Causal Reasoning

[koolaid]

Hi, I just want to confirm if my understanding of causal reasoning is correct...

Let's say "A causes B." I know that the fundamental assumptions are that
1. A always causes B
2. A is the only possible cause of B

I am a little confused when I think of the scenarios. Please let me know if any of the four scenarios below are incorrect.

1. If we know that A happens, we also know that B will happen. (Because A will make B happen)
2. If we know that A does not happen, I feel like there is still a chance that B can happen (by pure luck for example). Is this the right conclusion? Sure A makes B happen, but does this mean that B can only happen if A happens? What is the appropriate conclusion if all we know is that A does not happen?
3. This is similar to #2. If we know that B happens, do we know that A happened? My suspicion is that we can't draw a conclusion.
4. If we know that B does not happen, we know that A does not happen.

Thanks for your time!

[Adam Tyson]

Hey, Koolaid! (now, come breaking through the brick wall, please!)

Causal reasoning on the LSAT is a little bit different than causal reasoning in the real world. In the real world, I might say "cheese upsets my stomach," and you would not give me a hard time about it if coffee also upset my stomach. You would be okay with the idea that there could be many causes for one effect. You would probably also be okay with it if sometimes I had some cheese or some coffee and my stomach didn't get upset, right? It could still be true that those things upset my stomach, even if they don't do so every time.

In the LSAT world, we can behave a little differently. It's like we have stepped through a rip in the fabric of space-time into a parallel dimension, where it looks a lot like ours but sometimes (often) things are very different. In that world, if I say cheese upsets my stomach, you can treat me more aggressively. You can assume that I mean cheese always has that effect, that nothing else ever causes that effect, and the the effect never happens unless it is preceded by that cause. Because of all that, you can weaken my causal claim by showing that, just one time, something else (coffee, perhaps) caused me to have an upset stomach (that's an example of an alternate cause, one of our favorite ways to weaken a causal argument). You can weaken my claim by showing that just one time, I had cheese but no upset (cause without effect) or I had upset with no cheese (effect without cause). You can weaken my claim by just suggesting that maybe I have the cause and effect backwards - maybe when I have an upset stomach, it causes me to eat cheese? That's reverse cause and effect, a fun one. Or, you can weaken my claim by attacking any data, studies, or experiments that I used to come to my conclusion. Maybe other studies came up with different results, or my methodology was unsound, or I did a survey with an unrepresentative sample or biased questions, etc. We call those a data attack.

So, to your examples:

1. Correct! If the causal claim is valid, then when A (the cause) happens, B (the effect) must happen. If it doesn't, the original causal claim is weakened, maybe even disproved.

2. Not in the LSAT world! If the cause does not happen, then the effect should not happen. If it does, then the original causal claim is weakened, maybe even disproved. Effects don't just happen in the LSAT universe - they must be caused, or else the causal argument is no good.

3. Nope again! If the causal argument is accepted as valid, then whenever the effect happens, the cause must have preceded it. If not, the causal claim is no good.

4. Correct, if the causal argument is good. If the effect doesn't occur, the cause should not have occurred. If it did, then (you guessed it) the causal claim is weakened, maybe disproved.

For all of these reasons, causal arguments on the LSAT are extremely vulnerable. They come up a lot in weaken and flaw questions. You can also help fix those problems (strengthen) by, for example, eliminating an alternate cause, or identify an assumption made by the author (that there wasn't an alternate cause, or that they were not reversed, or that the data was reliable, etc.)

Remember, the LSAT universe isn't exactly like ours! Go ahead and be aggressive and unreasonable when dealing with them. Attack!

[koolaid]

Thanks Adam! I've got two more questions for you.

Question 1
Am I right to say that when an author presents causal reasoning, he/she believes that the cause and effect are in a "double-arrow" relationship? Just looking at the scenarios we discussed, it looks like if one occurs, the other must also. If one does not, the other must not have. It is an "all or nothing" scenario. I am assuming that the causal reasoning is always valid.

Question 2
I've just read the website supplement on sufficient and necessary causes.

What would a basic causal argument (like the one we discussed) be considered as?

To me, it appears that the cause is "sufficient" since the occurrence of the cause guarantees the occurrence of the effect. (I am assuming that the causal argument is always valid).

At the same time, the cause is also "necessary" because as seen in our scenario #3, the occurrence of the effect guarantees that the cause preceded it.

I guess I just described a "double-arrow" scenario again, where each part of the argument is both sufficient and necessary to one another.

What makes sufficient and necessary causes unique from the example we discussed?

Thanks again!

[Adam Tyson]

As to the first question, you are right about the author believing that there is a sort of "double arrow" relationship, although we do not use arrows to diagram causal reasoning. Arrows are reserved for conditional reasoning and formal logic. Don't get these mixed up! As to the validity of the causal claims, when we find a causal premise on the LSAT we almost always accept it as true, as with any premise. However, when we find a causal conclusion on the LSAT, we almost never treat it as valid! Causal arguments are inherently flawed, for all the reasons I mentioned before - alternate causes, cause without effect, effect without cause, reversed cause and effect, and data problems. That's what makes causal arguments so easy to deal with on this test - they are almost always awful, and thus easy to attack or to strengthen.

As to the second question, STOP! Before you go any further, heed my advice here - never try to mix causal and conditional reasoning! That's like mixing milk and orange juice - a bad idea. Causal reasoning is not conditional and conditional reasoning is not causal, and you should use different tools and strategies to tackle them.

Technically, yes, causal reasoning is also somewhat conditional, and the cause is sufficient to show the effect, but in the mind of the LSAT author the effect is also sufficient to show the cause, so it's that double arrow thing as you said. Don't try to apply conditional tools to a causal argument - apply causal tools instead (the five things I described for you in my prior post and above). Don't even think about applying causal tools to a conditional argument, either - that way leads madness and errors galore.

Keep those two separate. On those occasions when the authors mix them up, it's usually by giving us a conditional premise and then making a causal conclusion. When that happens, focus on the causal conclusion, not the conditional premise, because the conclusion is where the flaws will be.

I'll say it one more time, because it bears repeating: do not attempt to apply conditional reasoning tools to causal arguments or vice versa! Keep them away from each other!

Okay, I think (hope) I made my point. Good luck with that!

[koolaid]

Thanks Adam, you've been super helpful