stimulant

I was re-reading Alfie Kohn’s essay, What Does It Mean to Be Well-Educated? earlier today, after recommending it to an acquaintance, and something clicked:

Is it even possible to agree on a single definition of what every high school student should know or be able to do in order to be considered well-educated? Is such a definition expected to remain invariant across cultures (with a single standard for the U.S. and Somalia, for example), or even across subcultures (South-Central Los Angeles and Scarsdale; a Louisiana fishing community, the upper East side of Manhattan, and Pennsylvania Dutch country)? […] Some criteria are more defensible than others. Nevertheless, we have to acknowledge a striking absence of consensus about what the term ought to mean. Furthermore, any consensus that does develop is ineluctably rooted in time and place. It is misleading and even dangerous to justify our own pedagogical values by pretending they are grounded in some objective, transcendent Truth, as though the quality of being well-educated is a Platonic form waiting to be discovered. […] Who gets to decide what it means to be well-educated? Even assuming that you and I agree to include one criterion and exclude another, that doesn’t mean our definition should be imposed with the force of law – taking the form, for example, of requirements for a high school diploma. There are other considerations […]

When I first read this article, I confess I didn’t think very carefully about this question: I thought that raising the issue of who should be making decisions was an obvious problem (i.e. it is a very easy question to raise) to which I had no clean, elegant, answer¹.

I began considering the standardization movement as a political rather than pedagogical question and out popped the following claim: handling educational details and policies at the national level is undemocratic and imprudent.

The basic idea — and the one I’m most interested in discussing — is that scale is an essential element of democracy. Democracy isn’t simply giving problem-solving a public forum with the alleged opportunity for citizen input.² The scale at which you confront a systemic problem (e.g. education) should be closely coupled to the minimum scale that provides for independent solutions. What are those? I’ll get to that.

The basic assumption behind all this is that given a system, a problem, and a solution, there is a natural scale for implementing that solution. For instance, imagine bike theft were a big problem in cities nationwide. And assume that most stolen bikes stay within, say, fifty miles of their theft after being stolen. It wouldn’t make any sense to implement a national bike registry and anti-theft program, because the natural scale of the problem is a fifty-mile radius. On the other hand, drugs are trafficked across city, state, and country lines, and requires an infrastructure for sharing information across the systems at all those scales. Now, a national bike registry wouldn’t be so bad — that is, there isn’t a terrible overhead to its implementation (particularly when compared with a collection of city/state registreies). But with education, there are significant downsides to national standards from both an organizational and a pedagogical standpoint.

So to return to my first statement of this idea:

The scale at which you confront a systemic problem should be closely coupled to the minimum scale that provides for independent solutions.

What do I mean by “independent solutions?” Given a scale, you naturally partition a system. A nation into states into districts into cities into neighborhoods into blocks into houses. At each different scale, the granularity you define gives you a handful of subsystems (all the states, all the districts, … you get the idea). When you implement a solution at a particular scale, the question of whether it is independent asks whether this solution can be implemented without affecting other subsystems and their solutions.

Alternatively, you can think about the natural scale at which the problem occurs. For bike theft, a solution implemented in Albuquerque doesn’t affect an anti-theft program in Seattle. A spike in bike theft in Albuquerque doesn’t ripple out to Seattle. The problem doesn’t exist on a national scale³ and the solution shouldn’t be implemented on a national scale.

Theoretically, the federal scale is best-equipped to address problems that cannot be addressed within states (or at ever smaller scales). Unfortunately, the government frequently oversteps its bounds. And at other times, states are left to solve problems that really call for federal intervention.

Education isn’t one of these problems.

Education is a nationwide concern. But this doesn’t mean that solutions and reforms should happen on a national level any more than the fact that all states want garbage collection means that waste management should be a national issue.

I don’t think that anyone argues that tests are the best pedagogy. For the most part, they seem to be put forth as a tool for making pedagogy pragmatic: assessing what students “know”⁴ and comparing the systems that educate them. And there are plenty of people to highlight the detriment standardized testing does to education. I think that point is too obvious for me to belabor here.

But what I don’t think is obvious is how this notion of scale gives us tools to think about why the class of solutions that nationally standardized testing typifies are so egregiously ineffective (and undemocratic, to boot).

Democracy is (among other things) about compromise. Ideally, a compromise is as equitable as possible. For a compromise between two parties, this means that ideally, both parties leave equally [dis]satisfied.

When you move to larger and larger scales, it seems that compromises become less and less desirable because typically, you have a broader and broader distribution of ideal solutions.

To elaborate: with two parties, there are two potential solutions at the beginning of negotiations that one party would find perfectly satisfying. The process of negotiation and compromise involves mixing those together or finding alternate solutions that give-and-take appropriate ground.

As you increase the scale of your decision making, you add more and more “perfectly satisfying solutions” as you add more and more distinct parties. What this means, unfortunately, is that a single compromise satisfies fewer people less. Maybe a diagram will help.

Figure 1

Now, this isn’t a real diagram. You know how you can tell? Two points on the x-axis represent two solutions to a problem, and the closer those two points are, the more “similar” the solutions are. Whatever that means. Better yet? We’re assuming there are an infinite number of solutions.⁵

Anyway, geeky whining aside, the y-axis is a straightforward number: the number of people perfectly satisfied with a given solution. So this gives us a visual representation of satisfaction: the total area under the curve is the total satisfaction. This diagram depicts a perfectly neutral compromise: the solution is literally middle-of-the-road. But notice the difference between one, large group’s compromise (in blue, Figure 1) and the compromise of two, separate groups at smaller scales (in white, Figure 2).

Figure 2

Because the distribution of perfect solutions is narrower, a greater portion of those people involved leave more satisfied. Now imagine continuing this exercise at smaller and smaller scales, and take a look at Figure 3.

Figure 3

Why can’t we continue this indefinitely? Because problems have a natural scale.

All of this is theoretical wanking⁶. To reiterate: all of this is theoretical wanking. But I do think that there’s something important in choosing the right scale to solve a problem at. And I think that that is a discussion I haven’t heard when it comes to education.

To bring this back to education: people are constantly suggesting reforms. And there are a plethora of approaches, from progressive to reactionary. But why should we impose one solution on many, many systems when the problem doesn’t require it? NCLB allows for state standards, but setting standards is a much smaller problem than the big question it sweeps under the rug: is this the type of assessment we want, in the first place?

Now, what scale, if not national? I’m not sure. That opens up lots of questions, dealing with everything from what people want to resource allocation. And these questions are about how independent you can make your solutions. You don’t want a system that forces two groups of people with radically different visions (read: a wide distribution) to be forced to compromise. At the same time, you don’t want single schools to be able to get resources without input from other schools’ affected by budgeting decisions. Money given to a high school in DC doesn’t affect a preschool in Sacramento. Despite appearances, budgeting education nationwide is not a zero-sum game. More importantly, policymaking isn’t [a zero-sum game]. Essentially, this is a question of accurately gauging the size of a community, as defined by a common sphere of influence.

I don’t know how much of this is valuable and how much is empty theory, but I’m pretty convinced that the theme of matching scales of problems and solutions is a strong idea. Has anyone else heard arguments like this? Is this just dressing up an old idea?

1. outside of letting students decide [↩]
2. Or if it is, the definition is incomplete, and I would revise my statement to “handling educational details and policies at the national level is a bad idea.” [↩]
3. note that this is different than “happening” on a a national scale. Garbage collection is a nationwide problem, but that’s not the scale on which it makes sense. We can’t really afford to be vague about the word, “scale.” But I’m not going to try to define it, here. [↩]
4. Don’t get me started. [↩]
5. alternatively: the x-axis is a continuous space of solutions to a problem that have some metric [↩]
6. I even have the faux graphs to prove it [↩]