«why do you ask?» context and communication in the conservation task

10
European Jourrud of Psychology of Education 1987, Vol. II, n.~ 1, 73-82 9 1987,I.S.P.A. 73 ,~Why Do You Ask?,, Context and Communication in the Conservation Task Paul Light Caroline Gorsuch Julie Newman Uniuersity o/ Southampton, England Mo• of conservation tasks •225 to increase their 'social intelligibility' have frequer~tly been shoavn to improve young chit• per[ormance. Several studies in this area have involved introducing ah emphasis on fairness o[ distribution, but the significante of this factor has not been independently assessed. Another 1actor to which attention has recently been drawn concerns the child's understanding oŸ the experimenter's intentions in asking the conservation question. Perner (1984) has reported that young children responded more accurately avhen the question was asked by an experimenter who had not witnessed the earlier stages o-[ the procedure. This paper reports ah experiment in which these two factors the emphasis on [airness and the introduction oZ a naive experimenterm were maniputated indepen• One hundred and twelve 4-6 year olds were divided between Ÿ conditions and tested in pairs on a task involving conservation o[ dis- continuous quantity. The resutts offered clear evidence that emphastsing [airness through the device of a competitive game did increase the [requency of correct responses. The introduc- tion of a second experimenter to ask the conservation question also had a signfficant, ff more limited, Ÿ effect. Pos- sible social and cognitive processes underlying these results are discussed, Ia91 The various tests of conservation (e. g, Piaget & Szeminska, 1952 [1941]; Piaget & Inhelder, 1974 [1941]) are amongst the most conspicuous aspects of the Piagetian heritage in developmental psychology. Piaget held mastery of conservation to be a crifical element in the emergence of operational thought he even referred to conservation as the psychological criterion of operational structure (Piaget 1968 [1964] p. 121). Moreover, it has been widely accepted that the ac~ievement of conservation, along with other milestones of conceptual development, is a matter of considerable educational significance. WiChin psychology, the validity of the conservation ~est has long been the subj~ct of critical scrutiny (e. g. Braine, 1959). However, receŸ years have ,seen the emergence of a distinctive critical literature on cons,ervation which holds out We are indebtea to the headtear staff and children of Banister and Bassett Green First Schools, Southamp~on, England, who made this study possible.

Upload: paul-light

Post on 25-Aug-2016

212 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: «Why do you ask?» context and communication in the conservation task

European Jourrud of Psychology of Education 1987, Vol. II, n. ~ 1, 73-82 �9 1987, I.S.P.A.

73

,~Why Do You Ask?,, Context and Communication in the Conservation Task

Paul Light Caroline Gorsuch Julie Newman Uniuersity o/ Southampton, England

Mo• of conservation tasks •225 to increase their 'social intelligibility' have frequer~tly been shoavn to improve young chit• per[ormance. Several studies in this area have involved introducing ah emphasis on fairness o[ distribution, but the significante of this factor has not been independently assessed. Another 1actor to which attention has recently been drawn concerns the child's understanding oŸ the experimenter's intentions in asking the conservation question. Perner (1984) has reported that young children responded more accurately avhen the question was asked by an experimenter who had not witnessed the earlier stages o-[ the procedure. This paper reports ah experiment in which these two factors

the emphasis on [airness and the introduction oZ a naive experimenterm were maniputated indepen• One hundred and twelve 4-6 year olds were divided between Ÿ conditions and tested in pairs on a task involving conservation o[ dis- continuous quantity. The resutts offered clear evidence that emphastsing [airness through the device of a competitive game did increase the [requency of correct responses. The introduc- tion of a second experimenter to ask the conservation question also had a signfficant, ff more limited, Ÿ effect. Pos- sible social and cognitive processes underlying these results are discussed,

Ia�91

The various tests of conservat ion (e. g, Piaget & Szeminska, 1952 [1941]; Piaget & Inhelder, 1974 [1941]) are amongst the most conspicuous aspects of the Piagetian heritage in developmental psychology. Piaget held mastery of conservation to be a crifical e lement in the emergence of operat ional thought he even referred to conservat ion as the psychological cr i ter ion of operat ional s t ructure (Piaget 1968 [1964] p. 121). Moreover, it has been widely accepted that the ac~ievement of conservation, along with o ther milestones of conceptual development, is a mat te r of considerable educat ional significance. WiChin psychology, the validity of the conservat ion ~est has long been the subj~ct of critical scrut iny (e. g. Braine, 1959). However, receŸ years have ,seen the emergence of a dist inctive crit ical l i terature on cons,ervation which holds out

We are indebtea to the headtear staff and children of Banister and Bassett Green First Schools, Southamp~on, England, who made this study possible.

Page 2: «Why do you ask?» context and communication in the conservation task

74 P. LIGHT, C. GORSUCH & J. NEWMAN

the p romise of f resh insights in.to the processes of cognit ive development . This l i t e ra tu re is concerned wi th the social context of the conservat io~ tes t and the in te rp re ta t ive processes at vr within it.

In general te rms, research in this a rea has involved the modi f ica t ion of the conservat ion tes t p roeedure in ways designed to e l iminate potent ia l ly mis- leading fea tures and to t e n d e r ir more 'social ly intelligib•e' to the child (Finn, 1985). Most s tudies have involved modff icat ions to the way in which the rearrange- ment or t r ans fo rma t ion of material.s in hand,led in the course of the test . The essent ia l compla in t aga ins t the s t anda rd Piaget ian p rocedure is that , by drawing the chi ld 's a t t en t ion to his ac t ions as he t r ans fo rms or r ea r ranges the mater ia l s , the experimen.ter is in effect refer r ing nonverbal ly to one p a r a m e t e r th rough his act ions (e. g. lenght o,f rows in the case of n u m b e r conservat ion) while re fer r ing verba l ly to ano the r (e. g. number of counters),. Much of the chi ld 's d i f f icul ty may lie in an inabi l i ty to disentangle these conflicting referent ia l cues.

McGarr igle and Donaldson (1974) sought to d e m o n s t r a t e this by compar ing a s t a n d a r d p resen ta t ion of n u m b e r and length conservat ion p rob lems wi th a modifi~ed presen ta t ion , in which the t r ans fo rma t ion was achieved ' acc identa l ly ' th rough the agency of a naughty teddy bear . Subs tan t i a l ly h igher levels of cor rec t j udgemen t were ob ta ined f rom 4 and 5 year olds in the modi f ied condit ion. Subsequent s tudies using s imi la r p rocedures have general ly found s imi la r resul ts , though differences have not always been s ta t i s t ica l ly re l iab le (e. g. Hargreaves , Molloy & Prat t . 1982; Miller, 1982; Neilson, Dockrel l & McKechnie, 1983).

A sl ightly d i f ferent app roach to the p rob l e m of mis leading cues su r round ing the t r ans fo rma t ion has involved a t t empt ing to make the t r ans fo rma t ion appea r to be incidental to the proc,eedings r a the r than cent ra l to them. Light, Buck ingham and Robbins (1979) achieved this by test ing 5 yea r old chi ldren in pa i r s in the context of a compet i t ive game. The game requi red tha t the chi ldren s t a r t wi th equal amoun t s of ma te r i a l (in this case p a s t a 'shells ') . Fqua l quant i t ies of shells were p laced in two s imi la r glass beakers . However , j u s t a s the game was abou t to begin a dangerous ly sha rp chip was found in the r im of one beaker . This p rovided the ra t iona le for pour ing the contents into ano ther d i f ferent ly shaped beaker which ' jus t happened ' to be available. This condi t ion p roduced very much h igher levels of cor rec t conserving judgemen t than a cont ro l condi t ion in which chi ldren were test,ed in pa i rs bu t wi thout e i ther the game e lement o r the ra t iona le for the t ransformaf ion . Mit ler (1982) has rep l ica ted this s tudy successfully, as have Bovet, Parra t -Dayan and Deshusses-Addor (1981), who suggest that the game element , by drawing a t ten t ion to the issue of fairness, may have had an i m p o r t a n t pa r t to play.

Other s tudies have highl ighted the s t ruc ture of the d iscourse in the s t anda rd conserva t ion procedure , and in pa r t i cu l a r the repe t i t ion of the same quest ion, once Nefore and once a f te r the r ea r r angemen t of the mater ia l s . Norma l ly all ch i ldren respond cor rec t ly the f i rs t t ime the quest ion is asked, since the ma te r i a l s a re a r ranged in such a way as to make the equal i ty obvious. But when the quest ion is asked again af te r the r ea r rangemen t young chi ldren charac te r i s t i ca l ly change the i r answ~r. I t rnay be that the s imple fact of the repe t i t ion of the ques t ion leads some children to suppose tha t the i r f i rs t answer m a y have been wrong. Alternat ively, the fact tha t the ques t ion is r epea ted a f te r the rear range- men t may lead the child to suppose tha t the r e a r r a nge me n t must have been rel.evant to the qu,estion in some way. Rose and Blank (1974) showed tha t by s imply omi t t ing the ini.tial questŸ in a n u m b e r conservat ion tes t they could ob ta in s ignif icant ly higher levels of cor rec t responding to the a l l - impor tan t pos t - t r ans fo rmat ion question. Samuel and Bryan t (1984) have obtain, ed s imi lar resul ts not only for n u m b e r conservat ion bu t for several o ther conservat ion tasks .

Page 3: «Why do you ask?» context and communication in the conservation task

CONTEXT AND CONSERVATION 75

Berner (1984) h~'s reported an intr iguing study in which the impact of quest ion repeti t ion was al tered in a ra ther more subtle way, by having the two questions asked by different people. The task involved two toy animals which were to be given equal amount s to dr ink ' to make tt fair'. The ch'ildren (4 'to 6 years old) had first to fill two identical buckets to equal levels and then to pour the contents of these buckets in~o the animals ' dr inking troughs, which were of differing shapes They were questioned about the equali ty of amounts both before and after pouring. However, for half of the sample, immediate ly after the child had poured the dr inks into the troughs the exper imenter announced that he had to leave and asked another adult f rom the n.ext room to take over. Thus in this case the post- t ransformat ion question was asked by an experimenter who had not seen the water in the buckets pr ior to pouring. Significantly more children gave correct judgem,ents in this 'naive experimenter ' condit ion than where the original exper imenter remained throughout.

Perner suggest's tha t the naive experim,enter condit ion is more intelligible to the child because the incoming experimenter, having not seen the pre-transfor- mat ion array, has reasonable grounds for asking whether the two amounts ate the same. His quest ion can th.erefore be taken at face vaJue, a s a s t raightforward request for information. In the usual single experimenter condition, by contrast , the experin~tenter's purpose in asking the quest ion may be obscure to the child unless he or she is able to grasp that the quest ion is not so much an enquiry about the quanti t ies as a test of the child's unders tanding. Perner suggests that young children have difificulty in grasping such 'second order epistemic intent ions ' . In the faoe of their resul t ing confusion they may resort to simply comment ing on the changed appearance of the array, and thus produce 'noneonserving ' respons.es.

This study has deservedly at tracted some at tent ion in the recent li tera ture (e. g. Elbers, 198£ bu t it is not without its problems. One ra ther surpr is ing feature of the results was the high level of cor~ect judgements found even in the single experimenter condition, This may have arisen, as the authors suggest, f rom the emphasis on fairness ,of d is t r ibut ion which was common to both conditions. Another factor which may have contr ibuted is the wording of the post- t ransformation question: ,,Did you gire them the same?,,. The form and tense of this question may have biased the children towards answering it in terms of the events preceding the t ransformation. Such a bias could plausibly have been accentuated by the ~ntroduction of a second experimenter who had not been present dur ing the earl ier stages of the proceedings.

In view of these ambiguities, the present study was designed to allow a clearer assessment both of the influence of emphasising fairness and of the influence of the in t roduct ion of a second experimenter . These two eactors were manipula ted independently, and close a t tent ion was paid to the wording of the conservat ion question.

Method

One hundred and twelve chi.ldren, 56 boys and 56 girls, served a.s subjects for the experiment. Their ages ranged from 4 years 8 months to 6 years 4 months. They were drawn from two state schools in Southampton, England, the catch- ment areas of which covered a wide ~socioeconomic range. Each child was assigned to one of four experimental conditions. Both schooIs and both sexes were equally represented in all conditions. The mean ages of children in the four condit ions ranged f rom 5 years 4 months to 5 years 6 months. Within each

Page 4: «Why do you ask?» context and communication in the conservation task

76 P. LIGHT, C. GORSUCH & J. NEWMAN

condit ion each child was paired with a par tner of similar age from the same school class. Within each pair one of the chi ldren was chosen arbi t rar i ly in advance to be the .first one quest ioned thr~ughout the test.

The four condifions all involved a test of conservat ion of discont inuous quanti ty. Condit ion S / l E involved a ' s tandard ' context with no emphasis on fairness, and a single experimenter . Gondition S/2E involved the same s tandard context but with the in t roduct ion of a second experimenter. Condition G / l E involved a 'game' context which served to empha, sise fairness, with a single experimenter . Finally, condit ion G/2E involved the same game context bu t with two experimenters.

In all condit ions the chi ldren were collected in their pairs f rom the classroom by one of the experimenters (E') and ~aken ~o a quiet adjoining room where the other exper imenter (E) was wai t ing to conduct the testing. E' then left the room. The two experimenters , both female, took turns to play the role of E' and E.

In condit ion S / l E the chi ldren were presented at the outset with a pile of dried peas, and were a, sk, ed by E to help her to divide the peas into two equal piles. Then each child was ~sked, in predetermined order, whether the amounts of peas in the twc, piles vvere equal. When children judged that the two amounts were not equal, they were encouraged to make them so. When both agreed that the two amounts were equal, E produced two glass containers, one considerably fat ter and shorter than the oth.er, and said ,,Let's put the peas into these pots,,. When the two piles of peas had been put into the two pots, E asked: ,,Is there the same amoun t of peas in each pot?,,. Ir should be noted that this question, though somewhat inelegant, expl~citly referred to the present si tuation, with the peas in the pots~ ra ther t han to the previous si tuation. The quest ion was asked of each child in t u rn and then each child w~s fur ther asked: ,,How do you know?~,.

The procedure for condit ion S/2E was identical ~.n all respects up to the point at which the peas were t ransfered into the pots. At this precise moment , however, E' knocked .at the door and entered, saying that E was wanted on the telephone. E asked E' to take over and departed, leaving E' to ask the post- t ransformat ion quest ions using exactly the same wording as in the previous eondition.

The two 'game' condition, s began with the following preamble: ,,You are going to p l a y a game with these peas. I t is very impor tan t that you both have the same amoun t of peas so that the game will be fair. The winner will be the first one to move all their peas using a s t raw like this,,. E then demonst ra ted how to lift a pea by sucking it with a straw. The procedure from this point on was exactly as for the s tandard conditions, with the peas being divided into two equal piles before being trans,fered into the two pots. The wording of the post-trans- format ion quest~ons was exactly as described for the previous conditions. In condit ion G / l E these quest ions were asked by E, while in condi t ion G/2E the telephone in te r rup t ion was ~staged and the quest ions were asked by E'. I n botla condit ions the chi ldren then went on to play the game, using dr inking stravr to more their peas into plastic bowls.

Results

The number s of children judg~ng the amounts still to be equal after trans- formation ate shown in Table 1. As predicted, the overall percentage of correct vesponses is lowest in the S / l E condit ion (the closest to the s tandard Piagetian situation) and highest in the G/2E condit ion (the closest to that used by Perner, 1984).

Page 5: «Why do you ask?» context and communication in the conservation task

CONTEXT AND CONSERVATION

Table 1: Numbers of children in each of the four conditions offering correct (conserving) judgements

77

CONDITION 1ST CHILD 2ND CHILD TOTAL (%)

n - - 1 4 n = 14 n = 2 8

S / l E 2 3 5 (18 %)

S/2E 7 8 15 (54 %)

G / l E 10 6 16 (57 %)

G/2E 9 11 20 (71%)

Tableau 1: Nombres d'en… pr› un jugement de conservation correct dans chacune des 4 con•

The data were analysed using the procedure out l ined by Meddis (1984) for factorial analysis of frequency tables. This procedurr offers an al ternative to Chi Square and has the advantage that ir permits analysis of interact ion terms. Ah overall analysis on the totals (see Table 1) indicates the existence of highly significant differences among the conditions (H = 17.27, dl. = 3, p < 0.005; H is a measure of va¡ di.stributed as chi square). Comparison of the combined 'S' condit ions with the combined 'G' condit ions reveals a significant main effect for corrtex~ (z = 3.01, p < 0.005), more correct responses being obtained in the game contexts. Comparison of the combined ' l E ' condit ions with the com- bined '2E' condit ions reve~ls a significan~ main effect for factor also (z = 2.63, p < 0.005), more correct responses being obtained where the second experimenter was introduced. The in teract ion term i.s not statisticaUy significant (z=l.13, p>0.1)..

Howewer, since the children were in all cases tested in pairs the data are not strictly independent , and any tendency towards conformity wi thin the pairs could have the effect of magnifying the differences betwe~n the conditions. In actual fact there was evidence r such conformity only in the condit ions which involved the s tandard context. (S/ lE: z = 2.8, p < 0,005; S1/2E: z = 2.1, p < 0.05). In the light of this complicating factor ir might be more appropriate to adopt the conservative strategy of basing the analysis on the '~mcontaminated' responses of the first child quest ioned in each pair (see Table 1).

When analysis ,is restricted to these first-questioned subjects an overall analysis still indicates the existence of statistically signi~icant condi,tions dfffe- rences (H = 10.66 dl. = 3, p < 0.05). Moreover, compa¡ of the combined 'S ' with the combined 'G' condit ions revealed a significant ma in effect for context (z = 2.65, < I).005). However, the ma in effect a, ssociated with the other factor ( lE vs. 2E) failed to reach statistical significance (z = 1.06, p ~ 0.1). There were indications of ah interact ion between these factors, in that, as Tabl, e 1 shows, the in t roduct ion of the second exper imenter has the predicted effect in the s tandard condit ions bu t no effect at all in the game conditions. The interact ion term from the analysis, however, fails to reach an acceptable level of ,statistical significance (z = 1.59; p = 0.06). A local comparison of outcome between condit ions S / l E and S/2E does show a significant difference (z = 1.99, p < 0.05), more correct response being obtained where the second exper imenter was introduced.

The children's just i f icat ions for their responses were grouped into three categories based upon those used by Perner (1984). The 'uniformat ive ' category (e. g. ~I don ' t know,,, or ~,it jus t is,0 acounted 'for some 55 % of all responses. Responses making reference to the differing shapes of the two con tainers were classed as ' spontaneous compen, sation' (e. g. ~,that one is wider so the peas are

Page 6: «Why do you ask?» context and communication in the conservation task

78 P. LIGHT, C. GORSUCI4 & J. NEWMAN

more spread oub,). These accounted for 30 % of responses. The remain ing 15 % fell into the ' r e t rospec t ive reference ' category, which involved some reference back to the ea r l i e r equal i ty of the two piles of peas (e. g. , ,because we m a d e the piles bo th the s~.me,,). This f igure cont ras t s sha rp ly wi th Perner ' s f inding tha t approx ima te ly 75 % of justi, f ications fell into this ' r e t rospec t ive ' category.

Discussion

Whether ana]ysed on tbe bas is of the whole sample or more conservat ively on the bas is of the f i rs t child ques t ioned in each pair , it is c lear f rom the resul ts of this s tudy tha t the ' game ' context was signi.ficantly faci l i ta t ive of cor rec t conservat~on judgements . The level of cor rec t responding in the game contexts averaged 64 %, a level not d i ss imi la r to tha t found by Perner (1984) a.nd ear l ie r by Light, Buck ingham and Robbins (1979). Both of these previous s tudies had included an emphas i s on the need for fairnes.s in d i s t r ibu t ion bu t ne i the r of them al lowed objec t ive assessment of the influence of this factor . In the case ~f Light, Buckin~ham and Robbins , the in t roduc t ion '~f the garne context was confounded wi th t he in t roduc t ion of an ' inc identa l ' ra t iona le for t r ans fo rma t ion (the chipped beaker) . Present f indings lend some weight to ~the sugges.tion noted in the in t roduc t ion tha t the game context may have been at leas t a necessary ir not a ,sufficient condi t ion for the fac i l i ta t ion found in tha t s tudy.

I t has been sugges ted (e. g. Bovet et al, 1981) tha t the game context serves to d | s t r ac t ch i ldren f rom the conservat ion t a s k - - t h a t in the i r has te to get on with the game they may be rela t ively ina t tent ive to these 'p re l iminar ies ' . The chi ldren in the pres.ent s tudy gave every indica t ion of a t t end ing clos~ly to the proceedings, however , and indeed i t could be argued tha t the game context p rovides them wi th a s t rong incentive to a t tend. Viewing the m a t t e r in t e rms of the a p p e a r a n c e / r e a l i t y di,stinction (Flavell, Flavell , & Green, 1983; Russell , 1982) i t seems plausible to suggest tha t the context of the impending game might help to ensure tha t ch i ld ren unde r s t and the conservat ion quest ion in t e rms of objec t ive equali ty, wbereas in the s t anda rd context they m a y unde r s t and i t in t e rms of appearances .

The fac i l i t a to ry effects of placing an emphas i s on fa i rness of di, s t r ibu t ion may con t r ibu te s ignif icant ly towards the well-documented benef i ts of a l lowing chi ldren to in te rac t wi th one another in con~servation s i tua t ions (e.g. Doise & Mugny, 1984 [1981]; Perret -Clermont , 1980). In some cases the r equ i remen t to achieve 's shares ' has been expl ici t in such s tudies and, even where i t is not, the re may be a res idual tendency for ch i ldren tes ted in pa i rs o f smal l groups to see the conservation t a s k in t e rms of the e s t ab l i shment and main tenancr of equal ' shares ' (Light & Perret-Clermont , 1986). Recent work in this f ie ld has indeed emphas ised the impor t ance of ' social mark ing ' as an ingredient in peer fac i l i ta t ion studies, and the s i tua t ion where two chi ldren of equal s ta tus a te given equal shares can be seen in these t e rms as a social ly m a r k e d s i tua t ion (cf. Doise, 1985; Mugny, 1985). However, i t should be no ted tha t the absence of individual post- tes ts in the p resen t s tudy makes any d i rec t compar i son wi th the s tudies of Doise and colleagues difficult .

Besides the issue of ' f a i rness ' the ot•er pr inc ipa l focus of this s tudy was on the effects of in t roduc ing a second expe r imen te r to a, sk the conserva t ion quest ion. As was no ted in the in t roduct ion, the s tudy by Perner which a p p e a r e d to show such an effect was f lawed by ques t ionable wording. The word ing used in the p resen t s tudy was designed to reduce any tendency to in te rp re t the conserva t ion quest ion as bear ing on the p re - t r ans fo rma t ion ,state of affairs.

Page 7: «Why do you ask?» context and communication in the conservation task

C O N T E X T A N D CONSERVATION 79

The m a r k e d reduct ion in ' r e t rospec t ive reference ~ in the chi ldren 's jus t i f ica t ions for the i r j udgement s suggests tha t this change did have the des i red effect.

The changed word ing m a y have cont r ibu ted to a weakening of the effect of in t roducing the second exper imenter . Indeed, the resul ts showed l i t t le difference be tween condi t ions G / l E and G/2E, which were the condi t ions mos t closely resembl ing those used by Perner. Nevertheless, e l sewhere in the resul ts there were s t rong ~ndications tha t the in t roduc t ion of a second exper imente r did have a faci l i ta t ive effect. The overal l f requency of cor rec t judgements in the 2E condi t ions was ~significantly h igher than in the l E condit ions. This difference fai led to reach s tat is t ical signfficance when analysis was based on the f i rs t respondents only, bu t the s t rong indicat ions of ah in terac t ion be tween context and exper imente r manipu la t ions may help to expla in this. When the s t anda rd condi t ions are considered on the i r own, a signfficantly h igher level of cor rec t responding in the 2E than i r the l E condi t ion is once again apparent .

The resul ts concerning this in te rac t ion are not suff iciently c learcut to bea r the weight of much in te rpre ta t ion . However, we should at least note tha t the 'game ' context, while ra is ing levels of c o r r ~ t responding may also reduce the chi ldren 's sensi t iv i ty t,o o ther p rocedura l modif icat ions . I t may be r e m e m b e r e d tha t in the star~dard condi t ions we found signfficant evidence tha t the child ques t ioned second tended to conform with the response offered by the first- ques t ioned child. The same finding was not ob ta ined in the game condit ions. This issue of the poss ible interaction,s between dif ferent contextual modi ' f icat ions is one which may repay fu r the r study.

The resul ts d iscussed above af ford some suppor t to Perner ' s a rgument that chi ldren in conservat ion tes ts a re l iable to mis in t e rp re t the exper imenter ' s ' second o r d e r ep i s temic intentions ' . Perner sees this as ah aspect of a r a the r general cognit ive l imi ta t ion (cf. W i m m e r & Perner , 1983). However , Elbers (1986) has reccnt ly poin ted out tha t such ' second o rde r ' quest ions (in which the quest ioner is more concerned wi th probing the child 's level of under s t and ing than wi th the answer per se) are actual ly commonplace in the d iscourse between adul t s and young chi ldren, bo th at home and at school. He argues tha t even preschool chi ldren are well used to such quest ions occur ing in the context of ins t ruct ion, and moreover tha t they under s t and them for wha t they are. However, Elbers argues, wi th in the context (or 'metacont rac t ' ) of ins t ruct ion adul ts typical ly used such quest ions a longside a var ie ty of devices and responses �91 to suppor t and reirfforce the child~s unders tanding . The way in which such devices for suppor t ing under s t and ing opera te in bo th parent-chi ld and teacher-chi ld exchanges has been discussed in g rea te r detai l by Wood (1986). But as Elbers notes these suppor ta t ive devices a re no tab ly lacking in the s t anda rd con, servat ion si tuat ion. Using Rommetve~t 's (e. g. 1979) terminology, Elbers suggests tha t while the child may expect the no rma l ground rules of t each ing / lea rn ing s i tuat ions ( ' the me tacon t r ac t of ins t ruct ion ' ) to apply, the adu l t is in fact opera t ing wi thin the very dif ferent g round rules of the test ing s i tua t ion ( ' the me tacon t r ac t of examina t ion ' ) Elbers ,suggests tha t the child 's diff icult ies a r i se f rom this discre- pancy, and from the chi ld 's unfami l ia r i ty wi th the ground rules of the test ing si tuat ion. Clearly this i ssue of impl ic i t 'me tacon t rac t s ' underp inn ing exchanges be tween adul t and chi ld is of considerable significance in an educat ional context. As Mercer and Edwards observed in thei r d iscuss ion of the ground rules of the c lassroom: , ,Knowing and learni~g a re not psychological bu t social psychological processes. K:aowing and reasoning. . , can only be demons t ra ted , recognised, t aught and assessed th rough acts of communication~~ (1981, p. 38).

I t is pe rhaps i m p o r t a n t to note tha t any analysis couched in term,s of g round rules o r me tacon t r ac t s of communica t ion impl ies a widening of the

Page 8: «Why do you ask?» context and communication in the conservation task

80 P. LIGHT. C. GORSUCH & J. NEWMAN

coneept of context. The context of the conservat ion task, in these terms, is determined not only by the intra-experimental variables such as the materials , the ins t ruct ions and the exper imenter ' s actions, bu t also by the way in which these resonate with the child's preexisting experiences in related communicat ive situations. Elbers (1986) has gone some way towards ref ining our available descriptions of the various pat ternings of teacher-child exchange to which the child mus t become par ty dur ing the early ,school years. There is much more to be done in this direction before wr will be able to clŸ any more than a very superficial unders tand ing of the processes involved. Exper iments such as the present one offer o~e possible way of forwarding this endeavor.

Elbers, like Perner (1984), appears to accept that, given a sufficiently clear unders tand ing of the question, young children a te in fact perfectly capable of und.erstanding ccn, servation of quantity. Correct responses to the conservat ion quest ion obtained through the various modi, fications to the task are taken as indicative of this unders tanding. We have argued elsewhere for a ra ther different view (e. g. Light, 1986) suggesting that correct judgements offered in game-like or other 'socially intelligible ' contexts may often be based on much les s than full unders tanding. Certainly the just if icat ions obtained from children in thi,s s tudy would do noth ing to contradict this. But we have also suggested that the social processes which suppor t such precociously correct judgements may actually have a const i tut ive role in helping to form and con, solidate the child's unders tanding. The results of the present study do not allow us to choose between these inter- pretat ive posi t ions , bu t they do confirm the existence of two dist inct influences upon young children 's responses. All in te rpre ta t ions of such influences concur in directing our a t ten t ion towards the complexities of th, e social nexus within which the child's cognitive c o m p e t e n c e s are expressed, and in which perhaps they have their origins.

References

Bovet, M., Parrat-Dayan, S., & Deshusses-Addor, D. (1981). Peut-on parlr de precocit› et de r› dans la conservation? I: Precocit› Archives de Psychologie, 49, 289-303.

Braine, M. (1959). The ontogeny of certain logical operations: Piaget's formulation examined by non-verbal methods. Psychological Monographs, 73, No. 5.

Doise, W. (1985). Social regulations in cognitive development. In R. A, Hinde, A. N. Perret-Clermont & J. Stevenson-Hinde (Eds.) Social Relationships and Cognitive Development (pp. 294-30B). Oxford: Clarendon Press.

Doise, W. & Mugny, G. (1984). The Social Development of the Intellect. Oxford: Pergamon Press (originany published, 1981).

Elbers, E. (1986). Interaction and instruction in the conservation experiment. European Journal of Psychology of Education, 1, 77-89.

Finn, G. (1985). L'intelligibilite sociale de la tache. In G. Mugny (Ed.). Psychologie Sociate • Developpement Cognitif, (pp. 167-184). Berne: Peter Lang.

Flavell, J., Flavell, E., & Green, F. (1983). Development of the appearance -- reality distinction. Cognitive Psychology, 15, 95-120.

Hargreaves, D., Molloy, C., & Pratt, A. (1982). Social factors in conservation. British Journal ol Psychology, 73, 231-234,

Light, P. H. (1986). Context conservation and conversation. In M. P. M. Richards & P. H. Light (Eds). Children 01 Social Worlds. Cambridge: Polity Press; Cambridge, Mass.: Harvard University Press.

Light, P. H., Buckingham, N., & Robbins, H. (1979). The conservafion task as ah interactional setting. British Journa! oŸ Educational Psychology, 49, 304-310.

Light, P. H., & Perret-Clermont, A.-N. (1986). Social construction of logical structures or social construction of meaning? Dossiers de Psychologie, Universit› de Neuchatel, No. 27.

McGarrigle, 3., & Donaldson, M. (1974). Conservation accidents. Cognitton, 3, 341-350

Meddis, R. (1984). Statistics using ranks: a uniŸ approach. Oxford: Basil Blackwell.

Page 9: «Why do you ask?» context and communication in the conservation task

C O N T E X T A N D C O N S E R V A T I O N 81

Mercer, N., & Edwards, D. (1981). Ground rules for mutual understanding: a social psychological approach to classroom knowledge. In N. Mercer (Ed.), Language in School and Community, (pp. 30-46). London: Edward Arnold.

Miller, S. (1982). On the generalisability of eonservation. British ]ournal of Psychology, 73, 221-230. Mugny, G. (198.:;). (Ed.) Psychologie Sociale du Developpernent Cognitif. Bern: Lang. Neilson, I., Dockrell, J., & McKeehnie, ~'. (1983). Justifying conservation: a reply to MeGam5gle and

Donaldson. Cognition, 15, 277-291. Perner, J. (1984). The insincerity of conservation questions. Paper presented to the British Psyehologieal

Society D~:velopmental Section Annual Con/erence, Lancaster, England (September). Perret-Clermont, A.-N. (1980). Social lnteraction and Cognitive Development in Children. London:

Academic Press. Piaget, J. (1968). Six Psyr Studies. London: University of London Press (originally published,

1964). Piaget, I., & Inhelder, B. (1974). The Child's Construction of Quantttzes: Conservation and Atomism.

London: Routledge and Kegan Paul (originally published, 1941). Piaget, J., & Szeminska, A. (1952). The Child's Conception of Number. London: Routledge & Kegan

Paul (originally published i.n 1941). Rornmetveit, R. (1979). Deep strocture of sentences versus message structure. In R. Rommetveit

& R. M. Blakar (Eds.), Studies of Language, Thought, and Verbal Communication (pp. 17-34). London: Academic Press.

Rose, S., & Blank, M. (1974). The potency of context in children's cognition. Child Deve•opment, 45, 499-502.

Russell, J. (1982). Propositional altitudes. In M. Beveridge (Ed.). Chil• Thinking Through Language. (pp. 75-98). London: Edward Arnold.

Samuel, J., & Bryant, P. (1984). Asking only one question in the eonservation experiment. Yournal of Child Psychology and Psychiatry, 25, 315-318.

Wimmer, H., & Perner, J. (1983). Beliefs about beliefs: representation and constraining funetion of wrong beliefs in yotmg ehildren's understanding of deception. Cognition, 13, 103-128.

Wood, D. (1986). Aspects of teaclaing and learning. In M. P. M. Richards & P. H. Light (Eds.). Children of Social Worlds, (pp. 191-212). Cambridge. Polity Press; Cambridge, Mass: Hin-varal University Press

' P o u r q u o i m e p o s e z - v o u s e e t t e q u e s t i o n ? ' C o , n t e x t e et communication dan~s la t~ehe de c o n s e r v a t i o n

Les modiŸ des taches de conservation destindes accroitre leur ,,intelligibilit› sociale,, ont [rdquemment en-

train› une am› des per[ormances des jeunes en… Plusieurs ›225 dans ce domaine ont mis l'accent sur l'im- portanoe du facteur a›233 de la r› sans toutefois que l'importance de ce facteur ait ›233 encore ›233 s›233 Un autre facteur r› ›233 concerne l'intention que l'enfant attribue dl I 'exp› qui lui pose la question de conservation. Perner (1984) rapporte que la r› de jeunes enfants est plus souvent correcte quand la question leur est pos› par un exp› n'ayant pas assist› aux phases ant› de la proc›225 Cet article prdsente une recherche • laquelle ces deux f ac t eur s - -accen t mis sur l'›233 et introduetion d'un exp› n d i f - - o n t ›233 manipul› in•233225 112 enfants de 4 �91 6 ans ont ›233 r› dans quatre conditions exp› et confron- t› par paires, �91 des taches de conservation des quantit› dis- continues. Les r› montrent clairement que l'intro• d'une recherche •233233 les situations › identifi› ~ des jeu_x comp› entraine une augmentation de la fr› des r› correctes. L'introduction d'un deuxikme exp› mentateur au momen t du questionnement sur la conservation est › source d'un effet significatff (quozque plus … de Ÿ Les m› sociaux et cognitifs sous-tendant ces r› sont pr›233 et discut›

Key words: C o n t e x t , C o n s e r v a t i o n , C o m m u n i c a t i o n .

Page 10: «Why do you ask?» context and communication in the conservation task

82 P. L I G H T , C. G O R S U C H & J. N E W M A N

Received: August 1986 Revision received: November 1986

P a u l H e n r y L i g h t . Department of Psychology, The University, Southampton, England, S09 5NH.

Current theme of research:

Social factors in cognitive developmcnt

Most reIevant publications in the Ÿ of E• Psychology:

Light, P. H., & Glachan, M. (1985). Peer interaction and problem solving. Educational Psychology, 5, 217-225.

Light, P. H. (1985). The development of view.specific representation considered from a socio- cognitive standpoint. In N. H. Freeman & M. Cox (Eds.),Visual Order, (pp. 214-230). Cambridge Cambridge University Press.

Light, P. H. (1986). Context, conservation and conversation. In M.P.M. Richards & P. H. Light (F_,ds.).ChiIdren of Social Worlds, (pp. 170.190). Camb¡ Polity Press.

C a r o l i n e G o r s u c h . C/o Department of Psychology,, The University Southampton, England, SO9 S NH.

Jul io N e w m a n . C/o Department of Psychology,, The University Southampton, England, SO9 5 NH.