Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

IOURNAL OF EXPERIMENTAL CHILD PSYCHOLOGY 19, 233-240 (1975)

Preschool Children’s Understanding of the


Coordinators “And” and “Or”

Bo S. JOHANSSON AND BARBRO SJ~~LIN


Uppsula University, Swederz

The development of the understanding of the words and and or was studied.
Children in the age interval 2:0-7: 6 received two tests of word understanding,
varying in the degree to which the context of the test items contributed in deter-
mining the meaning of the connectives, and one test of spontaneous usage. The
results from the tests of word understanding showed that the context variable
facilitated small children’s responding, and that most reponses were correct at the
age of four and beyond. The results from the production test indicated that and
was used to express enummerations and or to express alternatives. The difference
between the linguistic and the logical meaning of the connectives was discussed.

When studying concept formation it is generally assumed that forma-


tion of concepts involves grouping or classifying on the basis of a series
of critical attributes (e.g. Bruner, Goodnow, & Austin, 1956). It has
been found that the rules for combining the attributes can be described
in terms of logic, and results from rule learning research indicate that the
conjunctive, and, rule is easier than the disjunctive, or, rule (e.g. Neisser
& Weene, 1962; Bourne & O’Banion, 1971). Other researchers have
tried to trace the relation between these rules and the words and and or
defined as logical connectives (e.g. Neimark & Slotnick, 1970; Neimark,
1970). To study the logical and and or Neimark (1970) used stimuli with
two clearly defined dimensions, each with two attributes. The logical
connectives seem to require this type of restricted test situation since
the meaning of these terms is defined in terms of a specific pattern of
presence/absence of attributes. A positive instance of the logical and
must exhibit both relevant attributes, a positive instance of the exclusive
or must exhibit one and only one of the relevant attributes, and a posi-
tive instance of the inclusive or one or both of the relevant attributes.
Dik (1968) in his linguistic analysis of coordinators took exception to
the definition of these terms as logical connectives. The linguistic and
and or have a much wider range of application than the logical connec-
tives. To use the logical connectives it is necessary to deal with expres-
sions to which truth values can be assigned, but there is no such restric-
tion on the use of the coordinators of natural language. Dik defines and
as a combinatory term “without any further specification of the particu-
lar relation holding between the members of the coordination” (1968, p.
233
Copyright fQ 1975 by Academic Press. Inc.
All rights of reproduction in any form reserved.
234 JOHANSSON AND SJijLIN

277). Or is defined as an alternative term indicating that there is a choice


or alternative between the members of the coordination. In logic, a dis-
tinction can be made between inclusive and exclusive or, but usually
only inclusive or is used (Tarski, 1965). In natural language or seems to
imply an alternative not that the coordinated members should be taken
in combination. The alternative or does not seem to be equal to the
exclusive or, however, since the truth table is necessary for the defini-
tion of the logical or but not for the definition of the linguistic or.
The purpose of the present study was to try to trace the development
of the understanding and usage of the linguistic and and or as defined by
Dik (1968). To find out if mastery of the linguistic meaning of these
words develop before or after mastery of the logical meaning, the data
obtained in the present study can be compared with the data obtained by
Neimark and Slotnick (1970). Two tests of word understanding were
constructed, the tests differed in the degree to which linguistic and ex-
tralinguistic cues contributed in determining the meaning of the connec-
tives. Sokhin (1969) has reported results indicating that previous experi-
ence of object relationships indicated by the command is of importance
for the child’s level of performance. On the basis of this result it was
predicted that an everyday linguistic and extralinguistic context should
facilitate the small child’s performance.

METHOD

Subjects
The subjects were 26 boys and 34 girls, aged 2 :0-7: 6, with five
children in each half-year interval. The children attended different Upp-
sala public nurseries and kindergartens. The children are best character-
ized as from a middle-class background.
Materials and Procedure
All tests used a felt board and eight pictures of each of the following
kind: toys, child clothes, food, furniture, and tools. In addition there was
a picture of a boy and a picture of a girl. The children were tested indi-
vidually in a quiet room. They were seated on a carpet with the felt
board in front and the pictures laid out beside the experimenter on the
floor. The experiment was introduced by a few minutes of talk to adjust
the child to the experimental situation and to control that the child had
understood the names of the objects .depicted. The children were first
tested with the Test of Spontaneous Usage, then with the Story Test,
and last with the Put up Pictures Test. No correction was given to
responses in any phase of the experiment.
In the Test of Spontaneous Usage the child was encouraged to tell a
UNDERSTANDING “AND” AND “OR” 235

story about the pictures he chose to put up on the felt board. The in-
struction was “Here is a felt board and a lot of pictures, which you can
put up on the board and play with as you like. And while you do this I
would like you to tell me a story about the pictures”. The purpose was
to make the child produce speech. The experimenter recorded all ut-
trances with usage of and and or and made a short description of the
context of the uttrance. The aim of the test was to obtain data about the
ages at which each connective started to appear and to make an analysis
of the meaning attributed to these words in the children’s spontaneous
speech. This task lasted for 20 min.
In the Story Test the child was told a story about two children per-
forming some everyday activities, with the connectives inserted into
some of the sentences of the story. Each connective appeared in three
sentences, each time with four different pictures. These four pictures
were laid out in front of the child and the remaining pictures were set
aside. The child was asked to put up pictures on the felt board to illus-
trate the content of the sentences with the inserted connectives. The in-
struction was “Now I will tell you a story about a boy and a girl, and I
will put up pictures to show you what they are doing, and sometimes I
will ask you to put up pictures too.”
Examples of sentences in the Story Test.
(1) Richard is going out to play. Here are his toys: he has a doll, a
dog, a bear, and a boat (The experimenter shows pictures of a
doll, dog, bear, and boat). He brings the bear and the boat. Show
me what he brought out!
(2) They stop to have something to drink, They have brought coffee,
tea, lemonade, and milk (The experimenter shows pictures of
coffea, tea, lemonade, and milk). Richard wants to drink lemonade
or milk. Show me what he drank!
In the Put up Pictures Test the children were asked to put up pictures
on the felt board in accordance with different commands. Four different
pictures were used for each command, the remaining pictures were set
aside. Each connective appeared in three commands each time with dif-
ferent pictures. The instruction was “Now I will ask you to put up dif-
ferent pictures on the board.”
Examples of commands in the Put up Pictures Test.
(1) Put up the doll and the dog!
(2) Put up the car OY the bicycle!
As can be seen from the examples, the commands in the Story Test
were integrated into a wide context consisting of a story with a theme
related to a small child’s everyday activities, with key parts of the story
236 JOHANSSON AND SJijLIN

illustrated by the experimenter. In the Put up Pictures Test. on the other


hand, the commands had no relation to each other or to any illustrated
situation. Thus, the Story Test was expected to be more loaded with
everyday contextual information than the Put up Pictures Test.

RESULTS AND DISCUSSION

Comprehension
The raw data were categorized into various response types. Rrel in-
dicates that the child refused to respond to the command and showed no
signs of understanding the task. Rail indicates that the child put up all
four pictures in front of him. RI,? indicates that the child put up either
the first picture mentioned or the last picture mentioned in the com-
mand. R,,, indicates that the child put up both pictures mentioned in the
command. R,,z is considered as the linguistically correct response to the
or commands, and R,,, is considered as the linguistically correct
response to the and commands. R3 indicates that the child put up no pic-
ture, a nonmentioned picture, or a mentioned and a nonmentioned pic-
ture. The R, responses all had the characteristics that the child pro-
ducing them followed the instruction and tried to solve the tasks. Each
response could be assigned to one, and only one of the response types.
Figure 1 shows the response type variation as a function of age. The
points on the x-axis represent l-year intervals except the first point
which represent the age interval 2 : 1-2 : 6. The results indicate that the
distribution of the Rrefr Rallr and the correct response appeared in the
same order with regard to age for both words in the two tests. Rref disap-
peared at about the age of 3. The children emitting this response showed
no signs of understanding the task. For example, to the command “Put
up the dog and the bicycle!,” they looked away or continued playing
with the pictures. These children had followed the experimenter’s in-
struction as regards the Test of Spontaneous Usage, and no child used
the Rref response on all trials in the two comprehension tests. Therefore,
this response is taken to indicate lack of understanding of a specific task,
not to indicate refusal to participate in the experiment.
The Rail response disappeared at about the age of 4 : 0. Children older
than 4 : 1 made only 17% of the Rail responses produced in the tests. It
seems difficult to interpret the Rail response to indicate that the children
treated and and or as equivalents, i.e., as enummerative words, since the
response was not restricted to the mentioned pictures but included men-
tioned as well as nonmentioned pictures. It seems preferable to interpret
the R,,, response as an effect of an incomplete development of the direc-
tive function of speech (Luria, 1968). According to Luria, a child below
the age of four has difficulties in subordinating his actions to the content
UNDERSTANDING “AND” AND “OR” 237

o--o RWf
“AN 0” - “1.2
- “l*Z
-63
THE ETlRY TEST YHE Wl UP PIElURES ,ES,

IIIE SIORY IESI WE PUT UP PICIUMS TESI

FIGURE 1.

of a command, expecially when there is a conflict between the content of


the command and the external situation. In the present experiment the
child had to confine his responding to one of the four pictures, but prob-
ably the nonmentioned pictures attracted the child’s attention and made
him forget the instruction.
After the age of 4 : 1 most children demonstrated understanding of and
and or. Comparing with Neimark and Slotnick (1970), this seems to in-
dicate that the linguistic and and or are understood much earlier that the
logical and and or. Neimark and Slotnick (1970) found that the logical
and was understood by children in the fourth grade and beyond and that
the logical or was not understood until high school age.
To find out if there was any difference in level of difficulty between
the two words, an analysis was made of the individual child’s responses
to the two words in each task. A correct response to a given word was
defined as linguistically correct responding on at least two of the three
trials for that word. The subjects were divided into three age
groups:2 : l-3 : 6, 3 : 7-4 : 6, and 4 : 7-7 : 6. The age interval 3 : 7-4 : 6 was
presented separately, since the greatest increase in correct responding
seems to have occurred in that age. The results in Table 1 show, with
the exception of the Story Test results for the 2 : l-3 : 6 children and the
238 JOHANSSON AND SJiiLIN

NUMBER OF CHILDREN IN THE THRI.L AGE GROUPS RESPONDING CORRECTLY (+)


ANDIOR INCORRECTLY (-) TO ,L\IM/ ~YD 0~ IN THE STORE TEST AND THF.
Pu I CP PICTLIRES TEST

Test

Story Test Put up Pictures Test

Response pattern

and/or and/or and/or and/or and/or and/or and/or and/o1

Age interval II + + + - - + - - + + + - - + - ~

2:1-3:6 15 4 2 5 4 4 0 0 II
3:7-4:6 10 6 0 1 3 6 4 0 0
4:7-7:6 30 21 2 6 I 24 4 2 0

Put up Pictures Test results for the 3 : 7-4 : 6 children, that very few
children responded incorrectly to one word but correctly to the other.
No clear difference in level of difficulty between the two words could be
detected. The same conclusion was reached even when correct re-
sponding was defined as linguistically correct responding on all three
trials for a given word. This conclusion was supported also by a compar-
ison of the number of or interpretations (R,,J of and with the number of
and interpretations (RI+*) of or. A total of 17 or interpretations of and
and 19 and interpretations of OY was produced in the experiment. This
absence of a difference between and and or may be compared with
results from studies of logical connectives (e.g. Neimark & Slotnick,
1970), logical rule learning (e.g. Bourne & O’Banion, 1971), and logical
symbol use (e.g. Youniss, Furth, & Ross, 1971), which all show that the
conjunctive rule is easier than the disjunctive rule.
An analysis of the relative difficulty of the two tests showed that in the
age interval 2 : l-3 : 6 signifiantly more correct responses (RI+, for and
and R,,z for or) were given in the Story Test than in the Put up Pictures
Test, as revealed by the Wilcoxon’s test (T = 7.5, N = 11, p < .Ol). For
the older children there was a difference in the opposite direction but not
significantly so. To find out if the result for the younger children was in-
dependent of test order, the experiment was replicated on 10 new
children in the age interval 2 : l-3 : 6, with reversed order of the compre-
hension tests. Also this replication showed a significant superiority of
the Story Test (T = 4, N = 8, p < .05). This is taken to indicate that the
context created in the Story Test facilitated correct responding for small
children in comparison with the more context-free Put up Pictures Test.
UNDERSTANDING “AND” AND “OR” 239

Production
The data from the production test consisted of the recorded uttrances
containing and and or together with a short description of the situation
in which the children had made the utterance. As regards frequency of
usage, the word and was used by 7 out of 10 children in the age interval
2 : l-2 : 2 and by all children in the remaining age groups. The usage of OY
was much more infrequent, in the age group 2 : l-2 : 2 only one child,
aged 2: 11 used the word, in the older age groups 30-60% of the
children used the word. The following are typical examples of usages of
and: The child was looking at the pictures and said “There is a car, and
a dog, and an orange, and . . . ,” the child was looking out the window
and said “There comes Eva, and Per, and Jan, and . . . .” This seems
to be examples of an enummerative usage of and; the word is used to
enummerate a series of objects or events that are related temporally
and/or spatially. Temporal usage of and (Johnson-Laird, 1968) was not
noticed until after the age of 3 : 6, and no usage of causal and (Johnson-
Laird, 1968) could be detected. Examples of usages of or: The child was
fingering at the pictures of toys, and said “They can play with this one or
that one.” Then the child took one of the pictures and said “They will
take this one.” The child is putting clothes on one of the child pictures,
saying “Is he going to have the red one or the blue one?” Then the child
made use of only one of the two pictures. In these examples the word or
seems to be used to denote a choice situation, no usage of the inclusive
or was observed. These usages of and and OY seem to be quite close to
the linguistic definitions of these words proposed by Dik (1968). Thus,
the results from the comprehension and the production test showed that
the children easily understood and used the combinatory and and the al-
ternative or.
When discussing the generality of his definitions of the meanings of
and and OY, Dik (1968) assumed that his definitions might also apply to
other languages than English. The fact that the presented tasks, which
were based on Dik’s definitions, were easily solved by Swedish speaking
children is taken to give some support to that assumption. Dik also
analyzed other meanings attributable to and and or. He concluded that
the combinatory meaning of and and the alternative meaning of or are
the basic meanings of these words. This means that the present experi-
ment has tested only one, but presumably a central one, of the different
meanings in which and and or can be used.
The main findings were that the linguistically defined tasks in the
present experiment were solved much earlier that the logically defined
tasks in Neimark and Slotnick’s (1970) experiment, and that no dif-
ference in level of difficulty between and and or could be detected.
240 JOHANSSON AND SJiiLIN

These findings are taken to indicate that the linguistic and and or should
be kept separate from the logical and and or. If there is a difference
between the linguistic and the logical connectives and if the linguistic
meanings are acquired prior to the logical meanings, children should
exhibit tendencies to attribute the linguistic meaning to the logical con-
nectives. Error data reported by Neimark and Slotnick (1970) and
Suppes and Feldman (197 1) show that a common interpretation of the or
tasks was to choose only one of the alternatives. This may be taken to
indicate that the children tried to solve these tasks on the basis of the
meaning of the linguistic or. However, a detailed analysis of the dif-
ference between the linguistic and the logical meanings of and and or
cannot be made here. Such an analysis requires data from an experiment
in which both the linguistic and the logical meanings have been tested in
similar experimental situations.
REFERENCES
Boume, L. E., & O’Banion, K. Conceptual rule learning and chronological age. Develop-
mental Psychology, 197 1,5, 525-534.
Bruner, J. S., Goodnow, J. J., & Austin, G. A. A study of thinking. New York: Wiley,
1956.
Dik, S. C. Coordination: its implications for the theory of general linguistics. Amsterdam:
North-Holland, 1968.
Johnson-Land, P. N. “8~“. Journal OfLinguistics. 1969,6, 115-l 17.
Luria, A. R. The directive function of speech in development and dissolution, part I. In
R. C. Oldsfield & J. C. Marshall (Eds.), Language. Harmondsworth, Middlesex,
England: Penguin, 1968.
Neimark, E. D. Development of comprehension of logical connectives: Understanding of
“or”. Psychonomic Science, 1970, 21, 217-2 19.
Neimark, E. D., & Slotnick, N. S. Development of the understanding of logical connec-
tives. Journal of Educational Psychology, 1970,61,45 l-460.
Neisser, U., & Weene, P. Hierarchies in concept attainment. Journal of Experimental Psy-
chology, 1962, 64, 640-645.
Sokhin, F. A. (Reported in D. B. Elkonin). Development of speech. In V. V. Zaporozhets
and D. B. Elkonin (Eds.), Psychology of preschool children. Cambridge, MA: MIT
Press, 1971.
Suppes, P., & Feldman, S. Young children’s comprehension of logical connectives.
Journal of Experimental Child Psychology, 197 1,12, 304-3 17.
Youniss, J., Furth, H. G., & Ross, B. M. Logical symbol use in deaf and hearing children
and adolescents. Developmental Psychology, 197 1,5, 5 1 l-5 17.

RECEIVED: January 18, 1974; REVISED: May 7. 1974

You might also like