Saturday, April 13, 2002
Posted to the LA Forum today:
---
Paul Staines wrote:
> Does anyone seriously think that the imperial system
is easier to understand than the metric system?
I think 'ease of understanding' is mostly a matter of familiarity.
> If you were starting from scratch would you design a
system where sub-units were sometime 1/12,
sometimes 1/14 and sometimes 1/16 of a unit? Or
possibly a multiple of 144? Why is there some mad
number of yards in a mile?
If I were starting from scratch I would certainly want to be able to halve a quantity repeatedly, so I'd consider multiples of 2, 4, 8, 16 ...
I might also want to be able to divide evenly by 3, so that would suggest 12 = 3 x 4
But it's strange that the same multiple isn't used all the way up a chain of units: eg, 16 drams make an ounce, 16 ounces make a pound, but 14 pounds make a stone, 8 stones make a hundredweight, 20 hundredweights make a ton ... It suggests that the units originated as being of useful sizes in different trades (the ounce for grocers, the hundredweight for millers, etc), and were then standardized and connected with each other by some unsung rationalizers unknown to history.
Having 14 pounds in a stone is pretty hard to explain, I admit. 14 is as inconvenient as ... well, 10.
> Is it easy to remember how main grains in an ounce?
How many ounces in a ton, how many pounds in a stone,
how many fluid ounces in a gallon?
Once again: just a matter of familiarity...
The lesson I draw from the complexity of the traditional systems is that such complexity is largely irrelevant to their usefulness. As shown by how reluctantly people give them up.
And also that it's very hard for social engineers (such as designers of systems of weights and measures) to know everything that's going to be important to the users of their creations.
It's interesting to speculate how successful the metric system would be in everyday life given fair competition with traditional systems, rather than receiving massive state promotion and coercion. Not very, I suspect. (In scientific uses, the case is different.)
