Document Sample
ft Powered By Docstoc
					   Comprehension of spatial terms of different

  syntactic roles for normal developing children

                      Cheng Man Wai

A dissertation submitted in partial fulfilment of the requirements

for the Bachelor of Science (Speech and Hearing Sciences), The

University of Hong Kong, April 30, 1998.

The      study     investigated      the         effects   of    syntactic   roles

(adjective/postposition/adverb) on the comprehension of spatial terms (±M

upper/on/upwards and T @ lower/under/downwards).                Sixty-seven normal

preschool children aged between 2;06 and 4;0 (grouped into three age groups :

2;06-2;11, 3;0-3;05, 3;06-3;11) participated in the study.          Subjects were

required to point at one object set among four options in response to the

auditory stimulus in each trial.   The number of correct trials for each type of

sentence structures was taken as a measure of the comprehension of

particular spatial terms. The results show that there was a significant effect

of 'age' but not of 'sentence type'.           Acquisition of spatial concepts and

comprehension of spatial terms are discussed in terms of semantic-syntactic

interaction, and of the syntactic differences between English and Cantonese.


     As stated by Clark (1993), "Spatial language provides an excellent

domain for studying domain-specific representations in language learning"

(P.2) due to its early appearance in language acquisition, and also to its

complexity in terms of conceptual and linguistic development.

     Since the early 70s, many studies investigated the development of

children's comprehension and the use of spatial terms.          Charney (1979)

aimed to find evidence for supporting Piagetian egocentricity in stages of

comprehending 'here' and 'there' (cited in Furrow, Murray & Furrow, 1985/86).

Clark (1973) tested for the 'semantic feature hypothesis' using spatial terms

which stated that one member is semantically simpler and thus is predicted to

be acquired earlier than the other within certain antonymic pairs such as

'above and below'.       Besides, a lot of works were done to look for the

strategies used by children before they truly acquire the spatial terms

linguistically.   It is believed that children as young as two make use of some

non-linguistic strategies to enhance comprehension.       As children are less

confident in understanding language than in understanding the extralinguistic

world, they may just neglect the linguistic information. The comprehension

may base on perceptual properties of the (stationary) object (Clark 1973), the

relation with everyday world, i.e. using world knowledge (Wilcox & Palermo

1974) as well as "the proper, or canonical relation between objects".

     Before Johnston & Slobin's (1979) cross-linguistic studies, emphasis was

mainly put on the universal conceptual influence on the development of spatial

term acquisition.     In spite of the different methodologies being adopted,

results of different studies (on different languages) came up with a more or

less universal sequence in developing the spatial term acquisition.       Johnston

&       Slobin       (1979)      stated        the      sequence          as

IN/ON/UNDER/BESIDE<BACKf/FRONTF/BETWEEN<BACK/FRONT.                             That

reflects the universality of cognitive development and its interaction with

standard communicative settings.     In 1979, Johnston and Slobin threw light

on the role of cognitive and linguistic complexity in language acquisition by

doing cross-linguistic studies on spatial terms.        They stated that since

"languages differ in the ways in which semantic intentions are mapped onto

surface utterances, cross-linguistic differences will appear in the rate and

sequence of the development of particular linguistic forms." They suggested

a number for linguistic factors which can 'promote or impede the child's search

for semantic regularity' within the limits established by conceptual growth.

These factors included (1) the position of the spatial terms (Postpositional vs.

Prepositional), (2) lexical diversity in expressing a particular notion, (3)

etymology, (4) the degree of morphological complexity of adpositions and (5)


    There were a few studies investigating the development of spatial terms in

Cantonese-speaking children. They contributed to support the notion of the

universal conceptual development.      As in Chin (1996), the developmental

sequence      in   comprehending     spatial    terms    was   as       follows    :

IN/ON<UNDER<BEHlND<IN           FRONT      OF<BETWEEN.             It   supported

Johnston & Slobbin's results.

    "Language studies of spatial representation in children have focused on

predicates which specify the location of objects" (Johnston, 1988).            Much

effort was devoted to the dimension of varied spatial notions by using

adpositions while little emphasis was put on the dimension of varied syntactic


     My interest to look at the syntactic aspect of spatial terms arose from

some clinical observations.      It was observed that some language-delayed

children performed differently in comprehending a certain spatial term in two

different sentence structures which are of similar syntactic complexity.        For

instance, a child could follow the command        'MMVLJLW      (Put on the table.)

while could not follow the command 'jjjj hM@£HF (Take the upper doll.).

The spatial term 'JtET plays as a 'postposition' in the first command while as

an 'adjective' in the second command.          These lead me to the research

question 'Do syntactic roles exert an effect on spatial term comprehension?'.

     There is no standardized protocol for spatial terms assessment for

Cantonese speakers. There is no agreed priority of linguistic contexts for the

pratising speech therapists to follow when conducting spatial term treatment,

though it is relatively common to apply spatial terms in the 'postposition' within

the place adverbials.

          In English, of linguistic contexts the spatial concept relating to a

particular position / space can be expressed by various surface linguistic forms.

In other words, different syntactic roles of a particular spatial position are

expressed differently.    On comparing the three examples, (1) The ball is on

the table.', (2) 'Look upwards.1, and (3) 'Point at the upper bed.', the terms 'on',

'upwards' and 'upper' on one hand express similar notion of concept on the

other hand play different syntactic roles.   'On' acts as a preposition, 'upwards'

acts an adverb while 'upper' acts an adjective.     In one dimension, the spatial

term 'on' appears at an early position in the developmental sequence with

different locatives.   In the other dimension, it may appear somewhere along

the developmental sequence with the spatial terms 'upwards' and 'upper'.                   It

is probable that children acquire the three different terms with a sequence due

to their difference in surface forms and thus different frequency of occurrence

as well as other concurrent factors.        It was found that children acquire 'on'

around two to two and a half years old (Johnston & Slobin, 1979) while acquire

'above' around three to seven years old (Durkin, 1981).            Both 'above' and 'on'

can act as prepositions, the rate of acquisition for the two terms are very

different in spite of their similarity in terms of syntactic roles and conceptual

involvement. Thus it's reasonable to believe that children acquire 'on', 'upper'

and 'downwards' at different rates due to the differences in surface forms as

well as syntactic roles.

     Similarly to English, the spatial concept about ' l i ' (the space or position

relating to lop' / 'up' / 'above') in Cantonese can be applied onto language

differently in terms of syntax.   It can act as a postposition as in 'feMWUzW

(The ball is on the table.), as an adverb as in lfM±M        ( L o o k upwards.), and as

an adjective as in     ^B±MM^        (Point   at the u
                                                         PP e r   bed
                                                                        ) - However, the

surface form of the spatial term itself cannot give information about the

syntactic role it performs, unless the syntactic structure of the whole utterance

is viewed as well because the surface linguistic forms of the spatial concept

are homonyms for all three underlying notions.       Base on the fact that children

prefer   one-to-one mappings between semantic concepts and surface

morphemes (Slobin 1977), "homonymity" should impose a measure of

linguistic difficulty on acquisition (Johnston & Slobin 1979). While the terms

do not differ with each other in the surface level, do the syntactic roles that the

spatial term play impose any measure of difficulty on encoding the spatial


     Johnston & Slobin (1979) states that there is always a duration of lag

between the emergence of communicable notion in a child and one's mastery

of the proper means for the encoding of the intention linguistically and the

temporal gap in acquisition is determined by language-specific structural

features which affect the ease of perceiving the relevant forms or discovering

mapping conventions between forms and contents.         "Hymonymity", using the

same surface forms to express different underlying notions, should impede the

acquisition of spatial terms.    On the other hand, the presence of surface

locative markers 'Hf increases the ease of encoding spatial terms.         Which

factor exerts greater effects onto the spatial term acquisition for normal

Cantonese-speaking children?

    I would like to focus on two positions, *±M' (on) and \ T ® ' (under) and

extend the investigation in the dimension of varied syntactic roles (adjective /

postposition / adverb).     In comprehending these two positions, it is not

required to take speaker or listener's perspective into account.          As the

'syntactic role1 should exert its effect in whichever position, positions that

involve less complication Con1 and 'under) than in front of' and 'between' etc.

are used.

    The objective of the study is to test the hypothesis that 'syntactic roles do

exert effects onto spatial term comprehension'.      That is the ease of spatial

term comprehension is different with different syntactic roles.   It also aims to

provide guidelines in doing spatial term assessment and intervention.



       Sixty-seven children aged between two year and a half and four years old

were selected randomly from two nurseries in the predominantly middle-class

areas (Shatin and Wang Tau Horn) in Hong Kong. All the children were

native speakers of Cantonese. They were considered to have no sensory or

developmental deficit by their teachers.     The children were divided into three

age groups at six-monthly intervals.        There were about twenty children in

each age group.       Table 1 shows the age and gender distribution of the


Table 1 Age & Gender distribution of the subjects

                            Number of                   Standard
 Group       Age Range                      Mean Age                 Boy /Girl
                             Subjects                   Deviation

   1          2;06-2;11         20             2;10       ±0.14        7/13

   2          3;00 - 3;05       24             3;04       ±0.15        12/12

   3          3;06-3;11         23             3;09       ±0.16        14/9

Overall       2;06-3;11         67             3;03       ±0.41        33/34


    Both the auditory stimuli and object stimuli were included in the

experiment.     For the auditory stimuli, two types of sentence structures were

involved in the priming task while three types of sentence structures were

involved in the experimental task.    For the object stimuli, toys that children

were familiar with and interested in were used. Table 2 shows all the object

stimuli and Table 3 shows examples of auditory stimuli for both the priming

task and the experimental task.

Table 2 Object stimuli for both the Priming Task & the Experimental Task

                                      Priming Task                   Experimental Task

Placed objects                Hats                   (2)      Shoes                  (4)

                              Flowers                (2)

[Reference objects            Tables                 (2)      Double-deck beds       (2)

                              Sofas                  (2)      Single-deck beds       (2)

Agents                        Dolls                  (4)      Dolls                  (4)

Table 3 Auditory Stimuli for both Priming Task & Experimental Task

                                      Priming Task

Type I               ft        %

SA structure         flower    at     (classifier)    table

                     The flowers are on the table.

Type II              &# m               fi
SVO structure        doll     point    (asp. marker)       table

                     The doll is pointing at the table.

                                   Experimental Task

Type A -

Adjective'           shoe     at     upper/lower      (classifier)     bed

structure            The shoe is on the upper/lower bed.

(SA structure)
TypeB -                            tff                  m          ±mrfm
'Postposition'       shoe          at    (classifier)   bed         on/under
structure            The shoe is on/under the bed.
(SA structure)

Type C -             &   •
                             f?      m            fi                ±[g/Ti

'Adverb' structure doll             point    (asp. marker)       upward/downward

(SVA structure)      The doll is pointing upwards/downwards.


    The experimental procedures were carried out in the rooms of the

nurseries.   The subjects were brought into the room one by one. A few

minutes were used to build up rapport with each subject through talking.

    The experimenter and the subject sat ninety degree to each other at a

table.   The subject was required to point at one object set among four options

in response to the experimenter's verbal instructions in both the Priming Task

and the Experimental Task.

    Vocabulary screening

     Before the investigation, the subjects' vocabulary was checked to see if

they knew the names of object stimuli including 'table', 'sofa', 'hat', 'flowers',

'bed', 'shoe' and 'doll'.         The experimenter placed the objects on the table one

by one and asked, 'What is it?' (t2 lr f £?).                 If the subject did not respond to

the question verbally, auditory-object matching would be done as no

production from the child would be required and measured. Three objects

would be placed on the table. The experimenter asked, 'Which is a table?"

 (iif@f£S^?).       All the sixty-seven subjects passed the vocabulary screening.

      Priming Task

     A priming task was done to check the subjects' comprehension in terms of

 syntax and auditory memory span.       It served as a control for the structures in

 the Experimental Task by controlling the syllabic length and syntactic

 complexity.    Only the subjects who achieved 80% criterion for both Subject-

Adverbial (Type I) and Subject-Verb-Object (Type II) structures would undergo

the Experimental Task, otherwise, the procedures would be suspended. All

sixty-seven subjects met the criterion and their results were used for analysis.

     There were two series of object sets and four object sets in each series.

There were two types of auditory stimuli corresponding to two different object

series.    There were four schedules for each type of auditory stimuli. One

trial was given for each schedule except that one of the schedule was tested

twice.     Table 4 shows the auditory stimuli and corresponding object series

used in the Priming Task.     Please refer to Appendix I (Recording Sheet) for


Table 4 Auditory Stimuli & corresponding object series for the Priming Task

                     Auditory Stimuli               Choices of Object Sets

1 Type I                                       1. The flowers are on the table, I

   SA       The flowers are on the table. 2. The flowers are on the sofa.

            where are they?                    3. The hat is on the table.

                                               4. The hat is on the sofa.

          The flowers are on the sofa,

          where are they?

          m%\        m - mmm
          The hat is on the table, where is


          mmmtm          > mmm
          The hat is on the sofa, where is it?

Type II   wmm             iftM^f??               1.   The doll is pointing at the

 SVO      The doll is pointing at the table,          table.

          which doll?                            2.   The doll is pointing at the


                                                 3.   The doll is watching at the

          The doll is pointing at the sofa,
                                                 4.   The doll is watching at the
          which doll?

          £&m.m           mm&m
          The doll is watching at the table,

          which doll?

          0*1             &m/^

          The doll is watching at the sofa,
               which doll?

         Experimental Task

         The procedures were similar to that of the priming task.        There were

three       types   of    auditory   stimuli   (Adjective /   Postposition / Adverb)

corresponding to three different object series. Two antonymic positions (e.g.

'on' versus 'under') were included in each type of structures and five trials were

given for each position. The sequence of items within each type of structures

was randomized.           After every five trials, the order of arrangement of the

object sets would be altered to prevent the subjects from rote-memorizing the

target positions.        Table 5 shows the auditory stimuli and corresponding object

series in the Experimental Task.

         The auditory stimuli would be repeated once if the subject did not respond

within five seconds in any trial.       No more than two repetitions were used in a

trial.     The experimenter would postpone presenting the auditory stimuli if the

subject did not attend to the task or pointed at any object set before the whole

auditory stimuli was presented.

         Brief instructions about the task were given before starting a new type of

auditory stimuli. They were as follows:

'RSSW — * — * H * E * E ! ! 5 t J I ' ^ O H f ± > ? ! # « °'
'Here are one, two, three, four, four sets of toys, listen carefully and point for


         For arrangement of the object sets within each series for both the priming

task and the experimental task, please refer to Appendix II.

Table 5 Auditory Stimuli & corresponding object series for Experimental Task

                     Auditory Stimuli                  Choices of Object Sets

 Type A                                          1. The shoe is on the upper bed

Adjective The shoe is on the upper/lower] 2. The shoe is on the lower bed.

            bed, which shoe?                     3. The shoe is on the left bed.

                                                 4. The shoe is on the right bed.

 Type B                                          1.   The shoe is on the bed.

   Post-    The shoe is on/under the bed 2.           The shoe is under the bed.

 position which shoe?                            3.   The shoe is beside the bed.

                                                 4.   The shoe is in front of the


 TypeC                                           1. The doll is pointing upward.

 Adverb The           doll      is       pointing 2. The doll is pointing downward.

           upward/downward, which doll?               The doll is pointing forward.

                                                 4. The doll is pointing backward.


     There were totally thirty items in the experiment.            Each item was

categorized, depending on the syntactic role that the spatial term played.

The conditions were 'upper7, lower', 'on', 'under', 'upward' and 'downward5.

The sequence of the items within each type of structures was randomized.

For details of the record form, please refer to Appendix.     Each response to an

item in the experimental task was scored correct if the subject pointed at the

object set that matched to the auditory stimuli in a closed-set of four.      Each

subject had a maximum possible correct response score of ten for each type

of structures, i.e. five for each position.


     The subject scored one point for each correct response.             Each subject

had a maximum possible correct response score of ten for each type of stimuli

structures (i.e. five for each position).

Main Effects and Interaction Effects

    Table 6 shows the mean numbers of correct responses under each type

of structures for all three age groups.           A 3(Age) X 3(Type) X 2(Position)

analysis of variance (ANOVA) was performed. Age was a between-subject

variable while Type and Position were within-subject variables.               Overall

performance of subjects in these groups differed significantly with Age and

Position, Age:[F(2,61)=10.11; p<0.0002]; Position:[F(1,61)=32.41; p<0.000].

Figure 1 shows the Age effect.       However, there was no significant Type effect

Significant interactions were found between Age and Type [F(4,122)=3.76;

p<0.0064] as well as between Type and Position [F(2,122)=12.18; p<0.000].

Table 6 Mean Scores for the Experimental Task

                        [                        Mean Scores                        "1

                               Type A                Type B          Type C          I
   Age Group
                        ! upper     lower         on     under   upward downward I

         1                  3.30    3.20         3.70     1.85    2.50       2.50

         2                  3.54    3.08         4.38     3.38    3.88       3.75

         3                  4.39    3.74         5.00     3.26   4.78        4.17

                                      Figure 1: Plot of Means
  Score                               AGEGROUP Main Effect
                                      F(2,64)=11 80;p<.0000
  4.4 .—.    _     -.                              ,                                 ,

  4.2                                                                          ^ - ^ °

  3.8                                                        ^   ^   ^
  36                                             ^       ^

  3.2                             ^    ^

                          ^ ^ ^
  28                °"^
  2.6 I                                              '
                 Group 1                       Group 2                          Group 3

       A post hoc comparisons using Tukey test was made to investigate the

Age and Position effects within each age group.                      Figure 2 shows a two-way

interaction between Age and Type.              It was found that, for the comprehension

of Postposition' structures ('on' and 'under') as well as 'Adverb* structures

('upward' and 'downward'), subjects in Group 1 performed significantly worse

(p<0.005) than those in Group 2 and Group 3.                             No significant difference

between the other groups for the task was observed.                           Figure 3 shows the

interaction between Position and Type.                   Results demonstrated that subjects

performed significantly worse in the 'under1 structure than in the 'on' structure

for 'Postposition' structures but no significant Position effect was found in

'Adjective' and 'Adverb' structures.

                               Figure 2: Age & Type
                                 2-way interaction
                              F(4,122)=3.76; p<0064





                                                                                 - o - AGE GP 1

                                                                                 •o   AGEGP2

          Type A-adjective       Type B-postposition       Type C-adverb         - + - AGE GP 3

                     Figure 3: Position & Type

  Score                    2-way interaction
                     F(2,122)=12.18; p<.0000
  2.8                                                                      ±M
  2.6                                                                      upper/on/upward
        Type A-adjective                          Type C-adverb     --Q-   T®
                           Type B-postposition                             lower/under/downward

Tukey test revealed that Position effect was significant for Group 3 subjects

only (p < 0.0005) while no significant Position effect for Group 1 and Group 2

subjects as shown in Figure 4.                   Figure 5 shows a three-way interaction

among Age, Type and Position. Tukey test revealed that Type effect was

significant in the position of 'fW                 ('bottom' / 'down') for Group 1 subjects (p <


                          Figure 4: Position & Age

    Score                    2-way interaction
I     V?





    2.5             o'                                                   -o- ±B                      1
                                                                              upper/on/upward        1

                Group 1            Group 2               Group 3              lower/under/downward   1

                             Figure 5: Age & Type & Position
                                     3-way interaction
                                  F(4,128)=.84; p<.4999

                                                                                    -O   AGEGP1
          1.5                                                                       - £ > AGE GP2
           TYPE         Postposition             TYPE        Postposition
              Adjective              Adverb        Adjective              Adverb    • O * AGE GP3
                        POSITION                             POSITION
                          ±H                                    Tffi

 Error Responses

 For each auditory stimuli (e.g. 'The shoe is on the bed.'), there were four

options of object sets. They included one target object set (The shoe is on

the bed.), one object set corresponding to the antonymic partner (The shoe is

under the bed.), and two object sets that did not relate to the target

antonymically (The shoe is in front of / beside the bed.). The error responses

were categorized into two types : (1) Antonymic (A) and (2) Non-Antonymic

(NA).    Table 7 shows the distribution of error types for all age groups and

types of structures.

Table 7 Distribution of error types

             % of error types out of total no. of errors (no. of occurrence) I

                  Adjective             Postposition         I          Adverb         I

I Group         A          NA            A          NA              A            NA    I

     1       24.6(17)   75.4 (52) 20.2(18)       79.8(71)        20 (20)    80(80)

     2       11.1 (9)   88.9 (72)     16.7 (9)   83.3 (45)       17.5 (10) 82.5 (47)

     3        2.3(1)    97.7 (42) j   2.5(1)     97.5 (39)        0(0)     100 (24)

Results showed that subjects were more likely to choose Non-Antonymic

members than Antonymic member in all age groups and sentence types.

The percentage of NA increased as age increased.         However, there was no

significant difference in error pattern among sentence types.


Type Effect

     There was no significant Type effect in the experimental task.          This does

not support that hypothesis that 'syntactic roles (adjective/postposition/adverb)

of spatial terms ( ± ® / T ® ) do exert effect on spatial term comprehension.'

     Do the results imply that the syntactic roles of spatial terms do not impose

any measure of difficulty in spatial term comprehension?            Or are there any

other concurrent factors that may counter balance with the effect of syntactic

roles of spatial terms?

     The spatial terms 'upper', 'on' and 'upward' on one hand differ from each

other in terms of syntactic roles, on the other hand they involve different

semantic concepts and represent different meanings.                Weist & Lyytinen

(1991) showed that 'the capacity to process monoreferential locative

relationships emerges before Preferential locatives which 'require coordinated

spatial relationships involving referent objects and a projective relationship',

and thus locatives like 'on' and 'in' emerge prior to 'front' and 'between'. In

the same sense, the spatial terms 'on' and 'under' should emerge before

'upper' and lower', which are also bireferential, respectively if semantic and

cognitive factors are considered.

     Quirk et al. (1972) stated that 'under' is conceptually more complex than

'on' since it is referred to a locative of relative position, i.e. it implies direction

in addition to locate the object continguous to the reference object'. This may

also apply to 'upper' and lower'.       The fact may impose certain measure of

difficulty for the comprehension.

     The position of spatial terms may also exert effect on the encoding of the

terms.    For 'postposition' and 'adverb' structures, the surface markers of

spatial concept occur at the sentence-final position that may facilitate the

accessibility of the spatial terms in terms of 'recency effect'.     The surface

markers of spatial concept in 'adjective' structures occur in a relatively

embedded position that may impede children's grasp of concepts.

     Moreover, the more components and more complex arrangement of

object sets for 'adjective' structures may decrease perceptual salency of the

placed objects.    The object series corresponding to 'adjective' structures

involved two double-deck beds and four single-deck beds while those

corresponding to 'postposition' and 'adverb' structures involved four single-

deck beds and four dolls respectively.

     For 'adjective' structures, lW=tf%±MH:W]M!iM' (The shoe is on the upper

bed.), the resolution of spatial concepts is more complex.     There is a spatial

relation between the placed object 'shoe' and the reference object 'bed'. And

there is an additional spatial relation specified by the term 'upper' or lower'

between the two beds of double-deck beds. Though the surface structures

do not demonstrate this complication in Cantonese, it may still cause more

struggling for children in selecting options at the underlying level when

comprehending 'adjective1 structures.

    Further analysis showed that Group 1 subjects performed significantly

better for the lower' structure in 'adjective* structures than the 'under'

structure1 in 'postposition'.   This finding supports idea that the ease of

comprehension is different for spatial terms of the same surface forms but

different underlying forms, though it can only be applied to certain position ( T

I®) and certain stage of development (the youngest group).             As different

language levels (such as semantic and syntactic) exercise mutual influence

without priorities (Crystal 1987), it is difficult to view the effects of one single

factor alone.

Age & Type

     As shown in Figure 2, subjects in all age groups did not differ significantly

in comprehending 'adjective' structures involving 'upper & lower'.        Group 1

subjects performed significantly worse than Group 2 and Group 3 subjects in

comprehending 'postposition' and 'adverb' structures, which involve 'on &

under' and 'upward & downward' respectively, while the performance in

'postposition' and 'adverb' structures between Group 2 and Group 3 subjects

had no significant difference.    The difference between Group 1 and older

subjects in 'postposition' and 'adverb' structures shows an effect of

development by sentence type for Group 1 versus Groups 2 and 3. And it

may indicate that acquisition of spatial terms of that particular syntactic roles

occurs during the period between 2; 10 and 3;04.      There was no significant

difference in comprehending spatial Adjectives, spatial Prepositions and

spatial Adverbs within any age groups, and a definite developmental trend

cannot be obtained.    However, from the results, it was possible that children

put different weight of emphasis in comprehending spatial terms of different

syntactic roles during the process of acquiring spatial terms since children

learn new usage as their repertoire of spatial concepts expands (Johnston


    In the earliest stage, when children have just started to establish the

spatial concept and increase their comprehension level in terms of syntactic

complexity and variability, they may find it easier to encode the spatial terms

that take the role of Adjectives. As children encounter more different types of

syntactic structures, they may just shift their attention to some new structures

and thus their ability of encoding spatial terms taking the roles of Postpositions

and Adverbs increases critically.       As children become more linguistically

mature, the linguistic capacity is not as limited as before, they can manage to

cope with spatial terms of varied syntactic roles at the same chronological

stage.      The meanings of surface lexemes to the children change

systematically across the age (Johnston, 1988).

Type & Position

     As shown in Figure 3, children performed significantly better in 'on1

structures than in 'under' structures for 'postposition' structures while Position

effect was insignificant for 'adjective' and 'adverb' structures.     The findings

were consistent with that in most of the previous studies.       Different theories

were proposed to explain the phenomenon.               They included perceptual

salency (Clark, 1972), praxic aspect (Van Geert, 1985) that is not applicable in

this study as no object manipulation was involved, as well as conceptual

complexity, for instance, 'under' is referred to a locative of relative position and

thus is conceptually more complex than 'on' (Quirk et aL, 1972). The two

positions in either 'adjective' structures or 'adverb' structures do not differ with

each other in terms of perceptual salency and cognitive concept. Thus no

significant Position effect was expected.

Age Effect

     It was expected that children in older age group performed significantly

better in the experimental task which involved the comprehension of spatial

terms     Both the linguistic and experiential maturity contribute to the

increased ability in comprehending spatial terms as age increases (Johnston,


     In Chin (1996), the author obtained a developmental sequence of

locatives in Cantonese-speaking children using tasks involving object

manipulation.    Results showed that children aged 2;03 to 2;09 achieved 'on'

with over 90% accuracy and children aged 3;03 to 3;09 achieved 'under' with

over 90% accuracy.     In my study, the accuracy rate in 'on' structures was

87.6% for children aged between 3;0 and 3;05.     The accuracy rate in 'under5

structures was 65.2% for children aged between 3;06 and 4;0. The great

discrepancy can be accounted by reduced 'perceptual salency' of placed

objects within a relatively complex object series. The factor of 'perceptual

salency' was even dominant over 'praxic effect1 (Van Geert, 1985).    In terms

of syntactic complexity, the auditory stimuli used by two parties were more or

less the same.     The auditory stimuli used by Chin (1996) were 'Verb-

Adverbial' structure while that used in this study were 'Subject-Adverbial'


                          Limitations & Further Research

     In the study, the possibility that children might use key-word strategies in

comprehending the auditory stimuli was not ruled out.          Whenever children

could pick up the spatial term from the auditory stimuli, what they might do was

to match the spatial term to any one of the object sets.      For instance, in the

'adverb' structures,     ' £> f? J | f± ± jff / T ® ' (The doll is pointing

upwards/downwards.),       children   may only     need to match the terms

'upward/downward' to the direction of dolls' hands without comprehending the

whole structures.    This factor can be eliminated if the object sets within a

series differ from each other by various 'agents' and 'actions' on top of 'spatial

relation'.   This modification is suggested for further research to look for

possible 'syntactic-semantic interaction' in spatial term acquisition.

     During the procedures of this study, only two object sets were targeted

within every object series of four options.       However, children have high

tendency to include all options within a selective task. And this may also

account for the high percentage of Non-Antonymic errors.         To eliminate this

task effect, it is suggested to include non-targeted auditory stimuli as well.

     Moreover, it is suggested to investigate the production of spatial terms in

Cantonese-speaking children in order to obtain a more comprehensive picture

for spatial term acquisition.

                                 Clinical Implication


     To define developmental stage more distinctively, it would be more

informational to use spatial terms that play the roles of 'postposition' and

'adverb' in doing the spatial term assessment           It is because there is a time

lag between Group 1 and older subjects in these two types of structures.

     Taking 'perceptual saliancy5 and 'chance level within selective tasks' into

consideration, it is suggested to adopt object manipulation in the procedures

during spatial term assessment provided that the children have normal motor



    As there is not a definite developmental trend among spatial terms '±:W

and ' T ® ' of 'adjectives', 'postpositions' and 'adverbs', there would no need to

consider developmental factor in choosing specific syntactic contexts.         But it

is suggested to treat a particular spatial term with a fixed syntactic role, so that

the others can act as the generalization goals (contexts).


    I would like to express my sincere thanks to the supervisors, the teachers

and the children of Hong Kong YWCA Lung Hang Nursery as well as Lok Fu

Rhenish Child Care Centre for their kindly support and consideration during

my data collection.

    I would like to send my special thanks to Dr. Valter Ciocca and Miss Diana

Ho and I am grateful to all the other lecturers of the Department of Speech and

Hearing Sciences for their precious advice and comments on my study.

    I would also like to thank my family, Mr. Stanley Wong, Miss Trinnae Pang

as well as my dearest fellow classmates, especially Genee and Amy, for their

psychological and spiritual support through comforting words and prayers.

    Lastly, I would like to thank God for His grace and guidance in my times of

being helpless and lost.


Balvin E. (1990) Locative terms and Warlpiri acquisition. Journal of Child
Language 17, 43-66.

Chen C. Y. (1978) Aspectual features of the verb and the relative positions of
the locatives. Journal of Chinese Linguistics 6. 76-88

Chin W.     (1996)    Comprehension of locatives in Cantonese children.
Dissertation for the Bachelor of Science (Speech & Hearing Sciences). Vol I.

Clark E. V. (1993) Learning the Language of Space. The Proceedings of the
Twenty-fourth Annual. Child Language Research Forum. 1-22.

Crystal D. (1987) Towards a 'bucket' theory of language disability: taking
account of interaction between linguistic levels. Clinical Linguistics &
Phonetics 1. 7-22.

Durkin K. (1981) Aspects of late language acquisition : school children's use
and comprehension of prepositions.      First Language, 47-59.

Furrow D., Murray P. & Furrow M. (1985/86) Spatial term use and its relation
to language function at two developmental stages. First Language 6. 41-51.

Geert V. P. (1985/86) In, on, under : an essay on the modularity of infant
spatial competence.    First Language 6, 17-30.

Hoogenraad R., Grieve R. R. & Murray D. (1977) On the young child's use of
lexis and syntax in understanding locative instructions. Cognition 5. 235-250.

Johnston J. R. (1988) Children's verbal representation of spatial location. In
Stiles-Davies J., Kritchevsky M. & Bellug U. (Eds), Spatial Cognition. Hallsdale,
NJ : Lowrence Erlbaum Associates Inc.

Johnston J. & Slobin D. I. (1979) The development of locative expressions in
English, Italian, Serbo-Croatian & Turkish. Journal of Child Language 6. 529-

Lee R. (1995) Young children's comprehension of the locatives [±M] 'on' &
[ T ® ] 'under'. Dissertation for the Bachelor of Science (Speech & Hearing
Sciences). Vol II.

Masterson J. J. & Kamhi A. G. (1992) Linguistic Trade-Offs in School-Age
Children With and Without Language Disorders. Journal of Speech and
Hearing Research 35. 1064-1075.

Tomasello M. (1987) Learning to use prepositions : a case study. First Child
Language 14. 79-98.

Wiest R. M. (1991) Spatial and temporal location in child language. First
Language 11. 253-267.

Weist R. M., Lyytinen P., Wysocka J. & Atanassova M. (1997) The interaction
of language and thought in children's language acquisition : a crosslinguistic
study. Journal of Child Language 24. 81-121.

Weist R. M. & Lyytinen P. (1991) The Development of Spatial Location in
Finnish. International Journal of Psychology 26 (3). 345-56.

Appendix I

Recording Sheet                                            Date of recording
D.O.B.    :         Age               Gp 1/2 / 3           Sex :      M/F

Priming Task
Control - Screening (criteria         80% accuracy)
       Type I                   Results                 Type II            Results
    SA structure                                    SVO structure
   (for Type A &                                     (for Type C)
The flower is on the                            The doll is watching
sofa.                                           at the sofa.
                                                &   ftm&m
The hat is on the                               The doll is pointing!
table.                                          at the table.

The hat is on the                               The doll is pointing]
sofa.                                           at the sofa.

The flower is on the                            The doll is watching
table.                                          at the table.
The hat is on the                               The doll is pointing
table.                                          at the table.       I

If ^ 80% for both Type I & Type II, Go to Experimental Task
If < 80% for either Type I or Type II, Suspend procedures

 Experimental Task

 Type A 'Adjective' structure
 Verbal instruction: H %±M[TMM^              ESS-'
                        The shoe is on the upper/lower bed, which shoe?
                   Child's choice                                    Child's choice
 Target    ±®       T®      £®                      Target    ±® 1 T® £ ®
           upper   lower     left           right             upper i lower    left   right
  ±®                                                 T®
  upper                                              lower
  ±®                                                 ±®
  upper                                             upper
  lower                                             lower
  ±®                                                 T®
 upper                                              lower
  T®                                                 ±®
 lower 1                                            upper 1

Type B 'Postposition' structure
Verbal instruction: WMM^=tM[TM        ' Slti^
                      The shoe is on/under the bed, which shoe?
                   Child's choice                                    Child's choice
           ±® 1 T® itt® M®                          Target
                                                              ±m 1 T®        F!U®
            on     under in front beside                       on    under in front beside
                            of                                                of
  T®                                                ±®
 under                                               on
 ±®                                                 T®
   on                                               under
 ±®                                                 ±®
   on                                                on
 T®                                                 ±®
 under                                               on
 T® ;
1 under                             L
                                                    under j                1            J
Type C 'Adverb' structure
Verbal instruction: £&ffi£±MniM.             ' M4M??

                    Child's choice                                 Child's choice

             ±m     T®               m®       Target
                                                        ±®         T®
                                                                           ..,1 f „ — / ' • *

              up-   down- for-       back-               up-   down-       for-                   back-
             ward   ward 1 ward      ward               ward   ward        ward                   ward
 ±m                                            T®
  up-                                         down-
 ward                                          ward
 TM                                            -h®
 down-                                         up-
 ward                                          ward
 ±M i                                          TM
  up-                                         down-
 ward                                          ward
 ±®                                            T®
  up-                                         down-
 ward                                          ward
 TM                                            ±®
down-                                         up-
 ward j                                       ward 1

Summary of Scores
                     Type A                    Type B          r          TypeC
                     Adjective               Postposition                 Adverb                          |
  Spatial                                                                                       down-
                upper       lower        on            under       upward
   term                                                                                         ward
 Score (5)

Total (10)

Appendix II      Display of object series

For Priming Task,

For 'Adjective' structures (Type A),
Appendix II      Display of object series (Con't)

For 'Postposition' structures (Type B),

For 'Adverb' structures (Type C),
Appendix III         Correspondence letters
                                                                         February, 1998

Dear Ms.

     I am a Year 4 student in Speech and Hearing Sciences at The University of Hong
Kong. I am writing to seek your help in a final year project. I am looking for
children in Nursery and aged from two to four years. I would like to investigate how
sentence structures may interact with children's understanding of spatial concepts. I
believe the results of this study may help speech therapists in doing language
assessment and setting intervention goals.

     I need to spend 15-20 minutes with each child.    The children will be required to
point at objects in response to verbal instructions.

     All the information I gain from the study will be kept confidential and no
children will be identifiable from my results recorded. I need, in all, eighty children.
I understand that there is a possibility that there may not be enough children from your
nursery but I would greatly appreciate as much help as possible from you.

     If you do kindly allow me to seek the children from your nursery, would you
please sign the reply slip with this and return it to me. I will be able to come at
almost any time your schools are open from 2nd February to 26th February, 1998.

     In case you may want to know more about my study, it may be helpful to mention
my contact numbers.       They are 23284823(home), 7389 0122(pager) and 2328

     Thank you for your kind consideration.      I look forward to hearing from you

Yours Sincerely,

Cheng Man Wai

                                               2 2 MAR 2006

                                                                February 1998

Miss Cheng Man Wai
Department of Speech and Hearing Sciences
3416 Cheung Yuen House
Chuk Yuen North Estate

Dear Miss Cheng,

     Our nursery *is willing / is not willing* to participate in the study of
development of spatial concepts in different sentence structures.

     *Please, delete appropriately

You may come on the following dates and times for particular kindergartens :

        Nurseries                           Dates                   Time

    Please specify above the dates and times that you would like the study to be

                                     Name of Coordinator