Child Development, zyxwvutsr August 1998, Volume 69, Number 4, Pages 994-995 Wanting to Get It Right: Commentary on Lillard and Joseph zy Alison Gopnik The articles by Lillard and by Joseph are an example of how apparently divergent empirical results may turn out to reflect interesting differences between children and adults. INTRODUCTION The history of these two articles (Joseph, 1998; Lil- lard, 1998; both in this issue) involves a departure for Child zyxwvutsrqpo Development, one we hope will set a helpful precedent for the future. Our field has been faced with a series of apparently contradictory results on topics of importance. Usually these are not simple failures to replicate a result, but involve relatively small modificationsof procedures that appear to lead to very different conclusions. These results often turn into extended debates about whether children do or do not have a particular kind of ability, knowledge, or concept at a particular age. The outcome, all too often, is simply a sense of experimental chaos that ends in researchers withdrawing from the topic in boredom (or even disgust). This is especially unfortu- nate because this way of phrasing the question, “Do children know zyxwvuts x at age y or not,” obscures the very developmental processes we are trying to study. For example, if children are genuinely involved in pro- cesses of conceptual change, we would expect their concepts to be both similar to and different from the concepts of adults, and we would expect those simi- larities and differences to themselves change as chil- dren develop. Those similarities and differences might be quite complex and subtle, and it might take a number of studies to let us see the whole picture. In the current case, faced with two papers on an important topic with apparently contradictory re- sults, we contacted the authors, and with their mu- tual agreement asked them to read each other’s pa- pers and come to some sort of resolution about the relevant features of design that might have led to the contradictory results. Both authors then submitted revised versions of the papers, taking into account the results of the other investigator. Obviously the process was more difficult and complex (and took longer!) than the usual editorial process. The authors Commentary on Lillard, “Wanting to Be It: Children’s Understanding of Intentions Underlying Pretense”; and Joseph, “Intention and Knowledge in Preschoolers’ Conception of Pretend.” did a truly impressive job of working to find a com- mon answer to the apparent contradictions;their per- sistence and sincerity are greatly to be commended. In this way the publication process can, we hope, it- self become part of the scientific enterprise, helping to decrease chaos and noise instead of multiplying them. In this instance, at least, the process has been strik- ingly successful in finding a common underlying thread that makes sense of apparently divergent re- sults. In fact, the diverging articles actually help to make more sense of the literature at large. In her orig- inal paper, Lillard mentioned, among other things, the possible difference between the ability of children to infer whether someone is pretending, given their mental state, versus inferring whether someone has a mental state, given that they are pretending. Joseph’s results suggest that this difference, in fact, makes a difference. Both Joseph and Lillard converged on this explanation in their revised papers. Children seem able to infer that someone is trying to sneeze if they are pretending to sneeze, but not that they must be trying to sneeze if they are pretending to sneeze. It seems that children’s concept of pretense involves an association with intentional action, but that for young children pretense can also include cases that simply involve “acting as if.” For them, pretense is often, even characteristically, intentional, but it is not neces- sarily intentional. For adults, to say that Moe was pretending to be a kangaroo but wasn’t trying to be a kangaroo would be incoherent and contradictory. For children, apparently, that may be odd or unusual; if Moe is pretending to be a kangaroo, it’s quite likely that he’s trying to be a kangaroo, but it’s eminently possible. A second important convergence in these results is that, even with very different methods, nei- ther Joseph nor Lillard find evidence for a representa- tional, as opposed to intentional, understanding of pretense before the age of zyx 4. zyxw 0 1998 by the Society for Research in Child Development, Inc. All rights reserved. 0009-3920/98 /6904-0019$01.00