Teaching Literature and Information Resolution

Recently there has been a lot of conversation about the value of the humanities. It seems that increasingly people have been questioning the value of disciplines like literature. If, as seems to be the popular perception, higher education exists primarily as institutions for job training and credentialing, what is the point of teaching literature? Reading Flaubert doesn’t train any job skills or increase anyone’s employability. And as the cost of higher education goes up, these seemingly irrelevant courses begin to just look like a frivolous expense ripe for cutting, at least from general education requirements. And, of course, the study of literature doesn’t generally lend itself well to machine-graded tests and large lectures, meaning teaching literature is largely unaffected by the efficiency gains technology might bring to higher ed, making its cost relative to other courses seem higher. So why teach literature, much less require a couple of literature courses of every student, as most universities now do?

Well, I tend to think that the study of literature is inherently valuable (although I admit I’m a bit biased there). The ability to analyze a text is a critical skill that is immediately transferable and applicable to many areas of life: a law, company memo, conversation with a boss or TV commercial can be picked apart and analyzed for meaning in the same way literature can. Indeed, the move to emphasize job training over the humanities and the analytical skills they endow is a little creepy to me; an attempt to turn higher education into an apparatus that produces workers with the skills to be productive but none of the analytical capacity to question. And the ability to write, which students learn in humanities courses, is a crucial communications skill that anyone with even a modicum of responsibility in the corporate world must have.

But those are the arguments already out there, and many people remain unconvinced. So I’d like to throw another idea about why literature is valuable into the ring. Let me preface this by saying, firstly, these are just my own musings. I am not an educational theorist, I’ve just taught a few literature classes. Secondly, I’m approaching this problem from the perspective of regional literature (or foreign literature) specialization, so that probably colors my perception.

That being said, I think literature is valuable because it bridges a major information resolution problem we have in understanding the world around us. And that statement requires some explanation. Say you have a photograph of something that is extremely low resolution, only one pixel, for example:

 1px

Well, we can tell this picture is… greyish. But other than that, the resolution is way too low to have any helpful information about what is being depicted. If you up the resolution several times, to nine pixels, we get:

9px

Despite have nine times as much information available, we still don’t have enough to get any meaningful idea about what is going on. Now we just know that there’s grey, and maybe some brown too. OK, so what if we include 20 times as much information, bumping it up to 196 pixels:

196px

Suddenly this is looking a lot more promising. We can see there’s something brownish and diagonal on a grey background. That’s a lot better than we had before. But we still can’t tell what is being depicted, although we might start taking wild guesses at this point. Let’s double the resolution again, to 400 pixels:

400px

Hmm, it’s becoming clearer, there’s definitely some coherent diagonal shape there, but we still can’t see what it is. OK, here’s the image with couple of orders of magnitude more information, at 40,000 pixels:

40000px

Aha! At this level of information resolution, it’s clear we’re looking at a praying mantis on a stucco wall.

I propose that, in our study and understanding of the world around us, we want to be at that last, clear, high information density resolution, but often we are at 400 pixels or 196 pixels or less.

If we were to paint a picture of the human species, a one pixel image wouldn’t be very helpful.

grid1px

Like the one pixel version of the bug on the wall, this picture would just blur all of humanity together and get a great big average. It might be helpful for an alien anthropologist trying to get its head around our strange species, but it’s not helpful for us. We know that people in different regions, for example, have very different cultures and perspectives. So what we do in universities is divide the world up into a few very broad regions and teach students about our world accordingly. This might be our nine pixel image:

grid9px

Different universities will have different arrangements, but usually there’s a department of East Asian Studies, Middle Eastern Studies, South American Studies, American Studies, a few European departments, etc. And often we teach classes accordingly; “Introduction to East Asian Religion,” for example. But as we saw in the bug picture, this isn’t a very helpful resolution. East Asia, for example usually refers to Japan, China, North Korea and South Korea. These four countries (along with Taiwan) have very different cultures, histories, politics, literary cannons, etc. Nonetheless, because of the realities of funding, university organization and other factors, this is often how we teach students about the world. Even if we break our fields of study/pedagogy into individual countries, this still only gives us the 196 pixel image:

grid196px

Which is a lot better, but still not great. After all, even if we’re restricting our focus to only “Japan,” we know there is great variance in culture, outlook, social structure, etc. between different regions, different generations, different subcultures and so on. To supply a visual metaphor, imagine that one single pixel in the 196 pixel image of the bug represents a single country. But we know there is a lot more information available there at a higher resolution; the single pixel is a very blurry average of all the information that is underneath:

196px-over

So we know teaching at this level is deeply flawed, but unfortunately that is how we must teach about most foreign cultures, with perhaps a few brief mentions of regional differences to pretend that we have acknowledged the complexity we are drastically reducing. We know that a teenage girl in Tokyo is vastly different from a war veteran in Okinawa, but out of necessity we pretend that we can encompass them all with the same essential characterization. In research, we tend to be a bit more specific: we study the culture of a particular period or region, or break up populations in to smaller and more specific chunks for social science studies. This might give us the 400 pixel image:

grid400px

Perhaps a little better, but as we saw above, this is still a gross approximation. At 400 pixels we couldn’t even see that the image was a picture of an insect. Even if we limit the focus of our study to, say, Hispanic women over the age of 55 living in the Pacific Northwest, that’s still a group of (probably, I’m just guessing) hundreds of thousands of people. Any two of those people might be completely different in how they view the world, vote, raise their children, read books, participate in popular culture, etc. You can narrow your focus more, but you’ll still be dealing with large groups and, consequently, low resolution. We know, of course, that the level of information resolution we really want to be at is this:

grid40000px

Full resolution, which in this case means the level of the individual. We know that within any given group of people, no matter how small, there will be great variety among individuals. Therefore any characterization that groups individuals, no matter how finely, is a low-resolution essentialization. We would very much like to be able to study and anticipate the world at full information resolution, at the level of the individual.

Unfortunately, this is impossible for research and teaching. The whole point of researching or teaching about the world is to increase our own or students’ understanding of it. For that, it is necessary to reduce, synthesize and narrate information. We could teach a class that related the worldview of however many individuals we could fit into a semester, maybe 100, but that would just be 100 unrelated data points. Such an endeavor would increase our understanding of those particular individuals, but not the larger world. For that, we must synthesize information to a lower resolution level. We might say instead that the 100 people living in some village generally think in such-and-such a way, view the world in such-and-such a way, form relationships in such-and-such a way. Of course, by doing so we blur together all the individual variance we saw at the individual level, but if we don’t do so we never get any high-order comprehension.

This is the challenge Big Data has cracked: Big Data can cull information from people at the individual level and aggregate it. However, Big Data can both get input from the highest information-resolution (individual) layer, gathering information from people’s social media posts and web searches, and output back to that same highest resolution layer in the form of highly targeted advertisements. Scholarship, on the other hand, can adopt (and in some cases has adopted) the techniques of Big Data to garner very fine input from the individual layer, but must output to lower resolution layers: in other words, we can gather very accurate information from every individual (rather than relying on sample polling), but we must synthesize that data in order to say something about a certain subset of Pennsylvanians, Pennsylvanians as a whole, Americans, etc.

bigdatavsscholar

Big data, of course, doesn’t have to bother creating a high-level narrative understanding of the data it collects, it merely needs to use it to create effective advertisements and target them precisely. Scholarship, on the other hand, in the form of research and teaching, exists only to create a high-level understanding, and therefore can necessarily never output to the high resolution layer. In other words, we can never output (teach about, produce scholarship) at the level of the individual, even though we know that is the most accurate and honest level of information density. We must always output to a low-resolution layer, even though we know describing things this way is highly flawed. And frankly, the results of doing this are not too bad. Even though we know we are producing generalizations, those generalizations can still be very helpful with, for example, functioning in a foreign country. It works… but not well. We teach students about “Japanese culture,” even knowing that most individuals are not described by the generalizations we teach. We may teach a “Japanese culture” that describes our teenage girl in Tokyo or our war veteran in Okinawa, but not both. Usually we teach a blurry average that hardly describes anyone real at all.

This is the problem that literature offers a solution to. Because literature is, with very few exceptions, relentlessly individual. Narratives focus on one person or a few people and delve deeply into their emotions, psychology, worldview, preconceptions, etc. Yet at the same time literature is a product that is consumed at a mass level. In other words, narratives are about a story in the high resolution layer, but are consumed at the level of the low-resolution layers scholarship must operate at. It gives us information on how a high resolution (individual) experience is appreciated at a low-resolution (mass) level, and consequently allows us to fill in a lot of the complexity that we gloss over by teaching generalizations.

How this works in practice: students in Japanese Studies learn, in Japanese anthropology, sociology or language courses, that Japanese culture is hierarchical and is characterized by greater respect for and deference to authority than Western cultures. But when they take Modern Japanese Literature, they read the enduringly popular Botchan by Natsume Sōseki, a novel about a brash young man that thumbs his nose at any and all authority and crashes through the social hierarchy of a small town like a bull in a china shop.  Students are often confused by this book, but that’s the point. Here is a story about an individual who is the polar opposite of the Japanese culture that students have learned about in their anthropology, sociology or language courses (at low resolution), yet it is probably the best-selling Japanese novel of all time. It is also taught in schools, and is therefore functionally state approved.

Of course, that doesn’t negate the things students learn in other courses. Literature does not create its own high level, low resolution generalizations. It would be wrong to say “the sociologists are incorrect, because Botchan shows Japanese are really individualistic at heart,” therefore just replacing one generalization about “Japanese people” with another. After all, some people might be reading Botchan with disgust, as a negative example! The things students learn in other courses aren’t wrong, it’s just that reality is a lot more complicated. And what literature does is exactly complicate those low resolution generalizations. It is a non-totalizing perspective on the totalizing tendencies of low-resolution scholarship. It allows back into our understanding the complexity that individuals bring to a culture or society: some will align themselves with sociocultural norms, some will oppose them, and some will do something completely unexpected. It allows us to teach an interpolated high resolution image of the world, and since it is impossible for us to teach (output) a true high resolution (individual level) image of the world, literature gives us our best chance to approach describing the world accurately.

 interpolatedinterpolatedbug

Interpolated high-resolution image (low resolution scholarship filled in by literature)

As I noted at the outset, this perspective is highly influenced by the fact that I am in a “foreign” and non-Western field. Nevertheless, I think this theory has broader applicability. Courses in history, for example, tend to (necessarily) teach a totalizing, essentializing view of whatever era they are concerned with, but reading the literature of that period can allow the complexity of the contemporary individuals to filter back into the consciousness of students.  Studies of different regions or ethnic groups within one’s own country benefit from literature in the same way.

Therefore if we concede at all that the purpose of higher education is to give students some greater understanding of the world, literature is an indispensable part of that education. Without it students are taught a fuzzy, low resolution image of the world, with entire nations or eras swallowed up in totalizing generalizations. With literature, students are exposed to a hint of the complexity of a society that consists of millions of individuals each with their own perspective and each pulling in their own direction. With literature, students get an understanding of the world that is closer to the highest density layer of resolution that is reality.

Related reading

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s