning
speakers don’t mentally form a plan of an utterance in its entirety before producing it:
(Ferreira & Swets, 2002)
12
incremental planning
rather, we plan out the later components of an utterance as we produce the earlier ones:
(Ferreira & Swets, 2002)
14
incremental planning
this advance planning can be compromised, e.g. by cognitive load or complexity:[����] [����] [����] [����]
…which has consequences for the way our language comes out.
(Ferreira, 1991; V. Wagner et al., 2010)
17
intersection
19
linguistic conditioning
many sociolinguistic variables are conditioned by surrounding elements of language: � � � � � � � �
(Guy & Boberg, 1997)
[t] Ø~
20
linguistic conditioning
many sociolinguistic variables are conditioned by surrounding elements of language: � � � � � � � �
(Guy & Boberg, 1997)
21
[t] Ø~
sibilant
linguistic conditioning
many sociolinguistic variables are conditioned by surrounding elements of language: � � � � � � � �
(Bailey 2015)
[��] [�n]~
22
linguistic conditioning
many sociolinguistic variables are conditioned by surrounding elements of language: � � � � � � � �
(Bailey 2015)
[��] [�n]~
23
velar
planning failure
many sociolinguistic variables are conditioned by surrounding elements of language:[� � � ] � � � � �
but if planning past the variable fails,
24
planning failure
many sociolinguistic variables are conditioned by surrounding elements of language:[� � � ] [ � � � � � ]but if planning past the variable fails,surrounding conditioning should fail to apply, resulting in a different distribution of variants.
25
planning failure
many sociolinguistic variables are conditioned by surrounding elements of language:[� � � ] [ � � � � � ]but if planning past the variable fails,surrounding conditioning should fail to apply, resulting in a different distribution of variants.
(M. Wagner 2011, Tanner et al. 2015)
26
case study
28
variable contraction of is
Yeah, Salzburg’s nice. Austria’s nice. Europe is nice!
29
30
subject length
Salzburg’s nice 1
The real estate out here’s been pretty good 4
About the only thing I can do mechanically with a– a car is put gas in it 12
0.0
0.2
0.4
0.6
0.8
1.0
5 10 15words in subject
short
long
'has', N = 431
0.0
0.2
0.4
0.6
0.8
1.0
5 10 15words in subject
short
long
'is', N = 1172
0.0
0.2
0.4
0.6
0.8
1.0
5 10words in subject
short
long
'will', N = 392
31
contracted
full
(MacKenzie,2012,2013)
robust subject length effect
subject length effect as planning?
[About the only thing I can do mechanically with a– a car][is put gas in it]
32
methods
33
previous work
The enthusiastic band pleased the very impatient crowd.
The pianist in the band pleased the very impatient crowd.
The man who started the band pleased the very impatient crowd.
• memorize the target sentence• produce it on command
(Ferreira, 1991)
34
picture description task
goals:• elicit copula sentences with subjects of varying
length and naturally variable contractionse.g. the cat {is ~ ’s} is on the left
the cat on the left {is ~ ’s} blue
• avoid expletive/pronoun subjects, which contract nearly categorically
• replicate the subject length effect demonstrated in corpora
35
picture description task
• participants were tested individually• participant • assistant (“Stooge”)• laptop + whiteboard / cutout objects• task: describe to the assistant how the objects
are arranged on your screen.
36
picture description task
37
picture description task
38
subjects
• 12 native speakers of British English• monolingual• undergraduate Linguistics/English Language
students at U of Manchester
39
successful elicitations
The cat’s on the left, lemon’s in the middle, fork is on the – yeah, perfect! Yeah.
Uh, there’s a football on the right, and then a football in the middle, and a football on the left, and the one on the left is blue.
40
unsuccessful elicitations
One cat on the left, one lemon in the middle, and a hand on the right.
We have one pink lemon, right in the middle. One the right we have one lemon, as well, and on the left we have another lemon.
41
Plan B: spot-the-differences task
42
43
Welcome to the speech study.First, study this scene and describe what you see for the recording.
When you’ve finished describing the scene, press the right arrow on your keyboard to go on.
44
Thebicycleinfrontoftheyellowhouse{is}upsidedown.
Thebicycle{is~‘s}upsidedown.
In sum…
• eliciting variable structures is difficult when variants are orthographically encoded! — more methodological development needed
• points of commonality between psycholing & socioling = ripe for further collaboration!
45
thank you!�
46
47
Laurel MacKenzie NYU
[email protected]@laurel_mack
Hilary WynneOxford
[email protected]@Hilary_Wynne