The Pournelle axes
and the modern political climate

April 2017

Sometime in the mid-80s, I read an article (reproduced at by science fiction author and science essayist Jerry Pournelle that seemed to make a lot of sense to me, but apparently not to anybody whose opinions carry public weight. It was a rewrite of a portion of his political science PhD thesis from several years earlier. The fact that it has been generally ignored says something not good about the modern American political climate.

Here is a summary, with an admixture of some of my own opinions. Many have noticed that the labels "liberal" and "conservative" or "left" and "right" often don't work well as designators of people's political opinions. The observation is typically dismissed as signifying nothing more remarkable than the fact of our cognitive diversity. People are complicated and therefore difficult to categorize. However, life is full of complex phenomena for which we have nonetheless come up with useful taxonomies. The problem with our political taxonomy, as Pournelle notes, is that we insist on using a single metric, compounded by our inability to agree on just what it is we're measuring. The solution, he suggests, it to get it into our heads that one metric is not enough. At the very least, we need two.

Of course that would complicate our political discourse, but the result would be a better match between discourse and reality. The obvlivion into which Pournelle's proposal has fallen could be a sign that almost nobody cares much these days about matching their discourse to reality. There does seem to be a lot of such indifference, but something else could be at work, too.

In mathematical terms, a one-dimensional metric is a scalar quantity while anything in two or more dimensions is a vector quantity. One big difference between scalars and vectors is that scalars can be ordered while vectors cannot. That means this: For any two things measured by a scalar, you can say that if they are not equal, then one of them is more than the other, or bigger, or greater, or heavier, or something else strictly analogous. In other words, some kind of ranking is possible. That is not possible with vectors. They can be equal or unequal, but if unequal, there is no "more than" or "less than." All you can say is that they're not the same. When you're talking about vectors, you can never say, if X is different from Y, that X is more or less than Y in any relevant sense. In particular, you cannot say that X is better or worse than Y.

This does not sit well with our compulsion to assign virtue to certain political inclinations. The metric of virtue is a scalar: You can have more or less of it, and you can know how much a person has by counting the number of correct political opinions they hold. This lends itself very nicely to a manichean us-versus-them mindset: Left is good, right is bad, or vice versa. It's a hard notion to give up, even if one wants to, and almost nobody does want to.


That (with minor revisions) was as far as I got the first time I tried to write this essay. I don't recall just when that was, but it was more than a decade ago, possibly even before 9/11. I set it aside to work on other projects, and then more or less forgot about it.

In the years since, the political polarization at which I hinted has gotten only worse. And in my judgment, which is shared by a handful of other people, the political left has been at least as responsible as the political right. I have not changed my mind about the inadequacy of any one-dimensional political spectrum to track people's political philosophies, but Pournelle's observations are consistent with the observed clustering of contemporary political advocacy around two positions, one in support of the Democratic Party's agenda and other in opposition to it. The mistake lies in the supposition that a person will affiliate with one group or the other because they are either more or less moral or more or less committed to critical thinking, or more or less anything else.

I did not want Donald Trump to become president, but those who said on Nov. 9 that his election represented a victory for racism were as out of touch with reality as they accuse Trump and his supporters of being. Trump's supporters included some people whose opinions I actually respect. (I guess I should mention at this point that, in my lexicon, respect does not entail agreement.) They were not motivated just by their hatred of Hillary Clinton. It is true that most of them couldn't stand her, but they also believed that Trump would make a good president. I suspect he won't, but my jury is still out and is going to stay there for a while.

I don't think he won by making the case that he was better qualified than Clinton to be the nation's chief executive. At any rate, he didn't convince me of anything like that. I didn't like her any better than I liked her husband, and I never voted for him. But I did vote for her because I thought she had the better credentials, and I started feeling a little sick when it became apparent in the early evening on Nov. 8 that she was going to lose the election.

A little sick, yes, but not afraid. I was disappointed, not scared. I did not think then, and think less now, that Trump was going to be a 21st-century American Hitler. His was not a victory for Nazism. It was not a victory for reason, either, of course, but then, reason was never in the running. Clinton was no more its champion than he was.

If there was a winner, it was the collective Howard Beale—all the people in the states that Trump carried saying, "I'm mad as hell and I'm not going to take it any more." And what were they mad as hell at? At being told that if they didn't like what the national government has been doing over the past two or three generations, then they were either evil or stupid if not both. And both parties have been saying this about their adversaries. The Democrats tend to prefer the accusation of evil, the Republicans of stupidity, but both will use either as the occasion arises. Trump won his party's nomination because the Republican rank and file were tired of being told that they were stupid, and he won the general election because the nation's rank and file were tired of being told that they were racists and misogynists.

Yes, Trump lost the popular vote, but so did Clinton in the sense that more people voted against her than for her. The primary campaigns of both parties showed that Clinton was not the preferred candidate of most Democrats and Trump was not preferred by most Republicans, and in the general election, a majority of the voters were against both of them. Whatever else this is, it is not government by consent of the governed, and this is something we need to fix before we try fixing anything else; because without this fix, we're going to lose everything else that really matters.

It won't be easy, because if there is one thing the political establishment does not really want, it is to have to please a majority of the electorate. If you can gain office and hold it even if three-fourths of your constituents don't like you, then you're not going to vote for any reforms that would make that impossible. And we're not justing talking about the electorial college. Mathematicians have known for a long time how easy it is for candidates with minority constituencies to win elections as they are conducted in most American jurisdictions.

The apparently obvious solution is to elect more people who do want reform, but how do you do that while working within a system that opposes reform? Well, we're still a democracy. You don't get into legislative office without winning elections, and you don't stay in office long enough to do something useful without continuing to win elections. (I'll have to put off my rant against term limits for another essay.) What it takes to win elections is mighty hard to predict, which is why political consultants get big bucks for claiming that they can pull it off and why, no matter how good they are, they sometimes get it wrong. It takes considerable understanding of human nature, and the relevant science is still really young.

The prevailing wisdom seems to be that it all comes down to telling the voters, as believably as possible, "I'd be surprisingly good for you" or else "My opponent would be bad for you." The problem is in that "believably." It doesn't matter what you tell people if they assume, before you even open your mouth, that you're a liar, and how many people these days don't assume that all politicans are liars? Most of the people who voted for Clinton didn't do it because they trusted her, but because they trusted Trump even less. A democracy where that can happen is a democracy in deep trouble.

We need a political movement that doesn't only advocate good ideas but can also persuade voters that they are good ideas, and not good just in some platonic sense but in a sense both utilitarian and obvious. Utilitarian, because voters will ask, "What's in it for me?" without caring what's in it for anyone else, and obvious, because they won't take "Trust me" for an answer. Ah, but isn't that "Trust me" just exactly what worked for Trump? No, what worked for him was everything else he said. His opponents were all saying "Trust me" as much as he was.

I agree with those who say that America is great now, and is no less so than it ever was. I also agree that some changes are necessary if we are to become even greater, though I disagree with many, both liberals and conservatives, about which changes would take us in the right direction. Those changes will happen when Americans collectively want them to happen, not before then. The challenge for reformers, then, is to get the people to want what they want.

Some will say: The people already want what we want, but our democracy has been subverted by the moneyed interests. And that is possible; but possibility is not probability, and I'm not convinced of its probability. Anyway, this essay is not about campaign financing as such. It is about the kind of thinking represented by both of our two major political parties. And that, finally, gets us back to Pournelle and his political axes.

He correctly observed that a single-variable metric is useless at best and, more likely, misleading. He proposed a two-variable taxonomy, which was definitely an improvement. But I disagree with his claim that two variables are enough. I can't find myself on his chart.

Yes, I'm generally sympathetic with the "various conservatives" that he put near the origin of his grid—keeping in mind that he first wrote this half a century ago, and most of the conservatives he would have had in mind are no longer alive. But he put them there by attributing to them two tendencies. One was a favorable feeling about the state. The other was a slight aversion to rationalism. I don't dispute the attribution, since it does seem accurately describe how most conservatives actually feel about the state and the use of reason. But it is inaccurate in describing me and certain other people whose political views seem close to my own. Fifty years ago, maybe there weren't enough of us to matter. But even fifty years ago, Pournelle should have included at least one more dimension, and that is the acceptance of scientific thinking. And no, that is not the same as rationalism. Science is a development of empiricism, which is conventionally treated as distinct from rationalism (

No government, regardless of how it comes to power and no matter how much it caters to its citizens' wishes, can govern well if it ignores the facts of human nature or denies that such facts exist. Those facts cannot be discovered except by scientific inquiry, and neither can their implications, if any, on public policy be determined except by the proper exercise of reason. Our current scientific understanding of human nature is of course meager. The relevant science is very new, and its subject is very complex. What we think we know at this point is quite tentative, but it is both wrong and dangerous to argue that it must be ignored or disputed just because it is inconsistent with certain political desiderata, whether liberal or conservative.

My position on Pournelle's grid would be near the top of the verticle axis and at the midpoint of the horizontal (around 3/4.9' using his numbers). I would not say that I enthrone reason, but I do it must be followed without exceptions, in the sense that we should believe there are no acceptable contradictions: If we think we see one and have failed to find a resolution, we need to keep looking. And I think the question of whether government is good or evil is irrelevant. It is certainly necessary, and it is certainly dangerous, and the correct attitude toward it is just to never forget either. We can't get rid of it, but it can get rid of us. Enough said, as far as I'm concerned.

But as I said, I would add an axis. To his statism and rationalism, I would add empiricism. He asks: How do you feel about the state and how do you feel about reason? I would also ask: How do you feel about any facts relevant to the exercise of political power, as those facts are revealed by currently accepted science? The answer at one extreme would be: We should ignore them, because science can't tell us anything about how the country should be governed. At the other extreme: All political decisions should be informed by the best available science that is pertinent to those decisions.

Pournelle's numbering system was a mistake. He was trying to do science, and you can't do science with scales like 1 to 5 paired with 1' to 5'. It would get even worse if a third axis had to be scaled 1'' to 5'' or some such. Each axis needs to be 1 to 5 or, probably better still, -5 to 5, since neutrality is intuitively best measured as zero. On the Shaver Axes, then, I would perhaps be in the neighborhood of 0,4.8,4.8.

No political party or other political movement that I know of is consistently anywhere on the empricism axis. The activists in both major parties either accept as authoritative or dismiss as irrelevant (or wrong) any science according to its consistency with whatever issue is momentarily up for debate. And at this stage of our cultural evolution, it is probably just as well that there is no Science Party or anything like it. If there were, there are several reasons for which I would not join it. For the time being, those of us who are consistently committed to witnessing the arrival of a government adequately informed by science and reason, while at the same time being averse to tyrannical excesses, will have to settle for spreading these memes as best we can.

Site home.

Philosophy index.