音韻障礙與句法障礙的關連性研究指導老師：張顯達研究生：許馨仁台灣大學語言學研究所

1

The Association of Phonological Disorders and Syntactic Disorders - A Study of Mandarin-speaking Children

音韻障礙與句法障礙的關連性研究指導老師：張顯達研究生：許馨仁

台灣大學語言學研究所

2

Developmental Phonological Disorders Definition -those who have phonological problems but

without evident causal origins such as speech problems secondary to mental retardation or cleft palate, yet their speech is often hard to understand (Grunwell 1991).

3

Phonological Disorders “Specific phonological disorders” (Grunwell

1991)

Coexistence of impairments in phonology and other aspects of language such as syntax and morphosyntax (e.g. Faircloth & Faircloth 1970; Leonard 2000)

4

Association-1

Faircloth & Faircloth (1970); Panagos & Prelock (1982) Spontaneous speech Production errors: sentences > isolated words control complexity in phonology and syntax More errors when processing demands increased

5

Association-2

Merino (1983) Morphosyntactic deficits often co-occur with

impaired phonology

Leonard (1989) phonetic feature: short duration & unstressed

syllable

6

Association-3

Bishop (1997) causal relationship source: receptive side of phonological

impairments

7Figure1 Schematic Representation of the Processing Components (Levelt 1993)

8

Research Questions Q1: Do children with phonological disorders

differ from those with phonologic-syntactic disorders in their phonology?

Q2:What dimension of phonological capacity or tests could best distinguish the two groups?

Q3: What syntactic problems are prone to co-exist with phonological problems in Mandarin?

9

Phonological Disordered Subjects Children with phonological disorders

34 children from three hospitals Age range: 5; 0~6;11 With developmental phonological disorders

Language ability Preschool Language Disorders Test (Lin & Lin 1993)

& Language Disorder Test for School Ages Cutting point: the 10th percentile

10

Normally Developing Controls Normally developing controls

31 children from four kindergartens age-matched group younger group

Language ability Preschool Language Disorders Test (Lin & Lin 1993)

& Language Disorder Test for School Ages Cutting point: the 10th percentile

11

PPVT-R & MLU PPVT-R

LZY (5th)and GWZ (9th) MLU

12

SubjectsTable 1 Subjects in Each Group

Groups Mean Age

Age Range

MLUw

Girl Boy Total

Phonological Disorders (PD)

69.70 5;1-6;9 2.72 14 6 20

Phonologic-SyntacticDisorders *(PSD)

71.64 5;0-6;11 2.47 9 5 14

Age 6 (NL6) 71.94 5;11-6;1 3.28 7 9 16

Age 5 (NL5) 59.87 4;11-5;1 3.09 9 8 15

* Specific language impairments

13

Tasks: Set I

-Phonological capacity Set 1 Language ability

1. Picture Naming Task Phonemic inventory

2. New Word Imitation Task & New Word Discrimination Task

Imitation and discrimination of minimal word pairs

14

Tasks: Set II

Set 2 Language Ability

1. Word Span Task Memory capacity with lexical supports

2. Nonword Repetition Task

Phonological memory capacity (without lexical supports)

-Memory capacity

15

Tasks: Set III

-Morphosyntactic & Syntactic capacityLanguage Ability

1. Classifier Elicitation Task & Classifier Learning Task

2. Sentence Comprehension Task

Morphosyntactic capacity

receptive language ability

3. Sentence Construction Task

Syntactic segmentation ability

Set 3

16

Tasks on Phonological capacity-1

Set 1 Language ability


2. New Word Imitation Task & New Word Discrimination Task


17

Picture Naming Task Show

Ask 這是什麼 ?

18

Picture Naming Task: Scoring 33 pictures for 42 target phonemes 8 tokens (4 in word-initial and 4 in word-

medial position) for each target sound were collected

One point was given for each target sound when 6 correct production out of eight were found

Maximal=42

19

Picture Naming Task: Results

NL6=NL5 > PD=PSD

Group N Mean Scores (total=42)

NL6NL5PDPSD

16 41.00 (1.21)15 38.86 (2.09)20 32.85 (5.09)14 29.64 (5.31)

Table 2 Mean scores in the Picture Naming Task

20

Variability-1

Multiple mismatches e.g.

/d/ (incorrect realization) /t/ /k/ (incorrect realization) /f/ (incorrect realization)

(Grunwell 1981; Zhu & Dodd 2000)

21

Variability-2

Alternation between a correct target and an error production was not included e.g.

/t/ (correct realization) /t/ /k/ (incorrect realization)

22

Word-based Variability Word-based variability (Zhu & Dodd

2000) e.g: /pingguo//bingguo/, /bingduo/ Total=33 One point was given for each word when more

than two types of error production of a word was found

23

Phoneme-based Variability Phoneme-based variability

e.g: /t//d/, /k/ Total=42 One point was given for each phoneme when

more than two types of error production of a phoneme was found

24

Variability Rating:Results

Word-based variability No significant difference was found between

the PD group and the PSD group Phoneme-based variability

PD < PSD PSD group at a more holistic stage

25

Tasks on Phonological capacity-2

Set 1 Language ability


2. New Word Imitation & New Word Discrimination Task


26

New Word Imitation & Discrimination Show

Imitation: 4 tokens for each word Discrimination: 4 times for each pair

bingbing dingding

27

New Word Imitation & Discrimination:Scoring

Total= 8 sets of minimal word pairs One point was given when 75% of

correctness was reached Maximal=8 Only syllable initial consonants were taken

into account in the Imitation task

28

New word Imitation: Results

NL6=NL5> PD=PSD

Group Number of Subjects Mean Scores in Set(total=8)

NL6 16 7.06 (1.38)

NL5 15 5.80 (1.52)

PD 20 4.05 (1.35)

PSD 14 3.86 (2.17)

Table 3 Mean Scores in the New Word Imitation Task

29

New Word Discrimination: Results

NL6>NL5= PD=PSD

Group Number of Subjects Mean Scores in Set(total=8)

NL6 16 6.62 (1.86)

NL5 15 5.00 (1.36)

PD 20 5.05 (1.87)

PSD 14 3.50 (1.60)

Table 4 Mean Scores in the New Word Discrimination Task

30

A Difference in Profile

0102030405060708090

100

New Word Production New Word Discrimination

(%)

Nl6NL5PDPSD

Figure 3 Proportion of Correctness in Imitation and Discrimination Task

31

Summary Picture Naming Task

NL6=NL5 > PD=PSD New word Imitation & Discrimination

Imitation: NL6=NL5 > PD=PSD Discrimination: NL6>NL5= PD=PSD

Higher variability PSD: general phonological system

Difference profile PD: motor level of articulation

32

Tasks on Memory Capacity-1





33

Word Span Task Recalling of spoken word lists ranged from

two to seven words e.g.

獅子、青蛙、斑馬、小熊 Six items were prepared for each length level When three correct repetitions out of six

were reached, move to the next length level

34

Word Span Task: Results

NL6= PD = NL5 > PSD

3.93

3.73

3.95

3.21

0 1 2 3 4 5

NL6

NL5

PD

PSD

Word (s)

PSDPDNL5NL6

(.89)

(.51)

(.59)

(.77)

Figure 4 Results in the Word Span Task

35

Tasks on Memory Capacity-2





36

Nonword Repetition Task Thirty-six nonwords, six in a set, were

repeated. e.g.

Examiner: bai3-sha 4 kang1-gu4 zhan4-dao1 One point was given for each correct

syllable Maximal=36

37

Nonword Repetition Task: Results

Group N Mean Scores (total=36)

Correctness (%)

NL6 16 18.56 (5.50) 51.55%

NL5 15 14.26 (4.62) 39.61%

PD 20 12.45 (5.09) 34.58%

PSD 14 8.78 (4.26) 24.38%

Table 5 Mean Scores in the Nonword Repetition Task

NL6> PD=PSD

38

Comparison of the Two Tasks Better lexical supports in the PD group

PD=PSD in nonword repetition task PD>PSD in the word span task

Two possibilities smaller lexical pool less efficient lexical access

39

Lexical factorsTable 6 Vocabulary size and Word span of the PSD group

Subject PPVR-R (percentile)

Word Span Subject PPVR-R (percentile)

Word Span

LZY *5 5 GAO 97 3

CAI 53 4 GWZ *9 3

SHU 87 4 LRW 29 3

LJR 27 4 ZHJ 16 3

WBK 55 4 LTQ 68 2

SON 23 3 CHY 18 2

LYZ 50 3 HYF 32 2

Correlation: r = .27, p <.05

40

Tasks on Morphosyntactic& Syntactic Capacity-1

Language Ability







Set 3

41

Chinese Classifiers Distribution -(Demonstrative)(numeral)CL (Noun) -full form: with a head noun -reduced form: without a head noun

Morphological properties -some classifiers never occur independently as a word

42

Classifier Elicitation Task Show

Twelve classifier, one general classifier and eleven specific classifier, were included.

One point was given for each target response

Ask: 這裡有多少公車 ?

( ) 四輛 / 台 / 部公車(X) 四隻公車(X ) 四公車

43

Classifier Elicitation Task: Results

NL6> PD= PSD

Table 7 Results in the Classifier Elicitation TaskGroup Target Responses

(SD)Mean Score in

Percentage

NL6 6.44 (2.63) 53.65 %

NL5 4.33 (2.82) 36.11 %

PD 3.30 (2.70) 27.5 %

PSD 2.00 (1.36) 16.67 %

44

Classifier Learning Task

這裡有多少電話 ? 這也是電話。這裡的電話有“兩具”。這裡有多少電話 ?

Step I

Step II Step III

Five specific classifiers with low frequency of use in Modern Chinese were taught

45

Classifier Learning Task:Results

NL6> NL5=PD >PSD

Table 8 Results in the Classifier Elicitation Task

Group Target Responses(SD)

Mean Score in Percentage

NL6 4.94 ( .25) 98.80 %

NL5 4.27 ( .96) 85.40 %

PD 3.55 (1.54) 71 %

PSD 2.71 (1.77) 54.2 %

46

Response Pattern: Categorizations

Types Example

1. use of specific classifier

2. use of the general classifier

3. use of inappropriate classifier 兩隻警察 ; 三個鞋子 ;

4.error classifier construction 一馬 ; 一顆兩簍 ; 一個罈酒

47

Response Pattern: Classifier Elicitation Task

0%

20%

40%

60%

80%

100%

Construction Errors 0 0 0.42 4.76Inappropriate classifier 4.17 10 5.42 12.5General classifiers 42.19 53.89 66.67 66.07Specific Classifier 53.65 36.11 27.5 16.67

NL6 NL5 PD PSD

Figure 5 Response Pattern in the Classifier Elicitation Task

48

Response Pattern: Classifier Learning Task

0%

20%

40%

60%

80%

100%

Construction Errors 0 4 4 12.86Inappropriate classifier 1.25 8 3 5.71General Classifier 0 2.67 22 27.14Target Response 98.75 85.33 71 54.29

NL6 NL5 PD PSD

Figure 6 Response Pattern in the Classifier Elicitation Task

49

Summary Classifier Elicitation Task

NL6 > PD=PSD

Classifier Learning Task NL6>NL5= PD>PSD More construction errors in the PSD group

50


Language Ability







Set 3

51

Sentence Comprehension Task 19 test items (adopted from Chang 1991) One point was given for each correct

response Maximal=19

52

Sentence Comprehension Task: Results

NL6=PD >NL5 > PSD

Group Group Mean(total=19)

S.D

NL6 18.25 1.00

NL5 15.60 2.26

PD 16.50 1.93

PSD 12.36 2.56

Table 9 Results in the Sentence Comprehension Task

53


Language Ability







Set 3

54

Sentence Construction Task Show

Expected Response: “ 爸爸開黑色的汽車”

“ 爸爸開汽車” “黑色的”

55

Sentence Construction Task: DesignTable 10 Levels of Construction in the Sentence Construction TaskStructure ExampleⅠ. Basic (NP+VP) Construction 看電視 / 爸爸Ⅱ. NP Level

ADJP +NP 爸爸開汽車 / 黑色的 Classifier+ NP 弟弟買皮球 / 三顆 Classifier +ADJP +NP 黃色的 / 妹妹畫一隻小貓Ⅲ. Clausal Level

Serial Verb Construction 爸爸梳頭髮 / 用梳子 ADVP 弟弟要去上學 / 今天

56

Sentence Construction Task: Results-1

Table 11 Results in the Sentence Construction Task

NL6=PD > NL5> PSD

Group Group Mean(total=12)

NL6 10.25 (1.18)

NL5 7.60 (2.47)

PD 8.35 (1.98)

PSD 4.29 (2.64)

57

Sentence construction Task: Results-2

0

20

40

60

80

100

Basic Noun+ VP NP Level CL Level

NL6NL5PDPSD

Figure 7 Proportion of Correctness in Each Level

58

The Factor of Position Basic (NP+VP) Construction 看電視 / 爸爸爸爸看電視 NP Level 爸爸開汽車 / 黑色的爸爸開黑色的汽車妹妹畫一隻小貓 / 黃色的妹妹畫一隻黃色的小貓 Ⅲ. Clausal Level 弟弟要去上學 / 今天今天弟弟 ( 今天 ) 要去上學

59

Summary Sentence Comprehension Task

NL6=PD >NL5 > PSD

Sentence Construction Task NL6=PD > NL5> PSD

60

Different components of disorders in phonology

Phonological differences the PSD group

Higher variability rating in phoneme production

Deficits in the general phonological system the PD group

a different profile Deficits at motor level of articulation

61Figure 8 Schematic Representation of the Processing Components (Levelt 1993)

62

Memory capacity the PSD group

Word span task & Nonword repetition task PSD < NL6

Less efficient storage and access the PD group

Word span task (PD=NL6) Nonword repetition task (PD<NL6)

63

Target 企鵝qi4-er2

海豚hai3-tun4

公雞gong1-ji1

大象da4-xiang4

Response ki4-er2 hai3-kun4 gong1-gi1 ga4-xiang4

Scoring correct correct correct correct

Scoring: Word Span Task

e.g.

Responses of CSW (PD) in the Word Span Task

64

Scoring : Nonword Repetition Task

Target ba3-gan1 chao1-dai3 ku4-shang4

Response ba3-gan1 chao1-gai3 ku4-kang4

Scoring correct-correct correct-incorrect correct-incorrect

e.g.

Responses of CSW (PD) in the Nonword Repetition Task

65

Morphosyntactic Capacity Classifier Learning Task

NL5= PD > PSD CL Elicitation Task & CL Learning Task

NL6> PD, PSD Two Possibilities

Avoidance Input frequency

66

Conclusion Receptive side of phonological disorders

would bring about difficulties to other aspects of language development

Since the children in the PD group do not show evident receptive side of phonological problems, it is possible for them to have age-expected syntactic capacity.

67

-END-

音韻障礙與句法障礙的關連性研究 指導老師：張顯達 研 究 生：許馨仁 台灣大學語言學研究所

Documents

音韻障礙與句法障礙的關連性研究指導老師：張顯達研究生：許馨仁台灣大學語言學研究所