“Stay alert”, says Boris Johnson. “Stay at home”, says Nicola Sturgeon. The instructions may not be mutually exclusive, for an alert person may, if sceptical, judge that it’s more sensible to stay at home than to venture back into the workaday world. That said, one might conclude that Johnson has more confidence in the good sense of people in England than Sturgeon has of people in Scotland.

As a Remainer who broke the habit of voting Conservative several years ago, and as someone who has thought of Johnson as at best an occasionally amusing but lightweight newspaper columnist – and at worst a downright dishonest one – I’m surprised to find myself, at this stage in the Covid-19 drama, thinking that he may now have got it right.

No doubt mistakes were made and the lockdown should perhaps have been imposed a week or two earlier. No doubt too that we were insufficiently prepared, and that management has been easier and more successful in a genuinely federal state such as Germany – though on the other hand this certainly hasn’t been the case in the USA. It also seems that we haven’t had the resources –or perhaps the will?- to pursue the policy of “test, trace and isolate” which has apparently been so successful in New Zealand and Australia.

Nevertheless, when all reservations have been made and caveats entered, Johnson’s route for a gradual return to normality seems sensible. Others are quite naturally doubtful. Writing in The Guardian, Paul Hunter, Professor of Epidemiology and Medicine at Oxford University, says: “if there is evidence from modelling that social distancing while at work – rather than sheltering at home – is sufficient for virus control, then let us see it…”

This may seem a fair request, though some of us may be sceptical enough to remark that evidence from modelling is not hard evidence. Models, as we know from the arguments about past epidemics or indeed climate change, are theoretical, predictions which tell us what appears likely to happen in certain circumstances. Yet future circumstances are by their nature uncertain. A model is a guide for a voyage into the future. It may prove to have been so well designed that it is a good guide, but then it may not.

Now, while there may be no evidence from modelling about the efficacy of social distancing while at work, there is a good deal of real life evidence, even if this may not have been collected and analysed. After all, while millions of us have been locked up, millions have also continued to work in hospitals and care homes, in retail and other service industries, in manufacturing, farming and construction. The police have been on the streets and in their cars. The Army has helped build emergency hospitals and deliver supplies. Buses and trains have continued to run.

Given that the number of deaths from the virus has been falling, with 40% of them in the last week or fortnight being in care homes, it is reasonable to assume, even in the absence of a persuasive model, that the millions who have been working have been practising social distancing sufficiently to have at least exercised some control over the virus. If this wasn’t the case, the number of cases would be much higher than it is.

The Prime Minister tells us to be “alert”. Scotland’s First Minister prefers the word “vigilant”. Neither is perhaps the best word, though “alert” is better than “vigilant” which has an unhappy association with vigilance committees, scapegoating, and ultimately lynching. It should be enough to tell us to be careful, a word that everybody understands, and to use our common sense.

At some point we have to edge back to normality. Other countries, a week or two ahead of us in the coronavirus stakes, are already doing so; Italy, as severely hit as Britain, is about to permit bars and restaurants to open. The Prime Minister has announced only a few modest steps. If people are sensible and take care, further relaxation should follow soon.

Of course, there is a risk, but there would likewise be a risk in following any model, even one produced in the University of Oxford. Finally, it might also be dangerous to maintain the lockdown as it is now. One already knows that it is not without its harmful effects. It would surely be even more dangerous to do so when we reach the point at which it no longer enjoys the support of at least half the country.