Memory is a very different thing from record. I used to keep a detailed diary, and I'm always struck that I didn't even mention some of the moments that stick in my memory - and did record many moments I don't now remember at all

While records are contemporaneous (in the sense that they're created at the time), memory is formed retrospectively based in part on the importance of a moment in a broader narrative (and we don't, of course, see that narrative at the time)

In other words, it's hard (if not impossible) to tell at the time what moments will be memorable for you in the future. It's not that either recording or remembering is more reliable than the other - they are two entirely separate activities that deal with the past differently

Recording is subject to a very severe bias, i.e. what someone considered to be important *at the time*, which is like asking an ant to describe an elephant

For example, in the mid-1990s I thought the world wide web was a stupid passing craze, like shellsuits...

I think we're all familiar with how unreliable memory can be. But while memory is a very different experience of the past from records, in some respects it's *more* reliable, insofar as it is better at identifying the most salient events

So the lesson is that oral history derived from memory isn't better or worse than conventional history derived from records - it's just different. And historians working with records must be careful never to project a kind of omniscience onto the creators of records

Those who were writing about stuff at the time usually had no idea how important, or unimportant, it was...

The specific example that brought this to mind was my memory of lying awake all night when the Iraq War began, listening anxiously to the radio. I don't even mention this in my diary. It didn't fit the image I was trying to record of myself, or something