I think you’re correct here. The author assumes you have the starting knowledge out of undergrad (or ability to self teach) and everything in grad school is sor

I think you’re correct here. The author assumes you have the starting knowledge out of undergrad (or ability to self teach) and everything in grad school is sorting and signaling. My view of school is more nuanced. Both undergrad and grad have a mix of signaling, sorting and learning. Since the financial barrier to entry to solving math problems is low, the easy problems are solved. To solve the hard ones, you need several more years of post-undergrad coursework to get to the edges of the field, and several more being mentored on how to push the edges. It’s hard to get there without the more advanced guided tour that grad school provides.