The Development of Time Concepts in Young Children: The Relations between Duration

and Succession
Author(s): Iris Levin, Etziona Israeli and Ephraim Darom
Source: Child Development , Sep., 1978, Vol. 49, No. 3 (Sep., 1978), pp. 755-764
Published by: Wiley on behalf of the Society for Research in Child Development

Stable URL: https://www.jstor.org/stable/1128245

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Society for Research in Child Development and Wiley are collaborating with JSTOR to digitize,
preserve and extend access to Child Development

This content downloaded from

149.132.77.163 on Tue, 28 Feb 2023 08:15:13 UTC
All use subject to https://about.jstor.org/terms
The Development of Time Concepts in Young
Children: The Relations between Duration
and Succession

Iris Levin, Etziona Israeli, and Ephraim Darom

Tel-Aviv University

LEVIN, IRIS; ISRAELI, ETZIONA; and DAROM, EPHRAIM. The Development of Time Concepts
Young Children: The Relations between Duration and Succession. CHILD DEVELOPMENT, 19
49, 755-764. 108 children from nursery school, first grade, and third grade were given 5 pro
lems measuring the concept of time, in which they were required to judge and explain whic
of 2 partially overlapping events started first, which ended first, and which lasted for a lon
time. 2 types of problems were used which presented time with and without the interferen
of movement: "still time" (the duration of 2 lights) and "linear time" (the traveling time
2 toy cars). In both types of problems, judging the succession of both beginnings and endin
was equally difficult and was easier than judging duration. Duration judgments were explain
predominantly by succession, whereas succession judgments were explained tautologically or b
mere "seeing." Overall, performance increased with age and decreased with interference.
recent model of the development of time concepts is further elaborated and compared w
Piaget's framework.

Major insights into the ontogenesis of comes to grips with their necessary logical re-
time concepts have been provided by Piaget lationship. Such a developmental sequence im-
(1969), who claims that the child's concep- plies that the two concepts are equally difficult
tualization of time is derived from that of and develop parallelly along the same quali-
space. Piaget's analysis focuses on the develop- tative stages, so that ultimately each can be
ment of two time concepts-duration and suc- derived from the other.
cession-as well as on the emergence of the
The purported parallelism between du-
logical relationship between them. The pur- ration and succession deserves reexamination
pose of the present paper is to reexamine the
on both logical and empirical grounds. On a
development of the two time concepts and the
functional relation between them. logical level, the understanding of the concept
of duration seems to imply an understanding
Piaget argues that succession and dura- of succession, because without a grasp of the
tion are perceived by the preoperational child successiveness of the beginning and end points
in spatial rather than in temporal terms. "Be- of an interval of time it would seem to be
fore" and "after" in time are confused with difficult to construct its duration. Moreover, on
"before" and "after" in space, and a longer an empirical level, Lovell and Slater's (1960)
distance implies a longer duration. At an inter- replication of Piaget's experiments suggests
mediate stage, the distinction between time that these concepts are not equally difficult
and space is partial, since it is made by ap- but, rather, that succession is easier than du-
proximately half of the children just for suc- ration. In addition, Levin (1977) found that
cession and by half just for duration. Only when children are asked to explain their du-
at the operational stage does the child grasp ration judgments they refer to succession com-
that both duration and succession are inde- parisons. However, it can be argued here that
pendent of spatial dimensions, and he thereby the occurrence of the opposite-explaining suc-

This research was supported by the School of Education, Tel-Aviv University. We would
like to thank Shlomit Nagar for her devoted assistance with various stages of the study, Shaul
Nagar for designing the instruments, and Sidney Strauss and Peter Hess for their critical read-
ings of an earlier version of this manuscript. Thanks are extended to Susan Schmidt for her
participation in editing the article. We also thank the Israel Ministry of Education, the staffs
of the nursery and elementary schools involved, and, especially, the children for their kind
cooperation. Requests for reprints should be sent to Iris Levin, School of Education, Tel-Aviv
University, Ramat-Aviv, Israel.
[Child Development, 1978, 49, 755-764. @ 1978 by the Society for Research in Child Development, Inc.
0009-3920/78/4903-0025$00.95]

This content downloaded from

149.132.77.163 on Tue, 28 Feb 2023 08:15:13 UTC
All use subject to https://about.jstor.org/terms
756 Child Development
cession judgments by duration-seems a priori
to be rather unreasonable. These arguments
may suggest that the acquisition of the con-
cept of succession constitutes a prerequisite
for the development of the concept of dura-
tion and, as such, must be acquired at an
earlier age. Thus, we hypothesize that the
child first acquires the concept of succession,
and only later does he grasp the concept of
duration and the logical relationship between
them. Hence, the discrepancy between the
concepts is expected to decrease with age
within the limits of the age groups examined
here.
one in front of the other, covering different
distances, the children should err with regard
to the duration as well as the succession of
the stopping points. However, the more primi-
tive concept-succession-should be more re-
sistant to interfering information than dura-
tion, contrary to Piaget's conclusion that du-
ration and succession are divorced from their
spatial meaning at similar cognitive develop-
mental levels.
The following five problems were used
(their graphic representation is shown in pa-
rentheses):

Problem 1 Completely synchronous durations

Problem 2 Unequal durations resulting from different end points in
time

Problem 3 Unequal durations resulting from different starting points

in time

Problem 4 Equal but nonsynchronous, partially overlapping dura-

tions

Problem 5 Unequal durations with the shorter one occurring inside

the span of the longer one

While the concept of succession is as-
sumed to be more primitive than that of du-
ration, there is no reason to anticipate a dif-
ference in the complexity of beginning versus
end succession. If, indeed, succession is a uni-
tary concept, it is reasonable to expect an iden-
tical level of difficulty for beginning and end
points. Furthermore, if the relationship of each
to duration involves similar mental operations,
it is unlikely that either one of these relation-
ships will surpass the other in difficulty. This
notion is in contrast to Piaget's (1970, chap. 3)
assumption that children grasp the relation-
ship between duration and end-point succes-
sion earlier than between duration and begin-
ning point order.
Piaget (1969) and his followers (Demp-
sey 1971; Lovell & Slater 1960; Murray 1969)
used various problems in which the child com-
pared two synchronous events, which inten-
tionally differed in their rate of activity. Levin
(1977) dealt with the rate of activity or
amount of output as irrelevant variables which
supply misleading cues to the child in assess-
ing duration. The ability to compare durations
and to explain the conclusion logically was
found to be dependent on the type and num-
ber of interfering cues. Analogously, it is pro-
posed here that the child's judgment of suc-
cession should also be vulnerable to misleading
cues. Thus, if two toy cars stop simultaneously,
The tendency of young children to con-
centrate on unequal rather than equal cues in
problem solving (Pufall & Shaw 1972; Pufall,
Shaw, & Syrdal-Lasky 1973) supplies the basis
for predicting which problem should raise spe-
cial difficulty in succession judgments. When
the young child is asked to compare a succes-
sion point which is equal (i.e., simultaneous),
while the other point is unequal (i.e., dispa-
rate), he should fail with this problem more
than with other problems pertaining to the
same concept. Thus, problem 2 should be most
difficult when beginning succession is judged,
and problem 3 should be most difficult when
end succession is judged.
With duration comparison, however, these
problems should not cause any special diffi-
culty relative to other duration problems, be-
cause the child will concentrate on the un-
equal point which eventually provides the clue
to the duration judgment. Here particular diffi-
culty will arise when the child is required to
deal with duration which differs on both suc-
cession points. We assume that the child who
knows how to compare durations which differ
in either beginnings or endings does not neces-
sarily know how to integrate both conclusions.
This will be especially true with conclusions
which cancel out each other (i.e., problem 4),
a process known as the mental operation of
compensation.

This content downloaded from

149.132.77.163 on Tue, 28 Feb 2023 08:15:13 UTC
All use subject to https://about.jstor.org/terms
 Levin, Israeli, and Darom 757

A note should be added about problem 4. by mere repetition of the judgment. (11) The
In this problem the conclusion that durations tendency to use succession as an explanation
are equal must rely on time estimation rather for duration judgment will increase with age.
than on succession comparisons only. But since
the present study focuses on conceptualization Method
rather than estimation of time, any answer
Subjects.-One hundred eight children,
that evidenced taking both points of succession
into account in a compensatory fashion washalf boys and half girls, from nursery school,
considered satisfactory. If, for example, a child first grade, and third grade served as subjects.
said that he could not tell which was longerThe mean ages in years and months of the
because he did not know whether it was the three grades were 5-1, 6-3, and 8-10. The chil-
difference in starting or in ending points that dren came from schools in an upper-middle-
took more time, he was considered to be cor- class suburban section of Tel Aviv, Israel.
rect. Hence, the most difficult problems for be- Instruments.-A light box and cars on
ginning succession, end succession, and du- tracks were used to present time without and
ration are expected to be problems 2, 3, and 4, with interference, respectively. The light box
respectively. consisted of two rectangular-shaped lights (3.5
The relative difficulty of the problems is x 3.0 cm)-green and orange-which were built
expected to apply to problems presented with 1.5 cm apart onto the front part of a metal box.
and without interference. Following Piaget's The cars on tracks were two blue identically
claim, that young children derive from "greater shaped cars with a blue or white roof, 10 cm
speed" the conclusion of "longer duration," it in length, which traveled along parallel tracks
can be concluded that problems which involve 1 m long, from the child's right to his left.
unequal (problems 2, 3, and 5) rather than Since it has already been shown that cars
equal (problems 1 and 4) speeds should be covering unequal distances caused great diffi-
more affected by the interference manipulated culty in comparing durations, here the tracks
here. Moreover, since speed is apt to bias du- were kept equal so as to determine whether
even under this mild interference condition
ration but not succession judgments and young-
er children more than older ones, the inter- motion biased time judgments substantially. By
action of type of problem and interference having the cars start off and finish side by
should be qualified by both concepts and age. side, it was possible to lessen the impact of
spatial succession on the assessment of tem-
To conclude, we hypothesize that: (1) poral succession, while the equal distance was
Beginning and end succession will be equally intended to reduce its effect on duration judg-
difficult, and both will be easier than duration. ment.

(2) The discrepancy between succession and

It seemed preferable to use different du-
duration is expected to decrease with age. (3)
The introduction of movement as interference rations foi the two instruments, since pretests
revealed that the children lost their attention
will increase problem difficulty for succession
and even more so for duration. (4) Question- more quickly with lights and that they needed
more time to perceive the events with cars. Car
ing the child about simultaneous succession in
durations lasted for either 4 or 7 sec and light
the presence of a disparate one will cause durations for 3 or 5 sec.
special difficulty; with duration judgment call-
ing for compensation will do so. (5) These In order to guard against a response set
special difficulties are expected to decrease for color, the colors of the lights and the car
with age. (6) For all concepts, the interfer- roofs corresponding to the correct answer were
ence of movement will be especially biasing as balanced as possible. Because of a technical
in problems with unequal rather than with difficulty in constructing the instrument when-
equal speeds. (7) The discrepancy between ever the cars differed with respect to duration,
equal speed and unequal speed problems is the car that was farther from the child went
expected to be larger for duration than suc- for a longer duration. If the cars had been of
cession. (8) This discrepancy is also predicted fixed colors, then color could have served as
to decrease with age. (9) Overall, an age im- a discriminating cue for assessing durations.
provement in performance is predicted. (10) Therefore, we changed the color of the cars
Duration will be explained by succession, while roofs for some of the problems by using trans-
succession comparisons will be accompanied ferable white tape.

This content downloaded from

149.132.77.163 on Tue, 28 Feb 2023 08:15:13 UTC
All use subject to https://about.jstor.org/terms
758 Child Development
The interview.-Each subject was inter- of duration would indeed be argued based on
viewed on three different days: once each to succession, while judgments of succession
compare the order of starting points, the order would be justified by reference to "seeing" or
of ending points, and the durations. All five to succession itself, that is, through tautology.
problems were presented in each interview,
once by the lights and once by the cars. The Procedure
beginning succession question for the lights
was: "Did the two lights go on together or Thirty-six boys and girls from each grade
level were administered the interviews in six
did one go on first?" For end-point succession,
the question was phrased identically, replacing completely balanced orders. Half of each grade
"go on" with "go off." For duration, the ques- and sex were presented with the lights before
tion was: "Did the two lights last for the same the cars during all three sessions, while the
reverse was true of the other half. With each
length of time or was one on for more time?"
Parallel questions were posed about the toy instrument, the five problems were presented
cars with minor changes adapted to the instru- in random order that varied for each subject.
Each session lasted about 15 min.
ment. The order of the two phrases within
each question was kept constant, as it was
found to have no significant effect on perfor- Results
mance in our previous study (Levin 1977). The effect of three control variables on
Following each judgment, the child was the level of judgment-the six concept-presen-
asked, "How can you tell?" so that the expla- tation orders, the two instrument presentation
nation for his judgment could be assessed. If orders, and sex-was examined in a series of
he answered something like "because that's two-way ANOVAs (age and control variable)
what I saw," he was encouraged to describe so as to rule out the possibility of differential
what he had seen.
effects in the three grade levels. As no relation-
Our analysis of the concept of succession ship was found between any of the control
implies that the request to explain one's judg- variables either separately or in interaction
ment about succession should entail difficulty. with age (a = .05), they were ignored in
further analyses.
If judgments of succession are based on the
child's conceptualization of his perception of Judgment.-Hypotheses dealing with prob-
the event, then he is faced with an unanswer- lem difficulty were tested by means of a four-
able question when asked to explain his judg- way mixed model ANOVA (grade x instru-
ment, and he may resort to the pronouncement, ment X concept x problem). The use of
"because that's what I saw." We nevertheless
ANOVA with a dichotomous dependent vari-
required the child to explain his succession able (0, 1) indicating whether an item was
judgments for two reasons: first, in order to correctly judged was argued to be legitimate,
preserve an exact parallel between the investi- since the F test has been shown to be robust
gation of succession and duration; and second, for dependence between means and variances
in order to provide evidence that judgments (Darom, Note 1). Table 1 presents mean judg-

TABLE 1

MEAN JUDGMENT SCORES ACROSS FIVE PROBLEMS ACCORDING

TO GRADE LEVEL, INSTRUMENT, AND CONCEPT
(N= 108)

LIGHTS LCARS

Succession Succession

GRADE Beginning End Duration Beginning End Duration

Nursery....... .92 .88 .50 .68 .82 .38
First.......... .96 .93 .71 .88 .89 .59
Third......... .97 .95 .84 .87 .97 .71
Average........ .95 .92 .68 .81 .89 .56
NoTE.-Score range 0-1.

This content downloaded from

149.132.77.163 on Tue, 28 Feb 2023 08:15:13 UTC
All use subject to https://about.jstor.org/terms
 Levin, Israeli, and Darom 759

ment scores across the five problems. In line ginning succession, end succession, and dura-
with expectations, performance increased with tion were expected to be problems 2, 3, and 4,
age, F (2,105) = 58.67, p < .001; was higher respectively. The data appear in table 2. In or-
with lights than with cars, F(1,105) = 59.32, der to clarify the significant interaction between
p < .001; and differed for the concepts, problem type and concept, F(8,840) = 56.00,
F(2,210) = 183.63, p < .001. In a further p < .001, contrasts were drawn between each of
analysis of the concepts, while no difference these problems and the mean of the other four
was found between beginning and end suc- problems related to the same concept and instru-
cessions, F(1,120) = 1.74, N.S., the average ment. The only insignificant contrast was for
of the two was found to be higher than du- beginning-succession problems presented by
ration, F(2,210) = 19.08, p < .001. lights (t < 1). The contrasts for end succession
and duration with lights, as well as beginning
As predicted, the discrepancy between
succession, end succession, and duration with
succession and duration (i.e., the difference
between the average of the two successions cars, were all significant: respectively, t (935) =
and between duration) did in fact decrease 2.66, p < .01; t(935) = 17.16, p < .001; t(935)
with age for both instruments. This trend was = 19.58, p < .001; t(935) = 7.66, p < .001;
reflected in a significant interaction between t(935) = 15.42, p < .001.
concept and grade level, F(4,210) = 10.93, These discrepancies-between the "most
p < .001. difficult" and other problems for a concept in
It was hypothesized that duration would a given instrument-were expected to decrease
be more vulnerable to interfering cues than with age. Thus, the significant interaction
succession. The interaction between instrument found between problem type and grade-
and concept was indeed significant, F(2,210) F(8,420) = 2.80, p < .005-was further ana-
= 9.03, p < .001, but additional analyses of lyzed. A contrast was drawn between the
simple main effects revealed that this inter- above-mentioned discrepancies among pre-
action was not altogether in the predicted di- school children and the discrepancies among
rection. As expected, the discrepancy between the third graders. The discrepancies were sig-
lights and cars was significant for duration- nificantly larger among the preschoolers than
t(106) = 2.54, p < .001-and insignificant for among the third graders for end succession
end succession (t < 1), but, contrary to expec- and duration presented by lights, as well as for
tations, it was significant for beginning suc-
beginning and end successions presented by
cession as well, t(106) = 2.85, p < .001. cars: respectively, t(350) = 5.15, p < .001;
t(350) = 7.25, p < .001; t(350) = 3.86, p <
Overall, different problems varied in their .001; t(350) = 9.28, p < .001. But, contrary
difficulty level, F (4,420) = 48.79, p < .001. to expectations, the age decrease was not sig-
Specifically, the most difficult problems for be- nificant for beginning succession presented

TABLE 2

MEAN JUDGMENT SCORES ACROSS GRADE LEVELS ACCORDING TO

INSTRUMENT, CONCEPT, AND TYPE OF PROBLEM (N= 108)

PROBLEM TYPEs

4. Both
1. Com- Differ 5. Both
INSTRUMENT AND pletely 2. Be
CONCEPT Synchronous Simultaneou
Lights:
Beginning succession ...... .97 .93 .96 .94 .94 .95
End succession ........... .99 .94 .86 .94 .89 .92
Duration................. .86 .75 .70 .27 .83 .68
Cars:
Beginning succession ...... .96 .34 .93 .98 .84 .81
End succession .......... ..97 .90 .71 .95 .94 .89
Duration....................... .88 .57 .55 .19 .61 .56
NOTE.-Score range 0-1.
a The figures in italics represent the scores on the problems predicted to be the next most difficult within each row.

This content downloaded from

149.132.77.163 on Tue, 28 Feb 2023 08:15:13 UTC
All use subject to https://about.jstor.org/terms
760 Child Development
with lights (t < 1) and for duration with cars turned out to be even more difficult than du-
(t < 1). This complicated pattern was reflected ration judgment-a result to be discussed later.
in the following interactions: for instrument,
It was predicted that the special interfer-
problem type, and age, F (8,420) = 3.72, p <
ence of unequal rather than equal speeds will
.05; and for these three factors and concept,
decrease with age. The aforementioned dis-
F (16,840) = 2.83, p < .001.
crepancies among preschoolers were found to
It was predicted that the introduction of be larger than among third graders for begin-
movement would interfere with time judgments ning succession and duration, t(1060) = 4.97,
more when unequal speeds rather than equal p < .001; t(1060) = 9.76, p< .001, but not
speeds are involved. Moreover, movement of for end succession (t < 1).
unequal rather than equal speeds was expected
to interfere more with duration than with suc- Explanation.-The proportion of the types
of explanations offered for the judgments of
cession judgments. In line with expectations,
beginning succession, end succession, and du-
a significant interaction was found between
ration were computed, and those which ap-
type of problem, instrument, and concept,
peared substantially (over 15%) are presented
F(8,840) = 15.47, p < .001. In order to fur-
in table 3 according to grade level. While du-
ther clarify the support of the data for the
ration was explained mainly by either begin-
above-mentioned hypotheses, a discrepancy
ning, ending, or by both succession points and
was calculated between lights and cars for
not by duration itself (only 2%), each succes-
problems of unequal speeds (problems 2, 3,
sion judgment was accompanied predominantly
and 5) and for the problems of equal speeds
by tautological explanations. This pattern ap-
(problems 1 and 4). The first discrepancy was
plied, in general, to both instruments and
found to be significantly larger than the second
within each grade level. With cars, judgments
for each of the concepts: t(10.70) =10.32,
were also rationalized, albeit less frequently,
p< .001; t(10.70) = 1.84, p<.05; and
by referring to speed and overtaking. No sta-
t(10.70) = 6.28, p < .001 for beginning suc-
tistical comparison was drawn between the dif-
cession, end succession, and duration, respec-
ferent types of explanations, since the number
tively. As expected, this discrepancy was
of explanations provided varied from problem
larger for duration than for end succession,
to problem and between subjects. However, it
but, contrary to expectations, it was not larger
was possible to examine the effect of age on
for duration than for beginning succession.
frequency of providing each type of explana-
The data in table 2 point to the source of
tion for each concept with a series of one-way
this result. The judgment of beginning suc-
ANOVAs (age).
cession was very easy even when the problems
were presented by cars, except for problem 2 For succession judgments, the only type
of explanation related to age was succession
of simultaneous starting off. In fact, this was
the only problem in which succession judgmentitself. The tendency to explain a succession

TABLE 3

MEAN FREQUENCY SCORES OF EXPLANATIONS ACCORDING

TO CONCEPTS AND GRADE LEVELS (N= 108)

CONCEPT BEING JUDGED

Beginning Succession End Succession Duration

Begin-
ning End
Begin- Begin- succes- succes- Over-
GRADE ning End Both ning End Both sion sion Both Speed' taking&
Preschool..... .41 .25 .07 .39 .43 .13 .44 .30 .18 .19 .20
First......... .66 .47 .28 .62 .61 .39 .72 .57 .47 .36 .23
Third........ .69 .38 .21 .43 .74 .29 .72 .74 .58 .40 .06
Average...... .59 .37 .19 .48 .59 .27 .63 .54 .41 .32 .16

NoTE.-Score range 0-1.

a Refers to car presentation only.

This content downloaded from

149.132.77.163 on Tue, 28 Feb 2023 08:15:13 UT6 12:34:56 UTC
All use subject to https://about.jstor.org/terms
 Levin, Israeli, and Darom 761

point by itself increased with age both for cepts, predominantly on succession, while suc-
beginnings and endings: F(2,105) = 10.08, cession comparisons themselves are based on
p < .001 and F(2,105) = 12.48, p < .001, re- the perceptual elements of the event and can-
spectively. In contrast, the tendency to men- not be reduced to a more primitive concept.
tion one succession point as explaining the In contrast to Piaget's claim that both
other-F(2,105) = 4.13, p < .02 and F(2,105)
concepts develop synchronically, in an inter-
= 4.56, p < .01 for beginning and end judg- individual variant order, the data show that
ments, respectively-as well as mentioning both
succession is acquired earlier and may serve as
succession points as explanatory of one- a mediator of duration. However, for the pre-
F (2,105) = 8.07, p < .001 and F (2,105) = operational child, though duration is concep
8.84, p < .001-first increased from preschool
tually related to succession, the coordination
to first grade but then decreased for the third
between the two concepts is fragile, since other
graders.
cues may be assimilated into the concept of
While succession judgments were ex- duration and may interfere with duration judg
plained mainly by succession, duration judg- ments. It has already been found that duration
ments were rarely (2%) explained by duration judgments of young children take into con
itself, and this marginal type of explanation sideration cues such as distance, speed, height,
was not found to be related to age. With de- rate of the blooming of trees, etc. (e.g., Berndt
velopment, there was an increase in explain- & Wood 1974; Levin 1977; Murray 1969).
ing duration judgment by succession-F (2,105) All these interfering cues share the character-
= 10.78, p < .001; F(2,105) = 24.13, p < istic that they are logically relevant to time.
.001; F(2,105) = 29.89, p < .001 for begin- Hence, it remains an open question whether
ning succession, end succession, and both, re- duration judgments are vulnerable to interfer
spectively-as well as by speed, F (2,105) = ence even by cues which are irrelevant to
4.08, p < .02. On the other hand, explanations time, such as-to give a wild example-inten-
pertaining to overtaking decreased from the sity of light.
first to the third grade, F (2,105) = 6.20, p < The coordination between succession and
.01.
duration, as revealed through duration judg-
The lack of any explanation decreased ments in problems with no interference, is an
with age for duration judgments, F (2,105) = achievement which takes place as early as first
3.68, p < .03, whereas for succession judgment grade (see also Levin 1977). Moreover, even
it decreased but then increased. This trend, preschoolers show some indication of this co-
however, was significant for beginning succes- ordination, since they quite often tend to men-
sion, F (2,105) = 6.05, p < .01, but not for tion succession in explaining their duration
end succession, F(2,105) = 1.54, N.S. judgments. On the other hand, the vulnerabil-
ity of duration judgments to cues such as speed
Discussion (in the sense that more speed implies more
duration) or distance (further means longer
duration) emerges in preschool and persists
The findings of the present study support
the contention that the acquisition of theeven
con-among third graders (Levin 1977).
cepts of duration and succession are two dis- The coordination of succession and dura-
tinct developmental achievements. Duration is
more difficult than succession, both withtion
andand the vulnerability of duration to inter-
ference must lead under certain conditions
without the interference of misleading cues,
(e.g., when one of two cars stops earlier but
and thus may be viewed as a later develop-
further) to conflicting conclusions about du-
ment. In addition, while succession is explained
ration. Such cognitive conflicts have already
mainly by reiterating the judgment, duration
been assumed by Piaget to be essential to the
is explained predominantly by reference to
enhancement of cognitive conceptualization in
order of beginnings, endings, or both. With
general. Berndt and Wood (1974) have al-
age, there is an increase for tautological ra-
ready shown in a highly interesting study that
tionalizations of succession and for explana-
such a conflict enhanced to a certain degree
tions of duration pertaining to succession as
the conceptualization of duration.
well as to speed. Thus, it may be contended
that duration comparisons involve the execu- The argument that young children tend
tion of mental operations on some other con-
to concentrate on unequal rather than on equal

This content downloaded from

149.132.77.163 on Tue, 28 Feb 2023 08:15:13 UTC
All use subject to https://about.jstor.org/terms
762 Child Development
cues was supported by two classes of findings: the present study nor any of the previous ones
one referring to the cue of succession and the supply an adequate opportunity to examine
other to the cue of speed. With reference to these theoretical distinctions.
the cue of succession, problem 2 of simulta-
Piaget claims that the distinction between
neous beginnings and unequal endings was time and space is achieved by 45% of the chil-
most difficult when beginning succession was dren for succession first and by 45% for dura-
judged, and problem 3 of simultaneous end- tion first (Piaget 1969, p. 83). The question
ings and equal beginnings was most difficult arises whether the discrepancy between the
when end succession was judged. It should be present findings of the superiority of succession
noted, however, that simultaneity (i.e., equal- over duration judgments and those of Piaget
ity) per se did not raise much difficulty even may stem from the systematic difference be-
for preschoolers, as long as the other point of tween the problems used in these studies. While
succession was simultaneous as well, as in
Piaget's subjects compared durations of move-
problem 1. The difficulty emerged when the ments along different lengths of track, our chil-
children were asked to judge the equal cue dren compared movements along equal length
when another unequal cue was present. tracks. In other words, Piaget's children were
In judging duration, problem 4 raised presented with two misleading cues, speed and
special difficulty. In this problem two unequal distance, while ours were confronted with only
cues (beginning and end succession) resulted one interfering cue, speed. Only a replication
in equal duration, a situation which called for with durations of movements along different
the mental operation of compensation. Here, distances can rule out the possibility that the
addition of a distance cue increases the diffi-
again, "sameness" (of duration) per se, when
presented in problem 1, was not sufficient to culty of succession comparisons relatively more
raise special difficulty even among preschool- than that of duration comparisons and thereby
ers. Sameness was difficult when presented by equalizes their level of difficulty in Piaget's
two unequal cues, which compensated for each experiments.
other.
Viewing succession as a more primitive
With reference to the cue of speed, the concept than duration led to three hypotheses
discrepancy in difficulty between car presen- which were partially refuted: the discrepancy
tation and light presentation was larger for in difficulty level between the concepts was
problems 2, 3, and 5 of unequal speed than expected to decrease with age; duration was
for problems 1 and 4 of equal speed. Accord- predicted to be more vulnerable to the bias of
ing to the present theoretical framework, the misleading cues than succession, thus leading
introduction of unequal speeds attracts the to a higher discrepancy between instruments
child's attention to the cue of speed at the for duration; duration was expected to be more
expense of the cue of succession. Thus, when vulnerable to the bias of unequal relative to
asked to reconstruct the order of succession, equal misleading cues than succession, thus
the child tends to be confused and may err. leading to a higher discrepancy between in-
When asked to judge duration, the distinction struments in the difference between unequal
required between "more" and "less" duration and equal speed problems. All three hypoth-
eses were confirmed when duration was com-
is confused with the salient distinction pre-
sented between "more" and "less" speed. pared to end succession but not to beginning
Hence, the event with more speed is attributed succession. Looking carefully at the data,
more duration. This analysis leads to predic- though, reveals that all three hypotheses hold
tion of the same mistakes of duration judg- true for beginning succession as well, except
ments as Piaget's (1969) claim that for the for problem 2 of simultaneous beginnings.
young child more speed implies a longer du- Problem 2 for beginning succession turned
ration in the sense that "speed buys time" out to be highly difficult-even for the third
(Papert 1973). But, while the argument pre- graders-when presented by cars. This level of
sented here is that various "mores" can lead
difficulty is surprising from another point of
the young child to the judgment of "more du- view as well. Problem 2 for beginning succes-
ration," Piaget argues that spatial dimensions,sion and problem 3 for end succession were
such as speed and distance, are for the child
predicted to be the most difficult problems
conceptually indistinguishable from time. for
It their concepts on exactly the same theo-
should be remembered, however, that neitherretical ground. Thus, both problems might

This content downloaded from

149.132.77.163 on Tue, 28 Feb 2023 08:15:13 UTC
All use subject to https://about.jstor.org/terms
 Levin, Israeli, and Darom 763

have been expected to have similar difficulty creased with age, while by "overtaking" de-
levels, which they did not. Problem 2 (begin- creased from the first to the third grade. Both
ning succession) was much more difficult than trends are consistent with Piaget's claim that
problem 3 (end succession), pointing again to "overtaking" is the primitive, qualitative notion
the same direction. It may be supposed that of speed and with Berndt and Wood's (1974)
the simultaneity in these problems is not as finding that "ahead" explanation decreased
easily perceived as is a disparate succession from ages 5 to 7.
but is derived by inference from the analysis of
The implication that even preschoolers,
the movement of the cars along the tracks. In under certain conditions, deal with certain
order to decide whether the cars started off
time concepts in a temporal rather than a
simultaneously (in problem 2), the subject
spatial frame of reference invites a reconsider-
could be aided by mentally reversing the move-
ation of Piaget's basic assumption that the
ment of the cars. The conclusion that the stop-
origin of time concepts are their spatial coun-
ping of the cars was simultaneous (in problem
terparts. Clark (1973) claims that the under-
3) is supported by the gradual shrinking of
the distance between the cars. Consistent with standing of time is based on spatial metaphor
and goes one step further to hypothesize that
the notion of irreversibility, it seems likely that
a reconstruction of a movement in reverse is in the sequence of psycholinguistic develop-
ment "spatial expression should appear before
more difficult to establish than an anticipation
time expressions, and . . . each term should be
of a future movement. Such an explanation,
acquired in its spatial sense first" (p. 57).
however, implies that simultaneity is riot mere-
These hypotheses were recently refuted in a
ly "given by perception" but may be concep-
study by Friedman and Seely (1976), using
tually supported by the coordination of time
concepts of succession that have both temporal
and space.
and spatial meanings (e.g., before, after, first,
Problems 2, 3, and 4 for beginning suc- and last). Children as young as 3 years of age
cession, end succession, and duration, respec- evidenced comprehension of the temporal
tively, were expected to raise special difficulty. meanings, no less and maybe more than the
In addition, the discrepancies between the spatial meanings. Furthermore, the tendency
"difficult" problems and their easier same con- to err and understand a spatial term as having
cept, same instrument counterparts were ex- a temporal meaning was more frequent than
pected to decrease with age. Both sets of hy- the reverse. These findings seem to provide
potheses were confirmed, except for beginning support from a new vantage point for the
succession presented by lights. Even preschool- present model of the development of time
ers had no difficulty in comparing starting concepts.
points of lights, thus canceling any predicted
difference either between problems or age Reference Note
groups. It is possible, however, that these
trends could emerge with a younger age group. 1. Darom, E. Comparison of several methods to
the analysis of categorical data in a split-plot
Unexpectedly, the tendency to explain design. Unpublished doctoral dissertation, Syr-
one succession point by the other or by both
acuse University, 1975.
succession points increased but then decreased
with age. This type of explanation, though it
may correctly describe the presented events, References
is logically irrelevant to the judgment. Thus it Berndt, T. J., & Wood, D. J. The development of
should, indeed, from a certain point on, de- time concepts through conflict based on primi-
crease with development. The assumption that tive duration capacity. Child Development,
third graders refrain from providing irrelevant 1974, 45, 825-828.
explanations could also explain why they failed Clark, E. V. What's in a word? On the child's
to provide any explanation at all for beginning- acquistion of semantics in his first language.
succession judgments more frequently than the In T. E. Moore (Ed.), Cognitive development
first graders. On the other hand, providing suc- and the acquisition of language. New York:
cession explanations for duration judgments Academic Press, 1973.
systematically increased with age, replicating Dempsey, A. Time conservation across cultures.
similar findings (Berndt & Wood 1974). The International Journal of Psychology, 1971, 6,
115-120.
tendency to explain duration by "speed" in-

This content downloaded from

149.132.77.163 on Tue, 28 Feb 2023 08:15:13 UTC
All use subject to https://about.jstor.org/terms
764 Child Development
Friedman, W. J., & Seely, P. B. The child's acquisi-
tion of spatial and temporal word meanings.
Child Development, 1976, 47, 1103-1108.
Levin, I. The development of time concepts in
young children: reasoning about duration.
Child Development, 1977, 48, 435-444.
Lovell, K., & Slater, N. The growth of the concept
of time: a comparative study. Child Psychol-
ogy and Psychiatry, 1960, 1, 179-190.
Murray, F. B. Conservation aspects of the concept
of time in primary school children. Journal of
Research in Science Teaching, 1969, 6, 257-
264.
Papert, S. The theory of knowledge and complexity.
In G. J. Dalenoort (Ed.), Process models for
psychology. Groningen: Academic Book Ser-
vices, 1973.
Piaget, J. The child's conception of time. London:
Routledge & Kegan Paul, 1969.
Piaget, J. The child's conception of movement and
speed. London: Routledge & Kegan Paul,
1970.
Pufall, P. B., & Shaw, R. E. Precocious thoughts
on number: the long and the short of it. De-
velopmental Psychology, 1972, 7, 62-69.
Pufall, P. B.; Shaw, R. E.; & Syrdal-Lasky, A. De-
velopment of number conservation: an exami-
nation of some predictions from Piaget's stage
analysis and equilibrium mode. Child Devel-
opment, 1973, 44, 21-27.

This content downloaded from

149.132.77.163 on Tue, 28 Feb 2023 08:15:13 UTC
All use subject to https://about.jstor.org/terms