Why do birds of a feather flock together? Developmental ...evansem/PolEvans.BJDP.pdf · the varied explanations given for human behaviours. Over the last 50 years informed explanations

89

British Journal of Developmental Psychology (2002) 20 89ndash1122002 The British Psychological Society

Why do birds of a feather flock togetherDevelopmental change in the use of multipleexplanations Intention teleology and essentialism

Devereaux A Poling and E Margaret EvansUniversity of Toledo USA

In two studies 6ndash12-year-old children (Study 1 N = 58 Study 2 N = 38) and adults(Study 2 N = 22) rank ordered intentional teleological and essentialist explanationsfor different behaviours of living-kind groups representing a range of biological kindsfrom plants to humans Overall humans elicited more intentional explanations insectsand plants elicited more essentialist explanations and intermediate taxa such asungulates elicited more teleological explanations Children made fewer fine-grainedtaxonomic distinctions than adults and the youngest children tended to rejectessentialism The 6ndash7-year-old children preferred to reason about living-kindbehaviours from an intentional and teleological perspective only towards the end ofthe elementary school years did children seem to incorporate a biological essentialismNeither adults nor children were exclusively bound to a particular mode ofexplanation but exercised lsquocausal flexibilityrsquo across different behavioural contexts

As the coach herds his little league team onto the bus afternoon shoppers flock to themall for a sale Meanwhile a fearless youngster hangs like a monkey from the junglegym in the playground Each of these fairly common events seems simple to imaginealthough none of them is true in a literal sense Kahn (1997) argues that the ease withwhich we are able to detect commonalities among humans and other animals reflects anatural human tendency to affiliate with nature the biophilia hypothesis Given adultsrsquoability to use their biological knowledge to draw systematic analogies between thebehaviour of humans and other animals some variant of this capacity is likely to be afeature in childrenrsquos thinking as well In two studies we focus on the development ofbiological knowledge during the elementary school years and into adulthoodSpecifically we investigate whether during this period developmental change andor stability characterize the use of three causal explanations considered to be central toan understanding of the animate world intention teleology and essentialism In

wwwbpsorguk

Requests for reprints should be addressed to Devereaux A Poling Department of Psychology Porter Hall Ohio UniversityAthens OH 45701 USA (e-mail polingd1ohiouedu)

particular we examine changes in the flexible use of these explanations in differentbehavioural contexts which we claim is a critical component of the capacity to detectcommonalities among living kinds

To date most of the research on childrenrsquos reasoning about the biological world hasfocused either on the preschool years (eg Carey 1985 1999 Wellman amp Gelman1998) or on older biology students (eg Brumby 1984 Greene 1990 Settlage 1994)and adults in traditional cultures (eg Atran 1995) We argue that it is important todescribe the transition between the intuitive ideas of preschoolers and the folk biologyof adulthood as the nature of this transition has not been well characterized Thistransition represents a critical shift as cultural- and school-based knowledge of thebiological world is systematically incorporated into childrenrsquos intuitive belief systems(Evans 2000 2001 Hatano amp Inagaki 1999) The naiuml ve beliefs of the elementary schoolchild represent both the elaboration of preschoolersrsquo constructions and the early stagesof the folk biology of the adult community Moreover characterizing the nature of thebiological understanding of this age group has important implications for schoolcurricula and other educational settings

First we briefly comment on what is known about the development of biologicalunderstanding particularly of causal explanations of the animate world during theschool-age years Then we elaborate on one feature which we call causal flexibility theflexible use of explanations in different behavioural contexts We contend that causalflexibility is not only a characteristic of adult thinking but is likely to be present in theschool-age years and perhaps earlier

Intention teleology and essentialismBy 6 or 7 years of age at least there is some agreement that most children haveconstructed a kind of intuitive lsquobiologyrsquo informed by vitalistic (Carey 1999 Inagaki ampHatano 1993) or teleological premises (Keil 1994 Kelemen 1999b) Evidencesuggests however that although an intuitive biology might be in place by the earlyschool-age years it does not fully mature until much later the reasoning of the 6- to 7-year-old therefore differs in an unspecified manner from that of older childrenrsquos andadultsrsquo biological reasoning (Carey 1996 1999 Hatano amp Inagaki 1999 Inagaki ampHatano 1993) In part this shift includes a refining of biological concepts such thatsome concepts become more differentiated as with the concept of lsquonot aliversquo shiftingto include distinctions between lsquodeadrsquo lsquoinanimatersquo lsquounrealrsquo and lsquonon-existentrsquo (Carey1999 p 298) Other concepts coalesce in particular early ontologically distinctconcepts of person animal and plant are increasingly likely to be classified under asuperordinate category living thing and deemed to possess common biologicalcapacities (Carey 1999)

In the studies reported here we investigate the latter phenomenon by assessingwhether there are age-related changes in the endorsement of different causalexplanations for the behaviours of humans species of increasing taxonomic distancefrom the human and plants Underlying the adult ability described earlier to detectcommonalities among living kinds should be an appreciation of the extent to whichgeneral causal principles can reasonably be applied to the behaviours of different livingkinds The modes of construal we examine intention teleology and essentialism are allimplicated in childrenrsquos reasoning about the world of living organisms (Wellman ampGelman 1998)

90 Devereaux A Poling and E Margaret Evans

Living organisms exhibit teleological characteristics in that they act purposefullymoreover they appear to be lsquodesignedrsquo to function that way (Allen Bekoff amp Lauder1998 Keil 1994) Vitalism with its emphasis on goal-directed behaviour (MorrisTaplin amp Gelman 2000) may be an early form of teleological reasoning (Carey 1999)or it could be the school-age childrsquos transitional mode between the preschoolerrsquosintentional reasoning and the functional or mechanistic reasoning of older children andadults (Hatano amp Inagaki 1999 Inagaki amp Hatano 1993)

Although often conflated in the literature teleological and intentional reasoningshould in principle be separable processes (Keil 1994) For example complex artifactssuch as thermostats and cars are often treated as if they are purposeful without anyonebelieving that they are also intentional Recent findings indicate that teleologicalreasoning is applied to natural inanimate entities (eg mountains rocks) as well as toanimate entities by preschool children (Kelemen 1999a) This broad application offunctional cause becomes more selective only towards the end of the elementary schoolyears (10-year-olds) being applied almost exclusively to living kinds by Western-educated adults (Kelemen 1999a 1999b) Kelemen argues that this promiscuous use ofteleological reasoning by preschool and young elementary school children may be anextension of the initially preferred intentional reasoning (1999a 1999b) For adultshowever the inference of intentionality requires additional cues such as similarity tothe prototypical intentional entity the human (Inagaki amp Sugiyama 1988) purposefulor goal-directed behaviour is necessary but not sufficient In conclusion it is not clearfrom the reported literature whether the reasoning of early elementary school childrenat least is sustained more by a type of teleological causality or by an intentionalcausality or by both It is likely however that intentional reasoning will be exhibitedmore by the 6ndash7-year-old than by the older child

Essentialism is a catch-all term encompassing a variety of explanations that have attheir core the idea that the behaviour or surface appearance of an organism isdetermined by its inner nature or essence Gelman and her colleagues havedocumented preschoolersrsquo readily invoked capacity to infer internal causes for animalmotion even when the animals were unfamiliar and carried by a person (eg GelmanS Coley amp Gottfried 1994) These sorts of studies indicate that children grasp somenotion of individual essence at least to the extent that they appeal to an internalist orlsquoinnardsrsquo principle (Gelman R Durgin amp Kaufman 1995 Massey amp Gelman 1988) orto the concept of innate potential (Gelman S amp Wellman 1991) Howeverpreschoolersrsquo essentialism differs from that of adults in that it does not incorporatemechanistic causality (Carey 1999)

Causal flexibilityThe hallmark of adult reasoning is arguably the capacity to shift preferences for aparticular explanation depending on the available evidence and the nature of thebehaviour and the target called here causal flexibility (see also Keil Levin Richman ampGutheil 1999 Keil amp Wilson 2000) Nowhere is causal flexibility more evident than inthe varied explanations given for human behaviours Over the last 50 years informedexplanations for behaviours as diverse as autism alcoholism schizophrenia and evengender differences in maths achievement (eg Eccles 1994) have ranged from largelypsychological to largely essentialist (ie genetic) to some synthesis of the two

91Why do birds of a feather

explanations These shifts can be said to represent changes in preference for theseexplanations informed by the latest research findings

Most research to date has focused on the ages at which particular types of causalexplanation are likely to be mastered even though Keil suggests that these reasoningstances are not used in an all-or-none fashion (1994 Keil et al 1999) Central to Keilrsquosthesis is the idea that foundational domains of knowledge (Wellman amp Gelman 1998)such as an intuitive biology or psychology may be initialized by intricate combinationsof abstract intuitive causal reasoning principles such as teleology intention andessentialism None of these modes of construal he argues are uniquely tied to adomain but they may be lsquofootholdsrsquo that allow humans to acquire more lsquoelaboratedbelief systems in a number of specialized domainsrsquo (Keil 1994 p 251) Teleologicalreasoning and essentialist reasoning for instance are not necessarily mutually exclusiveand can be combined creating a teleo-essentialist form of reasoning (Atran 1995) Forexample the apparent purposefulness of many animal behaviours such as commu-nication among bees is thought to be innate and dependent on a collective geneticendowment (eg bee-essence) Alternatively teleological explanations can becombined with intentional ones Children might believe that lsquobirds flock togetherrsquobecause they like each other which would be a straightforward psychological lsquodesirersquoexplanation Or they could believe that birds stay together for a purpose because theylike to be safe from danger which would be a teleo-intentional explanation

Preschool and early school-age children are potentially flexible in their thinking inthat they can discriminate between these explanatory modes (Gutheil Vera amp Keil1998 Schult amp Wellman 1997) but lacking the detailed biological and physicalknowledge of older children and adults younger children may use these explanations ina less discriminating fashion Gutheil et al (1998) argue that flexibility may be a crucialelement in preschoolersrsquo reasoning and they demonstrate that preschoolers can switchbetween biological and psychological explanatory modes though they may default tothe latter Furthermore such flexibility apparently increases with age 8-year-olds andadults are more sensitive to different contexts than the younger children (Inagaki ampHatano 1993)

In sum we contend that it may be less interesting to ask whether individuals areexclusively wedded to a particular mode of explanation at specific ages than to askwhether they exhibit causal flexibility distinct preferences for different explanationsacross different behavioural contexts We focus on two relatively neglected age-groupsin this research area the elementary school years and early adulthood

In the first study reported in this paper we asked child participants to judge whyhumans and other animals stay together in groups We concentrated on the tendency tolsquoflock togetherrsquo because it had two useful attributes that were central to our task Firstit functioned as a neutral behaviour that could conceivably elicit any of the abovereasoning modes intention teleology essentialism Secondly it was likely to focusparticipantsrsquo attention on the behaviour of the species as a whole rather than on asingle individual In the second study with adults and children we extended the rangeof behaviours to include questions about individual behaviours in particular thereasons why animals and plants breathe and why they dream In all studies the range ofbiological categories was based on taxonomic distance from the human with theexpectation that different taxons and behaviours would elicit different permutationsand combinations of these reasoning modes (eg Inagaki amp Sugiyama 1988) Overallthe contexts we investigate vary along three dimensions the kind of speciesrepresented the nature of the explanation and the kind of behaviour to be explained


Although like Gutheil et al (1998) we sample different behavioural contexts withour method participants can potentially display a greater degree of causal flexibilityGutheil et al (1998) found that preschool children demonstrated flexibility in thegeneralization of target behaviours (eat sleep etc) from humans to different speciesdepending on the condition In each condition a fixed explanatory context wasinvoked biological psychological or no context In contrast we ask participants tochoose which of three explanations best describes why species of different kindsengage in different types of behaviour

Our hypothesis was that older school-age children and adults would display causalflexibility which would be demonstrated in differing patterns of explanationpreference for different biological categories across different behaviours Specificallyteleological and intentional explanations would become increasingly disassociated as afunction of taxonomic distance of the targeted taxon from the human (Inagaki ampSugiyama 1988) Moreover older participants should apply an essentialist explanationto species as readily as young children apparently do to individuals Although childrenrsquosability to use intention teleology and essentialism to reason about animate behaviourhas been documented (see Carey 1985 Keil 1994 and Gelman S et al 1994respectively) less is known about the extent to which school-age children wouldemploy causal flexibility There should be an increase in flexible use of theseexplanations over the elementary school years as children acquire more detailedknowledge of the biological world (Carey 1999 Keil et al 1999) There may also bechanges in childrenrsquos preferences for these explanations if they shift from a reliance onan intention-based reasoning stance to a more mechanistic or functional one In the twostudies reported in this paper childrenrsquos and adultsrsquo preferences for different reasoningstances were explored We chose explanations that are considered characteristic of thethree different reasoning stances intention teleology and essentialism investigated inthese studies (Wellman amp Gelman 1998)

STUDY 1

In Study 1 school-age children 6ndash12 years of age were asked lsquoWhy do these Xrsquos staytogetherrsquo for groups of humans non-human mammals insects and plants We used anambiguous version of the essentialist explanation (see Gelman R et al 1995 GelmanS A et al 1994) lsquoThey stay together because each one of them has the same thingsinsidersquo This was designed to elicit lsquointernalistrsquo interpretations that vary according toparticipantsrsquo level of biological understanding in that it was deliberately content-neutral As such we hoped it would appeal both to older children (or adults) whomight grant lsquoessentialrsquo causal power to genetics and to 6-year-olds who might grantessential causal power to an inchoate innards principle This explanation is just one ofseveral versions of an essentialist explanation but one that seemed the most relevantgiven our concern (Gelman S amp Hirschfeld 1999)

Participants in both studies were asked to choose one explanation as the lsquobestrsquo out ofthree possible explanations The advantage of this method is that participants explicitlycompared the three explanations when considering their choice However a potentialshortcoming is that each explanation is not measured independently of the other twoTo offset this shortcoming analyses presented here focused only on the responsechosen as the lsquobestrsquo Additionally we believe this method is satisfactory given findings


from a set of preliminary studies (Poling amp Evans 1998) with children and adults1 Weuse the term lsquoexplanation preferencersquo guardedly to refer to participantsrsquo choiceswithout implying that participants would necessarily spontaneously generate thesesame explanations

Method

ParticipantsThe participants were 58 children ranging in age from 6 to 12 years grouped in threenon-overlapping age and grade ranges (1) 6ndash7-year-old group N = 20 (13 males 7females) age range = 61 years to 71 years (mean age = 67 years) grade range = Kndash1(2) 8ndash9-year-old group N = 19 (9 males 10 females) age range = 78 years to 94 years(mean age = 85 years) grade range = 2ndash3 (3) 10ndash12-year-old group N = 19 (11males 8 females) age range = 100 years to 129 years (mean age = 112 years) graderange = 4ndash7 The majority of participants were Caucasian two were African-Americanand two were multiracial

All participants in Study 1 were interviewed in their after-school or summerprogrammes The procedure lasted about 20 minutes To verify interviewer reliabilityand to check the accuracy of transcripts interviews with children were audiotapedwith the written permission of the parents and the verbal permission of the childParticipation certificates and stickers or pencils were given to each child whoparticipated in the study

Materials and procedureThe same materials and procedure used in the preliminary studies were also used in thecurrent study (Poling amp Evans 1998) One set of test cards consisting of 12 differentnaturalistic photographs taken from natural history magazines and books was usedduring the procedure The photographs depicted four biological categories humansungulates insects and plants with three exemplars from each category all groups wereportrayed in a neutral posture (standing sitting or lying down) in species-appropriateenvironments To ensure that the human exemplars were unfamiliar they were fromthree different non-western cultural groups All human groups were pictured outdoorsnear their homes where the homes ranged from concrete buildings to straw hutssurrounded by forest The nine animal cards consisted of pictures of ungulates (giraffeselephants zebras) insects (ants bees ladybirds) and plants (tulips snowdrops anddaffodils) To familiarize the participants with the procedure three practice cards wereused as a warm-up task before the test cards were presented The practice cardsdepicted large stationary land lizards a cheetah chasing a group of impalas (calleddeer) and a group of children playing in water

1 In two preliminary studies we investigated 8-year-old childrenrsquos and adultsrsquo explanation preferences for the groupingbehaviour of humans and a range of non-human animals (see Poling amp Evans 1998) The findings in those preliminary studiesare replicated in the studies reported here In addition we collected data from one adult sample using both the rankingmethod reported here and a rating method in which adults responded to each explanation independently The results fromthese two methods were similar however pilot testing with younger children indicated that they found the ranking methodeasier Therefore we choose the ranking method for the two studies reported in this paper


The cards were randomly presented to participants one at a time Participants wereasked for each picture lsquoWhy do these Xrsquos stay togetherrsquo For each card all threeexplanations were verbally presented one at a time in random order with computer-drawn visual mnemonics to represent each explanation (see Appendix) The threepossible explanations were intentional lsquoThey stay together because they like eachotherrsquo teleological lsquoThey stay together because they will be safe from dangerrsquoessentialist lsquoThey stay together because each one of them has the same things insidersquoFor each picture participants rank ordered the three explanations from the best to theworst using the visual mnemonics After participants had completed the procedure forthe three practice cards the 12 test cards were presented randomly one at a time Theprocedure was identical for practice and test pictures but only test pictures were usedin the analyses

ScoringFor each picture the explanation ranked as lsquobestrsquo was given a score of 1 The remainingtwo explanations were given a score of 0 Thus for each explanation for each picturethe possible range of explanation preference scores was 0ndash1 and for each explanationwithin each biological category the possible range of explanation preference scoreswas 0ndash3 Across all trials chance is 1 on each of three trials any one explanation couldbe chosen at chance levels one-third of the time Mean explanation scores (mean lsquobestrsquoscores) that differ significantly from chance are indicated by an asterisk on each of thefigures

Results

Overall analysisOverall the intentional explanation (like each other) was preferred for the humans theteleological (to be safe from danger) for the ungulates and insects and the essentialist(same insides) was preferred for the plants However the 6ndash7-year-old childrenpreferred the intentional explanation more and the essentialist explanation less thanthe 8ndash9- and 10ndash12-year-old groups

The explanation preference scores for the children were analysed in a 3 (age group6ndash7-year-olds 8ndash9-year-olds 10ndash12-year-olds) 6 3 (explanation intentional teleologi-cal essentialist) 6 4 (biological category human ungulate insect plant) mixed designanalysis of variance (ANOVA) with explanation and biological category as repeatedmeasures (see Fig 1 for means and standard errors) Analyses indicated a significantmain effect for explanation F(2110) = 963 p lt 0001 Additionally there was asignificant explanation 6 age group interaction F(4110) = 298 p lt 02 and abiological category 6 explanation interaction F(6330) = 4298 p lt 0001 Moreimportantly there was an explanation 6 biological category 6 age group interactionF(12330) = 206 p lt 02 To clarify the nature of the interactions analyses will bepresented first for the intentional followed by the teleological and essentialistexplanations across all biological categories and then the explanation preferencescores will be compared within each biological category


Intentional explanation (like each other)An age group (3) 6 biological category (4) ANOVAindicated significant main effects forage F(255) = 547 p lt 007 and biological category F(3165) = 2933 p lt 0001Bonferroni post hoc comparisons indicated greater preference for the intentionalexplanation by the 6ndash7-year-old group as compared to both the 8ndash9-year-old (p lt 009)and 10ndash12-year-old (p lt 005) groups In addition the intentional explanation wasapplied more often to the humans than to any other biological category by all agegroups (all ps lt 0001)

Teleological explanation (safe from danger)In a similar analysis a significant main effect for biological category was found for theteleological explanation F(3165) = 4282 p lt 0001 Bonferroni post hoc comparisons

Figure 1 Study 1 Why do they stay together Patterns of explanation preference across fourbiological categories for ages 6 years to 12 years (Chance= 1 plt05)


indicated that all age groups preferred the teleological explanation more for ungulatesthan for any other category (all ps lt 008) In addition the teleological explanation waspreferred more for insects than for humans or plants (all ps lt 0001)

Essentialist explanation (same insides)In a similar analysis a significant main effect for biological category was found for theessentialist explanation F(3165) = 5424 p lt 0001 An age group 6 biologicalcategory interaction was also found F(6165) = 267 p lt 02 Means indicate that theessentialist explanation was preferred more for plants than for any other categoryBonferroni post hoc comparisons indicated that the 10ndash12-year-old group preferred theessentialist explanation more for the plants than did the 6ndash7-year-old group (p lt 007)

Human categoryAn age group (3) 6 explanation (3) ANOVA for the human category revealed asignificant main effect for explanation F(2110) = 2067 p lt 0001 Bonferroni posthoc comparisons indicated that the intentional explanation was used more for thehumans than was the teleological explanation (p lt 01) and the essentialist explanation(p lt 0001) collapsed across all age groups In addition the teleological explanationwas used more for the humans than the essentialist explanation (p lt 01) From Fig 1it can be seen that essentialism was rejected (significantly less than chance) by all agegroups whereas teleology was at chance As in the second preliminary study the 8ndash9-year-olds did not appear to distinguish between the teleological and intentionalexplanations for the human both were at chance

Ungulate categoryIn a similar analysis a significant main effect for explanation was found for the ungulatecategory F(2110) = 4916 p lt 0001 Bonferroni post hoc comparisons indicated thatthe teleological explanation was preferred more for the ungulates than the intentionalexplanation (p lt 0001) and the essentialist explanation (p lt 0001) across all agegroups

Insect categoryIn a similar analysis a significant main effect for explanation F(2110) = 2004p lt 0001 and an explanation 6 age group interaction F(4110) = 26 p lt 05 werefound for the insect category Bonferroni post hoc comparisons indicated that theteleological explanation was preferred more for the insects than the intentionalexplanation (ps lt 0004) and the essentialist explanation (ps lt 002) by the 10ndash12-year-old and the 7ndash9-year-old age groups The 6ndash7-year-olds did not show a preference forany of the explanations for the insect category they choose intention and teleology atchance levels and rejected essentialism

Plant categoryIn a similar analysis a significant main effect for explanation F(2110) = 225p lt 0001 and an explanation 6 age group interaction F(4 110) = 33 p lt 02 werefound for the plant category Bonferroni post hoc comparisons indicated that theessentialist explanation was preferred more for the plants than the intentionalexplanation (ps lt 002) and the teleological explanation (ps lt 006) by the 10ndash12-year-


olds and the 8ndash9-year-old age groups The 6ndash7-year-old children however did not showa preference for any of the explanations for the plant category

DiscussionSeveral age-related differences in elementary school childrenrsquos preference patterns areseen in this study As predicted the 10ndash12-year-old children demonstrated an overallpreference pattern that included fine-tuned discriminations between the explanationstargeting different taxonomic groups For the 10ndash12-year-old children in this study theessentialist explanation was favoured for the taxon most distant from the human whichwas plants and the intentional explanation was favoured most for the humansMoreover for both the ungulates and the insects the 10ndash12-year-old children preferreda teleological explanation

The 8ndash9-year-olds employed a range of explanations similar to those of the older agegroup although their discriminations were less fine-tuned (see Fig 1) However for theplant category both the 8ndash9-year-old and the 10ndash12-year-old groups endorsedessentialism at above chance levels and this preference was greater for plants thanfor any other biological category For the 6ndash7-year-old age group there was nosignificant preference for any of the explanations for the plant category Arguably thepatterns of preference for the 6ndash7-year-old children in this study reflect someuncertainty about plants at this age (see also Hickling amp Gelman 1995)

In addition the 6ndash7-year-old children in this study showed a greater overallpreference for the intentional explanation than did the older two groups One possiblereason for this endorsement is a preference for the more appealing visual mnemonicrepresenting the intentional explanation (see Appendix) However this is unlikely as anexplanation as neither insects or plants elicited a preference for intention in this agegroup These results suggest an overall shift in childrenrsquos preferences for intention andessentialism from the early to late elementary school years

In summary these analyses indicate that in general intention is preferred overteleology for the human category and that in comparison with the two olderelementary school-age groups the 6ndash7-year-olds preferred intention at higher levelsacross all biological categories The teleological explanation was preferred more for theungulates than any other biological category by all age groups Finally the essentialistexplanation was preferred more for the plants than for any other biological category bythe two older age-groups

STUDY 2

In Study 1 we investigated childrenrsquos use of three explanations for a neutral groupingbehaviour which could plausibly be construed as psychologically or biologicallycaused In Study 2 we examined the use of the same three explanations for this neutralgrouping behaviour as well as two additional individual behaviours breathing anddreaming Breathing was judged to be a prototypical biological behaviour and dreaminga prototypical psychological behaviour (Flavell Green Flavell amp Groomsman 1997Miller amp Bartsch 1997) Whereas it may be questionable to ask children and adultsabout dreaming (or any psychological behaviour) for the plant category we feel it addsan interesting dimension to the study That is asking children and adults why plants


dream should force them to choose whichever explanation they would normally defaultto under ambiguous circumstances If there is no default explanation we can expectresponses to the dream question for the plant category to be at chance

Secondly we altered the wording of the essentialist explanation In Study 1 theyoungest children (6ndash7 years old) seemed reluctant to choose the essentialistexplanation even for plants This hesitancy could be due to young school-agechildrenrsquos difficulty in grasping the explanation In Study 1 the essentialist explanationrequired participants to agree that all members in the pictured group had the samekinds of things inside which made them all the same kind of thing which in turncaused them to stay together In this study we simplified the essentialist explanation tomake it less cognitively taxing for the youngest children Finally with this study weinvestigated whether results from children would replicate findings from the first studyand we included adults as a developmental end-point As we were most interested incomparing the youngest and oldest elementary school ages with adults we did notinclude an 8ndash9-year-old group in this study

Method

ParticipantsThe participants were 38 children grouped in two non-overlapping age and graderanges 6- to 7-year-old group N = 19 (6 males 13 females) age range = 61 years to71 years (mean age = 62 years) grade range = Kndash1 10- to 12-year-old group N = 19(7 males 12 females) age range = 100 years to 129 years (mean age = 108 years)grade range = 4ndash7 Twenty-two adults from the university and the community alsoparticipated The majority of the adult and child participants were Caucasian one wasAfrican-American four were Hispanic and one was Asian-American

Written consent was obtained from all adult participants and from parents of thechild participants Verbal assent was also given by each child Participants wereinterviewed in their homes or in the university laboratory with the procedure lastingabout 20 minutes To verify interviewer reliability and to check the accuracy of thetranscripts interviews with the children were audiotaped Small tokens were given toeach child who participated in the study

Materials and procedureThe same set of naturalistic photographs as used in Study 1 was used in Study 2 withthe exclusion of the insect category and the set of practice cards used in Study 1 Thesecards were excluded to reduce the length of the procedure since two additionalquestions were added in this study Thus we used nine photographs in this studyrepresenting three biological categories (humans ungulates plants)

The set of nine cards was presented twice with all cards being presented in randomorder one at a time for each presentation The first time the set was presentedparticipants were asked for each picture lsquoWhy do all these Xs stay togetherrsquo As inStudy 1 participants rank ordered the three explanations from the best to the worstusing the same visual mnemonics for each picture The intentional and teleologicalexplanations were the same as in Study 1 intentional lsquoThey stay together because theylike each otherrsquo teleological lsquoThey stay together because they will be safe from


dangerrsquo The essentialist explanation differed lsquoThey stay together because of what theyall have insidersquo

The participants were then presented with the same nine picture cards a secondtime In this presentation participants were asked two questions for each picture (1)lsquoWhy do these Xrsquos breathersquo and (2) lsquoWhy do these Xrsquos dreamrsquo The question order wasalternated so that the breathe question was asked first for one picture and the dreamquestion was asked first for the next picture and so on Again the participants rankordered the explanations from best to worst The explanations were as followsintentional lsquoThey breathedream because they like to breathedreamrsquo teleologicallsquoThey breathe because their bodies need air to stay aliveThey dream because they needto keep their minds activersquo essentialist lsquoThey breathedream because of what they allhave insidersquo

ScoringScoring was done in the same way as in Study 1 Chance was 1

Results

Overall analysisThe explanation preference scores for the children were analysed in a 3 (age group6ndash7-year-olds 10ndash12-year-olds adults) 6 3 (explanation intentional teleologicalessentialist) 6 3 (biological category human ungulate plant) 6 3 (question staytogether breathe dream) mixed design ANOVA with explanation biological categoryand question as repeated measures (see Figs 2 3 and 4 for means and standard errors)Analyses indicated a significant main effect for explanation F(2114) = 4451p lt 0001 Additionally there was a significant explanation 6 age group interactionF(4114) = 962 p lt 0001 a question 6 explanation interaction F(4228) = 5075p lt 0001 and an explanation 6 biological category interaction F(4228) = 3473p lt 0001 Finally there was a question 6 explanation 6 age group interactionF(8228) = 872 p lt 0001 and a question 6 explanation 6 biological categoryinteraction F(8456) = 2375 p lt 0001 To clarify the nature of the interactions andto permit direct comparisons with the results from Study 1 analyses will be presentedseparately for each question for the intentional teleological and essentialistexplanations in turn

Stay together questionAs was found in Study 1 the intentional explanation was preferred for humans theteleological explanation was preferred for ungulates and the essentialist explanationwas preferred for plants In addition the 6ndash7-year-old children preferred the intentionalexplanation more and the essentialist explanation less than did the 10ndash12-year-oldchildren and adults

The explanation preference scores were analysed in a 3 (age group young oldadult) 6 3 (explanation intentional teleological essentialist) 6 3 (biological categoryhuman ungulate plant) mixed design ANOVA with explanation and biologicalcategory as repeated measures (see Fig 2 for means and standard errors) Analysesindicated a significant main effect for explanation F(2114) = 1874 p lt 0001


Additionally there was a significant explanation 6 age group interactionF(4114) = 59 p lt 0002 and a biological category 6 explanation interactionF(4228) = 4721 p lt 0001 More importantly there was an explanation 6 biologicalcategory 6 age group interaction F(8228) = 201 p lt 05

Intentional explanation (like)An age group (3) 6 biological category (3) ANOVA on the scores for the intentionalexplanation indicated significant main effects for age F(257) = 1133 p lt 0001 andfor biological category F(2114) = 2107 p lt 0001 Bonferroni post hoc comparisonsindicated greater preference for the intentional explanation by the 6ndash7-year-old children

Figure 2 Study 2 Why do they stay together Patterns of explanation preference across threebiological categories for ages 6 years to adult (Chance= 1 plt05)


as compared with both the 10ndash12-year-old children (p lt 01) and adults (p lt 0001) Inaddition the intentional explanation was preferred more for humans than for any otherbiological category by all age groups (all ps lt 0001)

Teleological explanation (safe)A similar ANOVA on the scores for the teleological explanation indicated a significantmain effect for biological category F(2114) = 4855 p lt 0001 Bonferroni post hoccomparisons collapsed across age groups indicated that the teleological explanationwas preferred more for the ungulates than for any other category (all ps lt 0001) Inaddition the teleological explanation was preferred more for the humans than for theplants (p lt 0001)

Essentialist explanation (inside)A similar ANOVA on the scores for the essentialist explanation indicated significantmain effe cts for age F(257) = 981 p lt 0002 and biological cate goryF(2114) = 7193 p lt 0001 An age group 6 biological category interaction was alsofound F(4114) = 295 p lt 03 Bonferroni post hoc comparisons indicated that theadults preferred the essentialist explanation more for the plants than did the 6ndash7-year-old group (p lt 002) as well the 10ndash12-year-old children preferred the essentialistexplanation more for plants than did the youngest children (ps lt 004)

As in Study 1 age group (3) 6 explanation (3) ANOVAs were conducted for the staytogether question for each biological category in turn

Human categoryThis analysis for the human category revealed a significant main effect for explanationF(2114) = 2009 p lt 0001 Bonferroni post hoc comparisons indicated that theintentional explanation was used more for the humans than was the teleologicalexplanation (p lt 01) and the teleological explanation was used more for the humansthan the essentialist explanation (p lt 01) collapsed across age groups

Ungulate categoryIn a similar analysis a significant main effect for explanation was found for the ungulatecategory F(2114) = 8148 p lt 0001 and an explanation 6 age group interaction wasalso found F(2114) = 290 p lt 05 Bonferroni post hoc comparisons indicated that theteleological explanation was preferred more for the ungulates than the intentionalexplanation (ps lt 0001) and the essentialist explanation (ps lt 0001) for all agegroups though the effect was greatest for the adults

Plant categoryIn a similar analysis a significant main effect for explanation F(2114) = 1870p lt 0001 and an explanation 6 age group interaction F(4114) = 524 p lt 005were found for the plant category Bonferroni post hoc comparisons indicated that theessentialist explanation was preferred more for the plants than was the intentionalexplanation (ps lt 002) and the teleological explanation (ps lt 006) by the 10ndash12-year-olds and the adults The 6ndash7-year-old children did not show a preference for any of theexplanations for the plant category


Breathe questionOverall for the breathe question all three groups showed a strong preference for theteleological explanation (see Fig 3) for all biological categories The explanationpreference scores were analysed in a 3 (age group 6ndash7-year-old 10ndash12-year-old adult)6 3 (explanation intentional teleological essentialist) 6 3 (biological categoryhuman ungulate plant) mixed design ANOVA with explanation and biologicalcategory as repeated measures (see Fig 3 for means and standard errors) Analysesindicated a significant main effect for explanation F(2114) = 15754 p lt 0001Additionally there was a significant explanation 6 age group interactionF(4114) = 682 p lt 0001 and a biological category 6 explanation interactionF(4228) = 383 p lt 005

Figure 3 Study 2 Why do they breathe Patterns of explanation preference across three biologicalcategories for ages 6 years to adult (Chance= 1 plt05)


Intentional explanation (like)In an age group (3) 6 biological category (3) ANOVA a significant main effect for agewas found F(257) = 1156 p lt 0001 Bonferroni post hoc comparisons indicatedthat the 6ndash7-year-old children preferred the intentional explanation more than did the10ndash12-year-old children (p lt 0001) and adults (p lt 0001) However even the 6ndash7-year-olds chose intention at significantly less than chance levels (see Fig 3)

Teleological explanation (need)Significant main effects for age F(257) = 841 p lt 0006 and for biological categoryF(2114) = 659 p lt 002 were found Bonferroni post hoc comparisons indicated thatthe 10ndash12-year-old children and adults used the teleological explanation more than the6ndash7-year-old children did (ps lt 0001) In addition the teleological explanation waspreferred more for the human (p lt 006) and ungulate (p lt 002) when compared toplants

Essentialist explanation (inside)A significant main effect was found for age F(257) = 430 p lt 02 with Bonferronipost hoc comparisons indicating that the 6ndash7-year-old children chose the essentialistexplanation more than the 10ndash12-year-old children (p lt 0001) or adults (p lt 0006)However this finding should be qualified by the fact that the 10ndash12-year-old childrenand adults rejected the essentialist explanation and the 6ndash7-year-old grouprsquos preferencescore was not significantly different from chance Thus none of the age groupsdemonstrated a preference for essentialism

Dream questionOverall for all biological categories the 6ndash7-year-old children preferred the intentionalexplanation the 10ndash12-year-old children did not demonstrate any significant explana-tion preferences and the adults preferred the essentialist explanation

The explanation preference scores were analysed in a 3 (age group 6ndash7-year-old10ndash12-year-old adult) 6 3 (explanation intentional teleological essentialist) 6 3(biological category human ungulate plant) mixed design ANOVA with explanationand biological category as repeated measures (see Fig 4 for means and standard errors)Analyses indicated a significant explanation 6 age group interaction F(4114) = 1163p lt 0001 which is investigated in the following analyses

Intentional explanation (like)In an age group (3) 6 biological category (3) ANOVA a significant main effect for agewas found F(257) = 1431 p lt 0001 Bonferroni post hoc comparisons indicated agreater preference for the intentional explanation by the 6ndash7-year-old children whencompared with the 10ndash12-year-old children (p lt 001) and adults (p lt 0001) Inaddition the 10ndash12-year-olds preferred intention more than the adults did (p lt 0003)

Teleological explanation (need)There were no significant differences by age in the use of the teleological explanationfor the dream question


Essentialist explanation (inside)Asignificant main effect was found for age F(257) = 1950 p lt 0001 Bonferroni posthoc comparisons indicated that the adults preferred the essentialist explanation morethan the 10ndash12-year-old children (p lt 0001) and 6ndash7-year-old children (p lt 0001) didIn addition the 10ndash12-year-old children preferred the essentialist explanation morethan the 6ndash7-year-olds did (p lt 0001)

Figure 4 Study 2 Why do they dream Patterns of explanation preference across three biologicalcategories for ages 6 years to adult (Chance= 1 plt05)


DiscussionIn Study 2 children and adults were presented with different contexts in which theycould evaluate the behaviours of different biological groups These contexts included aneutral grouping behaviour (stay together) and two individual behaviours a biologicalbehaviour (breathe) and a psychological behaviour (dream) Results from this studyrevealed several differences related to age of participant type of behaviour andbiological category

Findings from this study replicate and extend those of Study 1 For the stay togetherquestion the teleological explanation (safe from danger) was preferred for the ungulatesacross all age groups In addition the intentional explanation was used more by the6ndash7-year-olds than the other age groups for all three of the behaviours used in this study(stay together breathe dream) Moreover the essentialist explanation (same inside)was preferred for plants more than for any other biological category by the adults andthe 10ndash12-year-olds However as in Study 1 the youngest children used essentialismless than participants in the other age groups in fact their use of essentialism did notexceed chance levels

For the prototypical biological behaviour breathing all age groups preferred theteleological explanation for all biological categories although in comparison with theother age groups the 6ndash7-year-old children used teleology less These results supportother findings that by early elementary school children have a clear understanding ofprototypical biological behaviours (see Wellman amp Gelman 1998 for a review)

On the other hand for the prototypical psychological behaviour dreaming therewas a significant age-related interaction in the overall pattern of explanation preferencewith the adults showing a strong preference for the essentialist explanation across allbiological categories and the 6ndash7-year-old children relying on intention For the 10ndash12-year-olds every explanation was at chance for all biological categories indicating thatthey did not find any explanation preferable However 10ndash12-year-olds were less likelythan adults to endorse intention and more likely than 6ndash7-year-olds to endorseessentialism This interesting pattern of results suggests that the 10ndash12-year-oldsrsquopreference pattern represents a transition from an earlier preference for intention to amore mature preference for essentialism

Given the ambiguity of the dream question especially for the plants we argue that6ndash7-year-olds were using intention as a default whereas adults used essentialism as adefault finding no other explanation acceptable (Gelman S et al 1994) Thetransitional age group the 10ndash12-year-olds though did not appear to have any defaultexplanation easily available These findings are consistent with the interpretation thatthere is a qualitative shift in the use of essentialist explanations over the elementaryschool years and into adulthood Overall the results from Study 2 provide furtherevidence that children and adults do exhibit flexibility in their use of explanationsacross different biological categories and behaviours

GENERAL DISCUSSION

A strength of the current studies is the targeting of taxa that differ in degree from thehuman which has revealed that even within the world of living things neither adultsnor children are wedded to any one mode of explanation (see also Gutheil et al 1998Schult amp Wellman 1997) However adults and older elementary school children


(10ndash12 years) were much more likely than younger children (6ndash7 years) to make fine-grained discriminations between biological kinds when applying these explanationsWhen placed in the broader context of related studies the current study lends supportto the position that reasoning about something as complex as the behaviour of animateentities requires the coordination of multiple cues (Keil amp Wilson 2000 Gelman R etal 1995) and no single explanation is likely to suffice

The pattern of explanation preferences shown in these studies indicates that bothstability and change characterize the development of a naiuml ve biology during theelementary school years Stable features include the capacity to demonstrate flexibilityadults and elementary school children alike endorse different explanations dependingon the behavioural context However change is apparent also as demonstrated by theoverall decrease in preference for intentional explanations and the increase inpreference for essentialist ones from the early elementary school years into adulthoodBefore considering how these results relate to previous conclusions about the nature ofa naiuml ve biology we examine in more detail the roles of causal flexibility andessentialism in the emergence of an intuitive biology

Causal flexibilityThese results support those of earlier studies and indicate that children approach theworld of living things with a multiplicity of explanatory systems each characterized bygeneral causal principles (Gutheil et al 1998 Keil et al 1999 Schult amp Wellman1997) Intentionality seems to be one of the most salient of these systems at early ages(Gutheil et al 1998) However this is not to suggest that the teleological or essentialistmodes were inconsequential Adults and the 10ndash12-year-old children were most likelyto favour essentialist explanations for the insects or plants Yet teleology emerged asthe favoured explanation for ungulates especially by the 10ndash12-year-olds and the adultsEven the 6ndash7-year-olds overwhelmingly preferred teleology for prototypical biologicalentities and behaviours (see Carey 1985) the non-human mammals (ungulates in thesestudies) and breathing When 6ndash7-year-olds endorsed intention it was most likely to befor the prototypical psychological entity the human or for behaviours for which noother explanation might have seemed appropriate such as dreaming (see also Flavell etal 1997 Miller amp Bartsch 1997)

These findings provide more support for the central thesis of this paper that causalflexibility is a hallmark of human reasoning capacities for both adults and children Thiscapacity is as necessary we claim for the expert as it is for novices as explanations forphenomena must be modified transformed or interconnected (Keil amp Wilson 2000)as new causal mechanisms and concepts come to light

Reworking essentialismWhy was the essentialist explanation favoured less often by children than by adults inthese studies As described earlier researchers have demonstrated that some types ofessentialist explanations are used by preschool children (aged 3ndash5 years) to explain thebehaviour of individuals (eg Gelman S et al 1994 Gelman R 1990) However inour studies it was not until 8 or 9 years of age that children showed a preference for theessentialist explanation and even then they reserved it mainly for plants Most studiesthough do not pit essentialism against the other stances Gelman S and Hirschfeld


(1999) have suggested that essentialist explanations are invoked in the absence ofcompelling alternatives

Nonetheless there are several potentially interesting explanations of these findingsall of which imply a possible shift in an understanding of essentialism from theelementary school years to young adulthood The lack of preference for essentialistreasoning among children in the present studies does not appear to be a result of theirfailure to understand the importance of lsquoinsidesrsquo (see Gelman S amp Wellman 1991Massey amp Gelman 1988) but rather to the way the children conceptualized this insightThe essentialist explanation lsquoall have the same things insidersquo could elicit multipleinterpretations as it was deliberately content-neutral Despite the visual mnemonicrepresenting essentialism some children in the first study initially denied theessentialist explanation with individualizing comments such as lsquoThey donrsquot all havethe same things insidersquo lsquothey eat different thingsrsquo lsquosome are boys and some are girlsrsquo Achange in the way the essentialist explanation was worded did not produce anyappreciable difference in the results in Study 2 Adults in contrast appeared to map theessentialist explanation onto a genetic or quasi-genetic mechanism

Carey (1996) proposed that lsquoessentialism like taxonomic structure derives from thelogical work nouns dorsquo (p 193) and argued that initially essentialism is driven by thedefault assumption that the identity of every entity picked out by a given noun isunchanged in the face of surface changes One possibility therefore is that the shift inschool-age childrenrsquos understanding of essentialism may be characterized as amovement from a linguistic essentialism at the level of individual members of acategory to a biological essentialism at the level of the group or species Our findingssuggest that young children have very little sense of a species essence which at leastfor the working naturalist is the basic biological unit (Mayr 1982)

An alternative proposal though is that 6ndash7-year-old children may not believe that thelsquogrouping behaviourrsquo of a living kind has any relationship to its internal structureNevertheless as adults did acknowledge this relationship (as did older children in thecase of plants) the implication is that younger childrenrsquos interpretation of a speciesrsquoessence is different from that of adults That which is granted lsquoessentialrsquo causal powerfor the expert or lay adult (probably genetic inheritance or an internal biologicalprocess) may differ qualitatively in its consequences from the relatively inchoateinnards principle attributed to 6ndash7-year-old children Moreover it seems likely that thenaturalistrsquos concept of a species essence which refers to a visible morpho-behaviouralbreeding unit (Mayr 1982) as opposed to a category essence may be motivated by adifferent understanding of essentialism (Evans 2001) More detailed investigations ofpreschool and school-age childrenrsquos and adultsrsquo beliefs about category and speciesessence are needed to disentangle this issue

Implications for the nature of a naotildeve biologyOverall these results provide evidence for a shift from a preference for teleo-intentionalcauses to a preference for teleo-essentialist causes for the behaviours of diverse livingkinds from the early school-age years into adulthood These data support and extendprevious proposals that 6ndash7-year-old school-age children reason about the world ofliving things from an intentional (Carey 1985 1996) or teleological stance (Hatano ampInagaki 1994 Keil 1994) while older school-age children and adults reason moremechanistically (Carey 1999 Inagaki amp Hatano 1993 Keil et al 1999) In particularhowever we believe these changes represent shifts in lsquodefault biasesrsquo (Gutheil at al


1998) These findings suggest moreover that over the elementary school years and intoearly adulthood the teleological and intentional explanations of young childrenrsquosintuitive theories become increasingly differentiated with a gradual incorporation of abiological essentialism

These findings raise questions about the nature of a folk or naiuml ve biology Carey(1985 1999) considered one of the characteristics of an autonomous biology to be thegrouping of different species from humans to plants under the rubric of a living kindClearly in some cases humans other animals and plants were treated identically asliving kinds with a need to breathe by the youngest children in our study 6ndash7-year-oldsand the adults Yet in different contexts adults and children differentiated betweenspecies and behaviours with 6ndash7-year-old children more likely to favour intentionalexplanations and adults more likely to favour essentialist explanations thoughimportantly both endorsed teleological explanations Is an autonomous biologyachieved when individuals demonstrate an understanding that prototypical biologicalbehaviours such as breathing do not appear to be subject to intentional control andserve a similar function for all living kinds If that is the case then a majority of 6ndash7-year-old elementary children can reasonably be labelled as biological thinkers On the otherhand this label may only be justified for individuals who reason flexibly about a rangeof different behaviours and species applying non-intentional functional and (internal)mechanistic biological causes only where appropriate In that case it is probably onlythose with biological expertise that qualify

Several investigators have expressed scepticism about the scope of the biologicalknowledge of the average lay adult in modern industrial societies such as the USA orJapan Evans (2000 2001) and Hatano and Inagaki (1999) provide evidence that adultslike children often default to a form of psychological reasoning when explainingbiological phenomena Atran (1999) demonstrates that the biological reasoning ofmembers of the Itzaj Maya forest culture is in many respects superior to that ofuniversity-educated US students while being comparable to that of US parksmaintenance workers Young elementary school childrenrsquos naiuml ve biology we contendmay not differ appreciably from that of the ordinary non-expert adult

In conclusion we maintain that neither school-age children nor lay adultsdemonstrate a sophisticated understanding of the biological world yet in both casesthey access broader explanatory principles involving intention function and internalmechanisms that become more specific and concrete as they learn more about nature(see also Gutheil et al 1998) Such knowledge is typically acquired (imperfectly) inbiology classes for those inhabitants of modern industrialized societies who havebecome divorced from the natural environment For inhabitants of indigenouscommunities this knowledge is embedded in their understanding of the local ecology(Atran 1999 Ross Medin Coley amp Atran 2001)

We believe that the these results support a position endorsed by Keil and hiscolleagues (1994 Keil et al 1999 Keil amp Wilson 2000) that foundational domains(Wellman amp Gelman 1998) are initially characterized by a distinctive mix of abstractintuitive causal principles such as the intentional teleological and essentialist stancesnot by concrete mechanisms Nevertheless these causal intuitions might well enablechildren (and scientists) to lsquoguess rightrsquo (Keil amp Wilson 2000) most of the time whenthey are trying to figure out how or why something happens in the absence of specificknowledge of the phenomenon in question


AcknowledgementsThis research is based in part on a thesis submitted by Devereaux Poling to the University ofToledo in partial fulfilment of the requirements for the MAdegree the second author supervisedthe thesis The authors are grateful to the children and parents from the Northwest Ohio area andYMCA after-school and summer programmes whose cooperation made these studies possibleThe authors also wish to thank Henry Buchtel Peg Smith and two anonymous reviewers for theircomments on earlier versions of this manuscript Portions of these data were presented at theAnnual Symposium of the Jean Piaget Society in Chicago IL (June 1998) and Montreal Canada(June 2000) and the Biennial Meeting of the Society for Research in Child DevelopmentAlbuquerque (April 1999)

References

Allen C Bekoff M amp Lauder G (1998) Introduction In C Allen M Bekoff amp G Lauder (Eds)Naturersquos purposes Analyses of function and design in biology (pp 1ndash27) Cambridge MAMIT Press

Atran S (1995) Causal constraints on categories and categorical constraints on biologicalreasoning across cultures In D Sperber D Premack amp A J Premack (Eds) Causalcognition A multidisciplinary debate (pp 205ndash233) Oxford Oxford University Press

Atran S (1999) Itzaj Maya folkbiological taxonomy Cognitive universals and cultural particularsIn D L Medin amp S Atran (Eds) Folkbiology (pp 119ndash205) Cambridge MA MIT Press

Brumby M N (1984) Misconceptions about the concept of natural selection by medical biologystudents Science Education 68 493ndash503

Carey S (1985) Conceptual change in childhood Cambridge MA MIT PressCarey S (1996) Cognitive domains as modes of thought In D R Olson and N Torrance (Eds)

Modes of thought Explorations in cognition and culture (pp 187ndash215) CambridgeCambridge University Press

Carey S (1999) Sources of conceptual change In E K Scholnick K Nelson S A Gelman ampP H Miller (Eds) Conceptual development Piagetrsquos legacy (pp 293ndash326) Mahwah NJErlbaum

Eccles J S (1994) Understanding womenrsquos educational and occupational choices Psychology ofWomen Quarterly 18 585ndash609

Evans E M (2000) The emergence of beliefs about the origins of species in school-age childrenMerrill-Palmer Quarterly 46(2) 221ndash254

Evans E M (2001) Cognitive and contextual factors in the emergence of diverse belief systemsCreation versus evolution Cognitive Psychology 42 217ndash266

Flavell J H Green F L Flavell E R amp Grossman J B (1997) The development of childrenrsquosknowledge about inner speech Child Development 68 39ndash47

Gelman R (1990) First principles organize attention to and learning about relevant data Numberand the animate-inanimate distinction as examples Cognitive Science 14 79ndash106

Gelman R Durgin F amp Kaufman L (1995) Distinguishing between animates and inanimatesNot by motion alone In D Sperber D Premack amp A Premack (Eds) Causal cognition Amultidisciplinary debate (pp 150ndash184) Oxford Clarendon Press

Gelman S A Coley J D amp Gottfried G M (1994) Essentialist beliefs in children Theacquisition of concepts and theories In L A Hirschfeld amp S A Gelman (Eds) Mapping themind Domain specificity in cognition and culture (pp 341ndash365) New York CambridgeUniversity Press

Gelman S A amp Hirschfeld L A (1999) How biological is essentialism In D L Medin and SAtran (Eds) Folkbiology (pp 403ndash446) Cambridge MA MIT Press

Gelman S A amp Wellman H M (1991) Insides and essences Early understanding of the non-obvious Cognition 38 213ndash244


Greene E D (1990) The logic of university studentsrsquo misunderstanding of natural selectionJournal of Research in Science Education 27 875ndash885

Gutheil G Vera A amp Keil F C (1998) Do houseflies think Patterns of induction and biologicalbeliefs in development Cognition 66 33ndash49

Hatano G amp Inagaki K (1994) Young childrenrsquos naive theory of biology Cognition 50 171ndash188

Hatano G amp Inagaki K (1999) Adevelopmental perspective on informal biology In D L Medinand S Atran (Eds) Folkbiology (pp 321ndash354) Cambridge MA MIT Press

Hickling A K amp Gelman S A (1995) How does your garden grow Early conceptualization ofseeds and their place in the plant growth cycle Child Development 66 856ndash876

Inagaki K amp Hatano G (1993) Young childrenrsquos understanding of the mind-body distinctionChild Development 64 1534ndash1549

Inagaki K amp Sugiyama K (1988) Attributing human characteristics Developmental changes inover and underattribution Cognitive Development 3 55ndash70

Kahn P H (1997) Developmental psychology and the biophilia hypothesis Childrenrsquos affiliationwith nature Developmental Review 17 1ndash61

Keil F C (1994) The birth and nurturance of concepts by domains The origins of concepts ofliving things In L A Hirschfeld amp S A Gelman (Eds) Mapping the mind Domain specificityin cognition and culture (pp 234ndash254) New York Cambridge University Press

Keil F C Levin D T Richman B A amp Gutheil G (1999) Mechanism and explanation in thedevelopment of biological thought The case of disease In D L Medin amp S Atran (Eds)Folkbiology (pp 285ndash321) Cambridge MA MIT Press

Keil F C amp Wilson R A (2000) Explaining explanation In F C Keil amp R A Wilson (Eds)Explanation and cognition (pp 1ndash19) Cambridge MA MIT Press

Kelemen D (1999a) The scope of teleological thinking in preschool children Cognition 70241ndash272

Kelemen D (1999b) Why are rocks pointy Childrenrsquos preference for teleological explanationsof the natural world Developmental Psychology 35 1440ndash1452

Massey C amp Gelman R (1988) Preschoolers decide whether pictured unfamiliar objects canmove by themselves Developmental Psychology 24 307ndash317

Mayr E (1982) The growth of biological thought Diversity evolution and inheritanceCambridge MA Harvard University Press

Miller J L amp Bartsch K (1997) The development of biological explanation Are childrenvitalists Developmental Psychology 33 156ndash164

Morris S C Taplin J E amp Gelman S A (2000) Vitalism in naiuml ve biological thinkingDevelopmental Psychology 36 582ndash595

Poling D A amp Evans E M (1998) Childrenrsquos and adultsrsquo explanations for the behavior ofhumans and other animal groups Jean Piaget Symposium on Language and LiteracyChicago June 1998

Ross N Medin D Coley J amp Atran S (2001) Cultural and experiential differences in thedevelopment of folk biological induction Manuscript in preparation

Schult C A amp Wellman H M (1997) Explaining human movements and actions Childrenrsquosunderstanding of the limits of psychological explanation Cognition 62 291ndash324

Settlage J Jr (1994) Conceptions of natural selection A snapshot of the sense-making processJournal of Research in Science Teaching 31 449ndash457

Wellman H M amp Gelman S A (1998) Knowledge acquisition in foundational domains In DKuhn amp R S Siegler (Volume Eds) amp W Damon (Editor-in-Chief) Handbook of childpsychology Vol 2 Cognition perception and language (pp 532ndash574) New York Wiley

Received 28 May 1999 revised version received 30 May 2001


Appendix Visual mnemonics for intentional (like) teleological (safeneed) and essentialist (inside) explanations


particular we examine changes in the flexible use of these explanations in differentbehavioural contexts which we claim is a critical component of the capacity to detectcommonalities among living kinds

To date most of the research on childrenrsquos reasoning about the biological world hasfocused either on the preschool years (eg Carey 1985 1999 Wellman amp Gelman1998) or on older biology students (eg Brumby 1984 Greene 1990 Settlage 1994)and adults in traditional cultures (eg Atran 1995) We argue that it is important todescribe the transition between the intuitive ideas of preschoolers and the folk biologyof adulthood as the nature of this transition has not been well characterized Thistransition represents a critical shift as cultural- and school-based knowledge of thebiological world is systematically incorporated into childrenrsquos intuitive belief systems(Evans 2000 2001 Hatano amp Inagaki 1999) The naiuml ve beliefs of the elementary schoolchild represent both the elaboration of preschoolersrsquo constructions and the early stagesof the folk biology of the adult community Moreover characterizing the nature of thebiological understanding of this age group has important implications for schoolcurricula and other educational settings

First we briefly comment on what is known about the development of biologicalunderstanding particularly of causal explanations of the animate world during theschool-age years Then we elaborate on one feature which we call causal flexibility theflexible use of explanations in different behavioural contexts We contend that causalflexibility is not only a characteristic of adult thinking but is likely to be present in theschool-age years and perhaps earlier

Intention teleology and essentialismBy 6 or 7 years of age at least there is some agreement that most children haveconstructed a kind of intuitive lsquobiologyrsquo informed by vitalistic (Carey 1999 Inagaki ampHatano 1993) or teleological premises (Keil 1994 Kelemen 1999b) Evidencesuggests however that although an intuitive biology might be in place by the earlyschool-age years it does not fully mature until much later the reasoning of the 6- to 7-year-old therefore differs in an unspecified manner from that of older childrenrsquos andadultsrsquo biological reasoning (Carey 1996 1999 Hatano amp Inagaki 1999 Inagaki ampHatano 1993) In part this shift includes a refining of biological concepts such thatsome concepts become more differentiated as with the concept of lsquonot aliversquo shiftingto include distinctions between lsquodeadrsquo lsquoinanimatersquo lsquounrealrsquo and lsquonon-existentrsquo (Carey1999 p 298) Other concepts coalesce in particular early ontologically distinctconcepts of person animal and plant are increasingly likely to be classified under asuperordinate category living thing and deemed to possess common biologicalcapacities (Carey 1999)

In the studies reported here we investigate the latter phenomenon by assessingwhether there are age-related changes in the endorsement of different causalexplanations for the behaviours of humans species of increasing taxonomic distancefrom the human and plants Underlying the adult ability described earlier to detectcommonalities among living kinds should be an appreciation of the extent to whichgeneral causal principles can reasonably be applied to the behaviours of different livingkinds The modes of construal we examine intention teleology and essentialism are allimplicated in childrenrsquos reasoning about the world of living organisms (Wellman ampGelman 1998)















STUDY 1





Method








Results




















STUDY 2





Method









Results




































GENERAL DISCUSSION






















References

























































STUDY 1





Method








Results




















STUDY 2





Method









Results




































GENERAL DISCUSSION






















References




















































STUDY 1





Method








Results




















STUDY 2





Method









Results




































GENERAL DISCUSSION






















References














































STUDY 1





Method








Results




















STUDY 2





Method









Results




































GENERAL DISCUSSION






















References













































Method








Results




















STUDY 2





Method









Results




































GENERAL DISCUSSION






















References














































Results




















STUDY 2





Method









Results




































GENERAL DISCUSSION






















References




























































STUDY 2





Method









Results




































GENERAL DISCUSSION






















References
























































STUDY 2





Method









Results




































GENERAL DISCUSSION






















References

















































STUDY 2





Method









Results




































GENERAL DISCUSSION






















References














































Method









Results




































GENERAL DISCUSSION






















References















































Results




































GENERAL DISCUSSION






















References











































































GENERAL DISCUSSION






















References







































































GENERAL DISCUSSION






















References































































GENERAL DISCUSSION






















References




























































GENERAL DISCUSSION






















References




















































GENERAL DISCUSSION






















References

















































GENERAL DISCUSSION






















References































































References

























































References



















































References













































References








































































Documents

Why do birds of a feather flock together? Developmental ...evansem/PolEvans.BJDP.pdf · the varied explanations given for human behaviours. Over the last 50 years informed explanations