References and Notes J. Anat. 116, 341 (1965); C. M.

Leonard, in which movement in a particular di-

Brain Res. 12, 321 (1969); Brain Behav. Eiol.
1. E. A. Spiegel, H. I. Wycis, H. Freed, C. 6, 524 (1972). rection is sustained and arrows for the
Orchinik, Am. J. Psychiat. 108, 426 (1951); 10. However, the dorsal periaqueductal gray frame in which movement terminates
L. Schreiner, D. McK. Rioch, C. Pechtel, region from which electrical stimulation elic-
J. H. Masserman, J. Neurophysiol. 16, 234 ited attack extended only to the level of the or changes direction. An acoustic sound
(1953); F. Eclancher and P. Karli, C. R. middle of the red nucleus. This is consistent reader permits location of the film
Seances Soc. Biol. Fil. 162, 2273 (1968); J. with the anatomic data; after lesions of
Physiol. (Paris) 61 (Suppl. 2), 283 (1969); R. J. thalamic attack sites, the degeneration traced frame numbers of the boundaries of
Bandler, Jr., Brain Res. 26, 81 (1971); M. to the dorsomedial portion of the central gray
Girgis, Int. J. Neurol. 8, 327 (1971); A. Siegel, substance could not be followed caudal to speech segments, such as phones, syl-
R. J. Bandler, Jr., J. P. Flynn, Brain Behav. the level of the middle portion of the red lables, and words (4). Frame-num-
Evol. 6, 542 (1972). nucleus. Electrical stimulation of the contintia-
2. M. F. MacDonnell and J. P. Flynn, Behavior tion of this dorsomedial central gray region bered sound films of the oscilloscope
31, 185 (1968). into the caudal half of the midbrain elicited display of the speech provide a check
3. The peak-to-peak threshold necessary to elicit affective defense and affective flight reactions.
attack was between 0.20 and 0.80 ma. This was determined by examination of the on the accuracy in locating speech seg-
4. M. Wasman and J. P. Flynn, Arch. Neurol. histology for four of the thirteen cats in
6, 220 (1962). which stimulation failed to elicit biting attack. ment boundaries in relation to frame
5. Further evidence that the attack elicited by Histology was not available for the other numbers. Since the same speech is
electrical stimulation of the thalamus is nine cats. Further, as stimulation along the
particularly dependent on vision is found in dorsal periventricular-periaqueductal pathway recorded again on the sound track that
MacDonnell and Flynn's report (2) that blind- proceeded into the midbrain, the distance at accompanies the film of speech visual
folding cats completely blocks the attack which visual cues were effective in eliciting
elicited by thalamic stimulation; when the approach diminished. Stimulation of sites 1 dislay, the investigator can both hear
same cats are blindfolded and stimulated at and 2 (Fig. 2) elicited an attack upon a rat and see the speech change points in
hypothalamic attack sites, they locate and placed 0.9 to 1.5 m from the cat, whereas
attack the rats. stimulation of sites 3 to 6 (Fig. 2) elicited relation to frame numbers.
6. R. P. Fink and L. Heimer, Brain Res. 4, 369 attack only if the rat was within 30 cm of
(1967); R. P. Eager, ibid. 22, 137 (1970). the cat. These procedures permit an analysis
11. R. W. Hunsperger, Helv. Physiol. Pharmacol. of the relation of body movements to
7. A. Siegel, H. Edinger, H. Lowenthal, Anat.
Rec. 175, 442 (1973). Acta 14, 70 (1956); A. Fernandez DeMolina
8. B. L. Hamilton, J. Comp. Neuirol. 149, 1 and R. W. Hunsperger, J. Physiol. (Lond.) each other and to speech in both
(1973). 160, 200 (1962); F. M. Skultety, Arch. Neurol. speaker and listener during interaction.
9. W. E. LeGros Clark, Brain 55, 406 (1932); 8, 608 (1963); S. B. Edwards and J. P. Flynn,
W. J. H. Nauta, ibid. 85, 505 (1962);
Brain Res. 41, 51 (1972). As a person speaks, several body parts
12. Supported by PHS grants MH-08936 and
and D. G. Whitlock, in Brain Mechanisms MH-25466 to J.P.F. We thank F. Gomes and are usually moving. Units of behavior
and Consciousness, J. F. Delafresnaye, Ed.
(Blackwell, Oxford, 1954), pp. 81-104; R. R. M. Groves for technical assistance. were observed: Several body parts,
Khalifeh, W. M. Kaelber, W. R. Ingram, Am. 25 June 1973 which might be moving in different
directions and with different speeds,
maintain those directions and speeds in
relation to each other for a brief time,
Neonate Movement Is Synchronized with Adult Speech: usually 0.04 to 0.16 second. This pat-
terning of movement appears to be
Interactional Participation and Language Acquisition panhuman in that it has been observed
in all films studied, including cross-cul-
Abstract. As early as the first day of life, the human neonate noves in precise tural films. Several of these quantal
and sustained segments of movement that are synchronous with the articulated configurations in infant behavior are
structure of adult speech. These observations suggest a view of development of circled in Fig. 1. These quantal forms
the infant as a participant at the outset in multiple forms of interactional organi- of microorganization of a speaker's
zation, rather than as an isolate. motion are isomorphic with the articu-
lated structure of his speech. This has
In the discipline of kinesics, methods ments-of an arm for example-can been called self-synchrony (2). Fur-
of microanalysis of sound films of hu- be repeatedly reexamined by varying ther, the configurational organizations
man communication have revealed that the number of frames scanned until the or "units" of the listener's body motion
key elements of interaction exist in the frame is located in which change in are synchronous with the speaker's
gestures, postures, and configurations direction occurs. All body parts de- speech. This has been called "inter-
of movement accompanying speech (1- tected to be moving-including brows, actional synchrony" (3). These svn-
3). These methods, applied to inter- eyes, and mouth-are systematically chronies are not readily detectable at
action between neonate and caretaker, analyzed in this fashion. Notations are normal communication speed, appear
have revealed a synchronization of in- made for each part: lines for frames to occur primarily in relation to speech,
fant movement organization with the
articulatory segments of adult speech
as early as the first day of life.
Table 1. Correspondence of infant movements with live speech. Baby E's motion segmentation
Frame-by-frame microanalysis of was compared with speech segmentation of the adult for the total sequence (study 1) and for
sound films of adult interaction has the first 336 frames only (study 2). In study 3, baby C's motion segmentation during 336
led to study of the organization of frames of silence was compared to the first 336 frames of speech segmentation on the
sound track of baby E.
events within intervals of 1 and 2 sec-
onds-the domain of microkinesics Estimated Breakdown of discrepancies by
(2). A Bell and Howell 16-mm pro- Total Total Agree- range of linguistic categories
jector (time-motion analyzer) is used frames ancies me agreement* Within Phone Word
to segment and transcribe body mo- a a) phone boundary boundary
tion (4). With this projector, film can 1 892 65 93 91.3-94.5 55 5 4
be manually transported, one frame at 2 336 21 94 90.7-96.4 14 5 2
a time, or series of frames can be 3 336 119 65 59.0-70.0 51 38 30
scanned and contrasted with each
e This column gives, for P = .025, the maximum risks of overestimating the lower limit and of
other. Each film frame has an identify- underestimating the upper limit for random samples having the percentage of misses in studies 1 to
ing number at the top (5). Body move- 3. Values obtained from (10).

It I ANU ARY 1974 99

Table 2. Correspondence of infant movements with recorded speech and nonspeech sounds. for study of correspondence between
Baby Total Total Agreement adult speech and infant movement,
Study Sound (%)
occurrences discrepancies since it is difficult to detect movement
4 Speech (word boundaries) A 146 19 87 in sleep or drowsy states. The amount
5 Disconnected vowels E 167 97 42 of speech during which infants stayed
C 124 51 59 in the awake-active state varied for
6 Tapping sounds E 27 15 44 each infant and ranged from several
C 34 16 53 sentences to lengthy utterances. Ap-
proximately 1 hour of material suitable
with regard to infant state and occur-
rence of adult speech was studied in
and are usually totally out of the inter- from a binomial population is 80 to detail.
actants' awareness. In contrast, micro- 99 percent at P = .025). In study 1, a frame-by-frame analy-
analysis of pathological behavior-for We studied whether or not such sis of the body motion of baby E was
instance, that of subjects with aphasic, "interactional synchrony" might exist performed on a section of a continuous
autistic, and schizophrenic conditions- between sound segments of adult sequence of 89 words spoken directly
reveals marked self-asynchronies (6). speech and points of change in the con- to the infant by an adult male (study
Delayed auditory feedback also mark- figurations of neonate movement. 1 in Table 1). This constituted 892
edly disturbs this self-synchrony. Video tapes were made for eleven nor- frames of verbal material at 30 fracmes
Research thus far indicates that in- mal neonates (sound film was subse- per second (pauses between phrases were
dependent judges can achieve a high quently made from the tape), and omitted). The infant was not looking
degree of reliability in microkinesic sound films were made directly for at the speaker, so eye contact could
segmentation. Four types of binomial five normal neonates. A total of ap- be ruled out. The comparison shows
reliability analysis were performed, proximately 5 hours of recorded infant correspondence between analysis of the
with 86, 90, 97, and 93 percent agree- behavior was obtained, and sections sound track of the adult's speech and
ment between independent judges (the in which the infants were in awake- the body movement of the infant. Fig-
estimated range for random samples active state were found most suitable ure 1 illustrates correspondence over
seven words transcribed by the method
of analysis described for adult behavior.
AN
For example, as the adult emits the
COME OVER kk of "come," which lasts for 0.07 sec-
FRAME NO. 12 13 14
1/30 SEC. 13093 2 3 56 8 9 10 I1 ond, the infant's head moves right very
PHONETICS K KA A m mo 0 0 0; V W. #r ir eV eV1 fl
HEAD Is _R slightly (Rvs), the left elbow extends
L. ELBOW fes _- E _ Ef :1 E us -- aEIR" -- slightly (Es), the right shoulder rotates
R. SHOULDER Ru RI
RRrs _
R. ELBOW upward (RU), the left shoulder rotates
L. SHOULDER Rax ;Rfvs ElIU JRU Is ,1RI.us outward slightly (ROs), the right hip
R. HIP ROf _RO Fe EKVs IFVS
i1 -- _1
--
v

rotates outward fast (ROf), the left

L. HIP Es _ REs 3 Rs _ROe_ E [E.Rl
& AI
_ _
RYes
-
_

L. FOOT LA AD hip extends slightly (Es), and the big

toe of the left foot adducts (AD).
SEE WHO'S
These body parts sustain these direc-
FRAME NO. 15 16 17 ' 8 19 l 20 21 22 23 :24 25 26 :27 28 29 l tions and speeds of movement together
PHONETICS A $ :Si ' a 0 f4 : u u u :z z z for this 0.07-second interval. This
HEAD
L. ELBOW
R. SHOULDER
forms a "unit" composed of the sus-
R. ELBOW tained relation of these movements of
L. SHOULDER RI RI ,'-c RIve the body. By means of the time-motion
R. HIP
L. HIP H .RIe
s
Fvve . Evs
e analyzer, the sustained movements in
L. FOOT b.RIve rEvs 1 this unit can be contrasted with those
O VER HERE
of the next unit, during the concluding
42 43 44 45 46 47
articulation of "come" (A mm), which
FRAME NO. 30 31 32 33 34 35 36 37 38 39 40 41
lasts 0.1 second. The left elbow in-
PHONETICS 0 0 V
VM
V r X e t a'a'rOf a Vr aS
HEAD R,Dv- Rf
~~~~ -_0 IP
creases speed, the right hip adds ex-
EYES
R. SHOULDER
C squeeze
I RD
tension, the left hip adds rotation in-
R. ELBOW
ROs
Cs- ward, and the big toe stops moving.
L. SHOULDER RITvs
MI r.T -sfu.)s
EvS
10 92W.-Lia
:R.U
0, .A. --
10 The previous head movement contin-
Evvs Evs
R. HIP
L. HIP ES ues. Movement of right and left shoul-
L. FOOT EP ders also continues but changes form
L. ELBOW iI so Fs -
w exactly at the end of "come." Thus, the
Fig. 1. Two-day-old neonate moving synchronously with adult speaking, "Come over infant responds organizationally to
an see who's over here." The transcription read vertically shows that the infant's adult speech: Two movement con-
configurations of movement coincide with the articulatory segments of the adult's figurations are sustained across the two
speech. Definition of descriptive notation: F, forward or flex (depending on body aspects of "come," kk and A mm,
part); H, hold; D, down; E, extend; C, close; RI, rotate inward; RO, rotate out- shoulder movement is sustained across
ward; AD, adduct; and U, up. Lower case letters refer to speed: s, slight; f, fast; and
vs, very slight. the total word kk A mm. The next
1(K0
SCIENCE, VOL. 183
word, "over," is articulated with three corded isolated vowel sounds is pre- devices, although this might be an ex-
major segments, ooo, vv, and iririr, sented in study 5, and the relation to treme interpretation of that position.
which are also accompanied by seg- recorded tapping sounds in study 6 The observations described in this
ments of infant movement exactly syn- (Table 2). report suggest that the richness and
chronous with them. This 2-day-old That this level of interaction is im- complexity of the language behavior
infant displayed segments of movement portant in relation to the more macro- system develops on the basis of an
synchronous with the adult's speech scopic level of regulation and organi- organized, "hierarchic" response of the
during the entire 89-word sequence. In zation of the infant-caretaker system infant to the organized pattern of adult
other words, this is a sustained and (8) is suggested by the observation speech. They suggest that infant motor
precise concurrence. Another 2-day- that correspondence between speech organization, entrained by these orga-
old infant sustained similarly synchro- and body movement in the neonate nized patterns for many mnonths after
fious movement throughout a series of appears when the infant is already in birth, may prepare operational formats
125 words of tape-recorded female movement. It is not usual that the for later speech.
speech (7). sound starts the infant moving from a Sustained synchrony of organized
A test was devised (study 3) to rule motionless state. However, when the correspondences between adult speech
out the possibility that such corre- infant is already in movement, points and neonate body movement at this
spondence was a random fitting-together of change in the configuration of mov- microkinesic level within epochs of
of patterns of infant movement with ing body parts become coordinated with less than a second raises issues about
patterns of adult speech. A film of points of change in sound patterns the nature of communication, and par-
awake-alert movement of baby C dur- characterizing speech. Thus, an optimal ticularly about the role of auditory
ing 336 frames of silence at 30 frames timing in relation to the state of the function in development (9). It sug-
per second, taken from another film infants' exposure to stimulation by gests that the "bond" between human
made with the same camera, was seg- speech may provide an essential char- beings should be studied as the expres-
mented and compared with the first 336 acteristic of the interactive experience. sion of a participation within shared
frames (36 words) of the speech spoken This study reveals a complex inter- organizational forms rather than as
to baby E (study 2). The frame-by- action system in which the organi- something limited to isolated entities
frame analysis of body motion of baby zation of the neonate's motor behavior sending discrete messages.
C (filmed during silence) was com- is entrained by and synchronized with WILLIAM S. CONDON
pared to the analysis of adult speech the organized speech behavior of adults Louis W. SANDER
on balby E's sound track. The corre- in his environment. If the infant, from Child Development Unit,
spondence obtained was compared with the beginning, moves in precise, shared Boston University Medical Center.
the correspondence of baby E's body rhythm with the organization of the Boston, Massachusetts 02118
motion to the same speech (studies 2 speech structure of his culture, then he
References and Notes
and 3 in Table 1). participates developmentally through
The precision of synchronization in complex, sociobiological entrainment 1. R. L. Birdwhistell, Kinesics and Context
(Univ. of Pennsylvania Press, Philadelphia,
Fig. 1 characterized the correspondence processes in millions of repetitions of 1970); A. E. Scheflen, Stream and Structure
of Communicational Behavior (Behavioral
between adult speech and infant move- linguistic forms long before he later Studies Monograph 50, Eastern Pennsylvania
ment for all 16 infants. Of these in- uses them in speaking and communicat- Psychiatric Institute. Philadelphia, 1965); Body
Language and the Social Order (Prentice-Hall,
fants, 14 were from 12 hours to 2 days ing. By the time he begins to speak, he Englewood Cliffs, N.J., 1972).
old and 2 were 14 days old. The cor- may have already laid down withln 2. W. S. Condon and W. D. Ogston, J. Ners.
Ment. Dis. 143, 338 (1966).
respondence occurred whether the adult himself the form and structure of the 3. , J. Psychiatr. Res. 5, 221 (1967).
speaker was present or whether the language system of his culture. This 4. W. S. Condon, Behav. Res. Methods Instrum.
2, 51 (1970).
voice came from a tape recorder. An would encompass a multiplicity of in- 5. J. Van Vlack, in Methods of Research in
Psychotherapy, L. A. Gottschalk and A. H.
audio tape containing American Eng- terlocking aspects: rhythmic and syn- Auerbach, Eds. (Appleton-Century-Crofts, New
lish, isolated vowel sounds, tapping tactic "hierarchies," suprasegmental York, 1966), pp. 15-24.
6. W. S. Condon and H. W. Brosin, in Schizo-
sounds, and Chinese language excerpts features, and paralinguistic nuances, phrenia: Current Concepts and Research, D.
was used as a stimulus. Two of the not to mention body motion styles and V. Siva Sankar, Ed. (P.J.D., Hicksville, N.Y.,
1971), pp. 812-837.
infants were held and the rest were rhythms. This may provide an empiri- 7. Interactional synchrony had been observed in
supine in their cribs. Chinese presented cal basis for a new approach to lan- a 6-week-old infant by W.S.C. in 1967. It
was also observed by W.S.C. and D. Stern
lo American neonates was associated guage acquisition. in a film case of a 3½/2-month-old infant with
with as clear a correspondence as was This perspective provides a descrip- mother. However, these were not systematical-
ly examined.
American English. Disconnected vowel tive frame and a conception of human 8. L. W. Sander, in Brain and Early Behavior,
R. Robinson, Ed. (Academic Press, London,
.and tapping sounds, however, failed to behavior and language acquisition dif- 1969), pp. 311-333.
show the degree of correspondence fering to some extent from traditional 9. R. B. Eisenberg, ASHA J. Am. Speech Hear.
Assoc. 12, 119 (1970).
associated with natural, rhythmic and current perspectives. While the 10. D. Mainland, Elementary and Medical Sta-
speech. The possibility that the adult learning theoretical model has empha- tistics (Saunders, Philadelphia, 1963), table 2,
p. 362.
speaker was synchronizing his speech sized response and operant interaction, 11. Supported by the Grant Foundation, New
with the infant's movements is thus it has not seemed totally adequate to York, and by NIMH research development
grant K5-20-505 (L.W.S.). Research was done
ruled out by the observation that in- the richness and syntactic complexity at Boston City Hospital and the lying-in divi-
sion, Boston Hospital for Women. We ac-
fants also synchronized movements of language behavior. The comtempo- knowledge M. La France who contributed
with speech from a tape recorder. rary alternative of transformational greatly as a judge in these reliability studies,
and P. Knapp and T. Cassel, who made
(study 4 in Table 2). The relation grammar approaches tends to empha- valuable critical comments.
between neonate movement and re- size innate acquisition and production 16 May 1973; revised 3 Auguist 1973
11 JANUARY 1974 101