False Consistency

The fallacy that you can avoid having to choose between two alternatives by arguing for both.

This temptation can arise when you are faced with two or more options which, though desirable in themselves, are inconsistent, or cannot be adopted at the same time as key defining objectives. For example, in the field of education, “access” and “excellence” are both things which any well-intentioned government might want, as are “tradition” and “change”, not to mention “sustainability” and “development”. The Utilitarian principle of “the greatest happiness for the greatest number” also has this flaw, which its chief advocate, Jeremy Bentham, recognised, later changing the formula to “the greatest happiness”. Sustainable development is sometimes taken, with unashamed faith in impossibility, to mean the maximum growth with the minimum environmental impact.^F7

It seems a pity not to be able to go for two good things with equal determination—surely they are twice as good as one—but that’s the way it often is, and communities will need to be able to make these difficult choices, particularly if trying to go for both means that they are likely to end up with neither.

A name for this problem is the “double maximand”: the aim is to maximise two aims, despite there being an inevitable trade-off between them. A popular way of presenting this is to call it a “balance”. Balance can, of course, be a good thing, but it can also be a plausible excuse for missing the point. For, example if someone told you to “run as far as possible as fast as possible”, you would be left wondering whether to jog for 26 miles or sprint for 100 metres, or to compromise between the two. Actually, those instructions leave you with freedom to do whatever you want, and if asked how you come to be lying in the grass listening to the skylarks, you could explain that you were striving to achieve a balance.^F8

Where there are two trade-off-related variables in an orderly system, one of them is the “control variable”; the other is the “residual”. You cannot control them both. If there are several variables, you may be able to control some of them, but not all. The rule is that if there are n variables, the most you can control is n-1. That is, in any system, there is always at most one “degree of freedom” fewer than the number of variables. “Degrees of freedom” are a complex idea and the only secure definition is that they are “the rank of a quadratic form”, but let’s not disentangle that here.

Think, instead, of a system which has two variables. Example: a nation with the intention of reducing its carbon emissions. It can either cap the quantity of emissions in line with climate science (e.g., through TEQs) and let the price of carbon-intensive goods and services adjust accordingly, or it can set a price (e.g., through carbon taxation) and let the quantity of emissions adjust accordingly. Control of the quantity is enabled by flexibility in prices; control of the price is enabled by flexibility in emissions. It cannot control both.

False consistency lets the moment go by when it would have been helpful to face up to a problem. T.S. Eliot called it “obnubilation, [that is] reconciling, hushing up, patting down, glozing over, concocting pleasant sedatives . . .”. And he added: “The sense of fact is something very slow to develop, and its complete development means perhaps the very pinnacle of civilisation.”^F9

And yet . . . Lean Logic has to be cautious in making the case for consistency. Some of the complex matters included in its conversation have not been sieved and sorted down to the level at which no inconsistencies remain. Nor, perhaps, should they be. The last, ruthless rendering-down of an argument in the interests of consistency is rich with opportunities for mistakes which may wreck the whole enterprise. Anyway, inconsistency is fertile: it admits limitations; it invites the other person to apply his or her mind as required by lean thinking. It suggests that maybe it’s time for a walk in ironic space.

Related entries:

Good Shepherd Paradox.