KHWiOR Volume II. pages 283-290

Social and Cognitive Aspects of the Development of Aggression in Infancy Marc A. Provost Universite du Quebec, Jrois-Rivieres, Quebec, Canada

This article ycldresst's Ihctnetical and mefhcHlolouical problfins in ihc sliith til a sion in infants and tfiddlers. It is arjjucd that tht" (KTurreiKT ofajjjjressive behavior in VTr> younf; children must be considered aj^ainst tht hackeroiiiul of their developing Mwial competence. As a result, evidence for Ihe infants' capacities to ati v.ith SfjecilK" intemions and Xhv inlants' knovvledj;i' of the other as a (iistinctive individual ha\in}> his own plans and goals are reviewed. Studies of eyrly siKial interacfiuns oi humans are then examined wilh lipeciaS emphasis on eiij-nitive development and social cognition. Finally, the problem of dominance in very vouni; children's liroups Js considered.

Key words: agj^ression, infancy, stKial competente. cognition, iistenliimaiity

INTRODUCTION c research on peer reladons ainong preschiKilers i^ burgeonniog. cinpirica! a and Iheoretical work on very cyrly hunnm peer reiation.s and more spccsficatly aggression is still rare. Socia! inieraetion in young animals has been well docunted. but the dala aie hardly reievan? (o human infanis* social behavior because animals usually exhibit motoric aod social capacities more readily eoiiiparabie •ith those of human pn^schoolers. It is largely because o\' the paucity of infornsation Maccohy | I'^SOi considered thai "aggression hcgms to be an issue only m liie hird and fourth year," This situation can probably be best explained by two factors. First, historiealiy, studies of soeial interactions in hunums have concentrated more HI preschool children :han on infants inoslK beeause ol' the observation facilities ivailable in day earc tenters. The second iactor is the difficulty oi' defining and nterpreting such behav or. .A.s a result, this article addresses theoretical and method slogjcal problems in iht study of aggression in infants and toddlers.

l''!|XT p r c ' - c i i i c u a t I h i ' .Si\!lt ISjcnnuil

Address reprint requcsis !n M,A

1985 Alan R. l.iss- Inc.

\k-uun;i IITMR

!ii(,"f:iiilKi; a

SsK':i,i\

i o r \i::v::i\>:i\ .Mi A i . ' i ' j i ' s M i K i .

Provi>sL Univorsili.' du Qu:-i>cc, i'row-Rivicrcs, Quebec. Canada

:iany authors |Feshbach. 1970; Hiode. 1974: Kagan. !974| we may detune aggression as behavior directed towards causing physical injury to anothei individual However, many have pointed out that this definiiion cannot be easily operatiiHialized and requires a distinetion, not always easy, between acts thai accidenuUly lead it* injury and acts that are intentional. This dist!ncti(>n prt^ves to be crucial when we address the speciSle question of aggression in infants. In addition, because pure aliack is rare and aggression is iisiinHy associateci wiih etenients of withdrawal, ethoiogists would prefer to speak of agooislic behav!!1. For our present eoncern, threat behavior is espeeially relevant, for it implies intentionaiity. ie. a distinction between goal^ ami means IPiaget. I952|. As stressed by Harding |1983]. if one wants to interpret a behavior a.s socially directed, the assumption irmst be made that an individual is attempiing to affec! another infentionai!y--that is. in the ease of aggression, lu harm intentionally In order to examine developmental ehanges in agonistic behavior, it is necessary Sc^ make inferences about whether or not the acEor is acting with intention. It Ls thus crucial to differentiate She aclual effecis ofthe infant's behavior from the intended effects. Harding [1983], raises three major dilTicullies in studying interactions in%'Ohing prelanguage infants; 1. Although nonverbal eomniunication is well docuiiieoteci in infanc> there i> ao conventional comniunicatiisn code m yomiii children~s interchanges lo mtcrpret, aithough some have suggested that such a code exists jStrayer and Strayer, 1978; Mimtagner, !978|; 2. The infant cannot corroborate our interpretation; 3. It is difficult to specify eriteria for the competence ol nifants to intend to Because of these methodological diffieulties, we caonoi hnu! ourselves to observable effects of behavior. In other words, the observation of a two-year-old ehild taking away a toy from a peer wilhimt any precautions should not ininiediately lead us to conelude this is an aggressive act. Presumably, for that reason Bludon Jones (1972) found that his measures of "readiness to behave aggressively" did not correlate widi the amount of taking. The seeond problem lo he raised here eoncerns the cognitive level of the chskl when acting with aggression, ie, the capacities ofthe uilant realty to eooiprehend the nature ofthe other and the other's goats and plans leg. Bowlby, 1969!- Piaget |I9S2| and Ruth Nielsen f !95i| elaimed that a child cannot fully understand tfie actual poini i)f view of the other, at least at more than a very crude level, before the preschool years. Bowlby 11969| and Ainsworth 119731 '*f'5*> phK-ed tiie period of goal-correeted partnership wiih mother al around three years oi'age. However. Duno and Kendrick f 19821 have reeently challenged these views in observing that hetore '} years, chsldrcn could already anticipate and respond to the feelings of their baby siblings. Dunn and Mumi [19851 have further documented in 14 month-old infants some instances of •'pragmatic underslimding" of what will upset their older sibling.

Ai^gressitH! In Infiincx

2H5

The question of wht-thcr young children have an inlcnlion to hurt remains lo be discussed. According to Kagan JI974|. if aggression Is defmed as behavior dircclfjd toward causing physical injury to another individual, this infers that a child cannot be aggressive until he has some "psychic inluilion" of injuring. Psychic inliiition is very loosely defined and "Ae may substitute the concept t>f intention jHarding, i'l83], Kagan goes on to stau, that the most predouimant agonistic behaviors exhibited by IH-28-month-oid children are pu.shing, .striking, and seizing another's property, fn his view, pushing without any threat occurs smiply as the ''desire to tnake an eflcti ujjon the world, to test his own competence."' This last interpretation is shared by Smith |N74] for whom "some aggressive behaviour did not have (such) an obvious niotivatit)n and may have been exploratory ill nature (such as pulling hair to see what happens),'" Furthennore, Blurton-Jones 11972] finds it "arguable that even though snatching things evokes crying t)r attack. Its motivation is not always aggressive."" For him. it may suggest a lack of response !(! these reactions, but "it need n()t indicate an aggressive response in the sense of a tendency to hit or use other physical vioience," These conclusions are corroborated somewhat by Gauthier and Jacques [i985| m a study of the ontogenesis of dominanee and affiliation in preschool children, 'fheir results showed that on y 6.7% of social encounters of 6 -:H-month-old infants were olan agonistie nature, whereas lhis proportion wa.s ^.2% for 2- ,^-year old children, Ihese prop(.)rtions are quite small compared to 54.4% and 57-7% respcctivelv tor affiliative behavior. However, the more sinking result was that, of these very small proportions. 9.4% was used in threat behavisn- !or the 2- and 3-year-old group while none of the time was spent in such behavior in the 6-21-nionth-old group. This lasr resull seems to be a fairly good indication of the lack of iiitentioii m agonistic behaviinin very young children The questions of th^' infants' capacities to act with specitle intentions and the infants" knowledge of !,he other as a distinctive individual having his own plans and goals iiidicate that agonistic behavior cannot be isolated from the rest of infants" behavioral organization. However, as stated by Kaiverboer il974|. ''as long as there !s no agreement about the definition and the operationali/ailion of aggression, diseus.sR)ri about its origins is fruitless." Kaiverboer goes on tt) state tliat questions eoneern-ing the origins of aggression iead to the very roots of the child's soeial development |see also Parke and Slaby. 19831. ^'hi-- tseeurronce of aggiressive behavior in young childreii must thus be :onsidered against the background of their developing social eompetence. Parke and Slaby 11983] suggest that the precursors of the soeial regulation of aggressive intciehanges are most likely to be observed in ibe very llrst peer interactions, namely in-^ant-infant interehanges. We will thus turn iiow to studies of social iiiteraclions and examine the few cnipirieal data on aggn;ssive behavior to be found in the literature, laymg emphasis on their relation to eogiiitive development and soeiai e(tgnition. F'rom an overall impression itf the literature on eariy peer interaetions, infants ean he portrayed as basically nonaggressive jLewis el al, i975|. Several authors have reported a low incidenee of agonisiie eneounters. Hay and Ross [I982| observed 5.1% of the time spenl in eonlliets: Rubinstein and Howes [1976| only a marginal 3%; Rrownlee and Bakeman [I981| 13 hard histings in 20 hours of observatitm; and Vin/ce 11971! and Diibon el. al 119811 reported virtually none. Yet some discrepant

results ean he iouml Holmberg | I9"71 and Maudiy and Nekula [i9'^9t consideax' %)% of toddlers' social interchanges as being "disruptive" and Bronson (I975| reporis quite a high ineideoce of agonislie eneounters. thotigh eautsoniug against niterpreinsg them as hostile behavior, However, these latter results sht>i!ki be analyzed eorieiirrendv %vn'n ,NOUie OIIICE features of irifani peer interchanges. Hartup 11983] notes ihat. on nieeling. interactivi elements seem lo emerge in a nii)re or le,S!; invariant order: nitants look at. reaeli, dml toueh each vHher. Several invesiigators have nsited that infani peer encounters r^teh seemed serious in lone, impersonally inotivated and directed tiHoards loys with bnnled aUeoEion to peers jMaudly und Nekula. 1939: Rubinsleui and Howes, 1976: MucHei and Brenner. t977|. lliese deseriptions are conststeni with obsenations niatie oh infants* exploration o( inanimate objects [Hull. I97O|. Ka\anaugli and MeC'jii 119K3; hypothesize ihas toddk^rs, in clTcct, are expkiring eontingeiicies in eru:ounter> ^ i d . peers and thai they are "m the proeess of determining what aeuon wiii ge^ a reaeiiun >laymate"' (p .SO). g the seeond year of lite, infants tend lo show mure gn*ss motor activiii lu. '• their envin>nniefU, They use OKM'e and rncire banging, puslnng,. rnilling g as pan of their explonnory reperton-e Thirs. -.ome hiituig seen ui Hie interaction of ehildren in then' seeond year eoiild be interpreted -AS a kind of expk-ra U)ry behavior [Smith. I974|. Several authors have deseribed iisfanls and knidlers as iniiiaily a\\aru oiiiy oi ih% toy. ignoring the presence of (sther eiiifdren. Hckerman and Wathiey 119791 aruue ihai a |x;rson\s eonlaeting a toy increa.ses iis attraetiveness. Hint | I9';?(l] bas shi5wri ihai exploratory behavior is increased by Ihe objeet's eapaeities to mve feedback lie, moveiuen?, noise, or light). An inanimate object eould thus beeome nsore interesting if manipulated by assother ehild. To gain acees> lo play material, infants or toddieoare often foreed lo take turns with s>ne ^mother fDunion. I983|- Thus, ihe scnnewiial aecidenSal peer interaetions esiahlished by objeei eont:ie! generate a soei;n experience thai helps the ehiki lo develop more advaneed ways of interaedng witb peers. This i^ supported by Mueller and colleagues [Mueller and Lucas, 1975: Mneller and Brenner 1975] who have argoed thai oianipulative play wilh inanimate ob|eeis is the predonii mmi aetivity during peer play sessions. Recently, some investigators have chaiienged this view, Vandell et ai | I9KO| foimd that ''soeialiy direeted behaviour (SDB)*" was the niosl frequent pallern obhcr\ed in their 6 - i2-nionSh-okl subjeeis. Beekn of one's Interest, can be shmvn through politeness or This leads to a last iheoretical point, that conecras ttxe iiuerpreiauon of one o! the HiHuediate consequence>. of aggression, namely dominance. This is of speeific interest here for dt)minance oien appears lo be an aspect o! effecliveness with peers. r>(Mninanee is usually seen as a relationship. It is thus arguable ihat dominance siruclurc in a group implies some stseial perception of others and social expectation of specific responses to one's tnvn behavior, Wi(h ihis perspechve. sludics ol .socid! dominance have examined forms of social exchange and described how patterns of behavior arc coordinated between two yidividuals as shcy {niiiicipate m soeia's inieraction (Slrayer, I98(l|. Stieh empha.sis upon the inleraehve conlext may eveotuuHy reveai teatures of soeial skills and of soeia 1 ink'ntion. Many studies | Vlissakian. I*-J8O; Strayer, !98(), 1981 j have described a linear dominance structure in stable preschool groups.' Likewise, Gauthier and Jacques 11985|, e>;plonrig the sociai organi/.alkm of groups of children from one- to 5-yeafs old, observed Hrsear dooiioanee structures m all these groups. H()we\er, agonistic encounlers were observed in 94% of all possible dyads in vhc i-yeiir-oid ,^roup. whereas they were observed in 34% of all dyads in £he 5 year old group. On he other hand, concurrent measures of affiliative behavior (aikni to show social reciprocity auiong children under 2 years of age. The authors believe these results to demonstrate ihe development of social identity ihnnsgh peer relaiitmships. Identification of social couiplenientary roles, they argue, is actively ^oll>iht by the infants with as many peer.-, as possible, but their idemilieation reqiiircs tewer interactions in ."^-year old ehildren. Linear donnnance strueture appears fir^i because the iinmctiiute consetjuenee of agonistic enct)iin!ers is more conspicuous than the consequences of affiliative interaetions However, (hesc results should be ir^ierpreted aganist the backgroiind of the method used hy s.'pauthier and Jacques. Their saiuple eonsisted of fivt; groups -inie per age ievei---or different si/,c: *). y. !.^. 15, anJ \9 ehiklren, respedively. The probabHif\ lo interact with ail possible peers was

therefore higher m nifants (n--9) and toddlers rn-:..-9i than in the 5-year oiils in T-^i ('arelully designed studies are still needed to traee more aecurak;ly ihe devek^pmeni of dominanee m. partieular and of social competence in general m infants. Quesnons such as the following need to be answered bcSi)re we have an adequate f/KHire of infants' social eompeienee: When does donunanee stan to be an issue m nilunis groups'? Him rigid i.s fhc linear strucmre in the wry young! How docs annpcimon and aggression interacS with cohesive behavior'^ How can infants' social Niatus be measured preeisely',' 1b conclude, we niay say that very Httle iS known aboui she onkJgencsis lA aggression in humans. Already, a large eorpus of data on presehool children has been gathered, but very little is known about infants" agonsstie behavit*!-, Misny problems of definition arc yet to be solved and carefully designed developmental studies lirc needed to traek down (he ontogenesis of agonistic behavior. The (iiost priHiiisifig eourse, it appears, is io relate Ihe developmcni of aggression to ihe ontogenesis ^^f sociai eompctence aud communieation; ie, the mutual understanding and agreeiiicni about what an act means IRieh ACKNOWLEDGMENTS This paper was prepared while the author was we al tne .VIeuieai Kesearci Couneil Unit on the Developmenl and integration ot' •iour, m (Cambridge, The author is grateful lo A. Tamplin kn reviewing the p; •er and to R.A. Hinde tor hi; advice an REFERENCES •\i^^^vunh MDS (!97.ir Ilk; dcvcinfinient o; rr, Vhtmon Vv (108i): "Scciai and uun nioHicr audclimcnL In C'aldwcH B. Ru. opnieril." Ncv, York. Nurt ciuU; H i'cdsl: "Review ot Child D!j\x'k)p!nen! l>uh(-.n DC, Johse I) I.c/me R o e a r c h . " Vn[, ^, Chicago- r n h c r s i l y H'L de^ (k'nasigc^ iiiirrc i'nlani> Chicago I'rcvs, pri-r^ii'res innico^ dc h •, i Beckci J (!977i, A barsimg anaiysis o!" (he dcvcl .''i 2'-)0 lintance 2 opsneni of jvei oriented bs^haviorh in mnc Dum: ,1, nioruh'(i|(i snfanrs. !.)evc!opnK-nl;il Ps>choloii\ and MA !3:4K] 4')!, ' "" vard Blunon Jones N i\'-yJ2} CulciixmiSh M dvAa child Diimi J, inicracuon. Irs Bkiruin Joncs N fed): ""f-.ths!!noK iniiiv s.-oiulicr and {he A Dpviil Svji.\k\ of Chikl Beha\KHn." Cun: s^via Cr\[\o hridjzc: Canmridjji.' (Jnivcrsilv P^L:^^ DcvcioptncfK 56,4K(i 4 9 ' . ISin\[[i\ J (1069) ".'Miachfik-ni und Lo\>- Vni, I lickcnnan C, \V:i!hk!v J ! l 9 7 7 r i"o>^Uacholcr!I " Lnndon; Hogarth Pres-; inieracUiM hcvAvcn intsint pi;ers. ( ;iiu p uimicni 4 8 !64s IMb. IJndgcs K (]'>H): A studyy of social devi-iopsneni mmrl\ mi'ancy. Child Dcvchpnicm 4:36 49. }v\hhiich S (19'/0)- Aggi^'wiDfL Li Broiisoii W'X." ()97^t: Developnienis in behavior (cdj: •'C:harniich:ters Misouaf or' wiih :ij!L' nKi!c> during tht; .second ycnv ui hie cholojiy." 2v,d cd, N\nJs Yoil: Wi!o In I cw^s M., Roscnsbluin l^A U'd.sj "The hianicns. I- {i97?i. Quelque> ?cindr'i Oniiiu-: of Behavior. lTicndy.h!p and Peer Re ^aii-se dc I;.' anr.r.mnkHUo^' OOD v iaiions,"' New York: John Wde) '••,-^ inii-'Tdclnm.--, e n l i c uausiW^on

tkownlee IR, Bakcinan R (1981 t: Hiftinji m ioddicr pcci" iiitcr;!ctiiin Child ])c\eionnieiii 32;!0?6--in-y '

dcncc de ia patoic S>r,i}!o\iuni d. i;on dc p.-yeliologie -Hicnidlqise dc raise "' Piin>, P l T p p ? ) 9 8^

If! tnfiuu:\' MaiTiCiU I (i'iS.^)- Pcul (in parjer d c rUc biiginriittg of friendship, hi Lewis M, Rn••.t)ci.ik"- cnlrc nounssons dc inoins dc hivA scnbium L (eds): -Friendship an(J Peer Relanidis, l-intarR-e. No. I ?. tions " Now York. isjuihior K, Ja^qLics M i I')S5K Dc^cklppi'mcnl Maccoby BF U9K0). -Social Development: Psy.Siiiiai' Aiialysc ^^an1Vcr^alL• df la dominancf,' ehologicii! Growth and Parent-child Relaiinoe? LU: \'•diTiliiiVion chc/ les LTifanis il'iigc pr^-s ship." New York: Harcourt Brace Jovanovich, voLiir^'. In IVLTiibiey Rl-., PJo^o^,l MA, Sin:\(.-r Maudry M. Nekula M (19.^9! Social relations I'f Scdsi: •"Hlholojiic ci Je\oloppcnk'iu di' 1' hi^tweeo children of the same age durmg [he cntani " P^ris: Slock, firs! Iwo years ol lite. .loiU'iiai i)\ Cleneral Psy (i..KKifni>ugh i-l. (I'>3I): '"Anj;c! in youn^i cliil Jiology S4:I"V-2I5, dren,'" Minncafxilis: Ihi vci.'-ity oi MinnewHa Missakian FA (!98()|: (lender differenees in agoisi-^iic behavior and domnianeo relations o( Syn Hufditig C"(i (19S3j' Ac! me wiili irslcnuoii. A anon eomnuinaliy reareii ehikheii. In Oniark IVaincAork lor exiiinniini tin.' licvclopnicni of i)R. Sirayer FF". Freednian DCi (eds)' "Domi lhe interiLion io conununicatc. In Feagans i,, nanee Relations "' New Yisrk' (jariand Press. (iarv'cy C. Golinkotf R (cdsk "The OriginN Montagtier H (l'J78): "1.'enfant e( la conimunicaand Growth of ('onmiunicaiitm." Neu York: iuni." Paris. Stock. AIbex, Mueiicr 1', ([97yi: (TixJtIlcrs f ioy,sj -- (anauionHartiip WW (i98Jl: Peer rclifioos. In Musscn PH ojiious soeial sysleml. In Lewis M. Roseiii(ed): '"Handb(K)k ot" Child Psychology," Voi bjuiii L (edsl: •'The Social Network of ihc IV, New York: John Wdty, Development Infant,'' \ e w York: Pienuin. H;-.v DK Ross HS |I9X2): The social nature o! Muelicr F, Brcnnei J (19771: The ongins ol ^ocuil early contlict, CliilJ Dcvelopincisi 53.I()S ii.^ skills and inieraciion among playgroup U)d diers. Child Development 4XS54 K6I, liuidc RA O'^'/'iV. The sUidy ol aggrL'ssion: T i 1975): A developmental anal!>e Witt J, Hartup WW ;edsi: "DcicnninanEs ysis ol peei' iiHeraetion among toddlers, in and Origins of Apgrc,s.-.ivr Behavior." "I hf Ha 1 ewis M, Rosonbluiti L (eds): "Frit-ntlship aiki gLie. Moiilon. pp ^ 2X. Pixr Relalions," New York. WiSey. pp 2232,i7, llolmbcrt; MC (1977). The lA-velopnicrLi of social inlerchangc paUeriis trom L' to 42 months- Nieben RF (1951): "'Le devclopi)€r?ient dc h socross-secfioiial iiiid short-lernt longitudinal e^abiiiic chez Fenfant." KetiehareL Delachaux :inaH'.se,s, Unpubl Lshcd t octoral disscvUitKHi, ct Nieslle. i'njvcisiiy of Noith t'aroiina at Chapel HilL Parke RD, Slahy R(J (i9K:() The deveiopnien^ ol Hutt C (1970): Sj>t'ciric and t iversivc t:>>plon)UO[i. aggression. In Msissen PH (ed): "Handbook ol in Reese HW. Ijpsict LI' Ich): "Advances jn Child Psychology." Vol !V, New York: John Child DevelopiiK-nt arsd Reha*K)ut." Vol 5, Wiley and Sons. New York: Academic Press. Piagsi J (!9.'i2j' ""La naissanec de rintciligeriee."' Jacohsnn i (1^8]). The role of inariirujtc objects Netieliiitel: f>elachaux el Niestie, m early peer inlcraeiion Child rX:velopniciil Richards MPM (1974)- The development of psy 52:!i!8-626. ehological esiinniunieatioii iii the fusl vs^iir or Kagan i (1974)- DevciDprnenia! and nielhodoiogi lile In Connolly JK, Bruoner JS (eds): " l h e cal considerations in the sludy of aggression. (ifowth of (Jtmipetcnce " Nev^.^ York. AcaIn \>c Witt i, Har(up. WW (ed.s!: -Derrrnii demic Press, Jiants and Origins o!" Aggre'-sive Behjvuir Rubcnstein ,f. i!ovv,;s C 11976}: The elTccis ot The Hague: Mtsiuon, pp 107-114 peers on ioddlcr interaetion with mother and toys. Child Development 47:>97 605, Kaiverhoei A (1974): Tlte method ol direct ohservation of behaviour: soioe reniaiks on its rele- Seutl JP. Frederscson F: (19^U: The causes oi vance in lhe study of aggr.-ssii^n. In De \V)ii J. fighting in mice ijrid rats Physiological Zooi Hamip WVv (cds|, "'!>cteniHn.in!s and Orignis opy 24:27.^ UI9. nj Aggressive Behaviour. ' The H;igue: Mou ,Shat/ M (iehiian R {I973i, The devt^iopnienr oi !tM!, pp 87 95. comrnuniciition skills, inoditlealions in sfjeech Ka>anaugii RD, McCall RB ,'!983|. Sociai infiuoj young ehiklren as d function of the hsteiier. csicmg anioiig two-year {^1 js: The roie of idTil • Monograph i?f the Society loi Research in Ciiild taiive and jntagonisiie behaviors Iiiiaiii Deveiopnicni .I.'SI.SK Behavior and Deve!i>pnieni 6.39 52. Smith PK (1974): Aggi-essiori in a preschool play l.vv^i^ M. Young (j, Bnxiks ,1, Michalson L i [9']'>)group: clfecSS of varying physical icsourees

nani^ and Oiigsos o^ Aggressive BehisvRM ~ The Hat!ue: MoulorK pp 97- 105. Spsl/. R t l % ^ l , "•[ liL' Mrs! Year of i tic ~ \ o w Yofk: Inlenmiiondl UniveisiKc^ Prc"-s Kiraycr M'n'^)S()), Social eajiogv of the prt'^chool peer group. In C'fiHins W,'\ (ed)- "[>cvelopn»cnt t>f CognJlioji, Ai'ku;! iirid Sor ¥\' M^JKI'r The or^'issii/ohon arki '.'IMVI-S-, nalioi! oi a'-y!or;;eUicijl rcliuions anKviig younj.', chiidren, A Iriokti-ieal vicu 'M soi'ial power,

]!rhirvii)ral Si'tcnce "-'.3 5 4*i SU"s>cr J Sluiver \-V ^07}^!• So; i:^l .\;ip:::s-.:^', .lod poiAi-r ref.ilions .y'iong preschool eh^ldrcii Aggressivd Bi',hinior - U " V i!