Our Questions shape the Universe we See

The way we look at the universe affects what things we notice. We build models of the things we notice. When a model describes everything we notice in a coherent framework, we treat it as describing the whole universe (and it becomes easy to forget that the model is not the universe). Thus the universe we see is shaped by the way we look at it.

The most widely visible form of this is the universal law of natural cussedness, which says that if anything can go wrong, something will. Scratch this `truth' a little and you will discover that it is consequent on the fact that we notice things going wrong - especially when they surprise us because we didn't realise the thing in question had the option of going wrong - and don't remember the (possibly numerous, but I don't remember them) moments when something could have gone wrong, but didn't. Selective memory begets Murphy's law.

Consider a Quiz

Somewhen I was I kid, I was accosted by some other kids on the way to school and asked `do you have a penis'. New word: `penis', so of course I had no idea. Saying so was the right thing to do, but I lacked the social awareness (having done so) to exploit that; they intimidated me into answering (and I guessed wrong) so that they could tease me. Quizes aren't allowed to do that.

There's a joke riddle that asks: what's brown and sticky. The correct answer is `a stick', because that's what makes it a joke, no matter how many things one may think up that are, indeed, brown and sticky. Quizes aren't allowed to do that, either.

Now quizzes live with a few other restraints too - more important to the participants than the above - to do with there being some fun to be had and reasonable doubt about who will win, else who but the winner would enter ? Quizzes allow one to abstain from answering; their other rule against my first illustration is that it has to be reasonable to expect the answerer to have some idea what the question means. My second illustration illustrates the need for definite answers - not only must the quiz master have a correct answer to the given question, but the question must be so phrased that it's the only right answer. Knowing that the questioner has abided by these rules gives enormous clues to anyone paying attention to the phrasing of the question; this is entirely dual to the way that an interrogator's choice of questions forces the answers into forms which conform to the interrogator's expectations. I quite routinely amuse myself, when the pub I'm in is having a quiz, by seeing how many of the questions I can answer by combining plausible speculation with this technique, to guess right answers simply because `any other answer would mean the question is unfair'.

When you ask a question, the form of the question makes presumptions which shape your ability to interpret anything as an answer. If your presumptions are skew to reality, even just a bit, you can expect the answers you get to be odd. The constraints of reproducible measurability shape Science's experiments; they work from a presumption of determinism, so made sense of such as is deterministic in our universe; having explored a rich seam extensively, it ran up against the universe's innate non-determinism - devising, along the way, means of measurement which capture the information which can be reproducibly measured and building models which describe systems deterministically as far as they can, then apply statistical methods later to deal with the actual chaos of the universe. Science has done very well thereby - but how has the phrasing of the questions affected the answers we can see ?

The classic `two slit' experiment

Take a look at the classical two-slit experiment: but start by looking at the ante-chamber to the experiment itself. A source of electrons is placed at one end of an accelerator which gets lots of the electrons up to a good speed, throws them at an obstacle at the other end with a single hole in it, thereby producing a beam of electrons coming out the other side of that hole (which might, admittedly, be a slit); enabling us to regard that hole as pretty much a point source of electrons with whatever energy they had shortly before reaching the plate's end. By designing the accelerator right (it helps to have both magnets and electrostatic fields involved) one can arrange for the electrons coming out of this source to all have (pretty much) the same energy as one another. Entering the first chamber of the experiment proper, we thus have a beam of electrons; all the electrons are coming away from a small hole, with (roughly) equal energy, in a spread of directons, mostly towards the far end of the first chamber - where they'll meet a barrier with two slits in it (and this time, they're definitely slits, though the details of why belong elsewhere).

The electrons that reach the vicinity of the two slits, which are close together, have all come from our point source, which is a fair way off, and have done so in straight lines; so they're pretty much going in the same direction as one another; and one arranges for the two slits to be equidistant from the source. Just as at the earlier barrier, plenty of the electrons are going to hit the barrier (or the walls of the chamber) and not make it into the second chamber of the experiment proper: the observation chamber. Only in so far as electrons make it into the second chamber do we notice them (whether by watching them light up a pattern on a phosphor, or by some more sophisticated gadgetry that spots individual electrons) and, of course, our observing aparatus may miss some of the electrons that do. The electrons that make it through the obstacle course and do get registered by the detectors form a pattern, the characteristic fringe pattern of interfering waves, rather than the simple pair of lines one might have expected, when regarding the electrons as `particles'.

In such a case, the observed `diffraction pattern' might plausibly be an artifact of the processes by which the aparatus (a) filters out the electrons which get stopped along the way and (b) ignores the electrons (if any) which miss the detectors.

Written by Eddy.
$Id: observer.html,v 1.3 2001/11/13 14:50:58 eddy Exp $