Defining a causal relationship

Peter

Prior to the LSAT I've never really considered why A causes B. I'm very unsure of my definition of a causal relationship and would like to put it out there for everyone to critique.

A causes B if and only if

1. B occurs whenever A occurs (A ->
2. B does not occur if A does not occur (B -> A)
3. A occurs temporally before B

For example, say we know that toxin T is strongly correlated with fish dying in large numbers. Say we also know that in the absence of toxin T, fish populations are always high. Lastly, we know that toxin T is always released prior to any die off of fish. Can we conclude toxin T _causes_ the decrease in fish populations?

Kristina

I would say that, no, we cannot conclude with certainty that "A causes B." We can only conclude that "A causing B" is more probable (given the conditions you cite) than it would be if these conditions weren't true. My 2 cents.

TC

Hi Peter,

First of all, for an explanation of causation, I would try to stay away from making a bi-conditional like if and only if.
The reason for that is if and only if introduces A<-->B relationship.

A causes B, means that everytime we have A as the cause, we get B as the effect.

For example, we know that if we introduce chemical T in the water, chemical T CAUSES the EFFECT of fish developing physical deformation.

T Physical D.
--------- -------> --------
Cause Effect

Does that mean that if we see a fish that has a physical deformation, that it was caused by T? Maybe, or maybe a thousand other things caused the physical deformation, so from the Effect alone it's hard to trace a cause without some extra evidence.

Now, the LSAT tries to imply several times that there exists a causation relationship between two things. This is 99% of the time wrong. There might be a strong correlation between the two things, but essentially, there is no causation. Why? Because it's extremely hard to prove causation, and even harder to prove it in a 3/4 line stimulus. There will be times where your job as an LSAT taker will be to specifically to show how this causation doesn't exist.

I really really, really, recommend you to review your lesson on http://7sage.com/lesson/causation/

Also, from your example, we can see that there is a possible correlation. But no, causation proved. An alternative cause is for example, that one of the employees puts poison in the water, right before releasing T.

Peter

Thanks all for the detailed responses

@NoRestForTheWicked, I probably should clarify the if and only if signals a definition here and not A <-> B relationship. I'm saying "A causes B" if and only if three conditions are satisfied.

You mention that if we introduce chemical T in the water, chemical T causes the effect of fish developing physical deformation. What are the conditions under which you can conclude the cause-effect relationship in that example?

What prompted my question here is PT45 S1 #12. It asks to weaken the conclusion that a causation relationship does not exist, which I interpreted (perhaps falsely) it means to strengthen the argument a causation relationship does exist. The answer gets at condition 2 showing B does not occur in the absence of A.

I guess the question comes down to: when can we conclude a causal relation exists? I often see LSAT questions that ask you to refute an argument that establishes these relations but it doesn't escape my imagination it can also ask you to make an argument for establishing them.

TC

Hey Peter,

Sorry, I thought you were talking about A is sufficient and B as necessary if and only if. I think you can see where I was coming from, since, If introduces sufficient, and only if, the necessary. Hence, A<-->B confusion.

It would be extremely difficult. I suppose, if we were to have a make-believe proof of causation, I guess it could be the following:
We have three cloned fishes that are identical genetically and physically from each other.
We have a closed environment where everything can be controlled, down to sodium, etc.
We introduce chemical T to the first fish -> it leads to physical abnormality in X amount of time.
We introduce chemical T to the second fish-> it leads to physical abnormality in X amount of time.
The time when chemical T is introduced to the first and second fish is the same.
We have proof that it is not the food that could lead to that physical abnormality.
We have proof that it is not the chemicals in the water aside from chemical T.
We have proof that they were not born predisposed to such a condition.

I guess even with all that, it still isn't enough to prove causation. But there is evidence of a very high correlation. The point I'm trying to make is how HARD it is to prove causation.

For example, in the scenario above, suppose one of the employees working on the research was being paid by a competing company to alter the results...
But every time you increase the correlation, your argument of causation strengthens.

PT45 S1 #12 is a great question!
Indeed, the answer is for you to show that causation is more likely, but keep in mind that causation has not been proven.

When an argument has a causal relationship in weakening questions, there are 3 ways to weaken that relationship
Provides an alternative cause
Provides an example of the presumed cause without the presumed effect
Provides an example of the presumed effect without the presumed cause.

The stimulus provides the claim that D is unlikely to be the cause, due to D being present in the environment and yet the fish recovered normal hormone levels.

Dioxin ---------Change in Hormones
_______ -----> _____________________
Cause ---------- Effect

To weaken this relationship, the author provides an example of the presumed cause without the presumed effect.

Dioxin------- NO Change in Hormones(after a short period of time)
_______ -----> _____________________
Cause ---------- Effect

Answer choice C says that dioxin was washed far down stream. Making Dioxin to not be present, therefore, undermining the argument that Dioxin did not have an effect. You see, the stronger you make that correlation, the stronger your claim of causation becomes.
Another way that the LSAT could have made it stronger would be to show the fish actually had not recovered normal hormone levels.

You should also review PT26, S2, Q12, and PT28, S1, Q5. Those two questions are similar in a way to this question.

Let me know if that makes sense! Hope that helps!

Peter

Hey thanks for the response! I realize my logical fault is confusing trying to strengthen a relation with trying to prove it.

TC

Glad I could help, Peter!

msdaisy16

I agree with TC Merlini. Causation is not the same as formal logic. Causation assumes something caused another simply because they happened at the same time. U shouldn't use arrows to draw out causation. You can use lines though such as C------E to represent cause and effect.

J.Y. Ping

I think our curriculum on causation is the least clear out of all the subjects. Don't get me wrong, it still kicks the "major test prep companies" collective asses out of the stadium. I just mean compared to our curriculum on Logic, Grammar, Games, etc.

This is because causation is a very unclear for me, philosophically. I think it does suffice for LSAT, but I'd like to have more clarity on the matter.

I'm pretty sure there are different types of causations that I'm failing to distinguish among.

Anyway, love the discussion here and if you guys have time, you should check out these articles:
http://plato.stanford.edu/entries/causation-probabilistic/
http://plato.stanford.edu/entries/causation-counterfactual/
http://plato.stanford.edu/entries/causation-process/
http://plato.stanford.edu/entries/causation-metaphysics/

TC

Hi JY, now that you provided those links and I was unfortunate enough to read some of them, you got me thinking.

If I say eating a whole pizza causes me to not feel hungry.
Someone goes and says eating a bucket of Icecream causes you to not feel hungry. Therefore, your theory of eating pizza is wrong.

In a weakening question, showing an alternative cause is usually considered to weaken a question involving a causal relationship.

However, the problem that I'm having is that, such as the example above, both could very well be the cause and showing an alternative cause would not weaken either theory.

It seems that the appropriate way to use an alternative cause to weaken a question should be something along the lines of this example:

I ate a whole pizza and that caused me to not feel hungry.
You also ate a bucket of Icecream, and that caused you to not feel hungry. Therefore, your theory of pizza is wrong.

(keep in mind that I understand that both examples are flawed, they are only being used to illustrate the issues)

So it seems that unless the second cause touches the occurrence of the effect that occurred on the first cause, it doesn't seem to weaken it.

Anyways, I guess it comes down to whether a causal relationship can be weaken, when the alternative cause doesn't occur within the occurrence of the first causation?

Please let me know your thoughts.

TC

TC

anyone?

Gracelover

Hello! I have another question regarding causation arguments! So is it an assumption to think that, just because a causation relationship has been established between A and B, that in every instance in which A occurs, B will automatically result? We cannot assume that just because a cause happens, an effect happens, correct? It's not like a conditional relationship...

Like say for example, say that a doctor tells you that eating lots of chocolate causes disease X.....but that doesn't mean that if you eat lots of chocolate you will get disease X, correct?

Rev_Lefe

I found surprising only a few discussions in this forum about counterfactual causation theory. In addition to the links JY mentioned, a feed on Wikipedia related to this topic was also helpful to me.
https://en.wikipedia.org/wiki/Causality#Theories

Here is another one made by a philosophical website:
https://www.iep.utm.edu/causatio/#SH4c

Hope these would help!