The impact of wordend phonology and morphology on stuttering

1
The impact of word-end phonology and morphology
on stuttering

• Chloe Marshall
• Centre for Developmental Language Disorders and
  Cognitive Neuroscience,
• Department of Human Communication Science,
• University College London
• UCL, 27th June, 2005

2
Acknowledgements

• Peter Howell and Jenny Hayes
• Heather van der Lely and John Harris

3
Aims of todays talk

• Are stuttering rates in English-speaking adults
  and children influenced by morphological and
  phonological complexity at the ends of words?
• Study 1 adult corpus data (UCLASS database)
• Study 2 non-word repetition task
• Study 3 elicitation task (inflected forms)

4
Background 1

• Definition of phonological complexity
• Based on what typically developing children find
  hard to acquire
• e.g. late-acquired consonants (dorsals,
  fricatives, liquids)
• e.g. multisyllabic forms
• e.g. consonant clusters
• Definition of morphological complexity
• inflected forms e.g. runs, walked, singing
• derived forms e.g. unreal, happier, sunny

5
Background 2

• Why should we expect phonological and
  morphological complexity to affect stuttering?
• Evidence that consonant clusters influence
  stuttering rates
• Howell, Au-Yeung Sackin (2000)
• Word-initially, adults
• Weiss Jakielski (2003)
• Children 60 115

6
Background 3

• English inflection (-ed, -s) adds a consonant,
  and therefore increases the phonological
  complexity of the word, often forming a cluster.
• e.g. wrap wrapped
• e.g. drive drives
• Confound between morphological and phonological
  complexity.
• How can they be teased apart?

7
Study 1 adult corpus data (UCLASS database)

• Contrasts 4 sets of words

phonological complexity
4
3
morphological complexity
2
1
1 phon simple, morph simple bad 2 phon
simple, morph complex died 3 phon complex,
morph simple month 4 phon complex, morph
complex looked
8
What we might find (1)
Phonological complexity influences stuttering
independently of morphology
9
What we might find (2)
Morphological complexity influences stuttering
independently of phonology
10
What we might find (3)
Interaction between phonological complexity and
morphological complexity
11
Details of study participants

• 16 males, diagnosed as PWS
• 16 47 years, mean age 25, sd 9
• Spontaneous material in response to questions
  from an interviewer, casual speech
• words stuttered for each speaker 2.61 - 20.33

12
Details of study words
Phonology word-shape is CVC or CVCC
Morphology past tense -ed, plural -s, third
person -s
13
Details of study selecting words from corpus

• Perl program identified words of shape CVC and
  CVCC
• Words hand-coded as being morphologically complex
  or not
• Irregular verbs removed by hand (e.g. had, got,
  has)
• Each word classified as stuttered (1) or not
  stuttered (0), based on such events marked in
  original transcriptions

14
Results 1
15
Results 2
16
Conclusions

• Presence of a cluster at the end of a word does
  not influence stuttering rates in spontaneous
  speech
• Nor is there an effect of inflectional morphology
• However data not amenable to individual
  analysis
• For certain speakers either type of complexity
  might have an effect

17
Study 2 Non-word repetition task

• Aim to investigate whether the presence of a
  consonant cluster at the word-end influences
  stuttering rates in absence of morphological
  complexity
• Advantage over spontaneous studies can elicit
  enough tokens of interest from all participants,
  therefore data amenable to individual analysis

18
Details of study participants

• 19 speakers
• Average age 14 years
• Divided into 2 groups
• Group 1 lt or 14 years (N12)
• Group 2 or gt 15 years (N7)

19
Details of study stimuli(Test of
Phonological Structure, van der Lely Harris,
1999)

• 2 experimental conditions
• No word-final cluster ket, prif (N 8)
• Word-final cluster kest, prilf (N 8)
• 1 filler condition
• Multisyllabic deppary, sapiffy (N 12)
• Responses scored according to presence (1) or
  absence (0) of stuttering

20
Results
21
Conclusion

• No impact of word-end phonological complexity
  detected in this task

22
Study 3 Elicitation task (inflected forms)

• Task designed to elicit 3rd person singular and
  past tense forms
• Aim to investigate whether the presence of a
  consonant cluster at the word-end influences
  stuttering rates in morphologically complex words

23
Details of study participants

• Same as study 2

• 19 speakers
• Average age 14 years
• Divided into 2 groups
• Group 1 lt or 14 years (N12)
• Group 2 or gt 15 years (N7)

24
Details of study stimuli

• 2 conditions, varying phonological complexity
• Condition 1 5 verbs whose inflected word-shape
  is CVC, e.g. weighs, weighed
• Condition 2 5 verbs whose inflected word-shape
  is CVCC, e.g. wraps, wrapped

25
Details of study procedure

• Experimenter
• Everyday I pour a drink.
• Participant
• Your friend also ________ .
• Yesterday you both __________ .

26
Results
27
Individual results
28
Conclusion

• A few individuals stutter on morphologically
  complex forms
• Only Participant 2 showed an effect of
  phonological complexity on stuttering

29
Discussion

• Some experimental evidence that phonological and
  morphological complexity influence stuttering in
  English PWS, but not for all speakers.
• Still to look at
• Syllabic suffixes (wanted, horses, washes)
• Derivational morphology (heavier, unreal)
• Much younger children what happens at age 3?

30
Reference
Marshall, C.R. (2005) The impact of word-end
phonology and morphology on stuttering.
Stammering Research, 1 375-391 http//www.stamres
.psychol.ucl.ac.uk/ My web page
www.ucl.ac.uk/DLDCN