Note: This post is mostly for future reference: something I can point students to if this topic comes up in class (as it sometimes does), although comments are of course welcome.
One of the features of our current age is a breakdown in representation. A few decades ago we were comfortable bunching people into various groups and representing them in various ways: “Americans believe such-and-such,” or “Japanese think in such-and-such a way.” Of course, this wasn’t limited to national groupings, we would say “East Coasters are such-and-such,” “women are such-and-such,” “men are such-and-such,” “Californians are such-and-such,” etc. But in our current age we tend to view this totalizing or essentializing impulse with increasing skepticism. After all, there is a huge amount of variation between individuals. Even neighbors or members of the same family have different priorities, belief sets or tastes. How can something that claims to represent a huge group of people, like “Americans” or “Japanese,” possibly be valid? Think of the fairly simple statement “Americans like cheese.” We might recognize that a lot of US cuisine features cheese, or that per capita cheese consumption in the US is higher than some other countries. But at the same time, we realize that not everyone who is American likes cheese, and we each probably even know several cheese haters ourselves. Therefore we have increasingly come to realize that such representative, totalizing statements are invalid, that even small groups are not uniform, and that we’re all defined individually.
However, this creates a problem in cultural studies, particularly foreign area studies. If no representation of another culture or country is valid, how can we possibly study other cultures or countries? If the only valid level of representation is the individual level, are such fields of study inherently flawed? But at the same time, we do recognize that there are important differences between cultures, and we need to learn about those differences, especially if we are planning to live and work in a foreign country.
How do we resolve this? One answer is to say that the perceived differences between cultures are really just differences in social ideals. For example, here’s a generalization I hear a lot when I teach: “Japanese are more deferential to authority than Americans.” Well, there may be some truth to that, but at the same time we know many Americans are plenty deferential to authority, and plenty of Japanese are not. One way to resolve this contradiction is to say that Japanese aren’t necessarily more deferential to authority, but deference to authority is a stronger social ideal in Japanese society. People who defy authority are punished (through social ostracization, lack of promotion at work, etc.) while people who don’t are rewarded, more strongly than they are in the US.
But this answer runs the risk of creating another kind of totalizing representation: that all people are really just the same, and only the social ideals we are forced to conform to are different. That is attractive in many ways; after all, it would mean that weird people in distant states/regions/countries aren’t really different from us, they’re just being forced to behave in ways that are strange to us because of pressure from different social ideals. But such a framework doesn’t allow for any real difference, which is the whole reason we discarded representation to begin with. So how can we acknowledge and study difference without creating one kind of totalizing representation or another?
To introduce a conceptual framework for doing so, I want first want to introduce something called a distribution graph. For those of you who are thinking “I majored in a cultural studies field because I was told there would be no math,” bear with me. We don’t need to actually crunch numbers to adopt this as a conceptual tool. For those of you who already understand distribution graphs, feel free to skip ahead.
A distribution graph is a graph that measures some score on the X-axis, and the number of people in a population with that score on the Y-axis. The classic example of this is test scores in a large lecture class:
This chart shows us that in this class, 20 people scored a 75 on the test, while 1 person got a score of 100 and one person got the lowest score. Falling down from 75, maybe 19 people scored 74, and 19 people scored 76, etc. Overall, this graph shows us that the majority of people in the course got Cs, a healthy portion received Bs or Ds, and a handful scored As and Fs.
This is called a normal distribution, and it is how we (generally) expect any score or attribute in a random population to be distributed: most people will be pretty average, and only a few people will score very low or very high. If we were to chart, say, the physical height of the students in our class, rather than their test scores, the chart should look pretty much the same: some people will be very tall, some will be very short, but the vast majority will be some shade of average height.
(Incidentally, this is what we mean by “curving” grades. Since we expect the distribution curve to look like this, when it does not that tells us that the test was either too easy or too hard, and grades are adjusted accordingly so they fit this normal curve.)
This is a powerful conceptual tool, since it gives us an overall “big picture” view of the class, but at the same time captures individual performance. If I had to represent the entire class above, I might say “the class was average,” which wouldn’t necessarily be incorrect, overall, but would fail to show that some individuals performed very well and some individuals performed very poorly.
Now, let’s apply this conceptual tool to the problem of cultural difference. Suppose, for the sake of argument, we could quantify (in other words, assign a score to) “deference to authority.” Like any characteristic in a random population, we would expect deference to authority to be normally distributed:
Here we see that a few people are exceptionally deferential to authority, a few people are exceptionally defiant of authority, but the vast majority are just somewhat deferential – average – in varying degrees. Now, take a statement like “Japanese are more deferential to authority than Americans.” That might make you think of something like this:
Here, the few Americans who are most deferential to authority are still less deferential than the few Japanese who are the least deferential. But, of course, this doesn’t make sense: as I said above, we may know many individual Japanese who are less deferential to authority than most Americans. I would argue that when we say things like “Japanese are more deferential to authority than Americans,” we’re really saying something like this:
This is a much more plausible picture. We can still see that there is some truth to the statement “Japanese are more deferential to authority than Americans.” After all, your Japanese teachers and sociology professors have been telling you that, so it’s not likely to be completely untrue. But at the same same time, this graph shows there is a huge amount of individual variance. Take a Japanese individual located here:
He or she is far less deferential to authority than the vast majority of Americans, indicated in lighter blue here:
You can see that this chart is a powerful conceptual tool that allows us to recognize broad differences between cultures, but at the same time capture the huge amount of individual variation within those broad trends. So while it is true that representation has broken down (we wouldn’t feel comfortable with a totalizing statement like “Japanese are deferential to authority” or “Americans are defiant of authority”), we can still talk about general cultural differences as long as we recognize that what we’re really talking about are slight shifts in distribution graphs like I’ve shown here. And, it should be noted, those shifts may be very slight indeed. I just made up the graphs above, since there is no way of actually measuring deference to authority in a meaningful way. For all I know the difference between the US and Japan may look more like this:
One final note: This conceptual framework should also show us that broad cultural trends or differences are not a meaningful way to analyze literature. Literature almost always focuses on a single or a few individuals. But, as we’ve seen, any single individual might fall anywhere on this distribution curve. Students are often tempted to ascribe a character’s traits to cultural trends they have learned about: “The protagonist is motivated by a typically Japanese deference to authority,” for example. But that particular protagonist might fall anywhere on the distribution curve, and unless he or she is on the extreme ends, he or she is probably just as differential to authority as a great number of Americans (or people from other countries, for that matter). Characters must be appreciated as individuals, with their own unique psychology, not as mouthpieces for their culture.