## LESSWRONGLW

Two relevant things.

First, the Epsilon Fallacy: the idea that effects are the result of many tiny causes adding up. In practice, 80/20 is a thing, and most things most of the time do have a small number of "main" root causes which account for most of the effect. So it's not necessarily wrong to look for "exactly one cause" - as in e.g. optimizing runtime of a program, there's often one cause which accounts for most of the effect. In the "logical-and" case you mention, I'd usually expect to see either

• most of the things in the and-clause don't actually vary much in the population (i.e. most of them are almost always true or almost always false), and just one or two account for most of the variance, OR
• a bunch of the things in the and-clause are highly correlated due to some underlying cause.

Of course there are exceptions to this, in particular for traits under heavy selection pressure - if we always hammer down the nail that sticks out, then all the nails end up at around the same height. If we repeatedly address bottlenecks/limiting factors in a system, then all limiting factors will end up roughly equally limiting, and 80/20 doesn't happen.

Second: the right "language" in which to think about this sort of thing is not flat boolean logic (i.e. "effect = (A or B) and C and D") but rather causal diagrams. The sort of medical studies you mention - i.e. "saliva is a risk factor for cancer but only if taken orally in small doses over a long period of time" - are indeed pretty dumb, but the fix is not to look for a giant and-clause of conditions which result in the effect. The fix is to build a gears-level model of the system, figure out the whole internal cause-and-effect graph.

3MoritzG6moRight, one could expand the clause indefinitely, that is kind of what I meant by "can only find what you are looking for". But that only means it is hard, not that it is bad to think that way. I do neither think of it as logic nor as causal diagrams nor Bayesian nor Markov diagrams but simply as sets of some member type that may have any number of features/properties/attributes that make them a member of some subset. When I wrote "A AND B" I wanted you to understand it as a dual logic clause, but only for simplicity. The way I really think about it is: attribute magnitude to impact function and then some form of interaction function that is neither only AND nor OR but possibly both to some degree. We have to deal with negative correlation in some way, I do not see how that is possible if it is always OR. They are nice on paper but I can not see how they are useful. To me they seem like some synthetic made up way to get the result, unfit to model the world. "If the world would not be as it is, it would be mathematically correct to do this." is so academic. As far as I understand it, the graph can not be cyclic. Since you do not know if the graph is cyclic and what factors are in the cycle you do not know which factors you must treat as an aggregate. The only directions known are those that go into the graph. There is only one joint probability for cases where there were multiple causal paths to one feature/property. Think of a hospital. Sick people go to hospitals, but sometimes people in a hospital will catch an infection that is only typical in hospitals. A= person is sick B= person is in hospital C= person has hospital infection C is a subset of A A causes B B causes C How do you work with that? "the fix is not to look for a giant and-clause of conditions [but] to build a gears-level model of the system, figure out the whole internal cause-and-effect graph" I thought that was what I was suggesting. Instead of stopping at: "It has to do with gears." keep

I think of "gears-level model" and "causal DAG" as usually synonymous. There are some arguable exceptions - e.g. some non-DAG markov models are arguably gears-level - but DAGs are the typical use case.

The obvious objection to this idea is "what about feedback loops?", and the answer is "it's still a causal DAG when you expand over time" - and that's exactly what gears-level understanding of a feedback loop requires. Same with undirected markov models: they typically arise from DAG models with some of the no... (read more)

# 9

Is there a name for and research about the heuristic / fallacy that there is exactly one cause for things? How come we do not look for the conditions that cause but for a cause?

I see this almost as often as the correlation = causation fallacy. When it comes in the form of "risk factor" it is ok if the factor is selective. But when it comes in the form of a general assumption about the world I find it simplistic. A risk factor is only a vague hint that needs to be looked at more closely to establish causation.

There also is this notion that multi causality is additive as would be the case if the probability for something would depend on this OR that happening but not this AND that.

A correlation of less than one may be random, but there might also be a hidden more selective cause/factor.

In medical news I keep hearing of risk factors for a condition. They find that there is a correlation between A, B and the studied disease. But how do we know that it doesn't take A and B and C to make it almost certain to develop that disease? I would like to know. C might be a common gene that is not even known.

Say it takes A and B. I really enjoy A, but I never do B, then why lower my life quality just because a study including people who also do B found that A is a risk factor? Risk factor is only a positive correlation. Eating and breathing have positive correlation to all diseases and the joke is, they come out with news about bad diets every year.

I keep hearing that A is a risk factor, then a follow-up study finds that there is no conclusive data for A being the problem, so A is cool again. But what if A and B is the problem and each alone is not harmful?

In the end this means that you can only find what you are looking for. (Kind of the big problem with science.) Looking for 1:1 correlation you will only find the low hanging fruit and the singular cause.

Whenever we find that some but not all who do/have A get Y we should look for additional factors, but this is not always done. As soon as A feels restrictive/selective enough the finding gets blown out of proportion. The reality might be that all who have A and B get Y, which would be a lot more informative. Who cares to know that breathing causes respiratory problems? Now that might seem silly and far fetched but how often have you heard that some common behavior is a risk factor?

# 9

New Comment

johnswentworth

### Dec 05, 2019

5

Two relevant things.

First, the Epsilon Fallacy: the idea that effects are the result of many tiny causes adding up. In practice, 80/20 is a thing, and most things most of the time do have a small number of "main" root causes which account for most of the effect. So it's not necessarily wrong to look for "exactly one cause" - as in e.g. optimizing runtime of a program, there's often one cause which accounts for most of the effect. In the "logical-and" case you mention, I'd usually expect to see either

• most of the things in the and-clause don't actually vary much in the population (i.e. most of them are almost always true or almost always false), and just one or two account for most of the variance, OR
• a bunch of the things in the and-clause are highly correlated due to some underlying cause.

Of course there are exceptions to this, in particular for traits under heavy selection pressure - if we always hammer down the nail that sticks out, then all the nails end up at around the same height. If we repeatedly address bottlenecks/limiting factors in a system, then all limiting factors will end up roughly equally limiting, and 80/20 doesn't happen.

Second: the right "language" in which to think about this sort of thing is not flat boolean logic (i.e. "effect = (A or B) and C and D") but rather causal diagrams. The sort of medical studies you mention - i.e. "saliva is a risk factor for cancer but only if taken orally in small doses over a long period of time" - are indeed pretty dumb, but the fix is not to look for a giant and-clause of conditions which result in the effect. The fix is to build a gears-level model of the system, figure out the whole internal cause-and-effect graph.