Choosing the Best Explanation Is Elementary, My Dear Watson

You arrive home one day to find the kitchen window broken and your stash of dark chocolate truffles dangerously depleted. Explanation #1: A chocolate-seeking criminal has broken in! Explanation #2: The neighbors' children threw a ball through the window, and your husband helped himself to a husband-sized serving of truffles.

Short of catching the culprit chocolate-handed, which explanation should you believe?

Sherlock Holmes sagely advises that when faced with competing explanations, you "balance probabilities and choose the most likely." So if past experience suggests that careless children and hungry husbands are common while truffle-motivated crime is rare, then you should go with explanation #2.

But explanation #1 finds an equally august champion in the form of William of Occam, whose well-known razor suggests that we choose the simpler explanation. In this case, that's the single, criminal act that explains both the broken window and the missing truffles, not the incidental conjunction of poor aim and a peckish spouse.

So whose advice should you follow?

Your preschooler's, of course. That's because recent research finds that 4- and 5-year old children are surprisingly savvy reasoners when it comes to choosing between competing explanations, navigating a middle course between Holmes's probability and Occam's simplicity.

In research conducted at UC Berkeley and MIT, Dr. Elizabeth Baraff Bonawitz and I asked 4- and 5-year old children to explain why a toy lit up and spun around. Children were first taught that blocks of different colors could be placed in the toy's "activator" to generate different effects: a red block made the toy light up, a green block made the toy spin, and a blue block (like our potential chocolate-thief) generated both observed effects, providing a simpler explanation for both of the toy's actions.

Now here's where it gets tricky. Children were also shown that there were different numbers of each kind of block. For some children, red and green blocks were only a little more common than blue blocks: 3 red and 3 green for 1 blue. For other children, red and green blocks were much more common than blue blocks: 18 red and 18 green for 1 blue. In this last case Holmes and Occam disagree, with probability pointing to the conjunction of a red block and a green block as the best explanation, but simplicity still pointing to the blue block.

Children's responses revealed surprising sophistication. While there was an overall preference for the simpler explanation, with a majority of children explaining why the toy lit up and spun around by appeal to a single, blue block, this preference was drastically reduced when the blue block was very rare, with the majority of children now appealing to the conjunction of a red block and a green block. In other words, children went with simplicity when there wasn't strong evidence for an alternative, but as evidence accumulated they followed its lead.

Holmes + Occam = one clever kid.

But what about adults? And what exactly makes one explanation simpler than another? Is taking simplicity into account the right thing to do, or a sign of human error? (And who did eat the chocolate truffles, anyway?) For more on these and other questions related to human curiosity and its consequences, subscribe to Explananda, my new blog through Psychology Today.