literature review
ShadowMe
192259188.pdf
UN CO
RR EC
TE D
PR O O F
Research Report
Children Learn When Their Teacher’s Gestures and Speech Differ Melissa A. Singer and Susan Goldin-Meadow
University of Chicago
ABSTRACT—Teachers gesture when they teach, and those
gestures do not always convey the same information as
their speech. Gesture thus offers learners a second mes-
sage. To determine whether learners take advantage of
this offer, we gave 160 children in the third and fourth
grades instruction in mathematical equivalence. Children
were taught either one or two problem-solving strategies in
speech accompanied by no gesture, gesture conveying the
same strategy, or gesture conveying a different strategy.
The children were likely to profit from instruction with
gesture, but only when it conveyed a different strategy
than speech did. Moreover, two strategies were effective in
promoting learning only when the second strategy was
taught in gesture, not speech. Gesture thus has an active
hand in learning.
People gesture when they talk. Even teachers routinely produce
gestures as they instruct children in both individualized tuto-
rials (Goldin-Meadow, Kim, & Singer, 1999) and the classroom
(Crowder & Newman, 1993; Flevares & Perry, 2001; Neill,
1991; Roth & Welzel, 2001; Zukow-Goldring, Romo, & Duncan,
1994). And children pay attention to those gestures, often
gleaning substantive information from gesture that cannot be
found anywhere in the teacher’s speech (Goldin-Meadow et al.,
1999). Gesture is thus present and salient in teaching situa-
tions. The question we addressed in this study is whether ges-
ture promotes learning.
Gesture conveys information in a visuospatial format
(McNeill, 1992). Previous work suggests that when visual im-
ages are presented simultaneously with a spoken message,
listeners do better at remembering the message (Baggett, 1984)
or solving problems with information contained in the message
(Mayer & Anderson, 1991). Because gesture is produced si-
multaneously with speech, it allows speakers to present visual
information at the same time as, indeed synchronized with, their
words (Kendon, 1980; McNeill, 1992; Morrel-Samuels &
Krauss, 1992). One might therefore expect gesture to be ef-
fective in getting a message across to learners, particularly
when gesture reinforces the message conveyed in speech. And,
indeed, researchers have found that spoken instruction pre-
sented with gesture promotes learning better than the same
spoken instruction presented without gesture (Church, Ayman-
Nolley, & Estrada, 2004; Perry, Berch, & Singleton, 1995;
Valenzeno, Alibali, & Klatzky, 2003).
However, the information conveyed in gesture does not al-
ways match the information conveyed in the speech it accom-
panies (Goldin-Meadow, 2003; Goldin-Meadow, Alibali, &
Church, 1993). For example, when giving a child instruction in
how to solve the problem 7 1 6 1 5 5 __ 1 5, a teacher ar-
ticulated the equalizer problem-solving strategy in speech: ‘‘We
need to make this side equal to this side.’’ At the same time, she
conveyed a grouping strategy in gesture: She pointed at the 7
and the 6 on the left side of the equation and then at the blank on
the right side (7 and 6 are the two numbers that, if grouped and
added, give the correct answer). The two strategies lead to
correct solutions yet do so via different routes and, in this sense,
constitute a ‘‘mismatch.’’ Gesture-speech mismatch occurs
when gesture conveys information that is different from (al-
though not necessarily contradictory to) the information con-
veyed in the speech it accompanies (Goldin-Meadow, 2003).
Gesture-speech mismatches are produced by speakers of all
ages and in a variety of tasks (Alibali, Bassok, Olseth, Syc,
& Goldin-Meadow, 1999; Garber & Goldin-Meadow, 2002;
Gershkoff-Stowe & Smith, 1997; Pine, Lufkin, & Messer, 2004;
Schwartz & Black, 1996; Stone, Webb, & Mahootian, 1991)
and are frequently found in teaching situations. In fact, teach-
ers spontaneously increase the number of gesture-speech
P S C I 7 8 6 B Dispatch: 1.12.04 Journal: PSCI CE: BlackwellJournal Name Manuscript No. Author Received: No. of pages: 5 Saravan/Anand Kumar
Address correspondence to Susan Goldin-Meadow, University of Chicago, Department of Psychology, 5730 S. Woodlawn Ave., Chi- cago, IL 60637; e-mail: [email protected].
PSYCHOLOGICAL SCIENCE
Volume 16—Number 2 85Copyright r 2005 American Psychological Society
B W U S P S C I 7 8 6 . P D F 0 1 - D e c - 0 4 1 5 : 1 1 1 0 6 3 1 5 B y t e s 5 P A G E S
o p e r a t o r = A n a n d K u m a r
UN CO
RR EC
TE D
PR O O F
mismatches in their instruction when teaching children who are
on the cusp of learning the task (Goldin-Meadow & Singer,
2003). But just because mismatches are found in teaching sit-
uations does not mean they are good for learning.
Why might one expect gesture-speech mismatch to promote
learning? Mismatching gesture allows speakers to add a second
problem-solving strategy to an instruction. Given that having a
variety of approaches to a problem is positively associated with
cognitive change (Siegler, 1994), it might be beneficial for
learners to be exposed to multiple problem-solving strategies.
Indeed, studies of teaching across nations have found that
students in Japan are exposed to more alternative methods for
solving math problems than are students in America, and they
learn more (Stigler & Hiebert, 1999)—although there is as yet
no evidence of a causal link between multiple approaches in
instruction and children’s outcomes.
We asked two questions in this study: (a) Does teaching
children more than one strategy for solving a problem facilitate
their mastery of the problem? (b) Does it matter whether those
strategies are presented in speech, in gesture, or in both speech
and gesture? To address these questions, we presented children
with one or two strategies for solving mathematical equivalence
problems and varied whether the spoken instruction was ac-
companied by matching gesture, mismatching gesture, or no
gesture at all.
METHOD
One-hundred sixty children (58 boys, 102 girls) who were fin-
ishing the third grade or beginning the fourth grade (ages 8–10)
in Chicago public and parochial schools participated. The
children were randomly assigned to one of six instruction
conditions. Each child participated individually in a pretest, an
instruction session, and a posttest. During the pretest, the child
was given a paper-and-pencil test containing six addition
problems and asked to explain at the chalkboard how he or she
solved each problem. If the child solved any of the pretest
problems correctly, he or she was eliminated from the study. A
second experimenter then gave the child instruction in math-
ematical equivalence using four additional problems. Two in-
structional trials were given on each of the four problems; after
each trial, the child was asked to solve and explain the problem
again. After the instruction period, the first experimenter re-
turned and gave the child a posttest comparable to the pretest.
Two problem-solving strategies spontaneously used by
teachers when instructing children in mathematical equiva-
lence (Goldin-Meadow et al., 1999; Goldin-Meadow & Singer,
2003) were used to teach the children: (a) equalizer, a strategy
highlighting the principle underlying the problem, and (b) add-
subtract, a strategy highlighting a procedure for solving the
problem. We determined the gestural equivalents of these
spoken strategies by examining the gestures children typically
produce when expressing equalizer and add-subtract in speech
(Perry, Church, & Goldin-Meadow, 1988; Goldin-Meadow,
2003, describes the general procedure used to attribute
meaning to gesture). Table 1 presents examples of the two
strategies in speech and gesture.
Instruction was organized around two factors, resulting in six
conditions (see Table 2). The first factor was the number of
strategies taught in speech: In half the conditions, children were
taught one strategy in speech, equalizer. In the other half,
children were taught two strategies in speech, equalizer and
add-subtract. The second factor was the relation between
speech and gesture: In the two no-gesture conditions, the ex-
perimenter produced no gestures during the instruction. In the
two matching-gesture conditions, she produced in gesture a
strategy that was the same as the spoken strategy it accompa-
TABLE 1
Examples of the Strategies Taught in Speech and Gesture
Sample problem : 6 1 4 1 3 5 — 1 3
Equalizer in speech : ‘‘We can add 6 1 4 1 3, which equals 13. We want to make the other side of the equal sign the same amount, and 10 1
3 also equals 13, so 10 is the answer.’’
Equalizer in gesture : Sweep with palm under left side of problem, drop hand, sweep with palm under right side.
Add-subtract in speech : ‘‘We can add 6 1 4 1 3, which equals 13. We then subtract the other 3 from 13 and get 10 as the answer.’’
Add-subtract in gesture : Point at the 6, the 4, and the left 3, then produce a flick-away gesture near the right 3.
Note. On all trials, the experimenter gave the correct answer in her explanation but did not write it in the blank.
TABLE 2
Design of the Six Instruction Conditions
Instructional trial on each problem
Number of strategies taught in speech
One Two
Speech Gesture Speech Gesture
No gesture
Trial 1 Equalizer — Equalizer —
Trial 2 Equalizer — Add-subtract —
Matching gesture
Trial 1 Equalizer Equalizer Equalizer Equalizer
Trial 2 Equalizer Equalizer Add-subtract Add-subtract
Mismatching gesture
Trial 1 Equalizer Add-subtract Equalizer Add-subtract
Trial 2 Equalizer Add-subtract Add-subtract Equalizer
86 Volume 16—Number 2
Learning From Gesture-Speech Mismatches B W U S P S C I 7 8 6 . P D F 0 1 - D e c - 0 4 1 5 : 1 1 1 0 6 3 1 5 B y t e s 5 P A G E S
o p e r a t o r = A n a n d K u m a r
UN CO
RR EC
TE D
PR O O F
nied. In the two mismatching-gesture conditions, she produced
in gesture a strategy that was different from the spoken strategy
it accompanied. There were approximately 27 children in each
condition (range: 24–31).
Table 2 displays for each condition the strategies taught on
the two instructional trials for each problem. In all six condi-
tions, the experimenter taught children the equalizer strategy in
speech on a problem’s first instructional trial. Depending on the
child’s condition, the experimenter produced either no gesture,
matching gesture (equalizer in gesture), or mismatching gesture
(add-subtract in gesture) along with the spoken equalizer
strategy. On the problem’s second instructional trial, children in
the one-strategy-in-speech conditions received the same strat-
egies in speech and gesture as they received on the first trial.
Children in the two-strategies-in-speech conditions were taught
their second spoken strategy (add-subtract) on the second in-
structional trial and, depending on their condition, received no
gesture, matching gesture (add-subtract in gesture), or mis-
matching gesture (equalizer in gesture) along with this second
spoken strategy.
Each session was videotaped, and the experimenter’s speech
and gestures were spot-checked to be certain they conformed to
the child’s condition. 1
Because no children solved any problems
correctly on the pretest, number of problems correct on the
posttest was taken as a measure of learning. Data were analyzed
using an analysis of variance with two between-subjects factors:
number of strategies in speech (one, two) and relation between
gesture and speech (no gesture, matching gesture, mismatching
gesture).
RESULTS
Figure 1 presents the number of problems solved correctly on
the posttest in each condition. 2
Children performed better on
posttest when taught one strategy in speech than when taught
two, F(1, 154) 5 4.25, p 5 .04. The relation between gesture
and speech also had an impact on learning, F(2, 154) 5 4.44,
p 5 .01, and did not interact with number of strategies taught in
speech, F(2, 154) 5 0.27, p 5 .76. Mismatching gesture was
significantly better as a teaching device than no gesture
( p 5 .01, Newman-Keuls) and was also better than matching
gesture ( p 5 .04). There was no reliable difference between
matching gesture and no gesture ( p 5 .39).
Surprisingly, teaching children two problem-solving strategies
in speech was significantly worse than teaching one strategy (i.e.,
adding add-subtract to equalizer was less effective than teaching
equalizer on its own), suggesting that children may have been
overwhelmed by the additional spoken strategy. Note, however,
that including a second strategy in instruction was effective when
that strategy was presented in gesture. Children in all of the one-
strategy-in-speech conditions were taught one spoken strategy;
however, those who received mismatching gestures along with
their one spoken strategy were actually exposed to a second
strategy—but only in gesture. The fact that children in this group
did so well on posttest, significantly better than children in all
other groups (even those exposed to two spoken strategies),
suggests that instruction containing a second strategy can indeed
promote learning—as long as that second strategy is produced in
gesture and not in speech.
Not only was gesture effective as a teaching device when it
conveyed information not found in speech, but it was also ef-
fective when it conveyed the same information packaged dif-
ferently. Note in Table 2 that within the two-strategies-in-
speech conditions, children given matching and mismatching
gesture were exposed to precisely the same information. The
only difference between the two groups was how the information
was packaged. In the matching condition, each of the spoken
strategies was presented along with the same strategy in gesture.
In the mismatching condition, each spoken strategy was pre-
sented along with a different strategy in gesture. The mis-
matching group performed better on posttest, suggesting that
mismatch in instruction can promote learning even when type
and amount of information are held constant.
Fig. 1. Number of problems children solved correctly after receiving instruction that contained either one or two strategies in speech and that was accompanied by no gesture, gesture matching the strategy in speech, or gesture mismatching the strategy in speech.
1 To determine whether the experimenter not only used the same words, but
also took the same amount of time to present the spoken strategies in each condition, we calculated time taken to deliver the strategies to 6 children ran- domly chosen from each of the six conditions; we found no differences in either the one-strategy condition, F(2, 17) 5 0.69, n.s., or the two-strategy condition, F(2, 17) 5 2.33, n.s.
2 Girls performed significantly better than boys on posttest (2.0 vs. 1.2),
F(1, 148) 5 3.99, p < .05, but gender did not interact with any of the other factors; data were therefore collapsed across genders for all analyses.
Volume 16—Number 2 87
Melissa A. Singer and Susan Goldin-Meadow B W U S P S C I 7 8 6 . P D F 0 1 - D e c - 0 4 1 5 : 1 1 1 0 6 3 1 5 B y t e s 5 P A G E S
o p e r a t o r = A n a n d K u m a r
UN CO
RR EC
TE D
PR O O F
DISCUSSION
Teachers gesture when they teach, and those gestures do not
always convey the same information as the speech they ac-
company. Gesture thus offers students a second approach to the
problem at hand. Our findings make it clear that children can
take advantage of the offer—children profit from gesture when
it conveys information that differs from the information con-
veyed in speech.
Previous research led us to expect that gesture would have a
hand in promoting learning. Producing speech with gesture
requires less effort on the speaker’s part than producing speech
without gesture (Goldin-Meadow, Nusbaum, Kelly, & Wagner,
2001; Wagner, Nusbaum, & Goldin-Meadow, 2004); gesture
might therefore be expected to reduce the effort that listeners
expend. Moreover, when children are at a transitional point in
acquiring a concept, they often find it easier to produce ideas
relevant to that concept in gesture than in speech (Church &
Goldin-Meadow, 1986); those same ideas might therefore be
expected to be easier to understand in gesture. However, in the
present study, gesture per se did not promote learning—only
gesture that conveyed mismatching information led to improved
performance.
Why might mismatching gesture be such an effective teach-
ing tool? We begin to tackle this question by examining the two
strategies used in instruction. These strategies represent dif-
ferent, albeit complementary, approaches to mathematical
equivalence. Add-subtract offers an algorithm for solving the
problem; equalizer articulates the principle underlying the
correct solution. Previous work has found that these two strat-
egies are not equally effective teaching tools—children are
more likely to succeed on mathematical equivalence problems
when taught the equalizer principle than when taught the add-
subtract algorithm (Perry, 1991). Interestingly, however (and as
is consistent with our findings), children do less well if taught
both principle and algorithm than if taught the principle on its
own (Perry, 1991). Offering children a step-by-step algorithm
seems to prevent them from attending to the principle they
could have profited from. Indeed, when asked to explain their
answers to the posttest problems, children in our study were
more likely to express the equalizer principle if it was the only
strategy taught in speech than if it and add-subtract were both
taught in speech (1.74 vs. 0.36 equalizer explanations produced
on the posttest per child in the one-strategy-in speech condi-
tions vs. the two-strategies-in speech conditions).
But gesture changes the picture. An algorithm detracts from
the principle it accompanies only when it is taught in speech—
not when it is taught in gesture. Algorithms presented in gesture
provide children with a step-by-step procedure to follow but
(unlike algorithms presented in speech) do not encourage
children to rely exclusively on that procedure. Perhaps because
gesture is not as explicit as speech, the information it conveys is
less intrusive than information conveyed in speech. In our
study, children expressed equalizer in their postinstruction
explanations more often if they had been taught equalizer along
with add-subtract presented only in gesture (2.54 equalizer
explanations in the one-strategy-in-speech condition with
mismatching gesture) than if they had been taught equalizer
along with add-subtract presented in speech (0.36 in the two-
strategies-in speech condition, with or without gesture)—and,
interestingly, also more often than if they had not been taught
add-subtract at all (1.27 in the one-strategy-in-speech condi-
tion with matching or no gesture). Thus, when an algorithm is
taught only in gesture, rather than overpowering the principle it
accompanies, it appears to enhance it and render it more ac-
cessible to learners.
Gesture seems to be good at making children aware of the
synergistic relation between principle and algorithm, perhaps
because it allows the two to be presented simultaneously and
not sequentially, as would be required by speech on its own. In
our study, not only did the equalizer principle benefit from
occurring simultaneously with the add-subtract algorithm, but
add-subtract also benefited from co-occurring with equalizer.
Children were more likely to express add-subtract in their ex-
planations after instruction when add-subtract was taught si-
multaneously with equalizer, that is, when it was taught in
speech (i.e., in the two-strategies-in-speech condition) with
mismatching gesture (1.77 add-subtract explanations), than
when it was taught in speech with matching gesture (0.81) or no
gesture (0.54). 3
Principles and algorithms are interconnected
approaches. Presenting them simultaneously—as only speech
and gesture can do—highlights the relation between the two
and seems, as a result, to facilitate learning.
Whatever the purported mechanism, the phenomenon re-
ported here remains robust. Presenting students two different
explanations—one in speech and one in gesture—facilitates
their mastery of mathematical equivalence problems compared
with presenting either one or two strategies in speech without
gesture. Given previous work establishing the breadth and
depth of gesture production across many tasks and ages, these
data open the possibility for a heretofore unappreciated tech-
nique to improve learning in and out of the classroom.
Acknowledgments—This research was supported by grants
from the Spencer Foundation and National Institute of Child
Health and Human Development (Grant R01 HD47450) to
S.G.-M. and by a Benjamin Bloom Dissertation Fellowship to
M.A.S. We thank Rebecca Nelson and Lorena Arevalo for help
in data collection.
3 Children expressed add-subtract 0.13 times in their posttest explanations
when it was not taught at all (i.e., one-strategy-in-speech condition with matching or no gesture), and 0.93 times when it was taught only in gesture and not in speech (i.e., one-strategy-in-speech condition with mismatching gesture).
88 Volume 16—Number 2
Learning From Gesture-Speech Mismatches B W U S P S C I 7 8 6 . P D F 0 1 - D e c - 0 4 1 5 : 1 1 1 0 6 3 1 5 B y t e s 5 P A G E S
o p e r a t o r = A n a n d K u m a r
UN CO
RR EC
TE D
PR O O F
REFERENCES
Alibali, M.W., Bassok, M., Olseth, K.L., Syc, S.E., & Goldin-Meadow, S.
(1999). Illuminating mental representations through speech and
gesture. Psychological Science, 10, 327–333. Baggett, P. (1984). Role of temporal overlap of visual and auditory
material in forming dual media associations. Journal of Educa- tional Psychology, 76, 408–417.
Church, R.B., Ayman-Nolley, S., & Estrada, J. (in press). The effects of
gestural instruction on bilingual children. International Journal of Bilingual Education and Bilingualism.
Church, R.B., & Goldin-Meadow, S. (1986). The mismatch between
gesture and speech as an index of transitional knowledge. Cog- nition, 23, 43–71.
Crowder, E.M., & Newman, D. (1993). Telling what they know: The role
of gesture and language in children’s science explanations.
Pragmatics and Cognition, 1, 341–376. Flevares, L.M., & Perry, M. (2001). How many do you see? The use of
nonspoken representations in first-grade mathematics lessons.
Journal of Educational Psychology, 93, 330–345. Garber, P., & Goldin-Meadow, S. (2002). Gesture offers insight into
problem-solving in children and adults. Cognitive Science, 26, 817–831.
Gershkoff-Stowe, L., & Smith, L.B. (1997). A curvilinear trend in
naming errors as a function of early vocabulary growth. Cognitive Psychology, 34, 37–71.
Goldin-Meadow, S. (2003). Hearing gesture: How our hands help us think. Cambridge, MA: Harvard University Press.
Goldin-Meadow, S., Alibali, M.W., & Church, R.B. (1993). Transitions
in concept acquisition: Using the hand to read the mind. Psy- chological Review, 100, 279–297.
Goldin-Meadow, S., Kim, S., & Singer, M. (1999). What the teacher’s
hands tell the student’s mind about math. Journal of Educational Psychology, 91, 720–730.
Goldin-Meadow, S., Nusbaum, H., Kelly, S.D., & Wagner, S. (2001).
Explaining math: Gesturing lightens the load. Psychological Science, 12, 516–522.
Goldin-Meadow, S., & Singer, M.A. (2003). From children’s hands to
adults’ ears: Gesture’s role in teaching and learning. Develop- mental Psychology, 39, 509–520.
Kendon, A. (1980). Gesticulation and speech: Two aspects of the
process of utterance. In M.R. Key (Ed.), Relationship of verbal and nonverbal communication (pp. 207–228). The Hague, The Neth- erlands: Mouton.
Mayer, R.E., & Anderson, R.B. (1991). Animations need narrations: An
experimental test of a dual-coding hypothesis. Journal of Edu- cational Psychology, 83, 484–490.
McNeill, D. (1992). Hand and mind. Chicago: University of Chicago Press.
Morrel-Samuels, P., & Krauss, R.M. (1992). Word familiarity predicts
temporal asynchrony of hand gestures and speech. Journal of Experimental Psychology: Learning, Memory, and Cognition, 18, 615–622.
Neill, S. (1991). Classroom nonverbal communication. London: Rout- ledge.
Perry, M. (1991). Learning and transfer: Instructional conditions and
conceptual change. Cognitive Development, 6, 449–468. Perry, M., Berch, D., & Singleton, J.L. (1995). Constructing shared
understanding: The role of nonverbal input in learning contexts.
Journal of Contemporary Legal Issues, 6, 213–236. Perry, M., Church, R.B., & Goldin-Meadow, S. (1988). Transitional
knowledge in the acquisition of concepts. Cognitive Development, 3, 359–400.
Pine, K.J., Lufkin, N., & Messer, D. (in press). More gestures than
answers: Children learning about balance. Developmental Psy- chology.
Roth, W.-M., & Welzel, M. (2001). From activity to gestures and sci-
entific language. Journal of Research in Science Teaching, 38, 103–136.
Schwartz, D.L., & Black, J.B. (1996). Shuttling between depictive
models and abstract rules: Induction and fallback. Cognitive Science, 20, 457–497.
Siegler, R.S. (1994). Cognitive variability: A key to understanding
cognitive development. Current Directions in Psychological Sci- ence, 3, 1–5.
Stigler, J.W., & Hiebert, J. (1999). The teaching gap. New York: Free Press.
Stone, A., Webb, R., & Mahootian, S. (1991). The generality of gesture-
speech mismatch as an index of transitional knowledge: Evidence
from a control-of-variables task. Cognitive Development, 6, 301–313.
Valenzeno, L., Alibali, M.W., & Klatzky, R. (2003). Teachers’ gestures
facilitate students’ learning: A lesson in symmetry. Contemporary Educational Psychology, 28, 187–204.
Wagner, S., Nusbaum, H., & Goldin-Meadow, S. (2004). Probing the
mental representation of gesture: Is handwaving spatial? Journal of Memory and Language, 50, 395–407.
Zukow-Goldring, P., Romo, L., & Duncan, K.R. (1994). Gestures speak
louder than words: Achieving consensus in Latino classrooms. In
A. Alvarez & P. del Rio (Eds.), Education as cultural construction: Exploration in socio-cultural studies (Vol. 4, pp. 227–239). Madrid, Spain: Fundacio Infancia y Aprendizage.
(RECEIVED 1/21/04; REVISION ACCEPTED 6/17/04)
Volume 16—Number 2 89
Melissa A. Singer and Susan Goldin-Meadow B W U S P S C I 7 8 6 . P D F 0 1 - D e c - 0 4 1 5 : 1 1 1 0 6 3 1 5 B y t e s 5 P A G E S
o p e r a t o r = A n a n d K u m a r
