Netflix and Algorithmic Canons
22 Nov 2014
After I experienced an interesting serendipity yesterday, I’m starting to think that we should view Netflix as a new form of cultural canon. Let me explain.
Two of my favorite websites, the webcomic Pain Train and The Toast, both published, in posts I linked to, jokes (and essentially the same joke) on the Peter-Falk-starring detective show, Columbo, on the same day. Now, as someone who’s defended Murder, She Wrote a lot, recently1, I’m always happy to see people championing middlebrow TV detective shows, but I suspect something more is going on here.
As someone who reads a lot of culture blogs, I’ve noticed recently that waves of think pieces and witty reminiscences on the same movies move in waves through a lot of these sites (places like Vulture, Pajiba, and Gawker). Further, as with the Columbo stuff yesterday, these waves seem to be tied to the availability of specific content on Netflix.
As has been trotted out a lot during debates over net neutrality, 30% of all North American internet traffic during primetime is Netflix videos. A lot of people watch Netflix and a lot of people watch a lot of Netflix. A related common complaint about Netflix, despite this high degree of viewership, is that the service often doesn’t have a lot of content that people want to watch.
You can imagine where this is going: people get bored and start watching Columbo because it’s available. Suddenly people begin to notice that Columbo episodes are really predictable and that Peter Falk’s character is really idiosyncratic. Then you have web content on Columbo, and, 38 years after it left the air, everyone’s (relatively) talking about Columbo (or whatever, the same thing happened when Netflix added Heathers to Instant).
This patten of cultural uptake and cultural conversation is known as “canon formation” in literary studies, and it’s the process by which works are selected from the sum total of all possible cultural works and hailed as “important” or “significant” or “great.” There are a whole host of problems with this process that I won’t get into here as well as a lot of interesting work being done on shifting values and the canon,2 but what interests me here is that, in a way, Netflix Instant constitutes a new popular canon.
The main thing that differentiates this new canon from an older cultural canon is that older canons were, nominally, determined by expertise (primarily through the editorship of literary anthologies and the subsequent shaping of literature survey syllabuses). Netflix Instant is a canon determined by the vagaries of both content-licensing agreements and recommendation algorithms. So this seems to be of a piece with the general death of expertise in our society, but I think it’s an interesting formation of the rise of an algorithmic approach to cultural curation.
Oh, and just one more thing: do we want to replace our notion of “great” or “classic” film and television with what is made available through the shifting whims of contractual law?
I’m only half joking about doing a media archeology of the show. It’s actually an awesome and insightful history of the adoption of word processors and other digital technology into our culture in the late 1980s and 1990s. ↩
Have any of you ever read any Thomas Wolfe? the author William Faulkner declared as America’s best? No? ↩