Somehow I’ve managed to forget my usual end-of-term post-mortem of the year’s lecturing. I think perhaps I’m only now recovering from 11 weeks of lectures, lab supervision, tutoring alongside a very busy time analysing Planck satellite data.

But a few weeks ago term ended, and I finished teaching my undergraduate cosmology course at Imperial, 27 lectures covering 14 billion years of physics. It was my fourth time teaching the class (I’ve talked about my experiences in previous years here, here, and here), but this will be the last time during this run. Our department doesn’t let us teach a course more than three or four years in a row, and I think that’s a wise policy. I think I’ve arrived at some very good ways of explaining concepts such as the curvature of space-time itself, and difficulties with our models like the 122-or-so-order-of-magnitude cosmological constant problem, but I also noticed that I wasn’t quite as excited as in previous years, working up from the experimentation of my first time through in 2009, putting it all on a firmer foundation — and writing up the lecture notes — in 2010, and refined over the last two years. This year’s teaching evaluations should come through soon, so I’ll have some feedback, and there are still about six weeks until the students’ understanding — and my explanations — are tested in the exam.

Next year, I’ve got the frankly daunting responsibility of teaching second-year quantum mechanics: 30 lectures, lots of problem sheets, in-class problems to work through, and of course the mindbending weirdness of the subject itself. I’d love to teach them Dirac’s very useful notation which unifies the physical concept of quantum states with the mathematical ideas of vectors, matrices and operators — and which is used by all actual practitioners from advanced undergraduates through working physicists. But I’m told that students find this an extra challenge rather than a simplification. Comments from teachers and students of quantum mechanics are welcome.

Given the vast amount of material in the QM course (at least the course taught by Prof. Dauncey) I think it would be difficult to spend enough time really emphasising dirac notation.

Although personally I only began to feel comfortable doing QM after sitting the Foundations of QM course. This is because it is easy starting from linear vector spaces moving to Dirac notation and then learning the rules of QM & the Schrodinger Equation.

Then the recipes for doing QM just seemed more natural than starting off from the wave picture. Eigenstates made more sense as you could make an analogy with a basis set of vectors.

Of course time spent on the formalism comes at a cost of time spent teaching the physics.

Also something which isn't emphasised in second year enough is that not just arrows in space form a vector space (i.e. functions and quantum states also form a vector space). So it takes a while (if ever) for it to click that fourier series and quantum states can be viewed as being decomposed into a set of orthogonal vectors/functions/states.

## KM

Given the vast amount of material in the QM course (at least the course taught by Prof. Dauncey) I think it would be difficult to spend enough time really emphasising dirac notation.

Although personally I only began to feel comfortable doing QM after sitting the Foundations of QM course. This is because it is easy starting from linear vector spaces moving to Dirac notation and then learning the rules of QM & the Schrodinger Equation.

Then the recipes for doing QM just seemed more natural than starting off from the wave picture. Eigenstates made more sense as you could make an analogy with a basis set of vectors.

Of course time spent on the formalism comes at a cost of time spent teaching the physics.

Also something which isn't emphasised in second year enough is that not just arrows in space form a vector space (i.e. functions and quantum states also form a vector space). So it takes a while (if ever) for it to click that fourier series and quantum states can be viewed as being decomposed into a set of orthogonal vectors/functions/states.