Available online at www.sciencedirect.

com

ScienceDirect

Explanation-seeking curiosity in childhood

Emily G Liquin and Tania Lombrozo

Children are known for asking ‘why?’ — a query motivated by metacognition [13]. For example, when prompted to
their desire for explanations. Research suggests that generate explanations (versus report observations) while
explanation-seeking curiosity (ESC) is triggered by first- learning novel causal rules, preschoolers are more likely to
person cues (such as novelty or surprise), third-person cues learn causal rules that are broad [14], are simple [15], and
(such as a knowledgeable adults’ surprise or question), and emphasize internal properties over perceptual similarity
future-oriented cues (such as expectations about information [16]; when prompted to explain aspects of a story, 5-year-
gain or future value). Once triggered, ESC is satisfied by an olds to 6-year-olds are more likely to extract its lesson [17].
adequate explanation, typically obtained through causal However, this research has rarely focused on ESC — the
intervention or question asking, both of which change in drive state that prompts explanation search.
efficiency over development. ESC is an important driver of
children’s learning because it combines the power of active Second, research on curiosity (including developmental
learning and intrinsic motivation with the value of explanatory prerequisites to experiencing and expressing curiosity [18,19])
content, which can reveal the unobservable and causal has been pursued in many contexts, including active word
structure of the world to support generalizable knowledge. learning [3] and question asking [20 ]. Among other benefits
[21,22], there is evidence that curiosity-driven exploration can
Address be as efficient as direct pedagogy for causal learning [23] and
Department of Psychology, Princeton University, Peretsman Scully Hall, lead to better memory for target informa-tion [24–29], and that
Princeton, NJ, 08540, USA
exploration efficiency relates to IQ in young children [30].
Corresponding author: Liquin, Emily G ([email protected]) However, this work has typically focused on curiosity directed
towards non-explanatory tar-gets, such as facts, object labels,
or object properties.
Current Opinion in Behavioral Sciences 2020, 35:14–20
This review comes from a themed issue on Curiosity (Exolore versus Given the epistemic power of explanation and the learn-ing
Exoloit)
benefits of curiosity, ESC is poised to be an especially
Edited by Daphna Shohamy and Ran Hassin important driver of children’s learning. While it is yet
unclear precisely how ESC diverges from other forms of
curiosity [31 ,32] or fits into existing taxonomies for
curiosity [e.g. Ref. 33], ESC is unique in its ability to drive
https://doi.org/10.1016/j.cobeha.2020.05.012 the discovery of unobservables and causal structure —
2352-1546/ã 2020 The Authors. Published by Elsevier Ltd. This is an important foundations for generalizable knowledge. For
open access article under the CC BY-NC-ND license (http://creative- these reasons, the study of ESC is likely to lead to
commons.org/licenses/by-nc-nd/4.0/).
important insights about how (young) human learners come
to know so much about the world given limited evidence,
time, and cognitive resources, with potential implications
Introduction for education and for the development of artificial learners
From early in development, children exhibit curiosity about [34,35], as well.
the properties [1] and names [2–4] of objects, and about
visual and auditory patterns [5,6]. Moreover, they often What triggers explanation-seeking curiosity?
seek explanations for their observations, and in so doing Children encounter far more unexplained observations than
construct intuitive theories that equip them to better predict they have the time or resources to pursue as targets of
and intervene upon the world [7,8]. Here we focus on inquiry. In fact, there is evidence that merely lacking
children’s ‘explanation-seeking curiosity’ (ESC), which we information is insufficient to trigger information search: in
define as an affective drive state [9] that motivates learning one study, 5-year-olds to 9-year-olds explored following
how or why something is the case [10]. We review recent some types of underinformative evidence, but not others
research addressing two questions: (though this may in part stem from a developing capacity to
(1) What triggers ESC? And (2) How is ESC satisfied? recognize whether evidence is informative) [36]. Thus,
both theoretically and perhaps empirically, we have rea-son
Our focus on ESC complements two existing bodies of to believe that ESC is selective, and that additional cues
research. First, research has considered the epistemic power make some unexplained observations call out for
of explanations. Seeking explanations supports learning in explanation more strongly than others [37,38]. We review
young children [10,11], and explanations can drive three categories of cues: first-person cues, third-person
subsequent exploration [12] and improve cues, and future-oriented cues (Figure 1). The evidence

Current Opinion in Behavioral Sciences 2020, 35:14–20 www.sciencedirect.com

 Explanation-seeking in childhood Liquin and Lombrozo 15

Figure 1

Causal
Intervention
Child sticks
magnets and non-
magnets together;
gives plants
Observation
different Explanation
kinds/quantities of Magnets exert a
Child observes Explanation- liquid force on metals,
magnet sticks to Seeking Curiosity but not non-
paper clip; child Affective drive state that metals; water
notes that plants motivates learning how or
This is selective, depends on: transports
need water to why something is the case nutrients to the
grow
Question plant
Asking
Child asks adult,
”Why did those
stick together?”;
First-Person Third-Person Future-Oriented “Why do plants
need water to
Cues Cues Cues grow?”
Child evaluates fit Adult surprised; adult Child evaluates
with prior beliefs asks, “Why did that expected utility of
(novelty, surprise…) happen?” acquiring explanation

Current Opinion in Behavioral Sciences

Model of explanation-seeking curiosity, including cues that trigger it (first-person cues, third-person cues, and future-oriented cues), and two
methods by which it is satisfied (causal intervention and question asking).

for these cues comes from studies of ESC as well as studies Third-person cues
of curiosity more broadly; we discuss the poten-tially Third-person cues rely on social information from others.
unique triggers of ESC in the Conclusions section. For example, infants use others’ surprised expressions to
guide visual exploration [46], and preschoolers spend more
time exploring a novel toy after viewing an experi-menter’s
First-person cues surprised expression while playing with it [47]. Adults can
First-person cues, such as surprise and novelty [39], are also highlight information through the use of pedagogical
those that relate unexplained information to an individual’s questions: questions with answers that are known to the
epistemic state. We refer to these cues as first-person asker, but posed to facilitate learning [48]. After a
because observations are not intrinsically surprising or pedagogical question from a knowledgeable teacher,
novel; they are surprising or novel with respect to a set of children spend more time playing with a novel toy and
expectations or prior experiences. Indeed, recent research discover more of its functions [49 ]. However, relying on
has clarified the first-personal (subjective) nature of these knowledgeable others can also suppress explo-ration —
cues. Infants preferentially explore objects that violate their direct pedagogy decreases preschoolers’ explo-ration and
expectations, whether those expectations are formed on the learning of non-target information [50], and preschoolers
basis of core knowl-edge [27], inferred rules [40], or (though not elementary schoolers) forgo exploration of
probabilistic information [41,42]. Furthermore, infants’ counterintuitive claims offered by reliable sources [51].
exploration after an expec-tation violation may be
specifically geared towards unco-vering an explanation for
that violation [43]. Future-oriented cues
Future-oriented cues concern expectations about how
One proposal is that curiosity is piqued not by maxi-mally acquiring information will serve the learner in the future. In
surprising or novel information, which may be too far one task, 4-year-olds to 5-year-olds were more likely to
beyond a learner’s grasp, but by moderately surprising or explore unknown rewards over known rewards only when
novel information, which presents the best opportu-nity for the unknown rewards would inform future choices [52].
learning [44]. Infants preferentially direct their visual While this could be explained by reward-seeking behavior
attention to patterns that are moderately predict-able [5, see (and not curiosity), there is evidence that 4-year-olds to 5-
also Ref. 6]. However, this ‘moderate information gap’ year-olds explore a causal system more than older children
hypothesis has been studied primarily in adult populations and adults, but fail to efficiently exploit their knowledge to
[e.g. Ref. 45], and has not been applied to ESC. gain rewards (EG Liquin et al., unpublished) [see also 53–
55]. These findings cast doubt on reward pursuit as

www.sciencedirect.com Current Opinion in Behavioral Sciences 2020, 35:14–20

16 Curiosity (Exolore versus Exoloit)

the sole motivation behind children’s future-oriented Causal intervention

exploration. Furthermore, children sometimes pursue Effective causal intervention in pursuit of explanations
information that may be relevant for future action even relies on the ability to perform informative experiments:
when this information will notimpactimmediate reward: in typically, interventions that isolate a single causal variable
one study, children displayed greater curiosity about coun- and control confounds. Early research suggested that chil-
terfactual outcomes that were under their control, relative dren’s ability to design informative experiments is limited
to outcomes they could not have caused [56 ]. until adolescence [82], but recent research has demon-
strated that preschool-aged children choose interventions
Future-oriented cues are in tension with approaches that define that disambiguate causal structure [83], use conjectured
curiosity as the pursuit of information as an end in itself [e.g. explanations to inform their later exploration [84], and
Ref. 44]. In fact, in both adults [57] and non-human animals selectively deploy different exploratory actions based on
[58], curiosity has been operationalized as costly information- their relative informativeness for the task [85 ]. Older
seeking without immediate benefit. We propose that curiosity children selectively deviate from controlled experiments in
is defined by the phenomenological experience of pursuing contexts where testing multiple variables at once is more
information for its own sake, even in cases where the efficient [86]. In more complex tasks, however, 3-year-olds
psychological function of this experience is instrumental. The to 6-year-olds fail to spontaneously create disambiguating
willingness to pursue information with no immediate benefit is interventions after receiving evidence that disconfirms their
good evidence that such a phe-nomenology is present. expectations [87]. Instead, they use intervention strategies
However, adults report experienc-ing curiosity even when like the ‘positive test strategy,’ which privileges positive
instrumental goals are at play, and this reported curiosity is in evidence consistent with a single target hypothesis [87–89].
part guided by the future utility of information [31 ,59–61]. However, some have argued that a positive test strategy
Our model predicts the same for young children (and couldbequite useful insomecontexts[90] orgiven particular
potentially for many non-human animals), but direct evidence learning goals [91].
for this claim will require moving away from paradigms that
rely on exploration as an index of curiosity, and Question asking
towardsparadigmsthat more directly measure the Causal intervention is not appropriate in all circum-stances;
phenomenology of curiosity. unobservable entities (e.g. germs) and inaccessi-ble entities
(e.g. the moon) cannot be readily manipu-lated, and thus
How is explanation-seeking curiosity question asking may be the only way for children to learn
satisfied? about explanations involving these enti-ties. Indeed, by the
ESC can be satisfied when the child acquires what they age of six, children ask fact-seeking questions about
judge to be an adequate explanation. Children prefer unobservable entities but directly explore observable
explanations that are simple [62] and general [63 ], and that entities [92].
contain a moderate amount of detail [64] that is
explanatorily relevant [65,66 ]. Children also prefer tele- Children begin asking questions before the age of two, and
ological explanations (which appeal to function or pur- the proportion of questions that are explanation-seeking
pose) [67,68] and explanations that appeal to properties increases with development, peaking around age three [93].
inherent to the thing being explained [69], though these While children begin to direct questions to appropriate
preferences decrease across development. When children sources in preschool [20 ,52,94], children’s ability to ask
receive a non-explanatory response rather than an explan- maximally informative questions to solve a specified
atory response, they are more likely to re-ask their original problem continues to develop during the pre-school years
question [70] (though children’s reactions to explanatory and beyond. In one set of studies [95], 3-year-old to 5-year-
and non-explanatory responses are moderated by socio- old children were asked to identify the most informative
economic status [71,72] and age [73]). The perceived question to discover the explanation for an event. With
quality of a provided explanation predicts subsequent increasing proficiency with age, children preferred
explanation-seeking behavior, as well [74 ]. questions that targeted an individual explana-tion when that
explanation was more likely than others, but chose
Often, children are not simply offered an explanation, but questions that eliminated several candidate explanations
must solicit an explanation, reason their way to one, or when all were equally likely. Other studies using similar
discover one for themselves. Children have many tools for methods have also shown age-related improvements in
eliciting information — such as pointing [4,75–77] and question-asking efficiency [96–98, but see Ref. 99], and
social referencing [78] — but these are unlikely to elicit that question efficiency can be influ-enced by certain
specifically explanatory information. Causal intervention structural supports [96].
and question asking, by contrast, are likely to be especially
useful for the pursuit of explanations because they allow To summarize, children use both causal intervention and
the learner to infer causal relations [79,80] and inquire question asking to seek explanations, with both strategies
about unobservable entities [81] (Figure 1). improving in efficiency over the preschool and early

Current Opinion in Behavioral Sciences 2020, 35:14–20 www.sciencedirect.com

 Explanation-seeking in childhood Liquin and Lombrozo 17

elementary school years, and with the former continuing to explanation [31 ]. Finally,satisfactionmethods maydiffer:
develop through adolescence. Additional research will be for example, causal intervention is more likely than point-
required to link these exploratory behaviors to the drive ing to elicit explanatory information.
states that motivate them over development, and to iden-
tify whether and when these behaviors are triggered by Lastly, research on decision making under risk [110] and
first-person, third-person, or future-oriented cues. For on how costs are weighed against the benefits of informa-
example, in tasks that involve seeking rewards, there are tion gain [53–55,111] will shed light on when we can
age-related changes in the extent to which exploration is expect children to pursue their curiosity. With adults, it has
random (the result of decision noise)or systematic(directed been proposed that information is itself rewarding, and that
towards uncertainty) [54,100], and in tasks that involve curiosity motivates learners to obtain this reward [112–
learning about categories, children direct their exploration 114]. In settings where children are more explor-atory than
differently from adults [101]. Future research is needed to adults [53–55], do children find information more
determine how and why exploration that is motivated by rewarding than adults, or are the costs associated with
curiosity and directed at acquiring explanations changes (in exploration less steep for young learners? If explan-atory
quantity or in nature) across development. information is indeed especially powerful for learn-ing, do
learners find the satisfaction of ESC especially rewarding,
Conclusions and correspondingly experience stronger curi-osity towards
Combining the motivational drive of curiosity with the explanatory targets?
epistemic power of explanations, explanation-seeking
curiosity is likely to be an important driver of children’s Despite these open questions, recent research has dramati-
learning, especially when it comes to the unobservable and cally deepened our understanding of children’s ESC.
causal structure of the world captured by intuitive theories. Through question asking and exploration, children actively
We reviewed evidence that very young children are pursue information that helps them make sense of the
selective in their ESC and preferentially seek expla-nations world, and they do so with increasing proficiency over the
when cued by first-person cues, third-person cues, and course of development. Motivated by explanation-seeking
future-oriented cues. Children’s ability to seek expla- curiosity, children come to know not just what events and
nations also develops, with causal intervention and ques- phenomena they can expect to encounter, but why such
tion asking playing critical roles. events and phenomena occur — providing a powerful
means for prediction, intervention, and understanding.
While much of our analysis is based on studies of actual
learners, many of the papers we review argue that fully Conflict of interest statement
rational learners should explore when a relevant cue is present, Nothing declared.
or that exploration should be pursued in a particular way to
resolve uncertainty, typically using Bayesian approaches. Acknowledgements
Bayesian models have provided a powerful method for We would like to thank Corey Cusimano for helpful comments on an earlier
draft. This work was supported by a grant provided by the Templeton
understanding active learning in development, but additional Foundation to TL and an NSF Graduate Research Fellowship to EL [grant
questions arise when combining these indi-vidual analyses into number DGE-1656466]. Any opinions, findings, and conclusions or
a single model: How should a rational learner weigh recommendations expressed in this material are those of the authors and do not
necessarily reflect the views of the Templeton Foundation or the National
competing cues to curiosity against each other? How should a Science Foundation.
rational learner choose between question asking, causal
intervention, and other available behaviors? And how do References and recommended reading
actual learners navigate these problems? Answering these Papers of particular interest, published within the period of review,
have been highlighted as:
questions will require building more complex models,
designing more sophisticated experiments, and better of special interest
characterizing children’s explora-tion in real-world of outstanding interest
environments [78,94,102–105,106 ,107]. Ultimately, however,
1. Smith LB, Jayaraman S, Clerkin E, Yu C: The developing infant
these efforts will improve our under-standing of learning creates a curriculum for statistical learning. Trends Cogn Sci
throughout the lifespan. 2018, 22:325-336.
2. Carvalho PF, Vales C, Fausey CM, Smith LB: Novel names
extend for how long preschool children sample visual
Future research may also provide insight into whether and information. J Exp Child Psychol 2018, 168:1-18.
how we should expect ESC to diverge from other forms of
3. Jimenez S, Sun Y, Saylor MM: The process of active word
curiosity. Explanation search often involves consultation learning. In Active Learning from Infancy to Childhood: Social
with and deference to experts [108,109], suggesting third- Motivation, Cognition, and Linguistic Mechanisms. Edited by
Saylor MM, Ganea PA. Springer International Publishing;
person cues may be especially powerful in triggering ESC. 2018:75-93.
Additionally, the future-orientedcues thattrigger ESCmay
4. Lucca K, Wilbourn MP: The what and the how: information-
focus on unique criteria — for example, how likely it is that seeking pointing gestures facilitate learning labels and
the received information will constitute a ‘good’ functions. J Exp Child Psychol 2019, 178:417-436.

www.sciencedirect.com Current Opinion in Behavioral Sciences 2020, 35:14–20

18 Curiosity (Exolore versus Exoloit)

5. Kidd C, Piantadosi ST, Aslin RN: The Goldilocks effect: human 25. Fandakova Y, Gruber M: States of curiosity and interest
infants allocate attention to visual sequences that are neither too enhance memory differently in adolescents and in children. Dev
simple nor too complex. PLoS One 2012, 7:e36399. Sci 2020:e13005.
6. Kidd C, Piantadosi ST, Aslin RN: The Goldilocks effect in infant 26. Ruggeri A, Markant DB, Gureckis TM, Bretzke M, Xu F: Memory
auditory attention. Child Dev 2014, 85:1795-1804. enhancements from active control of learning emerge across
development. Cognition 2019, 186:82-94.
7. Bonawitz E, Ullman TD, Bridgers S, Gopnik A, Tenenbaum JB:
Sticking to the evidence? A behavioral and computational case 27. Stahl AE, Feigenson L: Observing the unexpected enhances
study of micro-theory change in the domain of magnetism. infants’ learning and exploration. Science 2015, 348:91-94.
Cogn Sci 2019, 43:e12765.
28. Stahl AE, Feigenson L: Expectancy violations promote learning
8. Gopnik A: Explanation as orgasm and the drive for causal in young children. Cognition 2017, 163:1-14.
knowledge: the function, evolution, and phenomenology of the
theory formation system. In Explanation and Cognition. Edited by 29. Stahl AE, Feigenson L: Violations of core knowledge shape
Keil FC, Wilson RA. The MIT Press; 2000:299-323. early learning. Top Cogn Sci 2019, 11:136-153.

9. Kidd C, Hayden BY: The psychology and neuroscience of 30. Muentener P, Herrig E, Schulz L: The efficiency of infants’
curiosity. Neuron 2015, 88:449-460. exploratory play is related to longer-term cognitive
development. Front Psychol 2018, 9:635.
10. Lombrozo T: Explanation and abductive inference. In Oxford
Handbook of Thinking and Reasoning. Edited by Holyoak KJ, 31. Liquin EG, Lombrozo T: A functional approach to explanation-
Morrison RG. Oxford University Press; 2012:260-276. seeking curiosity. Cogn Psychol 2020, 119:101276 Characterizes
explanation-seeking curiosity in adults. Finds that expla-nation-seeking
11. Lombrozo T: Explanatory preferences shape learning and curiosity is potentially distinct from fact-seeking curiosity, is associated
inference. Trends Cogn Sci 2016, 20:748-759. with subsequent information-seeking behavior, and is related to
potential triggers capturing fit with prior knowledge, expecta-tions about
12. Danovitch JH, Mills CM: Understanding when and how future learning, and expectations about future export.
explanation promotes exploration. In Active Learning from
Infancy to Childhood: Social Motivation, Cognition, and Linguistic 32. Schwitzgebel E: Children’s theories and the drive to explain. Sci
Mechanisms. Edited by Saylor MM, Ganea PA. Springer Educ 1999, 8:457-488.
International Publishing; 2018:95-112.
33. Berlyne DE: A theory of human curiosity. Br J Psychol 1954,
13. Mills CM, Keil FC: Knowing the limits of one’s understanding: the 45:180-191.
development of an awareness of an illusion of explanatory depth.
J Exp Child Psychol 2004, 87:1-32. 34. Oudeyer P-Y: What do we learn about development from baby
robots? Wiley Interdiscipl Rev Cogn Sci 2017, 8:e1395.
14. Walker CM, Lombrozo T, Williams JJ, Rafferty AN, Gopnik A:
Explaining constrains causal learning in childhood. Child Dev 35. Oudeyer P-Y: Computational theories of curiosity-driven learning.
2017, 88:229-246. In The New Science of Curiosity. Edited by Gordon G. Nova
Science Publishers; 2018:43-72.
15. Walker CM, Bonawitz E, Lombrozo T: Effects of explaining on
children’s preference for simpler hypotheses. Psychon Bull Rev 36. Busch JTA, Legare CH: Using data to solve problems: children
2017, 24:1538-1547. reason flexibly in response to different kinds of evidence. J Exp
Child Psychol 2019, 183:172-188.
16. Walker CM, Lombrozo T, Legare CH, Gopnik A: Explaining
prompts children to privilege inductively rich properties. 37. Wong W, Yudell Z: A normative account of the need for
Cognition 2014, 133:343-357. explanation. Synthese 2015, 192:2863-2885.

17. Walker CM, Lombrozo T: Explaining the moral of the story. 38. Grimm SR: Explanatory inquiry and the need for explanation. Br
Cognition 2017, 167:266-281. J Philos Sci 2008, 59:481-497.
39. Berlyne DE, Frommer FD: Some determinants of the incidence
18. Ronfard S, Bartz DT, Cheng L, Chen X, Harris PL: Children’s
and content of children’s questions. Child Dev 1966, 37:177-
developing ideas about knowledge and its acquisition. In
189.
Advances in Child Development and Behavior. Edited by Benson
JB. Elsevier; 2018:123-151. 40. Wang S, Zhang Y, Baillargeon R: Young infants view physically
19. Sobel DM, Letourneau SM: Curiosity, exploration, and possible support events as unexpected: new evidence for rule
learning. Cognition 2016, 157:100-105.
children’s understanding of learning. In Active Learning from
Infancy to Childhood: Social Motivation, Cognition, and Linguistic 41. Sim ZL, Xu F: Infants preferentially approach and explore the
Mechanisms. Edited by Saylor MM, Ganea PA. Springer unexpected. Br J Dev Psychol 2017, 35:596-608.
International Publishing; 2018:57-74.
42. Sim ZL, Xu F: Another look at looking time: surprise as rational
20. Ronfard S, Zambrana IM, Hermansen TK, Kelemen D: Question- statistical inference. Top Cogn Sci 2019, 11:154-163.
asking in childhood: a review of the literature and a framework for
understanding its development. Dev Rev 2018, 49:101-120 43. Perez J, Feigenson L: Violations of expectation trigger infants to
Comprehensively reviews the research on question initiation, formulation, search for explanations. PsyArXiv preprint 2020 http://dx.doi.
expression, and response evaluation between infancy and elementary org/10.31234/osf.io/eahjd.
school. Discusses potential sources of individual differences.
44. Loewenstein G: The psychology of curiosity: a review and
21. van Schijndel TJ, Jansen BR, Raijmakers ME: Do individual reinterpretation. Psychol Bull 1994, 116:75-98.
differences in children’s curiosity relate to their inquiry-based
learning? Int J Sci Educ 2018, 40:996-1015. 45. Kang MJ, Hsu M, Krajbich IM, Loewenstein G, McClure SM,
Wang JT, Camerer CF: The wick in the candle of learning:
22. Gottfried AE, Preston KSJ, Gottfried AW, Oliver PH, epistemic curiosity activates reward circuitry and enhances
Delany DE, Ibrahim SM: Pathways from parental stimulation of memory. Psychol Sci 2009, 20:963-973.
children’s curiosity to high school science course
accomplishments and science career interest and skill. Int J Sci 46. Wu Y, Gweon H: Surprisingly unsurprising! Infants’ looking time
Educ 2016, 38:1972-1995. at probable vs. improbable events is modulated by others’
expressions of surprise. PsyArXiv preprint 2019 http://
23. Sim ZL, Xu F: Learning higher-order generalizations through dx.doi.org/10.31234/osf.io/8whuv.
free play: evidence from 2- and 3-year-old children. Dev
Psychol 2017, 53:642-651. 47. Wu Y, Gweon H: Preschoolers jointly consider others’ expression
of surprise and common ground to decide when to explore.
24. Bernacki ML, Walkington C: The role of situational interest in PsyArXiv preprint 2019 http://dx.doi.org/10.31234/osf.io/ ckh6j.
personalized learning. J Educ Psychol 2018, 110:864-881.

Current Opinion in Behavioral Sciences 2020, 35:14–20 www.sciencedirect.com

 Explanation-seeking in childhood Liquin and Lombrozo 19

48. Yu Y, Bonawitz E, Shafto P: Pedagogical questions in parent– that individual bear), but in physics by age 11. Suggests that ESC may
child conversations. Child Dev 2019, 90:147-161. be satisfied by different explanations across development.
49. Yu Y, Landrum AR, Bonawitz E, Shafto P: Questioning supports 64. Frazier BN, Gelman SA, Wellman HM: Young children prefer
effective transmission of knowledge and increased and remember satisfying explanations. J Cogn Dev 2016,
exploratory learning in pre-kindergarten children. Dev Sci 17:718-736.
2018, 21:e12696
Introduces ‘pedagogical questioning’ as a tool for teaching children 65. Chu J, Schulz L: Cognitive pragmatism: children flexibly choose
target information without decreasing their exploration of non-target between facts and conjectures. In In Proceedings of the 40th
information, in contrast to direct instruction. Provides a tool for introdu- Annual Conference of the Cognitive Science Society. Edited by
cing information without hampering curiosity. Rogers TT, Rau M, Zhu X, Kalish CW. Proceedings of the 40th
Annual Conference of the Cognitive Science Society Cognitive
50. Bonawitz E, Shafto P, Gweon H, Goodman ND, Spelke E, Schulz Science Society; 2018:226-231.
L: The double-edged sword of pedagogy: instruction limits
spontaneous exploration and discovery. Cognition 2011, 66. Johnston AM, Sheskin M, Keil FC: Learning the relevance of
120:322-330. relevance and the trouble with truth: evaluating explanatory
relevance across childhood. J Cogn Dev 2019, 20:555-572
51. Ronfard S, Chen EE, Harris PL: The emergence of the empirical Investigates whether children evaluate explanations by whether they
stance: children’s testing of counterintuitive claims. Dev are relevant to an explanatory goal, rather than merely by their truth.
Psychol 2018, 54:482-493. Children as young as 4 years old can select a relevant explanation as
52. Bonawitz E, Bass I, Lapidow E: Choosing to learn: evidence more helpful than an irrelevant but true explanation; however, the ability
evaluation for active learning and teaching in early childhood. In to identify relevant explanations in isolation improves across childhood.
Active Learning from Infancy to Childhood: Social Motivation, Suggests that the criteria for satisfying ESC may change across
Cognition, and Linguistic Mechanisms. Edited by Saylor MM, development.
Ganea PA. Springer International Publishing; 2018:213-231.
67. Kelemen D: Why are rocks pointy? Children’s preference for
53. Blanco NJ, Sloutsky V: Systematic exploration and uncertainty teleological explanations of the natural world. Dev Psychol
dominate young children’s choices. PsyArXiv preprint 2019 1999, 35:1440-1452.
http://dx.doi.org/10.31234/osf.io/72sfx.
68. Lombrozo T, Bonawitz EB, Scalise NR: Young children’s learning
54. Schulz E, Wu CM, Ruggeri A, Meder B: Searching for rewards and generalization of teleological and mechanistic explanations.
like a child means less generalization and more directed J Cogn Dev 2018, 19:220-232.
exploration. Psychol Sci 2019, 30:1561-1572.
69. Cimpian A, Steinberg OD: The inherence heuristic across
55. Sumner E, Li AX, Perfors A, Hayes B, Navarro D, Sarnecka BW: development: systematic differences between children’s and
The exploration advantage: children’s instinct to explore allows adults’ explanations for everyday facts. Cogn Psychol 2014,
them to find information that adults miss. PsyArXiv preprint 2019 75:130-154.
http://dx.doi.org/10.31234/osf.io/h437v.
70. Frazier BN, Gelman SA, Wellman HM: Preschoolers’ search for
56. FitzGibbon L, Moll H, Carboni J, Lee R, Dehghani M: explanatory information within adult-child conversation. Child
Counterfactual curiosity in preschool children. J Exp Child Dev 2009, 80:1592-1611.
Psychol 2019, 183:146-157
Finds that 4-year-olds to 5-year-olds pursue post-decision information 71. Kurkul KE, Corriveau KH: Question, explanation, follow-up: a
about alternative courses of action, preferentially after negative mechanism for learning from others? Child Dev 2018, 89:280-
outcomes and for alternatives they could have pursued. Suggests that 294.
children’s curiosity is selective to particular types of stimuli and ¨
potentially geared towards understanding undesirable outcomes that 72. U nlu¨tabak B, Nicolopoulou A, Aksu-Koc¸ A: Questions asked by
could be prevented in the future. Turkish preschoolers from middle-SES and low-SES families. Cogn
Dev 2019, 52:100802.
57. Lau JKL, Ozono H, Kuratomi K, Komiya A, Murayama K: Shared
striatal activity in decisions to satisfy curiosity and hunger at the 73. Woolley JD, Cornelius CA: Wondering how: children’s and
risk of electric shocks. Nat Hum Behav 2020 http://dx.doi. adults’ explanations for mundane, improbable, and
org/10.1038/s41562-020-0848-3. extraordinary events. Psychon Bull Rev 2017, 24:1586-1596.
58. Wang MZ, Hayden BY: Monkeys are curious about 74. Mills CM, Sands KR, Rowles SP, Campbell IL: “I want to know
counterfactual outcomes. Cognition 2019, 189:1-10. more!”: children are sensitive to explanation quality when
exploring new information. Cogn Sci 2019, 43:e12706
59. Dubey R, Griffiths TL: A rational analysis of curiosity. In In Finds that lower subjective ratings of explanation quality predict greater
Proceedings of the 39th Annual Conference of the Cognitive subsequent information search between the ages of 7 and 10 years.
Science Society. Edited by Gunzelmann G, Howes A, Tenbrink T, Supports a deprivation theory of curiosity, where learners experience
Davelaar EJ. Proceedings of the 39th Annual Conference of the curiosity when they judge that they lack information.
Cognitive Science Society Cognitive Science Society; 2017:307-
312. 75. Begus K, Southgate V: Infant pointing serves an interrogative
function. Dev Sci 2012, 15:611-617.
60. Dubey R, Griffiths TL: Reconciling novelty and complexity
through a rational analysis of curiosity. Psychol Rev 2020, 76. Begus K, Southgate V: Curious learners: how infants’
127:455-476. motivation to learn shapes and is shaped by infants’ interactions
with the social world. In Active Learning from Infancy to
61. Dubey R, Griffiths TL, Lombrozo T: If it’s important, then I am
Childhood: Social Motivation, Cognition, and Linguistic
curious: a value intervention to induce curiosity. In In Mechanisms. Edited by Saylor MM, Ganea PA. Springer
Proceedings of the 41st Annual Conference of the Cognitive International Publishing; 2018:13-37.
Science Society. Edited by Goel AK, Seifert CM, Freksa C.
Proceedings of the 41st Annual Conference of the Cognitive 77. Lucca K, Wilbourn MP: Communicating to learn: infants’
Science Society Cognitive Science Society; 2019:282-288. pointing gestures result in optimal learning. Child Dev 2018,
62. Bonawitz EB, Lombrozo T: Occam’s rattle: children’s use of 89:941-960.
simplicity and probability to constrain inference. Dev Psychol
78. Schieler A, Koenig M, Buttelmann D: Fourteen-month-olds
2012, 48:1156-1164.
selectively search for and use information depending on the
63. Johnston AM, Sheskin M, Johnson SGB, Keil FC: Preferences for familiarity of the informant in both laboratory and home contexts.
explanation generality develop early in biology but not J Exp Child Psychol 2018, 174:112-129.
physics. Child Dev 2018, 89:1110-1119
Finds that children prefer explanations at higher levels of generality in biology
79. Muentener P, Bonawitz E: The development of causal reasoning.
by age 5 (e.g. explaining a property of an individual bear by appealing to a
In The Oxford Handbook of Causal Reasoning. Edited by
property of all animals, rather than a property of all bears or
Waldmann MR. Oxford University Press; 2017.

www.sciencedirect.com Current Opinion in Behavioral Sciences 2020, 35:14–20

20 Curiosity (Exolore versus Exoloit)

80. Woodward JF: Making Things Happen: A Theory of Causal 97. Ruggeri A, Lombrozo T: Children adapt their questions to
Explanation. Oxford University Press; 2003. achieve efficient search. Cognition 2015, 143:203-216.

81. Harris PL, Koenig MA: Trust in testimony: how children learn 98. Ruggeri A, Lombrozo T, Griffiths TL, Xu F: Sources of
about science and religion. Child Dev 2006, 77:505-524. developmental change in the efficiency of information search.
Dev Psychol 2016, 52:2159-2173.
82. Kuhn D, Amsel E, O’Loughlin M, Schauble L, Leadbeater B,
Yotive W: The Development of Scientific Thinking Skills. 99. Meder B, Nelson JD, Jones M, Ruggeri A: Stepwise versus
Academic Press; 1988. globally optimal search in children and adults. Cognition 2019,
191:103965.
83. Lapidow E, Walker CM: Informative experimentation in intuitive
science: children select and learn from their own causal 100. Somerville LH, Sasse SF, Garrad MC, Drysdale AT, Abi Akar N,
interventions. Cognition 2020, 201:104315. Insel C, Wilson RC: Charting the expansion of strategic
exploratory behavior during adolescence. J Exp Psychol Gen
84. Legare CH: Exploring explanation: explaining inconsistent 2017, 146:155-164.
evidence informs exploratory, hypothesis-testing behavior in
young children. Child Dev 2012, 83:173-185. 101. Foster-Hanson E, Moty K, Cardarelli A, Ocampo JD, Rhodes M:
Developmental changes in strategies for gathering evidence
85. Ruggeri A, Swaboda N, Sim ZL, Gopnik A: Shake it baby, but only about biological kinds. Cogn Sci 2020, 44:e12837.
when needed: preschoolers adapt their exploratory strategies to 102. Bilal D, Gwizdka J: Children’s query types and reformulations in
the information structure of the task. Cognition 2019, 193:104013 Google search. Inf Process Manag 2018, 54:1022-1041.
Reports that children trained on a novel exploratory action (shaking a 103. Danovitch JH: Growing up with Google: how children’s
box to determine its contents) preferentially deploy this action when it is understanding and use of internet-based devices relates to
needed to reduce uncertainty. Suggests early developing competencies cognitive development. Hum Behav Emerg Technol 2019, 1:81-
underlying the design of effective causal intervention. 90.
86. Jones A, Bramley NR, Gureckis TM, Ruggeri A: Changing many 104. Lovato SB, Piper AM: Young children and voice search: what we
things at once sometimes makes for a good experiment, and know from human-computer interaction research. Front Psychol
children know that. PsyArXiv preprint 2020 http://dx.doi.org/ 2019, 10:8.
10.31234/osf.io/9qv5y.
¨ 105. Wade S, Kidd C: Cross-cultural differences in the influence of
87. Ko¨ksal-Tuncer O , Sodian B: The development of scientific peers on exploration during play. Cogn Sci 2018, 42:3050-3070.
reasoning: hypothesis testing and argumentation from
evidence in young children. Cogn Dev 2018, 48:135-145. 106. Willard AK, Busch JT, Cullum KA, Letourneau SM, Sobel DM,
Callanan M, Legare CH: Explain this, explore that: a study of
88. Nussenbaum K, Cohen AO, Davis Z, Halpern D, Gureckis T, parent–child interaction in a children’s museum. Child Dev
Hartley C: Causal information-seeking strategies change 2019, 90:e598-e617
across childhood and adolescence. PsyArXiv preprint 2019 Manipulated whether parents were directed to encourage exploration,
http://dx.doi.org/10.31234/osf.io/qukac. encourage explanation, or behave as normal in a gear exhibit in a
children’s museum. Children of parents who were prompted to encou-
89. Meng Y, Bramley N, Xu F: Children’s causal interventions rage exploration spent more time connecting gears, while children of
combine discrimination and confirmation. In In Proceedings of the parents who were prompted to encourage explanation spent more time
40th Annual Conference of the Cognitive Science Society. Edited spinning gears and discussing gear mechanisms. Underscores the
by Rogers TT, Rau M, Zhu X, Kalish CW. Proceedings of the importance of studying ESC in real-world environments.
40th Annual Conference of the Cognitive Science Society
Cognitive Science Society; 2018:762-767. 107. Callanan MA, Legare CH, Sobel DM, Jaeger GJ, Letourneau S,
McHugh SR, Willard A, Brinkman A, Finiasz Z, Rubio E et al.:
90. Navarro DJ, Perfors AF: Hypothesis generation, sparse Exploration, explanation, and parent–child interaction in
categories, and the positive test strategy. Psychol Rev 2011, museums. Monogr Soc Res Child Dev 2020, 85:7-137.
118:120-134.
108. Keil FC, Stein C, Webb L, Billings VD, Rozenblit L: Discerning the
91. Lapidow E, Walker CM: The search for invariance: repeated division of cognitive labor: an emerging understanding of how
positive testing serves the goals of causal learning. Language knowledge is clustered in other minds. Cogn Sci 2008, 32:259-
and Concept Acquisition from Infancy Through Childhood. 300.
Springer; 2020:197-219.
109. Wilkenfeld DA, Plunkett D, Lombrozo T: Depth and deference:
92. Fitneva SA, Lam NH, Dunfield KA: The development of children’s when and why we attribute understanding. Philos Stud 2016,
information gathering: to look or to ask? Dev Psychol 2013, 173:373-393.
49:533-542.
110. Rosenbaum GM, Hartley CA: Developmental perspectives on
93. Chouinard MM: Children’s questions: a mechanism for risky and impulsive choice. Philos Trans R Soc B 2018,
cognitive development. Monogr Soc Res Child Dev 2007, 72:1- 374:20180133.
129. 111. Rowles SP, Mills CM: “Is it worth my time and effort?”: How
children selectively gather information from experts when faced
94. Choi K, Lapidow E, Austin J, Shafto P, Bonawitz E: Preschoolers with different kinds of costs. J Exp Child Psychol 2019, 179:308-
are more likely to direct questions to adults than to other children 323.
(or selves) during spontaneous conversational act. In In
Proceedings of the 40th Annual Conference of the Cognitive 112. Gruber MJ, Ranganath C: How curiosity enhances
Science Society. Edited by Rogers TT, Rau M, Zhu X, Kalish hippocampus-dependent memory: the Prediction, Appraisal,
CW. Proceedings of the 40th Annual Conference of the Cognitive Curiosity, and Exploration (PACE) framework. Trends Cogn Sci
Science Society Cognitive Science Society; 2018:220-225. 2019, 23:1014-1025.
95. Ruggeri A, Sim ZL, Xu F: “Why is Toma late to school again?” 113. Marvin CB, Shohamy D: Curiosity and reward: valence predicts
Preschoolers identify the most informative questions. Dev choice and information prediction errors enhance learning. J Exp
Psychol 2017, 53:1620-1632. Psychol Gen 2016, 145:266-272.
96. Kachergis G, Rhodes M, Gureckis T: Desirable difficulties during 114. Murayama K, FitzGibbon L, Sakaki M: Process account of
the development of active inquiry skills. Cognition 2017, 166:407- curiosity and interest: a reward-learning perspective. Educ
417. Psychol Rev 2019, 31:875-895.

Current Opinion in Behavioral Sciences 2020, 35:14–20 www.sciencedirect.com