Confounded Age: Linguistic and Cognitive Factors in Age Differences for Second Language Acquisition
Ellen Bialystok York University Kenji Hakuta Stanford University
In spite of what we all learned in our first statistics course, we just cannot resist attributing causality to correlation. We have to remind ourselves every time we see two events contiguously linked in time and space that the most natural explanation for their co-occurrence, namely, that one causes the other, might simply be false. The assumption of causality is one of the basic tenets of commonsense logic: Spring rains lead to flowers, knocking over the juice container results in spilled liquid, and clicking the power button on a small handheld instrument causes pictures to appear on the television screen. We all know, too, that it is counterexamples that compel caution in assuming the interpretation of causality: Superstition notwithstanding, carrying or not carrying an umbrella has no causal consequence for local meteorological conditions.
How are we to discover the correct logical relation between two events that share patterns of occurrence? The simplest explanation, that one event causes the other, is often taken at the expense of details that do not fit easily into the interpretation but are overlooked, set aside, or discounted. Indeed, it was the final effort to deal with the inconsistencies in the Ptolemeic description of planetary motion that led to the overthrow of that explanation, but it took 14 centuries and