The prime minister likes to say that 'no deal is better than a bad deal'. Ministers pretend they'd walk away from Brexit negotiations. A substantial section of the British political class, from journalists to think tank bosses to politicians, seems to feel a frisson of manly excitement at the prospect of no-deal.
But once you start asking detailed question about exactly how that would work, things very quickly go downhill. Brexit secretary David Davis' Kafkaesque appearance in front of a Commons select committee today provided an object lesson in how that process works. During it, he all-but admitted that the government would never accept no-deal. Whenever the proposition is subject to scrutiny, it is chiselled away into non-existence.
What kind of no-deal?
Early on in the session, Davis differentiated between "various sorts of no-deal". We are once again through the looking glass in terms of language. In Davis' world, there are various types of nothing, each with unique qualities.
It soon became clear that when the Brexit secretary says no-deal, he actually means that there would be no trade deal. This would indeed be catastrophic for Britain, cutting it off from its largest trading partner. But it's only half the story.
With no agreement in place, Britain would also lose the ability to manage nuclear materials or aviation in the manner it does now, along with countless other legal and regulatory relationships which currently operate under the umbrella of EU membership. This scenario, though, was "so improbable it's off the scale", the Brexit secretary said. This means that when Davis says 'no-deal', he actually means 'a deal', just one that excludes the future trading relationship.
He isn't really threatening to walk away from the table. He's just threatening to not sign a trade deal. These are very different things.
How to prepare?
The Brexit secretary insisted there were "contingency plans" for what to do with customs in the event of a no-deal outcome. Work was being undertaken to figure out the infrastructure - roads, buildings, checkpoints etc - as well as the staff and IT systems that would be required.
We can only hope he's right. A recent Institute for Government report found civil servants had a "patchy" understanding of the customs process and needed to be massively levelled up in a huge cross-departmental project if they were going to be able to deal with the ramification of leaving the EU.
HMRC’s Customs Handling of Import and Export Freight can process about 60 million declarations a year. It will need to process an additional 200 million when we leave the EU. A planned upgrade, Customs Declaration Services, is only designed to manage 150 million. And even that, which was due to come online in January 2017, has been delayed because it is facing "significant issues".
But this is only half the story of being able to deliver on a no-deal threat. Because to do that you don't just need to plan for these things, you need to actually build them. You need to lease the buildings, expand the roads, set up the checkpoints, hire the staff, put in place the systems.
This is all cripplingly expensive, but more importantly it presents the UK government with an unsolvable logical puzzle: You don't know what you need to do until it's too late to do it. If there is a trade deal on tariffs but not country-of-origin, that entails one arrangement. If there are special measures on agricultural products, that is another. If there is some sort of deal on single market access, that's another. There are countless options. And the only way to know which one you need is to complete the talks. But if you wait until you complete the talks, you can't prepare for a no-deal outcome.
There is no solution to this problem, as point of logic. And because Davis is unable to defeat the reality of logic he was unable to offer one today.
What would it look like?
And once no-deal was activated, what happens then? Davis was asked today about the 57 impact assessments into the effect of Brexit which the government is trying to keep secret. We got no explicit details about their content, but there was an extremely revealing moment in which he felt the need to discount the entire concept of forecasting future trade flows.
The models for establishing future trade had changed completely, he insisted. High value products were becoming smaller and lighter - iPhones were more profitable than cars, blah blah blah - and that threw all the calculations out of whack.
Because the government is so intent on keeping these reports secret, this was one of the most accidentally informative things Davis could have said. After all, his views on the pointlessness of assessing future trade sit rather uncomfortably with the fact he actually commissioned those reports. He chose to do them and he presumably would have had significant influence over the models they employed. That suggests that he was not always of such a downbeat opinion about the concept of impact assessments.
So why berate it? Might it be because the impact assessments are catastrophically negative? That they paint a picture of a country about to impose a brutal form of impoverishment on itself? And that no-deal would be the worst possible variation on this policy? It just might.
Take those three points together: In just over an hour, Davis conceded that leaving without a whole range of technical agreements was "off the scale" improbable. He had no answer for the logical conundrum of customs planning. And he hinted that the government's own research into what it would entail is extremely bleak.
The truth is that Davis' talk of no-deal is a bluff. No-one in their right mind would ever countenance it. The trouble with this bluff is that everyone can see it: the MPs on the committee, the public watching on TV, and the negotiators in Brussels. No matter how elegantly he describes the angle of the barrel, everyone can see he is holding the gun to his own head.
Ian Dunt is the editor of Politics.co.uk. The new edition of his book - Brexit: What The Hell Happens Now? - is available to pre-order now.
The opinions in politics.co.uk's Comment and Analysis section are those of the author and are no reflection of the views of the website or its owners.