Because of the thoroughly enjoyable production of The Sorcerer which ate my life last week I’ve been thinking about the dramatic arts a lot. As I have said before I tend to think about things too much, and directing is one of those things.
It is common wisdom in our Gilbert and Sullivan society that two artistic directors are better than one, and I’d like to discuss why I believe that is.
There are, I believe, two main reasons. The first reason is a simple matter of editorial control; if I’m directing alone and I come up with a terrible and impractical idea (‘We’ll have everyone come on stage topless!’) there is no one who has the authority to tell me ‘Mark, that’s an awful idea.’ Hopefully such ideas are rare cases, but it’s very useful that someone else has the chance to veto any terrible ideas. In an equal two-person partnership this veto power obviously works both ways.
The second, and more interesting reason why directing teams of two or more is advantageous is the increase in incidence of interesting ideas. We can model the process of directing a show as a list of decisions for each moment (whether line of dialogue, or beat of music) and each element (lighting, choreography, scenery) in the show. Each such element-moment needs a single decision – should this scene be lit with a subdued blue, or a full white? Should the chorus do a box step, or stand still?
We can also observe that each of those decisions can have some measure of quality associated with it; some decisions are great decisions – one of those ‘once-in-every-show’ ideas that everyone agrees is majestically hilarious of heart-breakingly moving – some are terrible ideas – those that have cast members threatening to leave if they aren’t changed – and most are in the middle. These are of course subjective measures of quality and I’ll deal with that issue in a bit.
One could do an experiment with a pair of directors; lock them both in separate rooms with a pad of paper, and a copy of the script and vocal score of the show and ask them to come out when they have a complete list of their directions for the show. Imagine for a second that we have an oracle who can tell us completely reliably and objectively how good an idea is, and so using that oracle we can construct a graph of the quality of direction against progress through the show.
Each director will have a number of peaks and troughs along that graph as they have good, and then not so good, ideas. When they are brought together and are asked to come up with a single directing plan for the show then, assuming they can recognise good ideas and bad ideas accurately enough, the graph of the result should be the upper bound of the two individual graphs. In other words they compare their decisions for each element-moment and select the better of the two. As long as they have a different series of peaks and troughs along their graphs then the average quality of the resulting show will be demonstrably improved.
I said I’d return to the subjectiveness question, and it allows us to bring together the two theories posited above. The quality of a directorial decision cannot be easily calculated by any algorithm. It depends on personal taste, familiarity with the genre, demographic background, and so on. This is what we mean by subjective – the measure of quality depends on the person making the measurement. As a corollary to this, I suggest that a person’s estimation of the quality of their own ideas is likely to be both higher and more static than that of other people of those ideas.
In any case, there will be a theoretical underlying objective measure of an idea’s quality (even if it’s not in prectice discoverable). All else being equal we should expect that the more people are asked to (subjectively) judge the quality of an idea, the closer it will approach the true underlying (objective) quality. Therefore having more than one person involved in the directorial decisions both increases the incidence rate of good ideas, but also helps to more accurately identify those good ideas and discard the bad ideas.
If I get time I’ll add some pretty graphs to better illustrate what I mean, but that time is not now unfortunately.