Centre stage

A dream last night in which, for once in a while, I had a leading role. That is to say a leadership role, rather than just charging about on some mission or other. Various parts of the dream were reasonably clearly picked up from debris of the day before, but that is not the present concern.

So I was in some middle ranking role in some large organisation. There was a senior person who had been senior for a long time, was a respected and successful member of that organisation. But she - for it was a she, if otherwise only very vaguely delineated - was now not so successful. It was, in fact, time to go. But she did not know this and the rest of the organisation did not know how to tell her, how to ease her out onto the (exit) plank.

Which was where I came in. A meeting was held to talk over the situation, a meeting which I smoothly took over with a story about where we were and what we were going to do about it. How in fact we were going to ease her our onto the plank. And everybody very quickly bought into my story, my narrative. It didn't really matter whether the story was true or not, whether there was a better story. This was a story into which everybody could buy and the world could move on. Leaving me feeling rather pleased with myself.

Waking up, I went on to associate with all kinds of stuff in the real world.

So FIL, as another middle ranking person in a large organisation, knew all about taking over meetings. His story was that, most of the time, most of the other people at meetings did not really care what the outcome was. They just wanted an outcome which they could all sign up to and so move onto tea, biscuits and the date of the next meeting. So if you did care what the outcome was and turned up with a plan, with a story, and made a reasonable fist of presenting it, all the other people would just nod it through. You got what you wanted, that is to say what was right for the organisation, and they got their tea and biscuits. Everybody was happy.

And then there are the detectives in all the dramas we watch on ITV3. Or read in Simenon. After a few days on the case, the star detective has a few facts. His job is to weave a convincing story around those few facts. A story which convinces partly because it is a good story, partly because it is a predicting story and one can build belief in it by the quality of its predictions, by the good fit of new facts to the old story.

While opposing barristers in criminal trials have to build opposing stories for the jury to choose between. With the catch here being that we don't have the luxury of testing predictions. All we can do is build stories out of facts that are already to hand - and hope that that is enough for the jury to make the right choice.

Next I jumped sideways to the world of fitting functions to some points on a graph, the sort of thing illustrated in the snap above. In the days of paper and pencil, one just draw a nice neat function in red approximating to the data points in blue. And one could test one's function by getting some more data points and seeing how well they fitted. Whereas now, one has computers to do it for one, in this case Matlab, a product which I think I once tried but never got very far with. See reference 1.

The problem here being that it is all to easy to overfit. To get to a function which does a splendid job of fitting to the data points, but which is far too complicated. It has made function out of noise. Perhaps a problem which detectives and barristers would recognise.

And lastly I landed on (two dimensional) diagrams. The way that you can peer down your microscope at some messy and complicated bit of tissue, perhaps live and throbbing, and turning that very messy image into a nice tidy diagram. A diagram which everybody peering down the microscope can buy into. The problem here being that it is all too easy to buy into the story without bothering to peer down your microscope and see for yourself. To understand the big picture which the diagram is attempting to encapsulate. There should be a balance between the diagram pushing down from the top and your own observations, your own ideas, pushing up from the bottom

From where I associated to the story at reference 2. But it was time to call a halt and take some breakfast before doing anything else.

Reference 1: https://uk.mathworks.com/products/matlab.html.

Reference 2: https://psmv4.blogspot.com/2020/10/the-power-of-word.html.

Reference 3: https://psmv4.blogspot.com/2020/12/an-excursion.html. Not this one, which is about something rather different, interesting though it may be.