Why are nonnative L2 utterances longer than nativeproduced utterances

1
4aSC7 148th ASA meeting in San Diego 2004
Why are non-native (L2) utterances longer than
native-produced utterances?
2
James E. Flege Anders Højen Speech and Hearing
Sciences Division University of Alabama at
Birmingham
3
Abstract
Research has often but not always shown that
utterances produced by nonnative speakers are
longer than those of native speakers, even when
produced as rapidly as possible. The nonnative
speaking rate effect is known to be greater for
late than early L2 learners, and may disappear
for late learners when very short-duration L2
utterances are examined. This study examined the
duration of English utterances produced by native
English (NE) speakers and late native Spanish
(NS) learners of English (n20 each). For
normal-rate utterances, durations were
significantly longer for the NS than NE group for
syntactically complex sentences but not simple
sentences or noun phrases. However, the NS-NE
difference reached significance for all three
utterance types at a fast rate. In Experiment 2,
the NS group took longer than the NE group to
produce the digit string 1 2 3 4 5 6 7 as
rapidly as possible. Taken together, the results
suggest that differences in both the articulation
of L2 speech sounds and higher-order planning,
including the formulation of syntactic
representations, contribute to the nonnative
speaking rate effect.
4
The Phenomenon

• Research has shown that utterances are longer
  when produced by nonnatives than by native
  speakers
• This nonnative speaking rate effect is larger
  for late than early learners (Guion et al.2000
  Mackay Flege, 2004 Schmidt Flege 1996)

5
Background

• Previous research has used a variety of
  techniques to elicit L2 utterances
• Multiple productions of a single phrase (Schmidt
  Flege, 1995)
• Reading of L2 sentences (Riggenbach, 1991)
• Repetition of L2 sentences follwing an aural
  model (Guion et al., 2000)
• Reading of entire L2 paragraphs (Favreau
  Segalowitz, 1982)

6

• The L2 utterances examined in previous research
  have differed in a least 2 important respects
• Overall number of words
• Complexity of syntactic structure

7

• Elicitation technique/utterance type may cause
  the magnitude of the nonnative speaking rate
  effect to vary
• Production of phrase (a green pea) 12 times at
  normal vs. fast rate. Non-significant difference
  between native speakers and late learners
  (Schmidt Flege, 1995)
• Repetition of aurally modeled sentences (e.g.,
  Paul ate carrots and peas). Late learners
  sentences were significantly longer (about 15)
  than native speakers (Guion et al. 2000)

8
Research question
What accounts for the nonnative speaking rate
effect? 3 Hypotheses (not mutually exclusive)

• H1 Nonnatives volitionally choose to speak
  slower in the L2, perhaps to avoid being
  misunderstood as the result of their foreign
  accent

9

• H2 Nonnatives make phonetic adjustments to L1
  vowels and consonants in order to make them
  L2-like, or they have difficulty producing
  certain L2 segments. Prolongation in the
  articulation of L2 segments add up to
  differences in utterance duration
• H3 Nonnatives take longer than natives to
  specify the syntactic plan or organization of L2
  utterances

10
Experiment 1

• Examined the duration of English utterances
  produced by 2 groups
• Native English (NE) speakers
• Native Spanish (NS) speakers who learned English
  as adults
• Three levels of syntactic complexity
• Two speaking rates

11
Participants
Participants
Note Educ, total years of education AOA, age at
arrival in the U.S. LOR, length of residence in
the U.S. Ranges in parentheses.
12
Stimuli
Stimuli Procedure

• Speech materials
• Simple phrases (n6)
• Simple sentences (n6)
• Complex sentences (n6)
• Examples

a box of pencils Gary forgot a box of pencils The
boy saw that Gary forgot a box of pencils
13

• Utterances elicited using orthography and aural
  model (see utterances written and hear them via
  loudspeaker)
• Materials produced in fixed order (a,b,c) two
  times without instruction as to speaking rate
  (unspeeded condition)
• Procedure repeated, utterances now spoken as
  fast as possible (speeded condition)

14
Results Discussion

• Across both groups
• Noun phrases (M936 ms) shorter than
• simple sentences (M1663 ms) shorter than
• complex sentences (M2506 ms) (plt0.01)
• Speeded utterances (M1501 ms) shorter than
• unspeeded utterances (M1903 ms) (plt0.01)

15
Raw data graph
Star indicates significant native-nonnative
difference
16
results
Evaluation of hypotheses

• H1 Volitional choice hypothesis
• NS speakers produced significantly longer
  utterances (all 3 types) than NE speakers in the
  speeded condition 
• Conclusion reject H1

17

• H2 Segmental articulation hypothesis
• In the speeded condition, NS speakers produced
  significantly longer utterances than the NE
  speakers did even when minimal syntactic
  structure existed (noun phrases).
• Conclusion weak support for H2

18

• H3 Syntactic planning hypothesis For both
  groups, complex sentences were much longer than
  noun phrases. However, the magnitude of
  differences between the 2 utterance types was
  significantly greater for NS than NE speakers, in
  both the unspeeded speeded conditions (p lt .01)
• Conclusion H3 supported

19

• However, results could also be explained by H2
• If H2 is true and H3 is false, the percent
  native-nonnative difference should be constant
  across utterance types
• But table shows that both absolute and percent
  difference increases as a function of added
  syntactic complexity ( added words)

20
difference between NE and NS utterances
Conclusion Additional support for H3
21
Experiment 2

• In Exp 1, the native-nonnative difference emerged
  in the speeded condition when only minimal syntax
  was involved (phrases)
• This provided support for H2, but the result
  could also be explained by H3, because even the
  short phrases had some syntactic organization

22

• Aim
• To test H2, the segmental articulation
  hypothesis, using speech materials that had
  virtually no syntactic organization, viz., a
  string of digits

23
Stimuli and procedure
Stimuli procedure
The same participants produced the number string
1 2 3 4 5 6 7 in English as fast as
possible. Task repeated 3 times.
24
Result Discussion
25
Results 2
The durations of NS speakers digit strings were
longer than the NE speakers (plt0.01), even
though these digit strings had minimal (or no)
syntactic organizatoin Conclusion Accept H2
26
Conclusion

• Slower L2 speaking rate is not volitionally
  chosen native-nonnative difference persists (is
  even larger!) in speeded speech production (see
  also Mackay Flege 2004)
• Slower syntactic processing in the L2 contributes
  to the nonnative speaking rate effect The more
  complex the syntax, the larger the relative
  native-nonnative differences

27

• Prolongation of segments add up to differences
  in utterance duration. Possibly because L2
  speakers make phonetic adjustments to L1 vowels
  and consonants to approximate L2 norms. Or they
  have difficulty producing certain L2 segments.

28
Future research

• Effect of ease of lexical access?
• Implications for intelligibility?
• How is speed of L1 processing affected by L2
  learning?

29
References

• Favreau, M. and Segalowitz, N.S. (1982). "Second
  language reading in fluent bilinguals," Applied
  Psycholinguistics 3, 329-341.
• Guion, S.G., Flege, J.E., Liu, S. and
  Yeni-Komshian, G.H. (2000). "Age of learning
  effects on the duration of sentences produced in
  a second language," Applied Psycholinguistics 21,
  205-228.
• MacKay, I.R.A. and Flege, J.E. (2004). "Effects
  of the age of second language learning on the
  duration of first and second language sentences
  The role of suppression," Applied
  Psycholinguistics 25, 373-396.
• Riggenbach, H. (1991). "Toward an understanding
  of fluency A microanalysis of nonnative speaker
  conversations," Discourse Processes 14, 423-441.
• Schmidt, A.M. and Flege, J.E. (1995). "Effects of
  peaking rate changes on native and nonnative
  speech production," Phonetica 52, 41-54.
• Schmidt, A.M. and Flege, J.E. (1996). "Speaking
  rate effects on stops produced by Spanish and
  English monolinguals and Spanish English
  bilinguals," Phonetica 53, 162-179.