Abstract

When I first open a new document on my computer, as I have just done, the relationship between my key presses and what happens on the screen is rather loose. I press Enter a few times to move the cursor down the page a bit , and sometimes nothing happens for a while; or the cursor disappears for a few seconds before reappearing in its new position. After a minute or so, the machine seems to settle down, and then the correlation between taps on the keyboard and movements on the screen becomes much tighter and more consistent. Then, because taps and movements are tightly and predictably coupled, it hardly seems an inference at all to see the taps as the causes of the changes on the screen. While the computer is still 'waking up', I imagine that the taps are still the causes of the screen changes, but that the temporal relationship is lengthened and muddied by the fact that there is a lot else going on inside the machine while it is, so to speak, finishing dressing and cleaning its teeth. If, however, the machine was not just sleepy and preoccupied but seriously ill, then the relationship between taps and screen changes might become so bizarre and inconsistent that I would assume that causality had broken down. There may be causal connections inside still — the machine may still be operating in a completely deterministic fashion — but as I can no longer make sense of the relationship between input and output, I withdraw the imputation of causality