Veillard's Razor
If the conclusion relieves you, audit the reasoning.
Desire is a poor witness.
The rule
I call this Veillard’s Razor, and also the Bitter Razor because of what it usually costs. In plain words it is the comfort principle: when a line of reasoning ends exactly where I hoped it would, that arrival is a signal to go back and check the steps.
The rule does not say that comfortable conclusions are false. It says that comfort is a reason to look harder. Relief is information about me, not about the world, and I should treat it as such.
Why it works
Two tendencies stack. The first is that I oversimplify. Real questions have more moving parts than I can hold, so I compress them into a version I can reason about, and compression drops details. The second is motivated reasoning: emotion steers the search for arguments. I do not usually decide to believe what I want. I simply find the supporting arguments faster and stop looking once they are in hand. The justifications that come out sound rational, or close enough, which is what makes them hard to catch.
Cognitive dissonance adds a push in the same direction. When a conclusion hurts, holding it is uncomfortable, and I become less logical precisely at the moment when logic would cost me something.
The clearest example I know is death. It is far more comfortable to believe that something follows it than to accept that this is probably not the case. The comfortable belief has had every advantage in the history of human argument, and it would be surprising if that advantage left no trace in how people reason about it.
Comfortable beliefs are not the property of one camp. “The policy I like has no costs.” “My group is the one that is right about everything.” “The years I spent on this must have been worth it.” Each of those is a relief dressed as a conclusion, and each belongs to someone, often to me. The razor is useless unless it cuts in every direction, including toward the beliefs I would least like to lose.
How to use it
- Notice the relief. The feeling comes first and the reasoning follows it. When an argument lands and something in me relaxes, that is the cue.
- Name the wish. Say out loud, or write down, what I wanted to be true before I started reasoning. If I cannot name it, I am probably still protecting it.
- Rebuild the argument without it. Take the wish off the table and ask whether the premises still carry the conclusion on their own. Often a step that looked solid turns out to be the wish doing the lifting.
- Ask what evidence would change my mind. If nothing would, I am not holding a conclusion. I am holding a preference.
There is a corollary I find useful in debates. In most disagreements, one side is arguing against its own wish or its own tradition. That effort is weak evidence. Somebody who dislikes where the argument goes and follows it anyway is more likely to be arguing from knowledge and logic than from appetite. So I grant that side a modest presumption, which the rest of the evidence can overturn. Reality is not obliged to be pleasant, and people who have accepted that tend to be more careful.
Examples
Each example below has two readings. The sweet one is the conclusion as it first feels. The bitter one is the same case after the audit.
A policy with no downside.
Sweet I support a measure and read that it has produced only benefits. Relief.
Bitter Rebuilt without the wish: every policy moves resources from somewhere, so the question is who pays and whether the trade is worth it. The honest version of my support includes the cost.
The other side is simply dishonest.
Sweet My political group loses an argument in public, and the comfortable explanation is that the winners lied. Relief again, this time at not having to update.
Bitter Rebuilt: what would I have to concede if they were sincere and partly right? That conceded ground is usually where the real question lives.
The decade already spent.
Sweet I gave years to something and now wonder whether it was the right call. The conclusion that it was worth it arrives quickly and feels good.
Bitter Rebuilt: would I recommend the same path to someone starting today, knowing what I know? If the answer is no, the time is still spent, but at least I stop paying for it twice.
Where it fails
The razor is a heuristic. Like Occam’s razor, it does not prove anything. It lowers the weight I give to arguments that happen to land where I wanted, and then leaves the decision to the remaining evidence. Three limits matter.
Tradition often encodes evidence. A practice that has survived a long time has usually been tested against reality many times, and the fact that I find it comfortable does not erase that record. The razor tells me to check the reasoning, not to discard the conclusion.
Effort can be sunk cost. The person arguing against their own interest may have thought harder, or may simply enjoy looking brave. The presumption I grant them is modest for exactly this reason.
Bitter does not mean true. The reverse of a comfortable error is not automatically a truth, and a conclusion that hurts can be wrong in the ordinary ways. If I start preferring painful conclusions because they feel more rigorous, I have found a new wish and should audit that one too.
Compared with other razors
| Razor | What it tells you to do |
|---|---|
| Occam’s | Prefer the explanation that needs fewer assumptions. |
| Hanlon’s | Do not blame malice for what incompetence explains. |
| Hitchens’s | What is asserted without evidence can be dismissed without evidence. |
| Veillard’s | Audit reasoning that lands where you hoped it would. |
The first three judge a claim. This one judges the person holding it, starting with me.
Try it
Prior work
None of the pieces here are mine. Upton Sinclair wrote that it is difficult to get a man to understand something when his salary depends on his not understanding it. Leon Festinger described cognitive dissonance in the 1950s and showed how people reshape their beliefs to relieve the discomfort of holding contradictions. Ziva Kunda’s work on motivated reasoning laid out how our goals shape the way we gather and weigh evidence while we still experience ourselves as objective. Robin Hanson and Kevin Simler, in The Elephant in the Brain, argue that many of our stated reasons are built after the fact to serve interests we would rather not see. What I have added is a short rule and a trigger that is easy to feel.
I expect the razor to cut me more often than anyone else, and that is the point.