I've been thinking about canon formation a lot lately and the publication of the JIL last week has of course raised the anxiety level that I feel about the canon to higher levels. I'm taking the meaning of canon to be broad, perhaps even so broad that what I'm talking about constitutes a field, but because I also see fields as being largely constructed of texts, and acknowledge that some texts are more central to a field, I'll use the idea of canon.
So for me, the canon is the texts that you start with when embarking upon the study of a particular field. You can carve it up into larger and smaller bits, but in some ways, it is easy to see how some texts should be tackled first, while others might be read after one has an understanding of the period.
And let's face it, the canon makes it easier to put together a syllabus the first time you teach a course, because you've at least got some idea of what kinds of texts the students should encounter in your course if they're going to go on to take other courses. (This mostly applies to introductory level courses, but could be applicable at higher levels as well)
Of course, canon formation is fraught with problems, particularly because you can't include everything, and sometimes the things that get included, or even labelled as important, are there for reasons that one might (and should!) question.
So there are serious downsides to canon (or even field) formation. One of my greatest regrets in not pushing to stay in medieval studies is the loss of comfort in a fairly clearly defined canon or field. After all, the primary texts are all there (even if some have yet to be unearthed or discussed) and no more will be produced. It is also a well and clearly established field.
Now before you jump on me, I realize this is an overly simplified explanation of medieval studies and it does not do justice to the vigour and intellect in debates about what constitutes the field.
But what I would point out is that in comparison with the literature of the twentieth century, it is far easier to talk about "medieval literature" as some kind of coherent whole, even as its parts are lumped together under that label than it is to talk about the twentieth century in some ways.
For example, where does the twentieth century end? You could define the field chronologically (as many fields are) and say that the canon could include anything written between 1901 and 2000 (or 1900 and 1999 I suppose). But there's already some debate about the first decade or so of the century as to whether it is more closely related to the Victorian period in the previous century, or to the modernists in the later few decades. So where exactly do you draw the line?
One could legitimately draw it at 1901 at the beginning of the century since that coincides with Victoria's death, but that is more relevant for British twentieth century literature than American or Canadian, or Australian, or,... well, you get the idea. The first World War might be an appropriate place to begin since it involved a radical change in outlook on life, society, nationalism, technology etc. and involved many people from different parts of the globe.
You also then need to figure out where to draw the end of the century, and that's where I think things could get difficult. Certainly Y2K is a candidate and fits very nicely with the calendar, but the hype about Y2K seems in hindsight to be such a joke, that I think I might be reluctant to base the definition of the field I work in as being bookended on at least one end by a joke.
Of course part of the problem in figuring out the line that one might draw between this century and the next (assuming that even drawing lines roughly chronologically is something that we want to do) is that we have so little of the next century to work with. After all, there's only been a decade on the calendar of the twenty-first century, and I'm sure in 1909 it would've been pretty difficult to predict even two decades into the future, let alone further. So the end of the twentieth century could conceivably remain in flux for decades, barring something global and monumental enough to change prevailing ideas or attitudes toward one or more aspects of social interaction or development.
(That's just one of those really long qualified ways of saying, "I ain't got no idea")
Though I do have an idea. I've been thinking about 1989 as the actual endpoint for the twentieth century as a period. My sense generally is that there is a radical difference between the 1980s and the 1990s. Even as I say that, I realize that part of that sense of difference may simply be a by-product of my own life at that point - for most of the eighties I was a teenager while by the nineties, I was a parent and a working schlepp - but I think we can also point to some interesting changes in culture that were reflected in literature.
First, 1989 was the year the Berlin wall came down, which marked an end to a decades-old era and a real hope for the future, not just of Germany, but for world peace and all those good kinds of things. Although AIDS emerged in the early 80s, it really only caught the public's attention with a series of celebrity deaths in the early nineties. The 1990s also marked changes in the space race and consumer technologies with the personal computer becoming affordable.
In literature, it seems at once a little harder and a little easier to characterize a shift between the 1980s and the 1990s. Part of the difficulty in seeing difference is the proximity of the problem - it is always easier to categorize a literary movement from afar than from up close.* Whether this is a problem with categorization or just a reality is hard to say. But literary postmodernism, and cyberpunk as a genre, seem to have mostly died around this time, and if we think of cyberpunk not as a species of science fiction, but as the ultimate expression of a kind of postmodern sensibility in literature, then there seems to be a shift between the eighties and nineties.
I'm hesitant to try to define that shift or articulate what I see going on, partly because I think I am too close to it to see it clearly, partly because what came after will undoubtedly only become clearer as we get further and further into it, and partly because I think my personal history and transition from childhood to adulthood is caught up in much of my general sense of it.
But if you had to pin me down to say "where would you draw the line between a course in twentieth-century literature and twenty-first-century (contemporary) literature? I would say that 1989/1990 would be the place to put that boundary.
*The prognostications of writers in the late twentieth century about metafiction and experimental fiction forms, or even about the death of the physical book were made over and over and yet looking back, many have not come to pass, leading me to suggest that when you're too close to it, it's hard to make predictions about trends.
The other thing that's halfway is my job letters. Yes, I'm almost halfway through my application list. 
