I suppose the lesson is that you shouldn’t blog in the heat of the moment. In yesterday’s blog (Their game) I reacted to the news that a relative of  Abu (the man the police had thought was yesterday’s attacker on Westminster Bridge) had called Channel 4 News to say that Abu couldn’t have caused the mayhem at Westminster yesterday because he was in jail. I reacted in haste and without thinking clearly, forgetting that the attacker was dead – thinking instead that he had escaped. Since he was dead, the idea that the police would now go searching desperately for someone else to blame instead of him was nonsense.

I made this careless mistake partly because, like everyone else, I was upset at the events as they unfolded. And also, in the week that Martin McGuinness died, I thought of all the other times that the police, desperate to get “a result”, have arrested the wrong people, cooked evidence, and sent people to jail. The Birmingham Six and the Guildford Four sprang to mind. I also knew some of the people involved in the Carl Bridgwater case, where the wrong people were convicted of murder on manufactured evidence and served several years in jail before having their convictions quashed. And thus I blogged. But in this case I got it wrong and I apologise.

But the police got their identification wrong too and named a person who had not committed the attack. They now say the attacker’s name was Khalid Masood. If, of course, he had escaped, and the police had gone searching for Abu (we don’t know the basis of the original misidentification) they would presumably have quickly found him in jail. After that, the scenario I described yesterday would not have been such an unlikely one. While we all depend on the state authorities to protect us, we should also keep our eyes on them. Their main job is to defend power and we should not trust them overmuch.

What I forgot in yesterday’s furore were the words of the late, great Stephen Jay Gould, giving advice about publishing research (in his case, in the disciplines of evolution and palaeontology). Before you publish, he said, you should “examine, ponder, ruminate, observe again, obsess, reconsider” and then, and only then, “eventually publish.”

I seem to have missed all those steps yesterday, except the last one.