The Agency Problem

People like to anthropomorphize wherever possible, especially when they’re teaching someone else. The idea is that their student will learn a concept more easily if he imagines it in terms of another conscious decision maker. So they assign agency where it isn’t really due.

Of course, it’s possible to do the opposite: think of something as an unwieldy, automatic process when it’s actually controlled by interested agents.

In either case, it’s a dangerous — and I think prevalent — source of all sorts of misunderstanding. So I’ll give a few examples to make the point clearer.

The Path of Least Resistance

In high school, I was told that electrons choose the path of least resistance. Not only was this enough information to solve basic circuit design problems, but it stuck with me as the canonical example of nature’s remarkable efficiency. I latched onto the fact electrons were smart, that they had somehow figured out how to find the mathematically optimal path in a sea of choices.

It took me a couple of years to finally ask, “How do they know where the resistors are going to be?” Can electrons look ahead somehow? If they could, how the hell were they doing it, and did that suggest something special about spacetime they weren’t ready to teach us in AP physics?

Without getting into a lengthy and most likely poorly formulated discussion of quantum electrodynamics, let’s just take it for granted that my intuition about electrons was eerily inadequate. The truth is that they follow the path of least resistance, and the distinction is important.

An electron, like every other blind, unconscious, and tasteless mechanism of the natural world, just is. It doesn’t know what it’s doing. And when your AP physics teacher tells you, purposely or not, that it makes a choice, he obfuscates the facts: one, that the electron’s path is only probabilistic, and two, that a resistor changes the probabilities; how that happens is the salient issue that, unfortunately, you won’t learn until college (if ever).

The Pinky Toe Fallacy

A friend of mine recently suggested that humans are rapidly losing our pinky toes because, as she said, “We’re just not using them anymore.”

Future generations of humans may very well have no pinky toe. But it won’t be because we stopped using ours. My friend’s confusion stems (again) from the agency problem.

It is important to understand that our genes aren’t trying to make better survival machines (us), and nature isn’t picking and choosing which traits will reign supreme. Rather, the genes that “just happen” to make the best survival machines “just happen” to make it to the next generation, and the rest is history. Our brain doesn’t change our DNA — it’s a one-way street going the other direction.

Thus, the only way we’re going to have mutant four-toed humans is if the gene (or more likely, the set of variously activated genes) from which that fifth toe emerges declines in frequency in the gene pool. Which could happen in one of two ways:

  1. People with the pinky-toe gene start dying. The question is, does having a pinky toe somehow confer a survival-level disadvantage, especially given the benefit of modern medicine? Obviously not. The “a bunch of people with no pinky toe genes survive while all the people with pinky toes start tripping into oncoming traffic” scenario is about as likely as a pinky-toe-based genocide.
  2. The pinky toe gene gets cut up or crossed-over (in a mitotic mutation) in the entire population, pretty much simultaneously. In other words, if the gene disappears via a chance mutation — which it’s liable to do — in every single fetus — which it’s not — then yes, the next generation will have no more pinky toes. Otherwise, the surviving genes will be subject to (1), and more likely than not will prevail just as frequently as the no-pinky-toe genes (if there are such things).

The point is, just not using your pinky toe has nothing to do with its place in our DNA. You, the conscious human, are not the agent of evolution. Nor are genes, Nature, or the Flying Spaghetti Monster. It “just happens.”


The Impenetrable Fortress of Knowledge

Growing up I knew very little about where the stuff in textbooks came from. I didn’t realize that there were real human beings, fighting amongst each other for credibility, tenure, and money, who were writing what I understood as “the truth.”

I didn’t know that my Economics book was grossly simplistic, that my History book told a biased story by omission, that my Math book (which I studied in New Jersey) was written by committee and chosen by the Texas school board based on wish-washy criteria.

Only when I became an undergrad and got a first-person view of all the living, breathing, uncertain people competing in an imperfect economy of information did I understand what my homework really was. And each time I saw a theory updated or a result challenged, a professor working late in his office on a paper that might not get published, it chipped away at my naive notion that knowledge was stable and well-defined.

Understanding academia’s fallibility gave me a sense of duty to improve and refine that corpus of facts, theories, and fanciful imaginings I had previously taken for granted; it made me an agent.

The Big Corporation, Big Brother, and The Mob (briefly)

Every CEO, Senator, and citizen is a person, and the great woes of public policy would be a lot easier to solve if we all realized this.

Citizens would be reluctant to bestow so much power upon one man if they realized he was indeed a man. And a President would care more about his people if he saw them as such.

Concluding Remarks

The one thing to take away: agency is important. We should be wary that simple habits — from saying electrons “choose” paths and genes “want” to survive to forgetting books have authors and your fellow Americans have faces — can confuse, occlude, or otherwise impede understanding.




Avinash May 26th, 2007 at 4:34 am


There’s a bunch of other stuff like this. Take calculus for example. The actual application of it was severely limited when I started using it in school–so much so, that for the first few months I thought it was going to be as useless to me as when I did loci and arithmetic progressions.

I don’t think we’re ‘losing’ our pinky toes. Take a look at the image on this page. If you look at the bone structure, losing the pinky toe will (by association) probably cause instability.

I think natural selection is going to work against having a society of clumsy four-toed humans.

And it’s nice to see you finally used that electron question we had that conversation about in one of these entries.