Yesterday’s post recounted a conversation with Tabitha (5) in which she asked for a “math class” problem.

I focused the initial discussion on where she learned what constitutes a “math class” problem.

But there’s lots more in there that’s interesting.

It wasn’t just her affective response (rejecting the driving in a car context, asking for a naked number problem) that matters here. Notice the way she engaged with the two problem types.

When I relented and posed the question, *What is two plus three?* she guessed. I know that she guessed because she (a) took no time to process, (b) asked rather than told me her answer, and (c) it was wrong despite being within her grasp.

When I posed the exact same problem in a situation she could imagine (she and Griffy had been to the arcade just that day), she engaged quite differently. Her body position changed. She paused. Her fingers moved. Each of these is an indication that she was *thinking*. And when she had an answer, she stated it; she did not ask.

The central tenet of CGI and an important belief underlying the IMAP work is that children can use contexts to solve problems that they cannot solve abstractly. Here it is in action. 2+3 is meaningless to Tabitha right now. But her 2 tickets combined with Griffin’s 3 tickets? That’s got meaning.

The conclusion here is obvious, right? We start with contexts kids understand and can reason about (here, combining tickets). We move to the abstract mathematical representations (2+3). We don’t save arcade tickets for *after* the kid understands addition. We don’t wait for symbolic mastery before doing some applications.