When is a seven-hour play shorter than a one-hour performance? It's a question of rhythm

How long is a long play? This question, provoked by the opening of Robert Lepage's seven-hour vignette The Seven Streams of the River Ota and Harold Pinter's hour-long marathon Ashes to Ashes, is not one that a stopwatch will help you answer. The longest play I ever saw, for example, was a two-hour production of an existential drama in east Beirut. When the play was over I had to return to west Beirut, through checkpoints manned by heavily-armed men whose attitude to existential drama was likely to be equivocal, at the very best. I felt that every second of stage time added to their potential impatience. It was similar to the anxiety of missing the last bus home, with the added possibility that the conductor might shoot at you from the platform.

By contrast, the five-and-a-half hour production of Hamlet that I saw in Bucharest was a breeze. The naps helped, naturally, but there was something else besides - an air of rapt, unrestive attention in an auditorium so full that people were sitting in the aisles. When the curtain calls had finally ended it became clear what had distilled such patience.

The audience emerged from the theatre into what was a tedious commonplace for them but magically strange for a western visitor - a capital city as dark as a country wood. Here and there a dim gleam shone through thin curtains but there wasn't enough electricity for street lights, let alone neon, and there was nothing in the shopfronts worthy of illumination. Above the dark rooflines of apartment blocks and offices you could see the stars shining, undimmed by the glow-worm light of pocket torches, which people were using to pick their way home. Had the performance been even longer this audience would probably have been grateful; nothing else was waiting for their attention.

Strictly speaking, both of those examples represent trivial cases - or at least non-theatrical influences on our perception of duration. When it comes to internal adjustments of the sense of time, matters become a little more complex. While stage-time isn't clock-time, for example, it nevertheless obeys a kind of rhythm - Harold Pinter's famous pauses wouldn't even be detectable as such unless we were somehow aware that speech had arrived late, behind the beat. Any regular audience becomes trained to certain intervals and the machinery that governs those intervals may be buried very deep.

In his essay The Duration of the Present Moment, the Czech poet Miroslav Holub writes about the psychological finding that the clock of human perception seems to tick about every three seconds. Asked to reproduce an audible signal, most people will unconsciously extend anything that is under two seconds, become notably more accurate with signals between two and three seconds and then start to shorten signals over three seconds. Holub points out that an analysis of formal poetic metre seems to show an obedience to this internal metronome - what he calls a "carrier wave".

There are other suggestions, too: writing about the notorious slowness of Noh drama, William Empson builds his explanation of its effect round another bodily clock, arguing that the accompanying music governs your emotional responses: "A rhythm quicker than the heartbeat," he writes, "is one that you seem to control, or that seems controlled by some person: the apparently vast field of our music is always the frankness of the West, always the individual speaking up. Music based on rhythms slower than the heartbeat can carry a great weight of emotion and even of introspection ... but it remains somehow impersonal."

If theatrical speech obeys the law of the moment, there are larger-scale units that play their part, too - culturally reinforced rhythms of dialogue, scenes, even of incident. And different playwrights will use these rhythms in different ways. Extra duration, for example, offers the writer two opposed possibilities - the chance to get more in or to leave less out. They can, that is, either maintain the convention of artistic compression and use the extra time for more events or they can break the convention of ellipsis altogether and present an event "uncut". Lepage does both, accelerating to include large spans of historical time but occasionally slamming on the brakes to stretch out an important detail - as in the preparations for an assisted suicide. This means that the ratio of stage- time to imagined time is never fixed, something that may well help to diminish the audience's subjective sense of how long they have been there. In the Pinter, by all accounts, the ratio is more consistent - with events on stage seeming to take slightly longer than they would in life. Which may explain why some critics have experienced seven hours as rather shorter than an hour. The wise question, then, is not "how long is it?" but "is it syncopated or just late on the beat?".