If we have learned anything over the past 18 months it is this: that the first rule of politics - power must never be trusted - still applies. The government will neither regulate itself nor be regulated by the institutions which surround it. Parliament chose to believe a string of obvious lies. The media repeated them, the civil service let them pass, the judiciary endorsed them. The answer to the age-old political question - who guards the guards? - remains unchanged. Only the people will hold the government to account.

They have two means of doing so. The first is to throw it out of office at the next election. This works only when we are permitted to choose an alternative set of policies. But in almost every nation, a new contract has now been struck between the main political parties: they have chosen to agree on almost all significant areas of policy. This leaves the people disenfranchised: they can vote out the monkeys but not the organ-grinder. So voting is now a less important democratic instrument than the second means: the ability to register our discontent during a government's term in office.

Applying the first rule of politics, we should expect those in power to seek to prevent the public from holding them to account. Whenever they can get away with it, they will restrict the right to protest. They got away with it last week.

The demonstrators who have halted the construction of the new animal testing labs in Oxford...

I have read a lot of Fourth Amendment cases over the last few years, but today I learned something new: several courts of appeals have allowed the government to obtain and execute "anticipatory" search warrants. According to these cases, the government can get a warrant even if their case for probable cause hinges on some future event. If the future event occurs, the warrant becomes operative and they can execute the search. If the future event does not occur, then the warrant is not yet operative and they cannot execute the search.

... the problem with this is that a "future event" isn't necessarily a simple, clear-cut incident. It might be something that's unmistakably black-and-white, but it also might be something based on the suspect's behavior that's a bit of a judgment call.

And that's disturbing. The whole point of a warrant is that it prevents police from making their own judgment calls and requires them to make their case to a neutral judge if they want to execute a search. I wonder how long this has been going on and how common it is?

This sounds really bad in the abstract, but if the concrete manifestations are limited by normal legal standards of "reasonable" behavior, it might not be that bad.

Think of it: Policeman wants warrant to search when the impending meeting with Mendoza takes place (you know, when the drugs will still be there). Is it so wrong to give it to him? And to trust him to recognize that a meeting with Mendoza has (finally) taken place?

It's certainly not more epistemological discretion than we already grant our police, although we might want to reconsider.

Six years ago, the American sociologist Alan Wolfe published a strikingly important book. Entitled "One Nation, After All," and subtitled "What Middle-Class Americans Really Think About," it is an essential text for understanding the pulse of modern America. What makes it both important and essential is that Wolfe painted a picture radically at odds with the exaggerated perception, both in the US and abroad, of America as a nation of entrenched and embattled ideological extremes.

In fact, Wolfe argued that middle America was not so much a land of culture wars as of cultural pragmatism. "I have found little support for the notion that middle-class Americans" - a category within which three quarters of all Americans define themselves - "are engaged in bitter cultural conflict with each other over the proper way to live," he observed.

"Reluctant to pass judgment, [Americans] are tolerant to a fault," he concluded. "Not about everything - they have not come to accept homosexuality as normal and they intensely dislike bilingualism - but about a surprising number of things, including rapid transformations in the family, legal immigration, multicultural education and the separation of church and state. Above all moderate in their outlook on the world, they believe in the importance of leading a virtuous life, but are reluctant to impose values they understand as virtuous for themselves on others; strong believers in morality, they do not want to be considered moralists."

The essence of Wolfe's case is that the great wedge issues of the late 20th-century culture wars have simply shrunk in significance. The most important of these, as always, is affirmative action on race, where the supreme court has managed to strike a sensible compromise. Nor, he argues, does abortion still have the divisive potential of the past, though if a re-elected Bush attempts to nominate a supreme court dedicated to overturning the landmark pro-abortion Roe vs Wade judgment of 1973, that could change. Having won the political argument over what it calls partial birth abortion, though, Wolfe reckons the right is less angry than it was.

There's much about America in 2004 that bears this out. Over the past couple of months, the president has spent $50m on campaign ads designed to promote his opposition to gay marriage. As Wolfe's original research found, gay equality remains one of the issues on which middle America remains to be convinced; yet you would have to search long and hard to find many people who believe that gay marriage is the great dividing issue in America. At the margins, Bush's advertising may help to motivate some social conservatives to vote Republican, but mostly it has sunk without trace.

A possible explanation is that the polarisation of 2000 and 2004 is simply untypical - most US presidential elections are not nearly so close as the last one was and the next one promises to be. In that case, some special factor - the disabling effect of the Clinton scandals on the Democratic cause in 2000, perhaps, or the mistrust towards Bush's Iraq policy and his tax cuts this time around - may have made these two contests more impassioned than they might otherwise have been.

A second is that the practices of modern campaigning and media - giving voters a relentlessly inaccurate picture of the choices they face, presenting their own candidate in an unbelievably favourable light and their opponent in an equally unbelievably negative light - conspire to create a polarised contest between core electorates and to drive down participation. As US journalist Jack Germond says in his new memoir, the Republicans do not have a monopoly on such tactics - they just seem better at it.

There is, of course, a third possibility: that Wolfe's "one nation" theory is just wrong.