I don't often have personality crises, but when they strike I'm usually left paralysed. They tend to begin innocently enough, with the kind of question best left unanswered. Some things should just be pushed to the back of one's mind - life would be unendurable if we confronted them all the time. For example, the other day I asked a friend of mine whether telephones needed electricity to work or not. He didn't know the answer. Neither did I. I had no idea at all about how phones worked. In fact, the whole panoply of domestic appliances and consumer durables appeared to me as alien artefacts. Exactly what goes on inside my microwave, anyway?

The question preyed on my mind. How could I claim to be a real man if I could use, but not explain, my telephone? I panicked and called my engineer friend Fiona, to ask her. Now, I've always thought of engineers as being pretty retarded - the kind of kids who dismantled your Rubik's cube instead of solving it - but now I was unsure. As it turned out, she wasn't much help. It was Thursday evening, and Fiona rather sharply told me that she didn't want men calling her after Wednesday night for any reason - I mean did I think she was desperate or something, and she had to go now because she had a casserole in the oven and several attractive men waiting at the door.

I set to work on the issue myself, drawing on all nineteen years of my full-time education. Princeton had taught me well: instinctively reducing the problem to its essentials, I drew a picture of two tin cans connected by a long piece of string. No electricity there. I began to write a literature review to support my case. Citing a variety of authorities, I noted that phones don't plug in to power sockets, but in to special little sockets of their own (technically known as 'jacks'). The argument was clinched by the observation that phones continue to work during power failures (unless the murderer has cut the lines). The 'no-electricity' theory of phone operation was parsimonious in its assumptions, consistent with my common-sense intutions about the world, and - best of all - well founded in hard, empirically verifiable facts. Normal science swept forward with its customary elegance and strength.

A few awkward pieces of evidence remained. For one thing, my phone lights up when it rings. This posed a serious theoretical problem in itself. My early, functionalist theory argued that a light-up phone would give deaf people a way to know someone was calling them, but the limitations of this explanation quickly became apparent. Regardless, the light suggested that electricity was involved somewhere. An alternative explanation was that a previously unknown, luminiferous substance was responsible for getting the light to my phone. This idea fit much better with the theory as it stood. But there were other things. In addition to the light, I remembered once seeing a film where a woman gets electrocuted in the bath, with the assistance of her phone. (The cleverer-than-usual murderer deliberately failed to cut the lines.) For a while, I held out the possibility of integrating this stubborn fact with the received framework. Sadly, the discovery on the bottom of the phone of a sticker that read 'WARNING! Do not disassemble - Risk of Electric Shock' finally brought the whole thing crashing down. The theory was in ruins.

The ensuing paradigm shift was traumatic. Theoretical problems multiplied. I persevered. I rapidly worked out a new and sophisticated sub-atomic account that explained successful phonecalls in terms of spin, charm, and sex-appeal. Experimental validation was a problem, but a burst of insight led me to unplug the phone. The green light went off and the phone never rang. It doesn't ring very often when it's plugged in, either, but local experimental conditions explain this. The difference was clear, replicable and perfectly audible. The problem evaporated. I named the effect 'Bells Inequality,' sent the completed paper to Social Text and waited for tenure.