The most interesting thing about Dawson City: Frozen Time, by far, is the real-life history that enabled its existence. This is not to slight docu-essayist Bill Morrison, of the very highly-regarded 2002 Decasia (which I haven't seen), among a couple dozen other films that are, as I understand it, mostly rather like this: moving collages that combine archival footage with new musical compositions, equally adept at furthering an argument as creating an emotional state.

That's what I gather, anyway. With nothing to go on but Dawson City, I think I'd like to gather more evidence before declaring Morrison's aesthetic a conclusive success; it's a film that implies the existence of a considerably more satisfying version of itself, one that's often almost identical to the one we got, but without this film's inexplicably wrong-headed finale, which turns this very fluid, glassy avant-garde film into a talking-heads documentary.

Anyway, with all that frontloaded, I suppose I should get around to clarifying what Dawson City is all about, huh? Here's the real-life history I just mentioned: in 1978, the small Yukon town Dawson was the unexpected source of one of the most incredible finds in the history of film archiving. Several hundred reels of motion pictures, on highly flammable and unstable nitrate stock, had been buried as landfill in a swimming pool that had been buried almost fifty years prior, and in the permafrost found at that latitude, had somehow managed to survive more or less intact, save for water damage. Film archives in Canada and the United States set about the task of salvaging these reels as much as could possibly be done, to the tune of 533 reels by the time Morrison enters the picture. Dawson City is, in part, extracted from these reels, including shorts, feature, and news footage, which are here assembled into the primary documents by which Morrison tells us the history of Dawson from the 1898 discovery of gold in the Yukon (the height of Dawson's population and influence, as the territorial capital) up until the discovery of the film footage, and even a few months later.

Here's what Dawson City isn't: a poetic montage of archival imagery, creating an impression of life as it was lived a the dawn of the 20th Century through primarily emotional, not narrative, strategies. That's the film I was expecting; it's also the film that composer Alex Somers (who deserves to be considered an author of the film almost as much as Morrison) seems to have been under the impression he was called into score, because his music is droning, dreamlike, often transporting, sometimes goddamn irritating, and not really at all a good fit for what Dawson City is actually doing. The secret here is that this isn't really about imagery qua imagery at all: it's a chronicle of the years 1898-1924 or so, assembled only from those things which Morrison found in the footage, banishing all the rest of history to oblivion. In that respect, it's quite an interesting thing: more a work of historiography than of filmmaking, something like an exercise in dealing with the fact that we only ever get to "do history" with the bits and pieces of history that have accidentally survived the centuries or decades to come down to us. Frozen Time might be a damnable pun, but it's also a pointed reference to that essential truth about this material: it survives not through artistic merit, nor through sociological utility, but due to sheer fucking happenstance. Far more reels were dumped into the Yukon River than were buried, and they were presumably chosen largely on the basis of the order in which they'd been stacked into piles.

The flipside of this is that Dawson City is, God bless it, a little bit dry. The film contains a seemingly endless supply of onscreen titles, neatly clarifying where in history we are, what's happening in Dawson as it continues its steady, inexorable decline as the Klondike Gold Rush recedes from news to memory to myth, and occasionally what's happening in the world outside of Dawson - but only when there's film to support it (which is why the 1919 Black Sox scandal looms larger than the First World War in this telling of events).

That's a pretty interesting intellectual exercise, to be sure, but ultimately that's all it is. Aesthetically, Dawson City does one thing, and once you've figured out the thing that it's doing, the whole of the two-hour movie starts to feel a bit stretched out to absolutely no purpose whatsoever. Especially once Morrison runs out of frozen footage, and has to start drawing from other, more contemporary sources to get us from the late '20s up to the films' arrival in the archives. Whatever dreamy hold the film has comes to an abrupt, disorienting end, and the mildly dull feeling of watching a litany of dates marches to the forefront, as Dawson City ceases to be a tone poem about historiography and turns into a plain history lesson, of the sort you might encounter on any old PowerPoint slide in any old lecture hall.

I don't mean to belittle the film's strengths: the experience is a largely effective one. It's less transporting and overwhelming than it is intellectually stimulating, and I cannot imagine what force on earth would get me to sit through this a second time; but I am happy I sat through it once. If nothing else, the recovered footage is, itself, worthy of all the attention and hullabaloo; too battered and fragmentary to really "count", maybe, but the scars of the decades in the dirt and ice transform the footage into something more, or at least different, than other 1910s and '20s footage. It is a tribute to the film medium itself, and even if I can't honestly argue that we needed 120 straight minutes of such a tribute, I'm certainly happy to see this analogue survivor muscle its way into a digital world.