Tuesday, October 16, 2012
Important facts
Epsom market did not have anything very enticing in the way of apples on Saturday so on Monday I bought a pack of Jonagored apples from Belgium via our trusty Costcutter. Looked fine, flesh texture fine, flesh taste OK but skins bad. Rather tough and sour, at least to my palette. A serious failing as I prefer to eat my eating apples with their skins on. The balance is being cooked as I speak, with my consolation being that eating apples generally cook much better than cooking apples. From all of which I deduce that while Belgians might make the best chocolates and the best detectives in the world, they do not make the best apples. Nor, I note in passing, do the French. The best apples that I know come from England.
The fact that Golden Delicious was one of the ancestors of this apple may be the root of the difficulty. Another apple which I do not care for, mainly because of its soggy flavour but also because of the texture of its skin.
I collect another fact from a quite different sphere of endeavour from the TLS of 21st September. It seems that some academic has ruled that a civil war is 'a sustained military combat, primarily internal, resulting in at least 1,000 battlefield deaths per year, pitting central government forces against an insurgent force capable of inflicting upon the government forces at least 5 percent of the fatalities the insurgent sustain'. As a former statistician I should applaud the care with which this definition has been constructed, with this definer clearly having had a very clear idea about which conflicts he wanted in and which conflicts he wanted out. But if his idea is as clear as all that, what value is added by capturing that idea in a definition? Does the definition have predictive value? Does it add value? Does being so nice about scope advance the study in question? Are there so many civil wars knocking around that this sort of statistical nicety is helpful? As it would if, for example, we were trying to tie down the meaning of the word 'household' for the purposes of a census of population. There are far too many households to enumerate so we really do need to try to have some shared understand of what we mean by the word.
I think my answers to these questions are none, no, no, no and no.
The fact that Golden Delicious was one of the ancestors of this apple may be the root of the difficulty. Another apple which I do not care for, mainly because of its soggy flavour but also because of the texture of its skin.
I collect another fact from a quite different sphere of endeavour from the TLS of 21st September. It seems that some academic has ruled that a civil war is 'a sustained military combat, primarily internal, resulting in at least 1,000 battlefield deaths per year, pitting central government forces against an insurgent force capable of inflicting upon the government forces at least 5 percent of the fatalities the insurgent sustain'. As a former statistician I should applaud the care with which this definition has been constructed, with this definer clearly having had a very clear idea about which conflicts he wanted in and which conflicts he wanted out. But if his idea is as clear as all that, what value is added by capturing that idea in a definition? Does the definition have predictive value? Does it add value? Does being so nice about scope advance the study in question? Are there so many civil wars knocking around that this sort of statistical nicety is helpful? As it would if, for example, we were trying to tie down the meaning of the word 'household' for the purposes of a census of population. There are far too many households to enumerate so we really do need to try to have some shared understand of what we mean by the word.
I think my answers to these questions are none, no, no, no and no.
# posted by Jim @ 2:12 AM 
