Should Londoners be afraid?

If, as the Metropolitan Police commissioner, Sir Ian Blair, said, Thursday's attackers in London "meant to kill", and if, as he did not say, they were Al Qaida or its affiliates, then the whole episode is in one sense rather reassuring.

They may have meant to kill us, but they clearly weren't very good at it this time.

Yet though the attacks themselves caused rather fewer casualties than the recent opening of a new Ikea, there was one significant development for the terrorists on the Tube.

Drivers on the Piccadilly and Bakerloo Lines refused to work, even though their lines were unaffected by the incidents.

A significant part of the disruption to the service was caused not directly by the attacks, but by London Underground staff withdrawing their labour.

Nobody should criticise individual Tube drivers. They have to spend a lot more time down there than we do.

But their action, together with reports of panic and stampedes at some of the affected stations, is the first, embryonic evidence of something the terrorists signally failed to provoke a fortnight ago terror.

We were rightly proud of our defiance on 7/7, but defiance after the event is relatively easy. If, as seems entirely possible, Thursday signals an "event" of indeterminate length, a concerted campaign of attrition, our sang-froid will be much more seriously tested.

So we need to start thinking now about how to defend ourselves psychologically and physically.

Luckily, perhaps, terror, like beauty, is in the eye of the beholder. It is somewhat within our control. We cannot refuse to be bombed; but we can, to some extent, refuse to be terrorised.

We can construct defences in our own minds against terror. Not, perhaps, the simple "we-are-not-afraid" type of bravado, which sounds to me a bit like denial but a cool-headed, informed understanding of our adversaries' strengths and weaknesses. This would tell us that we should be afraid, but only a little.

Bonanza of clues

Thursday was not just an attack, but an opportunity our best opportunity yet to learn about the people who threaten us. The attackers were kind enough to leave a bonanza of clues for the police which should allow them to tell us what the attack actually was.

If the bombers were indeed from the same group as the 7/7 attackers, and were indeed mounting a failed follow-up mission, their seeming amateurism is rather good news.

It suggests that our enemies may, for the moment at least, be at the end of their resources. The bombers themselves may well be caught. They might well provide useful intelligence on whoever briefed them.

More broadly, Al Qaida may want to kill us. But the record that they simply do not have the capacity to make more than a handful of attacks a year and then nearly always in the Muslim world or Africa, where its support is widespread and governments are ineffective.

There have only ever been five successful Al Qaida inspired operations in the West. Thursday's attacks, if shown to be Al Qaida inspired, would be the sixth.

Most Al Qaida operations kill only small numbers of people, or none at all. Since 9/11, the average death toll in Al Qaida-linked attacks has been 32.

The average lethality of Al Qaida linked attacks has more than halved in the last two years. So far this year, including 7/7, the average number of deaths per attack has been 15.

Even the IRA, with its safe havens in South Armagh and the other nationalist ghettoes of Ireland and its longstanding sympathiser network, could mount no more than a few attacks on the mainland each year.

Al Qaida's task, with no host community here, is more difficult.

It is reasonable to assume that terrorists planning a campaign will make their most dramatic attack first, when the authorities' guard is down and the shock value is greatest.

Nobody should minimise 56 dead, but it was significantly less dramatic than we might have feared. Thank goodness it is Charles Clarke, rather than David Blunkett at the Home Office. Step forward that most unlikely heroine of the new London Blitz, Kimberly Quinn.