Lecture Three Orienting Question

1
Lecture Three - Orienting Question
Compare CRH and EXPLAN as accounts of stuttering.
2
Disfluency types

• Stallings (tend to occur
• on function words)
• Whole word repetition
• Pauses (silent and filled)
• Phrase repetitions

• Advancings (tend to
• occur on the first part(s)
• of content words)
• Part-word repetitions
• Prolongations
• Word breaks

3
Topics this talk will cover

• Context
• Difficulty conditions under which words should
  be affected by difficulty
• Speech rate
• Simulation
• How stuttering relates to fluent speech
• Differences between early form of stuttering that
  recovers and later form that persists
• Implications for treatment

4
Model assumptions

• Planning and execution processes are independent
• Difficulty (reflects planning) on current word
• Motor on previous word
• Needs a contextual unit that at least encompasses
  previous and current word

5
We use Phonological Word (PW) unit

• content word nucleus
• function words as satellites
• Main reason why we use
• has a single locus of difficulty (the content
  word) which can have easy words preceding it
  which are linked to it which are the elements in
  the model

6
What PW reveals about stuttering and how it
changes over development
Disfluency distribution in He hit him example
Problem word is hit See disfluencies on
hit or word preceding, not on following
function word He, he hit him He hhit
him He hit him, him Strenstrom and
Svartvik (1994) subject and object pronouns
7
Relationship between disfluencies on content
and on function word over age and fluency groups
on PW with initial function words (allows
influence on content word and the word prior to
it to be assessed). Occurs in either/or fashion
in PW less than 3 of PW have disfluencies on
both initial function words and content word
(significantly less than would have expected by
chance).
8
English
Young speakers stutter predominantly on function
words, older speakers on content words.
9
Note - does not use perceptual monitor to account
for different forms of fluency
10
Note - does not use perceptual monitor to account
for different forms of fluency
11
Note - does not use perceptual monitor to account
for different forms of fluency
12
Next look at ways of measuring how difficult
different content words are and influence of
rate in local context, starting with difficulty
13
Work on measures of difficulty - Index of
Phonetic Complexity
IPC scoring scheme. Factors are listed in the
left-hand column. The sounds which received no
points are indicated in the second column
(labeled no score) and the sounds which
receive a point each time they occurred are given
in the third column (labeled one point each).
Factor No Score One point each 1. Consonant
by Place Labials, coronals, glottals
Dorsals 2. Consonant by Manner Stops, nasals,
glides Fricatives, affricates, liquids 3.
Singleton Consonants Reduplicated Variegated by
Place 4. Vowel by Class Monophthongs,
diphthongs Rhotics 5. Word shape Ends with a
vowel Ends with a consonant 6. Word Length
(Syllables) Monosyllables, disyllables gt3
syllables 7. Contiguous Consonants No
Clusters Consonant Clusters 8. Cluster by
Place Homorganic Heterorganic
14
(No Transcript)
15
ANOPhS marked for difficulty
Marked when Onset presence - no onset
consonant is marked form Onset complexity
- marked when start with consonant
string Pre-initial s - s-consonant marked all
others not marked Rhyme - Presence of C after
V so word dif-rant f and n are rhymes
(marked) t in 2nd syllable is word end Word
end - ends in C Suffix - marked when have as
in cats and wouldnt Obstruent - these
marked Sonorants - these marked
16
(No Transcript)
17
Studies of difficulty in younger speakers (fluent
and speakers who stutter).

• Do children repeat initial function words in a PW
  more often when the content word is difficult
  than when the content word is easy?
• Hypothesis tested on fluent children (2-3 years)
  and CWS (up to 5 years)
• In both cases the following factors were excluded
  as possible confounding variables
• Frequency
• Number of unique function words preceding and
  following content word
• Length of function words preceding and following
  content word
• Lexical category of function words preceding and
  following content word
• MLU for PW
• Syntactic structure of PW

18
An independent samples t-test comparing these
means was significant, t (352) -2.63, p ? .01.
19
Speech rate as an execution factor

• Rate on preceding function words is also a
  determinant of whether plan for content word will
  be unavailable and, therefore, whether the word
  will be stuttered or not.
• Local effect of rate

20
Rate second line of model figure. Rate can lead
to execution getting ahead of planning Global
and local rate
21
Percentage of fluent and dysfluent items
categorised correctly by a connectionist Model
with 8 hidden units trained using back
propagation with/without duration of prior
function word as additional input to ANOPHS score
Without duration
With duration
22
Conclusions and on-going work

• We consider that EXPLAN accounts for stuttering
  and how it emerges out of normal
  fluently-developing speech.
• Problem at around teenage. Indicates what should
  be monitored (content words disfluencies)

23
Implications for treatment

• What mechanisms alter timing internal by
  repetititon, external (to planning and execution
  processes) using cerbellar inputs.
• Scanning work with Watkins and Davis that show
  cerebellar timing mechanism affected by AAF (a
  known fluency-enhancer).
• Lidcombe program our variant only time out
  content word disfluencies and reinforce function
  word disfluencies

24
On-going work

• Priming function word should increase chance of
  content word disfluency (arrive at it earler),
  priming content word should decrease (work with
  Savage).
• Pausing prior to a PW should operate like word
  repetition (work with Hayes and Samek-Lodovici).
• Gee and Grosjean (1983), Ferreira (1991) and
  Watson and Gibson (2004) have models which allow
  pauses to occur at onset of prosodic units like
  PW.

25
References Howell, P. (2004). Assessment of some
contemporary theories of stuttering that apply to
spontaneous speech. Contemporary Issues in
Communicative Sciences and Disorders, 39,
122-139. Howell, P. Akande, O. (2005).
Simulations of the types of disfluency produced
in spontaneous utterances by fluent speakers, and
the change in disfluency type seen as speakers
who stutter get older. In Disfluency in
Spontaneous Speech. ISCA Tutorial and Research
Workshop. J. Veronis E. Campione (Eds.).Pp
93-98. Howell, P. Au-Yeung, J. (2002). The
EXPLAN theory of fluency control and the
diagnosis of stuttering. In Pathology and therapy
of speech disorders. Pp. 75-94. E. Fava (Ed.).
Amsterdam John Benjamins.