I've been thinking about the "moderates vs. extremists" issue some more, and trying to punch as many holes as possible into my provisional definition of extremism: Within any group, the extremists are those who sacrifice principles held by the majority of the group in order to advance a single set of favored beliefs that the larger group considers of equal or lesser importance.
The two biggest problems that I've come up with so far are:
1) The difference between perception and reality. This is where you get into issues of people being seen as extreme who are actually fairly moderate and vice versa. This is actually a much larger issue than it appears to be on the surface. In order to really be able to distinguish the extremists from the moderates, you need to decide what question you're asking: are you talking about the people who really are extremists, or just those who the group in question views as extremists? Again, I'll bring in the issue of industrial regulation as an example. If massive deregulation is stated solely in terms of economic growth and the free market, then, by the above definition, it is a moderate position (as the majority of the group likes economic growth and the free market). However, if it is stated in terms of health, safety, and environmental impact as well as growth and the free market, then it can be seen as an extremist position (because the majority of the group may not like economic growth if it means dumping carcinogens into the water supply).
Then, of course, you have to address the question of objectivity and to what degree it is possible to achieve it. After all, depending on the level of objectivity it is possible to achieve, there is no real definition of an extremist; it's all a matter of point of view.
2) The essential uselessness of the conventional representation of the entire political spectrum as a continuum between anarchy/communism and totalitarianism/free market fundamentalism (or more commonly, communism vs. fascism) in determining moderation vs. extremism.
As number one is a rather complex issue (again, depending on what question you're asking), and raises a thesis-level philosophical question ("to what degree is objectivity possible?"), I choose to ignore it for now.
Number two, on the other hand, is something that's bothered me for awhile.
Setting aside the ridiculous nature of defining every conceivable political position based on where people sat in the French National Assembly two hundred years ago, dividing the political spectrum into "Left wing" and "Right wing" provides a seriously flawed view of any given person's political beliefs. It also presents a serious obstacle to identifying extremists.
To understand just how skewed a picture the Left-Right continuum gives of any particular person's beliefs, it is helpful to look at basic geometry.
Suppose you have two three-dimensional objects, a sphere and a cylinder. As long as they are viewed in three dimensions, the difference between them is obvious.
However, if one of the dimensions is removed, and the shapes are translated into two dimensions (in mathematics, this is called projection), you lose information related to the lost dimension, and the two shapes appear identical.
When the sphere and the square are projected onto plane xy, they can both be represented by circle A. What this means is that, given only the information available in plane xy, it is impossible to tell if the object is really a sphere or a cylinder.
While the information loss in the translation from three dimensions to two is bad, the transition from two to one is even worse.
Without that extra dimension, both triangle t and square s are represented by a single line segment, which comes nowhere near describing each shape.
Again, you can see how two different objects -- t and s -- are indistinguishable when they lose a dimension.
The implications of this for the one-dimensional conception of the political spectrum are obvious. Indeed, when it comes to the conventional method of dividing the spectrum, we are forced to overlay several different value ranges one over the other, the end result being two categories -- Left and Right -- which can, at times, give a seriously flawed picture of an issue.
Take, for instance, a person who supports large tax cuts but also strongly believes in gay rights. Because that person believes in both Right and Left wing positions, they would generally be classified as a centrist, as if gay rights and deregulation somehow canceled each other out.
The Left-Right axis forces everyone to shoehorn themselves onto the line somewhere with only one or two major issues, ignoring complexities and making positions that are very different appear to be the same.
The solution, then, is to add a dimension to the traditional Left-Right axis.
While there are a number of different ways that the Left-Right divide can be split into two dimensions, it seems the most useful way is to separate economic stance from social stance.
Instead of superficially lumping Communism and anarchy together in their opposition to laissez faire Capitalism and totalitarianism, a more useful representation would have two axes, one representing the range between Communism and laissez faire Capitalism, the other representing the range between anarchy and totalitarianism.
This method of dividing up the political spectrum is more accurate than a single Left-Right axis, while not being so complex as to cause confusion. Rather than simply being "Left wing" or "Right wing," a particular issue can in this way be described as belonging to quadrant one, two, three, or four.
This has the benefit of dividing up groups that are often incorrectly lumped together. For instance, libertarians would tend to be classified as quadrant II, because of their dual beliefs in personal freedom and the free market. A lot of non-libertarian Republicans, on the other hand, would be quadrant I, because of their increased emphasis on issues like stricter drug laws and additional powers for law enforcement.
It also removes certain contradictions in the extremes of the Left-Right axis (Anarchic Communism? How exactly does that work?)
The dual axis also correctly separates Stalinism, which belongs in quadrant IV, close to corner d, from what we now call progressivism, which would fall into quadrant III.
All of this is relevant to the question of moderates vs. extremists because it allows us to more easily spot those who tend toward one or two of the four extremes than does the standard Left-Right axis. After all, there is no good reason why someone who supports budget-busting tax cuts should be labeled as a moderate just because they also happen to believe in gay rights. The two positions are not related, and should not be made to look as though they are by an artificial model.
This is actually another one of the key problems with today's Democratic party, one which can be at least partially attributed to the way we see the political spectrum: no matter how corporate-friendly a Democrat is, they will always be considered a moderate because of their stances on social issues. This flawed perception has allowed the Party to move more and more to the Right (or, more properly, to migrate from quadrant III to quadrant II) without much outcry until just the last couple of years.
Could this kind of alternative classification scheme catch on? Probably not. The old scheme is too firmly entrenched in people's minds for there to be a chance of it changing for at least a generation. However, it can be a useful tool, for people from all sides of the spectrum.
And, really, what other endeavor can simultaneously remove the rhetorical burden of Stalinism from liberalism's back and stop "gay Republican" from being a contradiction in terms?
Sidenote: The quadrants are labeled counterclockwise from the upper right because that is how they are labeled in geometry (trig makes a lot more sense that way). Also, the corners are labeled a through d for reference purposes only (to avoid confusion). The letters have no significance and can be safely disregarded. Additionally, the three lines in the third image are labeled 3a, 3b, and 3c because when I made the image, I thought I was going to label the images themselves figure 1, figure 2, figure 3, and figure 4. (you know, "Please refer to fig. 3...") So the line labels refer to the fact that they are each a part of figure 3. However, when it came time to insert them into this post, I decided that labeling the images figure 1, figure 2, etc. was way too pretentious. So I didn't.
Sidenote the Second: For intellectual snobs who dismiss this as first year poli-sci stuff: get over yourselves. I'm just trying to get everyone on the same page here, and come up with a good answer to the moderate vs. extremist question. This seemed like a pretty good way to do it. For the record, I don't claim to have invented this particular scheme; I've never seen it before, but that probably says more about my experience in political science than it does about political science itself. Anyway, the Left in general doesn't seem to be winning hearts and minds with the way things are currently classified, so I figured what the hell.