D. W. Griffith was a product of the stage, an actor and a failed playwright. But he was also a product of the Biograph years, where he honed his craft as a filmmaker in short self-contained stories, which often have an anecdotal quality. In his feature work Griffith frequently used stageplays written by others as source material -- he never mastered this formal discipline in stories he wrote by himself . . . but this is a crucial failing only if you think movies need to be tightly-plotted narratives with an overall structure which the individual scenes all serve and to which they are subordinated.
This ideal became the Hollywood norm, but Griffith was at his best when he didn't follow and didn't need to follow it -- which is why his later films, when he was trying to fit in to the standardized studio style, are so inferior to his earlier work.
The Birth Of A Nation is a rambling, disjointed film narratively -- more like a collection of tales than a unified story in its own right -- and is least satisfying when it narrows its focus in its final episodes to the melodramatic mechanics of its theatrical source. Only the pure cinematic beauty and power of the Clan ride redeems it from this reductive derailment of its epic expansiveness.
Intolerance of course takes this narrative expansiveness to wild extremes, but even Griffith's great small films, like Broken Blossoms and True Heart Susie, have an anecdotal quality. There may be a heart-stopping final action climax or melodramatic denouement, but the films as a whole don't build towards it with the kind of precision and economy and momentum we have come to expect from popular movies since the onset of the studio era.
This is a criticism one could also level at Huckleberry Finn, which lies somewhere between the delightful, rambling yarn-spinning of Life On the Mississippi and the tauter formula fiction of Tom Sawyer. It is a criticism one could level at The Odyssey, too -- and the Bible. All of these works use narrative formulas, with a more or less developed overall structure, but proceed episodically, like a series of related tales told by the fire over the course of many evenings.
One can see why the studios resisted this sort of storytelling in movies. It's too hard to predict in advance how movies made this way are going to turn out -- they depend too much on the instincts and the genius of the storyteller and they lend themselves too much to improvisation. Griffith's style of anecdotal epic was still fresh in the mind when one of his truest disciples, Eric Von Stroheim, tried to emulate it in darker tones in Greed. From his perspective, the experiment of Greed probably didn't look that outrageous -- Griffith's method had, after all, led to astonishing success both critically and commercially. Greed was longer and grimmer, but followed the same loose-knit narrative strategy.
Thalberg, a corporate functionary with taste, but a corporate functionary first and last, really had to destroy the film -- not just as a warning to profligate directors but as a signal that the days of Griffith's method were over. Enter Rupert Julian and the era of the sensibly-made, pre-visualizable film. That era produced its own kind of treasures, but I think one of the reasons we are attracted to the silent era is because it was the last time the ancient voice of the storyteller could be heard it all its eccentric, iconoclastic, unclassifiable glory.
Its echoes took a long time to die out. It was last heard clearly, I think, in The Godfather, Part II, with its parallel storylines that reflected each other elliptically and suggestively rather than according to some formal narrative dialectic. It's a messy film, on one level, but unified by the passion and conviction of the storyteller's voice -- and the same is true of Griffith's messy masterpieces.