Applicability of formalisations -- how bollocks...

EDIT (...is natural science?)

I noticed the following weird phenomenon: while commuting, it's fairly common to spot people recognisable from earlier parts of the trip, often getting off the tube car in front of me, for example, after having been spotted waiting with me on the platform. Since I see lots of people and don't remember most of them, these 'familiar' people are presumably a random sample of the people I see, distinguished and admitted into consciousness only on seeing them the second time. Except they're not a random sample: the familiar people are disproportionately fit women. This would be an unimpressive phenomenon if, on my first sighting, it entered my consciousness that 'she's fit!', but it doesn't. It's only that she's apparently more likely to strike me as familiar on the second sighting.

Already, my comments suggest that a deeper analysis of this situation would be statistical in nature, involving an application of the formal ideas of events and conditional probability. Estimates of the probability that a randomly selected passenger is a fit woman might be invoked and compared to the proportion of fit women among the re-spots. The point is that the 'natural' thing to do would be to take a pre-existing theory (probability theory) which operates on very general objects (events) and use it to model the specific situation in an attempt to explain the specific phenomenon.

If I'm watching ripples on the surface of a pond, I might wonder why the water, acting under some influence, is behaving as it does. The natural course of action has historically been to observe the radially symmetric shape of the ripples and look at their cross-section, and find a differential equation which models the motion of the water in the vertical direction, solve it, and get a solution that says 'everything' about the waves.

Explanatory science almost always follows this process: some essential feature of the phenomenon in question is abstracted by noticing that the specific objects in question are share properties of the general objects in some formal theory. The specific objects are 'mapped' via some 'isomorphism' to the objects in the formal theory, which is then used to generate some conclusions about the general objects. We deduce that these conclusions hold in some form for the specific objects of our phenomenon because the formal theory is 'isomorphic' to the specific situation.

Now, the conclusions we draw about, say, light, are very similar, formally, to the conclusions we draw about our ripples, because the formal theory is very much the same.

Another example is Gauss's law for gravitational and electric fields: literally changing a few words or symbols renders two qualitatively different phenomena formally identical.

Historically, the formal machinery for studying some phenomenon tends to develop from observations of that phenomenon and slowly become more general and theoretical as history progresses. Eventually, someone uses it to analyse some other phenomenon, rendering two different parts of the physical world 'isomorphic' because the second shares enough properties with the first to be subject to analysis by the same machinery. It shouldn't be much of a surprise that formal machinery designed around data from some observed phenomenon explains and makes predictions about that phenomenon, but when the same machinery is found to be applicable somewhere else, I am surprised and have some misgivings.

How much of our knowledge of the second phenomenon is coloured by our use of pre-existing theoretical knowledge to study it? Have we abstracted the wrong properties of the objects involved? Does the theoretical understanding we gain really tell us anything useful about the qualitative properties of the phenomenon? If I didn't know probability theory existed, would I even have noticed what I did about fit birds on the tube?

Certainly the fact that probability theory is part of my intellectual toolkit made me impose a point of view on what I observed. An observation amounts to placing higher value on some data than on others -- when this bias results from comparing data to theoretical machinery we already have, do we give the world an invented interpretation? Can this sort of question even be answered by abstract reasoning without introducing some logical model that is internally consistent but equally open to questions about interpretation?

I hope Sherlock 'Twist Theories To Suit Facts' Holmes doesn't have to turn over in his imaginary grave.

You probably just need to get a good piece of ass. That should clear things up. Make your investigations more experimental and less theoretical, and you won't be so paranoid about theory.

Originally posted by DoctorScribbles
You probably just need to get a good piece of ass. That should clear things up. Make your investigations more experimental and less theoretical, and you won't be so paranoid about theory.

Originally posted by royalchicken
Hey rec slut, if I rec this post will you provide the good piece of ass 😛?

Originally posted by royalchicken
EDIT (...is natural science?)

I noticed the following weird phenomenon: while commuting, it's fairly common to spot people recognisable from earlier parts of the trip, often getting off the tube car in front of me, for example, after having been spotted waiting with me on the platform. Since I see lots of people and don't remember most of them, th ...[text shortened]... lock 'Twist Theories To Suit Facts' Holmes doesn't have to turn over in his imaginary grave.

Originally posted by Starrman
And you were complaining about not meeting girls? Jeez! Put the maths down and step away with your hands in the air! Put the maths down and step way...

Originally posted by DoctorScribbles
I'm retired from that business. You may want to check with ark13, who is currently the self-proclaimed Pimpin Wizzalrus, and as luck would have it, he's in da hizzouse.

Originally posted by royalchicken
He pimps? He doesn't merely pirp?

Originally posted by royalchicken
Hey rec slut, if I rec this post will you provide the good piece of ass 😛?

Originally posted by DoctorScribbles
You probably just need to get a good piece of ass. That should clear things up. Make your investigations more experimental and less theoretical, and you won't be so paranoid about theory.

Originally posted by royalchicken
When was I complaining? I was making an observation. What do you do on the tube? Drool over page 3 of the Sun like the rest of your lame countrymen 😉?