It is natural to assume that, if a majority votes No in the referendum on Britain’s EU membership, we shall then leave. It is not automatically so. After the vote, we would still be members. The government would then — morally at least — be mandated to negotiate Britain’s withdrawal. In theory, unlikely though it may currently seem, the EU could try to block this. Even assuming that it did not do so, the eventual terms of the withdrawal would not automatically be agreed by Parliament and would not necessarily correspond with the wishes of those who voted No. The context for our vote will be David Cameron’s presentation of a package secured with partners to persuade us to vote Yes. There is no negotiated package offered for those voting No. It could well be that, as so often happens in other EU countries who vote the ‘wrong’ way, there will be a second referendum. The pro-Yes government could say, ‘You, the public, voted No. We’re afraid that the following terms for withdrawal are the best we can secure. Are you sure you want them?’ It might turn out — particularly if the first result had been close — that we were not sure, after all. In short, as in the referendum on Scottish independence, there is a ‘What would actually happen?’ problem which is much more dangerous for the side which wishes to change the status quo. Alex Salmond could never answer the ‘What will happen to the pound?’ question. There will be equivalents, at least as grave, thrown in the path of the No campaign. At present, those of us who incline to getting out are the victims of a false logic, which states that because EU membership is bad, British independence is automatically better. Nothing is automatic here: every single point has to be answered, every possibility thought through.

‘Rolf Harris has written a ghastly song about his victims,’ said Justin Webb on the Today programme. I’m sure he is right. But why do we have to be told how we are supposed to react to anything said or done by a convicted sex abuser? The BBC doesn’t start an item by saying ‘King Jong-un has made a ghastly speech’ or even ‘Isis has released a ghastly video.’

Funny how, in all the excitement about the 800th anniversary of Magna Carta, there has been hardly a word about its first clause. This says that ‘the English Church is to be free, and is to have its rights in whole and its liberties unharmed’. You could argue that the whole history of England turns on this struggle, on what people mean by the freedom of the church. Catholics would say it was strangled by Henry VIII. Anglicans would argue that the eventual settlement, though creating a society where church and state were far from separate, guaranteed church liberty from foreign (i.e. papal) control and protected it through Parliament. Anyway, it mattered; no other liberty was secure without this one. Otherwise, the crown or — as we would now think of it, the executive — exercised ultimate power over beliefs and ideas (and land, come to that). As the Queen, the Archbishop of Canterbury and the Prime Minister gathered at Runnymede this week, a Church of England clergyman was suing the Diocese of Southwell because it won’t let him keep his licence to officiate as a priest because he has married a man. If the employment tribunal finds in his favour, the first clause of Magna Carta doesn’t mean much. Most people won’t mind, perhaps, but I doubt if its first clause is one of those, like clause 33 which removes riverine fish weirs, that doesn’t matter any more.

Clause 31 is good: ‘Neither we [i.e. the King] nor our bailiffs shall take wood belonging to another person for castles, or for our other affairs, unless with the consent of him whose wood it is.’ Again, this restriction on state power has much decayed since 1215.

The other great anniversary this week is, of course, that of Waterloo. I attended a very magnificent Waterloo Dinner in the National Gallery, given by the philanthropist Michael Hintze. We were eloquently addressed by Lt-General Sir Barney White-Spunner, who has just written a book (Of Living Valour) about the battle, based on contemporary accounts by participants, and by Peter Snow of swingometer fame. Superb dioramas of the charge of the Scots Greys and the struggle for La Haye Sainte were laid out on the dining table. They were built to fire off tiny explosions, but health-and-safety regulations in the gallery prevented this. Nothing could daunt our courage, however. After a Chassagne-Montrachet (Château de le Maltroye La Dent De Chien 2007), Leoville Poyferre 2001 and Yquem 1998, I had the most distinct and vivid memories of how bravely I had fought at Waterloo.

In a few weeks, I shall have finished the second volume of my three-part biography of Margaret Thatcher. I am now at the checking and revising stage — 3,000 endnotes to be made shipshape, 2,000 quotations to be cleared with interviewees, 300,000 words to be re-imagined as if read fresh. This involves the exchange of scores of emails every day. The question arises, ‘How did enterprises of this kind ever happen before computers?’ The answer, I think, is that the sources used were much narrower than they are today. Authors were extremely dependent on where archives physically were. Interviews were a rarity. Press, radio and television records were restricted, unobtainable or nonexistent. International exchange was much harder. We live in a golden age for historical research. It won’t last. My own work depends heavily on the fact that those involved in government kept excellent, confidential, largely truthful, written records. Now, because of emails and the Freedom of Information Act (FOI), they daren’t. No. 10 Downing Street has a system, quietly introduced just before FOI, in which all emails self-destruct after three months unless ‘actively saved’. For a brief period in history, a window opened and I have been able to climb through it. Now it is closing again.