So not all that surprising from the latest. But one part jumped out at me in particular.
In response to the question, "Regular folks don't get the distinction between certified teachers and qualified teachers - why the teachers' union wouldn't let Einstein teach physics to high school students because he wasn't certified," Duncan responded, "Isn't all that matters that our children learn? That teachers give students knowledge?[emphasis mine]"
A couple of things that made my blood boil here. One, the question is a total red herring. (For many reasons.) Two, the answer reflects a complete lack of understanding about how children learn and what makes a great [i.e. "qualified"] teacher. As it happens, I just finished reading a study, Effects of Teachers' Mathematical Knowledge for Teaching on Student Achievement from Hill, Rowan & Ball that shows that the most significant predictor of student achievement isn't teacher's knowledge of mathematical content, but their knowledge of the pedagogy of teaching math - as they put it, "mathematical knowledge for teaching." In other words, we have to know not just what we're going to teach, but how to do it.
Which brings me to my next point. "...teachers give students knowledge..." Mr. Duncan. Tabula Rasa is so 17th century.
Perhaps it's not all that surprising that Duncan likens education to Pavlovian experiments. Students = Dogs? Sure. Why not.