Add a comment

Earlier this week I received an email that asked:

How do you identify which part drives the key change?

I thought I’d reply publicly, since my correspondent might not be the only person who has ever wanted to know about this. And it’s one of those questions that at one level has a simple general answer, and at another opens more complex questions.

So, if we’re going for the one-line answer, I’d say:

Whoever sings the root of the dominant chord that immediately precedes the new key.

I can’t remember who that would be in the arrangement mentioned in the email (Nancy Bergman’s ‘At Last’), but it’s a reasonable generalisation and will probably help render most key changes cleanly.

But if I were planning to stop and think about the question a bit longer, I wouldn’t want to be quite so simplistic about it.

For one thing, whilst there may be one chord that gives you a leg up into the new key, the actual process of changing from one key to the other can vary in both length and route taken. Sometimes arrangements take a whole bar to pivot into a new key, whilst others do a last-minute handbrake turn. (‘Million Dollar Baby’ (arr: Mitchell, Waesche & Hopkins) in the BHS catalogue of published music turns on a dime on the last quaver of the bar, for instance.)

This means that it is not just the voicing of the last chord before the new key that affects the approach, but the lie of the lines in the whole process. Depending on how the arranger has gone about it, one or more parts may have a strong directional feel pulling the music into the new key, and/or a certain part may act as an anchor point around which the others can lever themselves up (or indeed down) into the new tonal world.

This observation in turn opens the question: what does it mean to ‘drive’ the key key change? There are all kinds of interesting assumptions in there about the expressive purpose of changing key. It assumes that it entails a distinct increase in forward energy, a ramping up of excitement. And of course, this is quite often the case – the assumptions in the question are good enough generalisations that there is a one-line answer that is okay in a quick-and-dirty, will work in enough circumstances to be useful kind of way.

But there are also enough exceptions to the ‘driving’ feel of key changes and enough variations in the different roles the parts can take to make them work that the one-liner isn’t enough. Key changes don’t only intensify – sometimes they relax, sometimes they transform. Moreover, if you find that a key change isn’t just ‘singing itself’ but is needing some analytical work to unpick what’s going on, then the chance are it’s because its expressive purpose isn’t simply forward oomph, but some more distinctive emotional journey. So it’s precisely at the point when you feel the need to ask this question that the one-line answer is likely to be least useful.

(Now I think of it, ‘At Last’ has a lift of a whole tone that has something of an ecstatic effect – a feel that the world has changed as a result of the love that has at last been found.)

So, in practical terms, how do you figure out the needs of the particular key change you are working on in the context of the particular voices you have singing it? As so often, the answer is lies in duetting the parts. You need everyone to experience the specific relationship between each pair of parts as they traverse the move into the new key: not just who they themselves are particularly working with at any one moment, but what other musical processes are going on around them. Then you’ll find the key change isn’t just a matter of one part driving (with the others, implicitly, as passengers), but of all parts cooperating to make it happen together.