speech priming: evidence for rate persistence in unscripted speech

15
This article was downloaded by: [University of Illinois Chicago] On: 28 October 2014, At: 11:56 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK Language and Cognitive Processes Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/plcp20 Speech priming: Evidence for rate persistence in unscripted speech Melissa K. Jungers a & Julie M. Hupp a a The Ohio State University – Newark , Newark, OH, USA Published online: 03 Apr 2009. To cite this article: Melissa K. Jungers & Julie M. Hupp (2009) Speech priming: Evidence for rate persistence in unscripted speech, Language and Cognitive Processes, 24:4, 611-624, DOI: 10.1080/01690960802602241 To link to this article: http://dx.doi.org/10.1080/01690960802602241 PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/ terms-and-conditions

Upload: julie-m

Post on 28-Feb-2017

215 views

Category:

Documents


2 download

TRANSCRIPT

Page 1: Speech priming: Evidence for rate persistence in unscripted speech

This article was downloaded by: [University of Illinois Chicago]On: 28 October 2014, At: 11:56Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registeredoffice: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Language and Cognitive ProcessesPublication details, including instructions for authors andsubscription information:http://www.tandfonline.com/loi/plcp20

Speech priming: Evidence for ratepersistence in unscripted speechMelissa K. Jungers a & Julie M. Hupp aa The Ohio State University – Newark , Newark, OH, USAPublished online: 03 Apr 2009.

To cite this article: Melissa K. Jungers & Julie M. Hupp (2009) Speech priming: Evidence forrate persistence in unscripted speech, Language and Cognitive Processes, 24:4, 611-624, DOI:10.1080/01690960802602241

To link to this article: http://dx.doi.org/10.1080/01690960802602241

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the“Content”) contained in the publications on our platform. However, Taylor & Francis,our agents, and our licensors make no representations or warranties whatsoever as tothe accuracy, completeness, or suitability for any purpose of the Content. Any opinionsand views expressed in this publication are the opinions and views of the authors,and are not the views of or endorsed by Taylor & Francis. The accuracy of the Contentshould not be relied upon and should be independently verified with primary sourcesof information. Taylor and Francis shall not be liable for any losses, actions, claims,proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to orarising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Anysubstantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,systematic supply, or distribution in any form to anyone is expressly forbidden. Terms& Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Page 2: Speech priming: Evidence for rate persistence in unscripted speech

Speech priming: Evidence for rate persistence in

unscripted speech

Melissa K. Jungers and Julie M. HuppThe Ohio State University � Newark, Newark, OH, USA

Previous research has shown evidence for priming of rate in scripted speech.Two experiments examined the persistence of rate in production of unscriptedpicture descriptions. In Experiment 1, speakers heard and repeated primingsentences presented at a fast or slow rate and in a passive or active form.Speakers then described a new picture. The timing of their productionsreflected the timing of the prime. In Experiment 2, participants heard but didnot repeat the priming sentences. As in Experiment 1, their picture descriptionsreflected the rate of the priming sentences. These experiments demonstratepersistence of rate, an acoustic dimension related to prosody, in unscriptedspeech. The persistence in timing may provide a social advantage and help tocoordinate conversation.

Keywords: Prosody; speech production.

There is a long tradition in the speech literature of examining not only what

is said, but also the way it is said (Ferreira, 1993; Lehiste, 1973). Acoustic

variations in stress, pitch, and timing inform the listener of sentence meaning

beyond the word level and may influence the listener’s future productions

(Jungers, Palmer, & Speer, 2002). The goal of the current study is to examine

whether listeners persist in the rate of previously heard sentences when they

produce unscripted picture descriptions.

Prosody has been described as the suprasegmental features of speech such

as pitch, timing, and loudness (Cutler, Dahan, & Donselaar, 1997). Prosodic

emphasis influences the interpretation of sentence meaning (Speer, Crowder,

& Thomas, 1993), and intonation helps listeners disambiguate the meaning

Correspondence should be addressed to Melissa Jungers, Ohio State University � Newark,

1179 University Dr., Newark, Ohio 43055, USA. E-mail: [email protected]

We would like to thank Kristin Vanover, Whitney Sims, and Ashley Shaw for their assistance

in collecting data.

LANGUAGE AND COGNITIVE PROCESSES

2009, 24 (4), 611�624

# 2009 Psychology Press, an imprint of the Taylor & Francis Group, an Informa business

http://www.psypress.com/lcp DOI: 10.1080/01690960802602241

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s C

hica

go]

at 1

1:56

28

Oct

ober

201

4

Page 3: Speech priming: Evidence for rate persistence in unscripted speech

of ambiguous sentences (Lehiste, 1973; Lehiste, Olive, & Streeter, 1976; Price,

Ostendorf, Shattuck-Hufnagel, & Fong, 1991). The placement and duration

of pauses provides another perceptual cue to sentence meaning; speakers’

pause patterns tend to correlate with the syntactic structure of a sentence,with longer pauses near important structural boundaries (Lehiste et al.,

1976). Although syntax influences prosody, there is a hierarchical structure

for prosody that is distinct from the syntactic representation (Beckman,

1996; Ferreira, 1993; Warren, 1996). Shattuck-Hufnagel and Turk (1996)

propose that syntax is just one of many factors influencing prosody,

including semantics, pragmatics, length, and rate. For example, when

speakers are asked to produce sentences at a slow rate, they show pauses

at lower levels in the prosodic hierarchy than when producing fast sentences(Gee & Grosjean, 1983).

Prosody and other non-verbal cues are involved in communication. What

happens to these extralinguistic features once an utterance is finished? The

approach of normalisation assumes that listeners ignore the acoustic details

in order to understand the message. On this account, listeners transform the

physical speech signal into a standard representation of the sentence devoid

of prosodic details (Pisoni, 1997). This approach is plausible because

listeners can comprehend highly variable speech that differs in pitch rangeor dialect. According to this approach, the acoustic details of a production

will not be retained.

However, several studies suggest that acoustic features of speech are

incorporated in memory for language. Sentences are recognised more

accurately when they are presented with the same prosody at learning and

test (Speer et al., 1993). Listeners use extralinguistic information, including

talker’s speaking rate, to accurately identify previously presented words

(Bradlow, Nygaard, & Pisoni, 1999). The rate of production affects listeners’abilities to recall items produced by different speakers. Listeners show better

recall for those items produced at the same rate in both familiarisation and

test than for items presented at different rates from familiarisation to test

(Nygaard, Sommers, & Pisoni, 1995). These findings suggest that extra-

linguistic cues such as rate influence memory for speech contents.

If extralinguistic information is retained in memory, it is possible that

such cues could influence future productions. One aspect of speech that

persists from perception to production is syntax. When listeners are asked torepeat a sentence they had heard and then produce a description of a picture,

they tend to use the same syntactic form as in the original sentence to

describe the scene (Bock, 1986). In addition, bilingual speakers who are

primed with a syntactic form persist in the syntax across languages (Loebell

& Bock, 2003). Since this persistence extends beyond adjacent sentences and

influences the syntactic structure of future sentences, structural persistence is

thought to relate to implicit learning of sentence construction (Bock &

612 JUNGERS AND HUPP

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s C

hica

go]

at 1

1:56

28

Oct

ober

201

4

Page 4: Speech priming: Evidence for rate persistence in unscripted speech

Griffin, 2000). Further, this persistence does not depend on production of

the primes and is present across many lags (Bock, Dell, Chang, & Onishi,

2007). As mentioned previously, syntax is just one of several factors

influencing prosody. Thus, it is reasonable to assume that other components,such as speech rate, may function like syntax and also show persistence.

There is already some evidence for rate persistence in scripted speech.

Kosslyn and Matt (1977) played a recording of two male speakers: one

speaking at a fast rate and one at a slow rate. Then the participants read

aloud a passage they were told was written by one of the speakers. The

participants imitated the rate of the speaker who supposedly wrote the

passage, although they were not explicitly instructed to do so. It is possible,

however, that subjects may have associated each written passage with aparticular speaker and felt an expectation to reproduce the rate of that

speaker. Similarly, Jungers et al. (2002) found evidence for prosodic

persistence in scripted speech. In one experiment, listeners heard a fast or

slow prime sentence followed by a written target sentence matched for length,

lexical stress pattern, and syntactic structure. When participants read the

target sentences aloud, their produced target rates were influenced by the

primed rate.

The study by Jungers et al. (2002) suggests that people persist in the globalprosodic dimension of tempo when reading sentences aloud, but the study

has some limitations. All of the prime and target sentences were in the active

form and matched in stress pattern and length. The current study seeks to

determine if speakers will persist in the rate when they produce an unscripted

picture description. The goal is to examine whether speakers’ production

rates will be influenced by priming sentences. Previous literature has

demonstrated syntactic priming in picture description tasks, but rate priming

has not been demonstrated in this ecologically valid way. If rate primingserves a communicative function in conversation, it is important to

demonstrate that the effect is not limited to scripted speech.

Thus, Experiment 1 asked whether speakers persist in the rate of sentences

they hear when they repeat the primes or create their own picture

descriptions. Both active and passive transitive sentences were used so that

results could be generalised across syntactic form. A final memory

recognition task was used to reduce demand characteristics such that

participants would focus on their memory for the sentences instead of onthe specific acoustic details of the sentences. By concentrating on their

memory for the sentences, participants were less likely to be aware of the rate

manipulation and less likely to consciously alter their speech.

Experiment 2 examined whether rate persistence is due to articulatory

priming. In this case, articulatory priming refers to the physical practice of

producing the sentence. In Experiment 1, the speakers repeated the priming

sentence before describing a new picture. To rule out the possibility that

SPEECH PRIMING 613

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s C

hica

go]

at 1

1:56

28

Oct

ober

201

4

Page 5: Speech priming: Evidence for rate persistence in unscripted speech

priming results were due to simply having physically practiced the particular

rate, in Experiment 2, speakers listened to but did not repeat the priming

sentence. A similar memory test followed. If priming is simply due to

practice articulating sentences at particular rates, then the rate persistenceshould be weaker in Experiment 2 than in Experiment 1. However, if rate

persistence is due to an internal representation of the timing aspects of the

sentences, then similar results will be observed in the two experiments.

EXPERIMENT 1

The goal of this experiment was to determine if speakers persist in the rate

they have just heard when they describe a new picture. Listeners heard primes

that were presented at a fast or slow rate. On each trial, participants heard a

sentence that described a picture on the screen. They repeated that sentence

and then described aloud a new picture. Their productions were analysed fortiming.

Method

Participants. Twenty-four adults who were native English speakers at alarge university in the Midwest participated in the study. Participants

received course credit in an introductory psychology course.

Apparatus. Participants heard language stimuli over AKG K270 head-

phones, and their voices were recorded to CD using a head-mounted AKG

C420 microphone. The auditory stimuli and the pictures were presented

using PowerPoint on a personal computer.

Materials. The experiment consisted of 24 transitive sentence primes

that described the action in a picture. The experiment also included 24

additional target pictures that subjects described aloud. Pictures included

cartoon characters involved in an action. The prime sentences were recorded

by a female speaker who listened to a metronome to create the tempi. The

slow was created with stressed syllables set to 60 bpm (beats per minute) and

the fast was set to 120 bpm. The sentences sounded natural, with the fast

sentence primes averaging 0.24 s/syllable and the slow sentence primesaveraging 0.40 s/syllable. Half of the sentences described the action in active

form while the other half described the action in passive form.

A 24-item memory test on the prime sentences included five types of

items: same tempo-same syntax, same tempo-different syntax, different

tempo-same syntax, different tempo-different syntax, and foils. There were

12 items related to the study and 12 foils. The foils contained some of the

same words used in the original sentences so that participants could not use a

614 JUNGERS AND HUPP

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s C

hica

go]

at 1

1:56

28

Oct

ober

201

4

Page 6: Speech priming: Evidence for rate persistence in unscripted speech

single word or phrase to remember a sentence. The foils also were balanced

for syntactic form and tempo.

Design. The independent, within-subject variable was rate (slow, fast).

The rate was blocked across trials; participants heard slow sentences

followed by fast sentences or vice versa. The order (first block/ second

block) was a counterbalancing variable. It was necessary to block the fast

and slow stimuli so that participants would be unaware of the rate

manipulation. The same sentence was heard at a fast tempo for half the

participants but at a slow tempo for the other participants. The syntactic

form was randomised across trials; an individual sentence was either active

or passive for all participants.

The dependent variable was the timing produced by participants. Sentence

timing was analysed by two independent coders who determined the

beginning and end of each sentence both by listening to and by looking at

the waveform in Adobe Audition.

Procedure. The participants first described aloud three pictures pre-

sented on the computer screen to practice using the apparatus. On each

experimental trial, participants listened to a prime sentence describing the

picture on the screen, repeated that sentence, and then produced a

description of the next picture. See Figure 1 for an example of the trials.

The first trial served as a practice trial for the procedure. The participants

were instructed to pay careful attention to the sentences because they would

be asked to recognise them later. There were a total of 24 experimental trials,

presented in two blocks of 12 trials. Between blocks, participants completed

a paper and pencil questionnaire about their music and language back-

ground.Following the picture description task, participants listened to a 24-item

memory test. On each trial, the sentence was heard twice. On paper, the

participants circled ‘yes’ or ‘no’ to whether the sentence described a picture

they saw during the experiment.

Results

Production task. Preliminary analyses revealed no effect of syntactic

form or blocking order (whether fast or slow block presented first), so they

were eliminated from subsequent analyses. The sentences were produced

without a script, so differences between the exact choice and number of

words may result. In order to account for variations in picture descriptions,

an analysis of syllable timing was conducted by dividing the total utterance

time by the number of produced syllables. A repeated measures ANOVA with

repetition condition (repeat, describe) and speaking rate condition (fast,

SPEECH PRIMING 615

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s C

hica

go]

at 1

1:56

28

Oct

ober

201

4

Page 7: Speech priming: Evidence for rate persistence in unscripted speech

slow) as factors showed a significant effect for speaking rate condition by

subject, F1(1, 23)�10.2, pB.01, l2�0.31, and by item, F2(1, 46)�39.54,

pB.01, l2�0.46. As seen in Figure 2, speakers produced faster sentences

following the fast prime than following the slow prime. Planned comparisons

revealed that this effect occurred both when participants repeated the prime

and when they described the new scene (pB.05). There were no differences in

rate for repetition condition, F1(1, 23)�0.06, ns, F2(1, 46)�0.24, ns, and no

Figure 1. Sample trials for Experiment 1.

616 JUNGERS AND HUPP

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s C

hica

go]

at 1

1:56

28

Oct

ober

201

4

Page 8: Speech priming: Evidence for rate persistence in unscripted speech

interaction between repetition condition and speaking rate condition,

F1(1, 23)�1.8, ns, F2(1, 46)�1.9, ns. There were no differences in rate

between the first and second halves of each block and no interaction between

the block halves (first half, second half) and rate condition (fast, slow). This

suggests that participants were not slowing or speeding systematically across

the items within each block.

Memory recognition task. An ANOVA on percent ‘yes’ response by type

of item (from study or foil) showed a significant difference, with subjects

most often responding ‘no’ to the foil items, thereby demonstrating a high

degree of overall accuracy both by subjects, F1(1, 23)�113.6, pB.05,

l2�0.83, and by items, F2(1, 23)�26.9, pB.05, l2�0.55. A repeated

measures ANOVA on percent ‘yes’ response to the items from the study with

tempo (same as study/new tempo) and syntactic form (same as study/new

form) as factors showed a significant effect for speaking rate condition, with

participants more likely to respond ‘yes’ to items in the same tempo,

F1(1, 23)�14.5, pB.01, l2�0.39, but no effect for syntactic form, F1(1,

23)�1.1, ns, and no interaction between tempo and syntactic form, F1(1,

23)�2.5, ns. An ANOVA by item on percent ‘yes’ response to the items from

the study with tempo and syntactic form as factors showed the same trend as

the analysis by subject, with participants more likely to respond ‘yes’ to items

in the same tempo, F2(1, 11)�3.64, p�.09, l2�0.31, but no effect for

syntactic form, F2(1, 11)�0.18, ns, and no interaction between tempo and

syntactic form, F2(1, 11)�0.42, ns. By category, the percent ‘yes’ responses

were: same tempo-same syntax�79%, same tempo-different syntax�65%,

Figure 2. Experiment 1. Timing by syllable for repeated and created picture descriptions.

SPEECH PRIMING 617

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s C

hica

go]

at 1

1:56

28

Oct

ober

201

4

Page 9: Speech priming: Evidence for rate persistence in unscripted speech

different tempo-same syntax�49%, different tempo-different syntax�51%,

and foils�20% (expected response ‘no’).

Discussion

This experiment demonstrated that participants encode timing information

when they listen to and repeat a sentence. This timing persistence was seen

for both active and passive sentences. The lack of systematic difference in

tempo between the first and second halves of each block suggests a local

priming effect which reflects the immediate processing of the prime instead

of a cumulative global effect. The memory test reveals that participants were

most accurate in their response to items identical to those heard before.

Participants were more likely to respond ‘yes’ to items with the same tempo

as heard in the study, despite the fact that the question was whether the

sentence referred to a picture they had seen in the study. The syntax and

timing should not influence this decision, but the participants’ responses

suggest that timing was encoded in memory. The main finding is that when

participants were primed with a fast sentence, they produced unscripted

picture descriptions that were faster than if they were primed with slow

sentences. However, it can be argued that this rate priming was dependent

upon articulation of the timing information. Experiment 2 examines whether

this effect was due to articulatory priming by having the participants listen to

but not repeat the priming sentence.

EXPERIMENT 2

The goal of Experiment 2 was to determine if speakers persist in the rate of a

sentence they have just heard when they describe a new picture, even when

they do not repeat the priming sentence. As in Experiment 1, listeners heard

primes that were presented at a fast or slow rate. On each trial, participants

heard a sentence that described a picture on the screen and then created their

own description of a subsequent picture. If speakers show rate persistence as

in Experiment 1, this would argue for priming based on an internal

representation and not on practice producing the rate of the previous

sentence.

Method

Participants. Twenty-four adults who were native English speakers at a

large university in the Midwest participated in the study. Participants

received course credit in an introductory psychology course. Participants

had not participated in Experiment 1.

618 JUNGERS AND HUPP

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s C

hica

go]

at 1

1:56

28

Oct

ober

201

4

Page 10: Speech priming: Evidence for rate persistence in unscripted speech

Apparatus and materials. The equipment and materials were the same as

in Experiment 1.

Design and procedure. The design was the same as in Experiment 1. Theparticipants first described aloud three pictures to become accustomed to the

equipment. On each experimental trial, participants listened to a prime

sentence describing the picture on the screen and then produced a

description of another picture. Unlike Experiment 1, the participants did

not repeat the priming sentence aloud. The first trial served as a practice

trial. There were a total of 24 experimental trials, presented in two blocks of

12 trials. Following the picture description task, participants listened to a 24-

item memory test. On each trial, the sentence was heard over headphones

two times. On paper, the participants circled ‘yes’ or ‘no’ to whether the

sentence described a picture they saw during the experiment.

Results

Production task. As in Experiment 1, preliminary analyses revealed no

effect of syntactic form of prime sentence or blocking order (whether fast or

slow block presented first), so they were eliminated from subsequent

analyses. A repeated measures ANOVA on syllable timing showed differences

between new sentences produced after fast and slow primes by subjects, F1(1,

23)�8.59, pB.01, l2�0.27, and by items, F2(1, 23)�4.86, pB.05, l2�0.18. As seen in Figure 3, speakers produced faster sentences following the

fast primes. There were no differences in rate between the first and second

halves of each block and no interaction between the block halves (first half,

second half) and rate condition (fast, slow). This suggests that participants

Figure 3. Experiment 2. Timing by syllable for picture descriptions following fast and slow

prime sentences.

SPEECH PRIMING 619

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s C

hica

go]

at 1

1:56

28

Oct

ober

201

4

Page 11: Speech priming: Evidence for rate persistence in unscripted speech

were not slowing or speeding systematically across the items within each

block.

Memory recognition task. An ANOVA on percent ‘yes’ response by typeof item (from study or foil) showed a significant difference both by subjects,

F1(1, 23)�75.9, pB.05, l2�0.77, and by items, F2(1, 23)�19.7, pB.05,

l2�0.47, with subjects most often responding ‘no’ to the foil items thereby

demonstrating a high degree of overall accuracy. A repeated measures

ANOVA on percent ‘yes’ response to the items from the study with tempo

(same as study/new tempo) and syntactic form (same as study/new form) as

factors showed a significant effect for speaking rate condition, with

participants more likely to respond ‘yes’ to items in the same tempo, F1(1,

23)�12.3, pB.01, l2�0.35, but no effect for syntactic form, F1(1, 23)�0.02, ns, and no interaction between tempo and syntactic form, F1(1, 23)�3.09, ns. An ANOVA by item on percent ‘yes’ response to the items from the

study with tempo and syntactic form as factors showed the same trend as the

analysis by subject, with participants more likely to respond ‘yes’ to items in

the same tempo, F2(1, 11)�2.78, p�.13, l2�0.26, but no effect for

syntactic form, F2(1, 11)�0.01, ns, and no interaction between tempo and

syntactic form, F2(1, 11)�0.50, ns. By category, the percent ‘yes’ responses

were: same tempo-same syntax�74%, same tempo-different syntax�65%,

different tempo-same syntax�47%, different tempo-different syntax�54%,

and foils�23% (expected response ‘no’).

Discussion

This experiment demonstrated that participants encode rate information

when they produce new sentences to describe pictures. When participants

were primed with a fast sentence, they produced unscripted picture

descriptions that were faster than if they were primed with slow sentences,

even though they did not repeat the prime. This suggests that the rate

priming observed in Experiment 1 was not due to articulatory priming

(the physical practice of producing the primes). It is possible, however,

that participants engaged in tacit motor activity without actually moving

the articulators. As in Experiment 1, the results point to a local priming

effect since there were no systematic differences between the timing in the

first and second halves of each block. Also as in Experiment 1, the

memory test reveals that participants were most accurate in their response

to items with the same tempo as heard before, despite the fact that the

question was whether the sentence referred to a picture they had seen in

the study.

To rule out an awareness of the rate manipulation as an explanation for

the rate persistence results, an additional 12 participants from the same pool

620 JUNGERS AND HUPP

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s C

hica

go]

at 1

1:56

28

Oct

ober

201

4

Page 12: Speech priming: Evidence for rate persistence in unscripted speech

participated in the study and were asked several questions after the

completion of the study. These questions included: What was the goal of

the study? What were the sentences like? All 12 participants responded that

the goal of the study was to test their memory for sentences. Some indicated

that the sentences differed in syntactic form, but no participant indicated

anything about the timing manipulation. Although this may seem a

surprising result, it is in line with other research that suggests conversation

partners are unaware of their convergence in speech and only notice large

and conversationally inappropriate differences (Street, 1990; Street & Brady,

1982). These additional participant responses along with the instructions to

focus on the upcoming memory test during the task lend confidence that the

results are not due to demand characteristics.

GENERAL DISCUSSION

The central finding of this study was the rate persistence, both with and

without repetition of the prime. The rate differences did not relate to the

syntax, but listeners still persisted in rate for both active and passive primes.

Past studies have shown rate persistence in scripted speech (Jungers et al.,

2002), but this is the first instance of prosodic persistence in an unscripted

task (for similar results with children, see Hupp & Jungers, in press). In the

first experiment, listeners persisted in the timing of picture descriptions after

repeating a priming sentence. In the second experiment, listeners again

produced picture descriptions that were similar in rate to the prime

sentences, even without the repetition of the prime. Thus, the priming effects

are not due to physical practice with articulating the priming sentence which

would support the notion of an internal representation of the timing

information. This effect of persistence was small relative to the difference

in primed rates. This may be due to the fast prime more closely resembling

speakers’ natural rates of speech than the slow prime. This study extends

previous work by eliciting picture descriptions; rate priming in these

unscripted productions lends support to the ecological validity of this effect.

Why might it be useful to persist in timing? Continuing the timing from

speaker to listener may help to coordinate conversation. Conversation

partners increase their comprehensibility by accommodating to each others’

speech patterns (Giles & Powesland, 1975). Such speech convergence has

been observed for extralinguistic features including utterance length, vocal

intensity, and pause length (Giles, Coupland, & Coupland, 1991). Com-

munication is facilitated because the conversation partners more efficiently

process the message when the extralinguistic features match (Berger &

Roloff, 1980). Implicit learning of paralinguistic norms may be helpful as a

listener begins to incorporate the particular speech patterns of a region or

SPEECH PRIMING 621

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s C

hica

go]

at 1

1:56

28

Oct

ober

201

4

Page 13: Speech priming: Evidence for rate persistence in unscripted speech

dialect. For example, rate differences exist between American and New

Zealand English (Robb, MacLagan, & Chen, 2004).

What mechanism might explain this rate persistence? One possibility is

that conversation partners’ syllable rates become entrained via oscillatorsin the brain (Wilson & Wilson, 2005). At the neural level, there is

evidence for mirror neurons that may control coordination (Dale &

Spivey, 2006). Another possibility is that rate is one of a number of levels

that are aligned automatically in dialogue (Pickering & Garrod, 2004).

Such a mechanism assumes that the comprehension and production

systems align at each level (syntactic, lexical, semantic) via priming. The

comprehension system may use an emulator that covertly imitates and

leads to predictions about the current input, allowing rapid processing(Pickering & Garrod, 2007). Although we believe that the current results

arise from straightforward and automatic priming of speech structures in

speaker and listener, we acknowledge that a full explanation of alignment

in dialogue should allow for the influence of social-motivational factors.

For example, the degree of phonetic convergence in a conversational task

is determined by the role of the participant and the sex of the pair of

talkers (Pardo, 2006).

These experiments demonstrate persistence of rate in unscripted speech.Thus, acoustic features are retained from perception to production, arguing

against a normalisation process. It is possible that this priming is temporary.

Further studies can examine the duration and strength of

the prime. The persistence in timing may provide a social advantage and

help to coordinate conversation. Such coordination may arise from an

automatic process in dialogue in which speaker and listener align their

speech at multiple levels.

Manuscript received March 2008

Revised manuscript received November 2008

First published online January 2009

REFERENCES

Beckman, M. E. (1996). The parsing of prosody. Language and Cognitive Processes, 11, 17�67.

Berger, C. R., & Roloff, M. E. (1980). Social cognition, self-awareness, and interpersonal

communication. In B. Dervin & M. J. Voight (Eds.), Progress in communication sciences.

Norwood, NJ: Ablex.

Bock, K. (1986). Syntactic persistence in language production. Cognitive Psychology, 18, 355�387.

Bock, K., Dell, G. S., Chang, F., & Onishi, K. H. (2007). Persistent structural priming from

language comprehension to language production. Cognition, 104, 437�458.

Bock, K., & Griffin, Z. M. (2000). The persistence of structural priming: Transient activation or

implicit learning? Journal of Experimental Psychology: General, 129, 177�192.

622 JUNGERS AND HUPP

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s C

hica

go]

at 1

1:56

28

Oct

ober

201

4

Page 14: Speech priming: Evidence for rate persistence in unscripted speech

Bradlow, A. R., Nygaard, L. C., & Pisoni, D. B. (1999). Effects of talker, rate, and amplitude

variation on recognition memory for spoken words. Perception and Psychophysics, 61, 206�219.

Cutler, A., Dahan, D., & Donselaar, W. van (1997). Prosody in the comprehension of spoken

language: A literature review. Language and Speech, 40, 141�201.

Dale, R., & Spivey, M. J. (2006). Unraveling the dyad: Using recurrence analysis to explore patterns

of syntactic coordination between children and caregivers in conversation. Language Learning,

56, 391�430.

Ferreira, F. (1993). Creation of prosody during sentence production. Psychological Review, 100,

233�253.

Gee, J. P., & Grosjean, F. (1983). Performance studies: A psycholinguistic and linguistic appraisal.

Cognitive Psychology, 15, 411�458.

Giles, H., Coupland, J., & Coupland, N. (1991). Context of accommodation: Developments in

applied sociolinguistics. New York: Cambridge University Press.

Giles, H., & Powesland, P. F. (1975). A social psychological model of speech diversity. London:

Academic Press.

Hupp, J. M., & Jungers, M. K. (in press). Speech priming: An examination of rate and syntactic

persistence in preschoolers. British Journal of Developmental Psychology.

Jungers, M. K., Palmer, C., & Speer, S. R. (2002). Time after time: The coordinating influence of

tempo in music and speech. Cognitive Processing, 2, 21�35.

Kosslyn, S. M., & Matt, A. M. (1977). If you speak slowly, do people read your prose slowly?

Person-particular speech recoding during reading. Bulletin of the Psychonomic Society, 9, 250�252.

Lehiste, I. (1973). Phonetic disambiguation of syntactic ambiguity. Glossa, 7, 106�122.

Lehiste, I., Olive, J. P., & Streeter, L. (1976). Role of duration in disambiguating syntactically

ambiguous sentences. Journal of the Acoustical Society of America, 60, 1199�1202.

Loebell, H., & Bock, K. (2003). Structural priming across languages. Linguistics, 41, 791�824.

Nygaard, L. C., Sommers, M. S., & Pisoni, D. B. (1995). Effects of stimulus variability on

perception and representation of spoken words in memory. Perception and Psychophysics, 57,

989�1001.

Pardo, J. S. (2006). On phonetic convergence during conversational interaction. Journal of the

Acoustic Society of America, 119, 2382�2393.

Pickering, M. J., & Garrod, S. (2004). Toward a mechanistic psychology of dialogue. Behavioral and

Brain Sciences, 27, 169�225.

Pickering, M. J., & Garrod, S. (2007). Do people use language production to make predictions

during comprehension? Trends in Cognitive Science, 11, 105�110.

Pisoni, D. B. (1997). Some thoughts on ‘normalization’ in speech perception. In K. Johnson & J. W.

Mullenix (Eds.), Talker variability in speech processing (pp. 9�32). San Diego, CA: Academic

Press.

Price, P., Ostendorf, M., Shattuck-Hufnagel, S., & Fong, C. (1991). The use of prosody in syntactic

disambiguation. Journal of the Acoustical Society of America, 90, 723�735.

Robb, M. P., MacLagan, M. A., & Chen, Y. (2004). Speech rates of American and New Zealand

varieties of English. Clinical Linguistics and Phonetics, 18, 1�15.

Shattuck-Hufnagel, S., & Turk, A. E. (1996). A prosody tutorial for investigators of auditory

sentence processing. Journal of Psycholinguistics Research, 25, 193�247.

Speer, S. R., Crowder, R. G., & Thomas, L. M. (1993). Prosodic structure and sentence recognition.

Journal of Memory and Language, 32, 336�358.

Street, R. L., Jr. (1990). The communicative functions of paralanguage and prosody. In H. Giles &

P.W. Robinson (Eds.), Handbook of language and social psychology (pp. 121�140). Chichester,

UK: John Wiley & Sons.

SPEECH PRIMING 623

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s C

hica

go]

at 1

1:56

28

Oct

ober

201

4

Page 15: Speech priming: Evidence for rate persistence in unscripted speech

Street, R. L., Jr., & Brady, M. (1982). Speech rate acceptance ranges as a function of evaluative

domain, listener speech rate, and communication context. Communication Monographs, 49,

290�308.

Warren, P. (1996). Prosody and parsing: An introduction. Language and Cognitive Processes, 11,

1�16.

Wilson, M., & Wilson, T. P. (2005). An oscillator model of the timing of turn-taking. Psychonomic

Bulletin and Review, 12, 957�968.

624 JUNGERS AND HUPP

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s C

hica

go]

at 1

1:56

28

Oct

ober

201

4