Facts Stimulate Reasoning
Words serve as tools to emphasize particular ways of thinking that were already available.
Again from Gelman's Learning from Others:
From their earliest use, words—especially count nouns—seem to serve as "placeholders" for children. Studies conducted with children ranging from 13 months through the preschool years demonstrate that children treat objects that receive the same noun label as if they have common, nonobvious properties. They do so in two respects: First, labels enable dissimilar objects to be treated alike, as having properties in common [e.g., upon learning that a blackbird feeds its young mashed-up food, children are more likely to extend that property to another bird (e.g., a flamingo) than to a superficially more similar nonbird (e.g., bat)], and second, labels promote inferences regarding nonobvious features, such as internal parts, functions, and other nonvisible behaviors. Waxman & Markow (1995) propose that count nouns are "invitations" to children to form categories and look for deeper correlates: Common labels lead even infants to search for commonalities; distinct labels lead children to search for differences. Even 9-month-olds (who are not yet producing speech) are more likely to attend to relevant within-category similarities when they hear two different items labeled with the same word. . . .
The lexicalization [labeling] effect is not one of words creating concepts where none had belonged, but rather involves words serving as tools to emphasize particular ways of thinking that were already available.
And it's definitely worth seeing how this emphasis on sameness and difference in children's reasoning from labels forms a large part of the foundation of Engelmann and Carnine's Theory of Instruction:
The learning mechanism that we postulate has two attributes:
The capacity to learn any quality that is exemplified through examples (from the quality of redness to the quality of inconsistency).
The capacity to generalize to new examples on the basis of sameness of quality (and only on the basis of sameness).
These attributes suggest the capacities that we would have to build into a computer that functions the way a human does. Note that we are not asserting that these are the only attributes that a human possesses, merely that by assuming the two attributes we can account for nearly all observed cognitive behavior. . . .
There is no sharp line between initial learning and generalization. The rule-construction of the learning mechanism is assumed to begin as soon as examples are presented. In formulating a rule, the mechanism does nothing more than "note" sameness of quality. Once the mechanism "has determined" what is the same about the examples of a particular concept, generalization occurs.
