In philosophy it’s often said that there are ‘lumpers’ and ‘spliters’. Lumpers try to unify discrete kinds under one category while spliters argue for maintaining (and insisting on more) distinctions between kinds. When it comes to critical thinking, I tend toward the former. Especially at the end of a semester, I always find myself obsessing about the best way to distill an entire course into as few basic principles as possible. Some textbooks explicitly do this (see Robert Shanabs and Sharon Gould’s excellent TLC method) while others approach critical thinking from the point of view of discrete modules.
Here is my latest distillation.
RRAR! Method: Critical Thinking for the Digital Age
Preliminary Step: Set Up
Before we begin any evaluation we need to put the subject of our evaluation into premise-conclusion form. I’m not going to fully explain it here but if you’re interested, here is the unit from the beta version of the textbook I’m writing (I’m still editing it so don’t scream at me about layout and the typos n’ stuff!). Basically, identify the conclusion (i.e., what the author is trying to persuade you to believe) and the main premises (the reasons and evidence used to support the conclusion). List the premises with the conclusion at the bottom.
(P1). The activities and decisions that most affect your well-being require you to be able to think well in order to make the right choices.
(P2). Critical thinking is a systematic method for thinking well.
(C). Therefore, you should study critical thinking if you want to increase your well-being.
Step 1: R=Reliability of the Source
When we approach an article, video, meme, and so on, our first step should be to evaluate the reliability of the source. Dismissing an argument outright based on its source is an instance of the genetic fallacy, so we should be careful not to do that. However, if an argument comes from a source known to be heavily biased or unreliable this tells us that we need to be extra skeptical during our investigation. Importantly, this means we should be on the look out for slanting by distortion or omission and the fallacy of confirming evidence (see the third section in the link).
Step 2: R=Relevance of Each Premise
In the context of arguments, the definition of relevance is the degree to which a premise increases the likelihood of the conclusion being true. Relevance comes in degrees. To understand the concept of relevance let’s look at some common fallacies: The argument from tradition and the naturalistic fallacy. They are both fallacies because their main premise is irrelevant to the conclusion.
Example 1: Women should stay home and raise the children since that’s what they’ve always done.
(P1) Women have always stayed home and raised the children.
(C) Therefore, woman should stay home and raise children.
Notice that even if (P1) is true it doesn’t meaningfully increase the likelihood of the conclusion being true. What women have done is the past has no bearing on what they should do now. Someone might point to other reasons (e.g., having mammary glads) for which women should raise children. But that’s a separate argument–whatever you think of it. Merely pointing to what women used to do isn’t on its own relevant to what they should do now.
If you’re not convinced, let me give you another example using the exact same reasoning (appeal to tradition)
Example 2: Humans have always murdered and raped therefore humans should murder and rape.
(P1) Humans have always murdered and raped.
(C) Therefore, humans should murder and rape.
Again, while (P1) is probably true it isn’t relevant to whether we should murder and rape now. It doesn’t meaningfully increase the likelihood of the conclusion being true. Some bleeding-heart liberals might even suggest there are reasons against murdering and raping [GASP!]. Some traditional human behaviors are good, some are bad, and there’s everything in between. Merely knowing that something was done traditionally doesn’t tell us either way whether it’s good or bad or whether we should do it.
Example 3: This snack is natural therefore it’s good for you.
(P1) This snack is natural.
(C) Therefore it’s good for you.
Whether something is natural or not doesn’t tell us whether it’s good for us. There are probably more poisonous things in the world than non-poisonous, so merely knowing that something is natural doesn’t increase the probability of it being true that it’s good for us.
A more advanced way of evaluating relevance is to identify the enthymeme but that’s another lesson. We’ll just stick to basics here.
Step 3: A=Acceptability of the Premises
By ‘acceptable’ I mean something close to ‘true’. Suppose it turns out that all the premises in an argument are relevant to the conclusion. That doesn’t mean a hoot if they’re all false! In critical thinking I don’t like to use the word ‘true’ for many reasons which I’ll skip over here. Instead I use ‘acceptable’. Here I mean simply that a premise would be accepted to a reasonable audience without further evidence. At Step 3, I apply the reasonable person test to each premise.
If we answer “not sure”or “there could be disagreement” to a premise then we get on our google machine and investigate. Also, this is where the Reliable Source criteria comes in: If the source of the argument is known to be unreliable or heavily biased, we should–nay! must!–verify each premise. The reasonable person test won’t suffice.
Step 4: R=Relative to What?
Step 4 is going to be applied at all stages of the evaluation. It makes me cringe to say this but with respect to a lot of things, “everything is, like, relative maaaaaaan.”
With respect to the source of the argument, reliability is relative. Suppose Source A is considered to be reliable. It contains an argument that X is false. However, I encounter Source B that argues that X is true. The relative reliability of B and A will inform my evaluation. Even though A is a reliable source B could be more reliable, just like it could be less so. All things being equal, I should go with B over A if B is more reliable relative to A.
Relevance also needs to take into account relativity. Suppose an argument presents relevant evidence in favor of a conclusion. I need to weigh that evidence against the relevance of the evidence against the conclusion. For example, there might be a preclinical trial that shows that X cures cancer. Pre-clinical trials have very small sample sizes and rarely have control groups or blinding. They are low quality evidence. However, there’s a Phase II trial (blinding, control group, larger sample size) that shows X doesn’t cure cancer. The strength of the evidence that X cures cancer is weak compared to the evidence against the claim. The Phase II evidence is more relevant to the conclusion relative to the pre-clinical trial. Claims rarely have all and only evidence in one direction. To repeat, I must consider the relevance of positive evidence relative to the relevance of negative evidence.
The same goes for acceptability. Some premises will be more easily accepted by reasonable people than will other premises.
Both relevance and acceptability require we apply the concept of relativity in another respect. Very often arguments (and conclusions) will make claims that include words like increase, decrease, good, bad, effective, ineffective, cheap, expensive, risky, beneficial, harmful, and so on. In order to even interpret claims that contain these words we must know the appropriate comparison class.
For example, if I say that the stock market increased, before I can even evaluate whether that’s relevant or acceptable I need to know relative to what? To yesterday? An hour ago? Ten years ago? To the Japanese stock market? To the bond market? To interest rates?
If a policy causes some people to pay higher taxes I need to know relative to what? Relative to last year? 40 years ago? Relative to another group? Which group? It’s vitally important to know what the comparison class is. Without it we can’t evaluate either relevance or acceptability.
I’m thinking about creating a worksheet for students that looks like this for each argument they must evaluate:
Set Up: Put the Argument into Premise-Conclusion Form
Step 1: Reliability of the Source
Score: /7 1=Very low reliability 7=very high reliability
Explain why you gave the source the score you did:
Step 2: Relevance
For each premise assign a relevance ranking of low, medium, high then in a sentence explain your ranking. Identify any claims that might be comparative and identify the comparison class or write “ambiguous”.
P1. Low/Medium/High because:
P2. Low/Medium/High because:
P3. Low/Medium/High because:
P4. Low/Medium/High because:
*If premises are low relevance, their acceptability won’t matter. A true but irrelevant premise doesn’t increase likelihood of the conclusion being true.
Step 3: Acceptability
For each premise state whether it is acceptable, unacceptable, or unsure. If unsure because of language problems look for contextual clues. If unsure because you don’t have enough information, google it then reassess. Cite your sources. If unsure because of ambiguous comparison class, try to identify the author’s implied comparison class.
P1. Acceptable/Unacceptable/Unsure because:
P2. Acceptable/Unacceptable/Unsure because:
P3. Acceptable/Unacceptable/Unsure because:
P4. Acceptable/Unacceptable/Unsure because:
Step 4: Relative to What?
With respect to the conclusion, identify the correct comparison class. For example, if the conclusion is that a certain policy is bad, compared to what alternative policies? Make the appropriate comparison of both costs and benefits.
Well, there you have it. The most recent incarnation of a critical thinking system based on as few principles as I can get away with. If for every argument you apply these four steps, you’ll soon find yourself to be a beast of critical thinking, RRAR!