Thursday, 18 June 2020

6,638 Miles of Graded Roads

In fact, I want to do a little codicil (edit: OK, maybe not "little") regarding the number 6,638.  More than anything other lyrics on the album, I love this lyric.  That there are 6,638 miles of graded roads in Halifax is a fact.  It is empirical, it is precise, it is quantifiable, presumably it does not care about my feelings.

It's almost certainly untrue as of this writing (as Chris Morris might put it, there's no evidence for it, but it is a fact).  Presumably whatever reference book Col. Hampton read it from was accurate at the time.  I suspect that Col. Hampton may perhaps have ellided some critical context, that perhaps there were 6,638 miles of graded roads in _Nova Scotia_ at some point.  Since Col. Hampton, in common with most people who will quote empirical facts at you, does not cite his source, this is difficult to verify.  It is made even more difficult by whatever his source is is far less remembered, far less important, than Col. Hampton's setting of that fact to a memorable, repeated melody on a cult rock and roll album of the early 1970s - try to find out how many miles of graded roads in Canada and all sorts of people will parrot Hampton's assertion back at you.

It's not enough to say that 74% of statistics are made up (possibly, pace Brass Eye, of "nasty stuff").  6,638 miles of graded roads - that's probably not a made-up number.  What is more important to me is that the fact that there are 6,638 miles of graded roads in Halifax is utterly and completely irrelevant.  I am hard-pressed to think of any circumstance under which knowing the number of miles of graded roads in Halifax could be even remotely useful.

When I say I have a lot of "useless knowledge" in my head, people think I'm putting myself down and feel compelled to disagree with me.  "There's no such thing as useless knowledge," they say, and me being disinclined to argue the point I generally do not bring up the number of miles of graded roads in Halifax.  I do want to state for the record that there is such a thing as useless knowledge, that something being a fact does not make it useful.  That a large amount of what I was taught in school was parroting useless statistics, was rote memorization, and that the way many of us learn to disagree is based around this, is based around regurgitating "facts" at each other and employing our critical thought exclusively to find new and interesting ways of rejecting the "facts" that don't agree with our preconceived biases.

I'm passionate about this professionally, in fact.  I am a data analyst, and everything I see in data analysis is about "big data".  The more data, the thinking goes, the better.  Data quantity is, after all, easier to implement then controls on data quality.  Most of what I do at work is determining what information to _ignore_, what to declare out of scope, what I decide is irrelevant for the purposes of my work.  We're all really good at doing this, but what we're not good at, usually, is _acknowledging_ this.  When I declare something out of scope, I always say that I am doing it arbitrarily.  This is a bit of an exaggeration, but not much of one.  Without axioms, without a mutually agreed upon dialectic framework, any data points I can bring up are no more useful than the number 6,638.  Most of the work I do involves data structures, or dialectic frameworks if you prefer.  First I establish the framework and then I attempt to convince the other people who will actually be using my data of its validity and usefulness.  This can be difficult.

No comments:

Post a Comment