Tuesday, June 11, 2013

Prediction, Confidence, And Language

The news is all bad, as usual, but it's also getting pretty old, so I thought I might spend a few pixels on an old irritant of mine. Perhaps it's occasionally bothered you, too: the predilection far too many people have for "predicting" the past...and getting it wrong.


There are a few words in the English language that defy precise categorization. One of my favorites is should. Let's start here: What part of speech is it?

Various dictionaries list should as a verb, the "implicative" form of shall. I suppose that's one way to view it, though to me it looks more like an adverb. The spectrum of its uses adds ambiguity to its place in the lexicon:

  1. The moral-ethical application: "You should do this."
  2. The forward-prediction application: "That should result in this outcome."
  3. The boy-am-I-an-idiot application: "Hm, that should have worked."

Each of these applications involves a knowledge context. In the first case, the speaker implies a knowledge of moral or ethical constraints, whether positive (obligations) or negative ("thou shalt nots"). In the second case, the speaker has imperfect confidence in his understanding of the current situation; he thinks his recommendation will lead to a particular outcome, but will allow, perhaps grudgingly, that he might be wrong. These are both perfectly comprehensible uses of should, and unobjectionable from most standpoints.

The third case is among the most important of all indicators of a lack of understanding. After all, what does the speaker really mean?

"According to my grasp of the situation in which I operated, I expected a particular outcome from my work. But I didn't get it. I don't know why I didn't get it, and it's near to pissing me off!"

The speaker is irritated by reality itself. By denying him the results he expected, it demonstrated that his "grasp of the situation" was, at best, incomplete. The dissonance he's experiencing over that is stressing him somewhat.

Upon hearing that third use of should, a prudent observer will retreat out of flailing range to await developments.


A departed friend, when he heard a colleague say "that should have worked," liked to ring in, "And here we are once again in the Land of Should, where all programs work on the first try." This is exactly correct:

If your understanding of a situation is correct,
And your application of that understanding is appropriately and accurately executed,
You will get the outcome you seek with perfect reliability...
But how often is that, really?

"That should have worked" is an erroneous attempt to predict the past. It tells no one anything of importance. The impulse to say such a thing is educational to him who feels it. He'll be best off for not succumbing to it, but he should draw the lesson nonetheless.

All right, all right. Given the high intelligence of the typical Gentle Reader of Liberty's Torch, all that "should" be obvious at this point. But there's one aspect of it that deserves further exploration: a forward-going prediction that "should" be closely scrutinized, specifically because of its application to politics and current events.


"A strong sense of confidence is always misplaced." -- James Hogan

Politics and economics are full of "shoulds." This is quite appropriate: the dominant feature of both realms of thought is ignorance. Therefore, he who deigns to make a prediction in either field is expressing confidence in his grasp of human nature, the dynamics of power, the laws of economics, and the bloody-mindedness and intransigence of ordinary individuals.

If you encounter such a person, cross the street and walk quickly.

We do not know enough to make firm predictions -- that is, predictions of a specific stimulus yielding a specific outcome at a specific future time -- about political or economic systems. It's likely that we cannot know and therefore never will know enough to do so. And indeed, you will note that even the very best oracles on these subjects routinely omit to state a time-to-completion of the processes they delineate. It's clearly the path of wisdom not to be too certain of anything in these domains.

It's possible to have good confidence in the outcomes of certain political and economic processes, if all the relevant aspects of the context to which our knowledge is applied can be kept from changing. But one who predicts must be mindful of the likelihood that the context embeds features of which he is not aware, and even more mindful of the uncontrollability of processes that interact with human desires, fears, and beliefs. Physicists insist on replicable experimental conditions. We who prognosticate about politics and economics know that we can never have them.


The power-seeker often speaks with an air of confidence to which he is not entitled. He propounds his nostrums as if their efficacy could not be more certain. It doesn't matter whether he truly feels that confident or whether his confidence is entirely rhetorical. What he wants is to win the trust of those listening...and tragically, an air of confidence, even when it's utterly mistaken, is a very attractive thing.

Americans might be coming to understand this. In the scandals of recent weeks, we've heard a lot, almost all of it from Democrats and other power-mongers, about what "should" be done to correct the abuses about which we've all heard. Most of it has been laughable or self-serving, because it's been done before and it brought us directly to the present messes. But you can bet your bottom dollar that we'll be deluged with "shoulds" -- and worse, "wills" -- before the furor dies down, for a simple reason:

Political egotism trumps both logic and experience.

So it has ever been, and so it will always be...no matter what you think "should" be the case.

No comments: