Whilst I write up all my full-day or full-evening coaching sessions and workshops here (for the combined purpose of reflecting on them for myself and the enesembles, and for sharing what we learned), I don't always write up shorter sessions. An hour by Skype has a different rhythm to it from a 2-hour+ intensive. It tends to be more about sorting out details and consolidating partly-grasped areas of development than breaking new ground.

But sometimes one of these sessions will throw up something that is really asking to be written about, either for the practical techniques involved, or for what it can teach us about how people think musically. Or, in this case, both.

Reasonably often, I get emails from people who have just started a new barbershop quartet (or, less frequently, chorus), asking for advice on finding music to sing. So I'm writing this so I can do a thorough reply which I can send out repeatedly, rather than writing a new sketchy reply to each new request.

So, the first thing to say is, if you wanted someone to say, 'Here, you should sing this, this and this,' you are asking the wrong person. I just don't store large lists of songs in my head like some people do, I have research skills instead. But I'm not going to spend hours doing song research for you, since you could do that yourself and cut out the middle man.

There is a rehearsal technique that emerged during the early months of Magenta’s existence that we have continued to use because it is rather effective, and I have been finding it useful to reflect on why it works. Its primary purpose is for memory work - for getting a piece that is basically learned off the page and securely into our brains - but it seems to have all kinds of benign unintended consequences.

The technique works like this. We sing the whole piece through with the music two or three times in succession (two if we’ve already run through it during the rehearsal, three if we’ve only looked at patches so far) very quietly and very fast, then drop the music and sing it at the correct volume and tempo from memory.

The pertinent elements of the technique are, I think:

That's a very poncy-sounding title isn't it? It's a classic example of starting off with a simple, practical question, and discovering that miscellaneous bits of cultural theory lodged in my brain from past research projects are actually quite helpful in thinking through the answer. The title only comes later when it's time to write it up...

So, the question that started this all off is: how can we, as performers, remember to do all the stuff in performance that we have prepared in rehearsal? There are all kinds of things that an ensemble will have considered in their performance preparation, and that the members 'know' to do, but you find yourself half-way through a song and realise that you're not doing something you should be, or are doing something you shouldn't (through ignorance, through weakness, through your own deliberate fault...).

Something that all musicians have to cope with, whether in their individual practice or working with ensembles, is fixing passages that 'always go wrong' (sometimes with the addendum, 'however much we practise them'). There are two issues to deal with here, neither of which can be fixed in isolation:

• the ingrained pattern of actions that routinely pull the music off-piste
• the negative emotional response associated with this moment - what Oliver Burkeman has called the 'psychological flinch', or 'ugh response'.

Comedian Jo Brand used to talk of an agony aunt who had received a letter from a man whose girlfriend considered him inadequately endowed. (Bear with me, this metaphor becomes relevant shortly...um, no pun intended...) She wondered who these women were making such comments - surely they should realise that if you observe that it's small, it just gets smaller...and smaller…

This sequence comes to mind whenever I'm working with singers and someone gives the remark that something sounds tentative or lacking in confidence. They may be right (they usually are), but it is the kind of observation that will elicit exactly the opposite response than the one needed.

A reasonably common conversation I have with directors when working with them on their technique is to suggest that they could usefully stop mouthing the words to the music they are conducting. They very rarely ask why (it is generally known to be a good idea), but they do object that it is very difficult. Well, I’m not going to argue with that.

But it’s probably worthwhile reflecting both on why people find it hard to stop doing this, and why they can become better directors if they do. It’s not so much that it’s a bad thing (though it can introduce specific technical flaws), but that it limits what you can achieve with your singers.

I have been thinking recently about the instructions that teachers and coaches give to refine what their students are doing, and how often you see what are essentially corrective instructions getting muddled up with how-to-do-things instructions.

Commonly remarked-upon examples include:

• Tuning of major 3rds: sometimes (well, quite often), people sing major 3rds a bit flat, and so are asked to raise them a bit to bring them in tune. This gets translated into a mistaken belief that major 3rds need to be sung really high, whereas in fact the justly tuned 3rd is slightly lower than the equal tempered one.
• Posture for singing: some people tend to slump a bit forward and collapse the chest as a matter of habit in their posture, and so are asked to raise their chests for a good singing posture. This gets translated into a general instruction to ‘raise the chest’ which, for the people who weren’t particularly slumped can result in their distorting their posture, narrowing the back and adding all kinds of extra bits of unnecessary tension.