While watching John Flynn's 1980 Defiance recently, I noted a storytelling pattern that has been honored almost without exception by commercial action films since the dawn of cinema. The duration and intensity of action scenes are generally allowed to vary somewhat over the course of a story, but filmmakers are expected to fashion a big action climax according to certain specifications:
1. The duration of the final scene is expected to be substantial. In most genres, a simple confrontation is not enough: the battle generally is segmented into multiple parts, if for no other reason than to achieve great length without tedium.
2. Whether or not any other action scene in the film has contained much suspense, the final scene generally should drag out a few moments in which the hero is on the brink of extinction, even though the audience usually cannot be expected to doubt a favorable outcome.
3. If at all possible, the final confrontation should come down to a hand-to-hand battle between the chief hero and the chief villain, no matter how military or large-scale the offensive.
A week prior to watching the Flynn film, I noted the same three elements in the climax of Hugo Fregonese's 1953 Blowing Wild, a considerably better film than Defiance. I also recall mentioning this pattern in a review of William Friedkin's 2003 The Hunted, a strong film made from an unambitious script. I name these few examples off the top of my head; I trust that the reader will acknowledge the dominance of this template, which I will call "the rule of escalating action."
The problems with the rule of escalating action are obvious. One can perhaps argue that it enforces a modicum of good dramatic practice; but too often the items on this laundry list are in conflict with the needs of the movie or with common sense. And, of course, any narrative structure that becomes a rule, however sound, is an obstacle to surprise and invention. Nonetheless, the pattern is going strong after a century, and probably precedes cinema in some form. I don't believe that it is merely a habit that has been retained out of commercial superstition: it's too old and too powerful to be an unmotivated sign.
There's an underlying principle that sheds light on this phenomenon. Fiction can always be considered on two levels: internally, according to the needs of the world being depicted and of the people who inhabit it; and externally, in terms of the audience's reactions, which are crafted according to laws of drama. With many issues of fiction - not just the rule of escalating action - we can observe that the prevailing approach, followed slavishly by conventional works and substantially even by most adventurous works, involves harmonizing the internal level of the fiction, by force if necessary, with a known and desired pattern on external level.
The implication of this convention is that a well-made film would be designed so that internal and external logic are worked out at the same time with the same gestures to generate the standard action climax in an organic fashion: no mean feat, but a valid goal. And the rule of escalating action, which becomes bothersome when this perfect structure cannot be achieved, is the result of a kind of automatism, a need to impose a default dramatic shape regardless of where the internal needs of the film universe might take the story.
(For another issue of fiction that involves subordinating the internal level to the external, look in the middle of this 1984 article I wrote for the L.A. Reader, where I discuss the rules governing audience mourning for the death of characters with different levels of billing.)
Obviously the rule of escalating action is a matter of statistics: some people are irked when a well-known, conventionally fleshed-out dramatic shape trumps internal logic; but the paradigm has enough support to flourish over the long haul. It's not surprising that hand-to-hand conflict should have an appealing symbolic clarity, nor that we should enjoy the same dramatic flow in a movie that we like in a sports event. And it would be too hasty to conclude that the internal existence of the film universe carries little weight: audiences are notorious for docking films when they perceive internal conflicts, even minor ones. It's easy to imagine many viewers finding the climactic Tommy Lee Jones-Benicio del Toro knife fight in The Hunted silly, and at the same time not really wishing for a more plausible but unconventional ending.
I'm willing to speculate only that there seems to be great comfort for many viewers in this kind of canonical dramatic structure - a comfort that is increased by, but is not entirely due to, its historical repetition and familiarity.