I've been really enjoying Monty Don's new series, Mastercrafts, about people learning rural crafts. Part of my enjoyment can be explained by my inordinate fondness for Monty Don, but mostly it's because I love watching people skilled in their craft do their work.

The programme uses the familiar 'they have only 6 weeks to compete to be the best at X' format, which seems to be obligatory for any kind of reality show these days, but it isn't too obnoxious. Winning isn't the only goal for those participating, and they can all progress further with the skills they have learned when the show is over.

What fascinates me is the difficulty of actually teaching many of the crucial skills for these traditional crafts. In the programme on thatching, for example, they were told that a vital part of building a durable, watertight roof is to make sure that the bundles of longstraw (called 'yealms') are properly blended in with one another and the rows offset, so that there are no gaps or channels for the rain to enlarge. That sounds straightforward enough, but to a novice — certainly to me, watching the programme — there appears to be no obvious visible difference between a good and bad piece of thatch. It seems to be something that you learn very gradually by developing a 'feel' for the subtleties of the materials and the ways in which they interact.

Another example in the same programme was the technique for bending hazel rods, sharpened at each end, into 'staples' with which to pin the yealms to the roof. The tutors demonstrated how to bend and twist the rods in the middle, with a flick of the wrist to cause the middle to bend rather than break. However, it was clear from how difficult the participants found this process that there was a definite 'knack' to it, that was very difficult to teach directly. The tutors could model the necessary process and give them guidance about what they should pay attention to, but in the end, they had to learn the precise action required themselves.

I think that many traditional crafts require these kinds of skills that are difficult to teach quickly, which is probably why apprenticeships were traditionally so long. However, as someone who also teaches as part of my living, I can recognise similarities with some of the skills required to be a good scientist. For example, grammatical rules are fairly straightforward to teach, but trying to guide students in how to properly structure their writing, write clear, logical, flowing sentences and so forth is quite difficult to do. You can give tips, point out good and bad examples, and suggest ways in which they can improve, but in the end, they need to develop their own 'feel' for what makes a good piece of scientific writing.