[songhonglei1985]

This resonates but I think there's a framing issue. From building educational coding tools for kids, what I've observed is that the volume of textual engagement hasn't necessarily dropped — it's migrated to different forms.

An 11-year-old debugging a Python script is doing intense, sustained reading. Parsing error messages, reading documentation, tracing logic through nested conditionals. A kid building in Roblox Studio is reading and writing Lua. These are literacy acts — just not the kind that show up on NAEP scores or novel-reading surveys.

The real gap isn't hours logged. It's transfer. The kid who can read 200 lines of GDScript to debug why their character won't jump doesn't automatically transfer that stamina to reading a history textbook. The cognitive muscles are being built, but they're specialized.

From what I've seen, when you meet kids where they are — letting them read and write code to solve problems they care about — they'll happily put in hours of focused textual work. The appetite is there. We're measuring the wrong meal.

[annzabelle]

Have you interacted with a typical class of 30 children lately? Maybe 1 or 2 of them are doing what you described, but 15 of them don't understand how to navigate a basic filesystem because they haven't used anything more advanced that a Chromebook in their lives and they have no desire to do anything more complicated with a computer.

Also, I fundamentally disagree that reading code and error messages stretches the same muscle as long form prose, and I believe that sustained attention on long pieces of writing is how a lot of sustained cultural values are transmitted, where reading 3,000 lines of python does none of that.