Deductive vs Inductive

What are you trying to say?

To this point we have been dealing with what is called Deductive Reasoning. However, there is another method, called Inductive Reasoning, that appears on the LSAT, and your ability to see the distinctions between the two is important in understanding your task.

The difference between Deductive and Inductive reasoning lies primarily in the potential degree of certainty of the conclusions reached and the methods by which these conclusions are derived. Both are critical to various fields, including logic, science, mathematics, and philosophy, but they serve different purposes and are used in different contexts.

Deductive Reasoning

Definition: Deductive reasoning is a logical process in which a conclusion follows necessarily from the given premises. If the premises are true (which they always are on the LSAT) and the argument is valid, the conclusion must also be true.

Characteristics:

Certainty of Conclusions: Deductive reasoning provides conclusions that are logically certain, not just probable.
Structure: Its rules often follow a conditional, "if-then" pattern. If the facts are accepted as true, then the rules given mean that the conclusion logically follows.
Example: All birds have feathers (rule premise) and a sparrow is a bird (fact premise). So, a sparrow must have feathers (conclusion).

Inductive Reasoning

Definition: Inductive reasoning involves drawing a general conclusion from a set of specific observations or instances. Unlike deduction, the conclusion of an inductive argument may be probable, based on the evidence, but not guaranteed.

Characteristics:

Specific to General: It uses specific observations to claim broader generalizations and theories.
Probabilistic Nature: Conclusions reached through inductive reasoning are probabilistic, meaning they are likely but not certain.
Generates Hypotheses: It is often used to generate hypotheses and theories based on observed data.
Example: For the last ten years, strawberries have been available in local markets from June to August (observation). It is currently July (known fact about the world). Therefore, strawberries are likely available in local markets now (hypothesis).

Key Differences

Nature of Conclusions:

Deductive reasoning leads to conclusions that are logically certain, provided the premises are true. Inductive reasoning, however, results in conclusions that are reasonable and probable, based on the evidence.

Flexibility with Evidence:

Inductive reasoning is open to being adjusted or refined as new data becomes available, whereas deductive reasoning is fixed and depends entirely on the initial premises being correct.

A simple way to think of the difference is to think about the difference between Aristotle and Sherlock Holmes (not necessarily the historical versions of these two but rather our pop culture version...apologies to the historians in the room!)

Aristotle was a philosopher. He sat around with his friends, relaxing in his toga, opining on the great mysteries of life, and trying to reach logical conclusions about what was going on. He was seeking deductive truths: what can I know based on what I already know?

Sherlock was an investigator. He skulked around in dark corners, finding clues, observing behavior, and attempting to determine who committed the crime. He didn't know for sure (unless of course he got a confession out of them) but the more information he gathered the more likely it was his hypothesis was correct.

OK, so why does any of this matter on the LSAT? It matters when the conclusion is causal.

Causation

It’s not just a flaw, it’s a way of reasoning

Smoking cigarettes causes cancer.

These days, this is not a very controversial stance. There’s no doubt in anyone’s mind as to this relationship. But there is something in this claim that is different from the deductive ones we’ve spoken about.

Does every smoker get cancer? Did every cancer patient smoke? Given that the answer is no, of course, what’s going on?

When we say “A causes B” we don’t mean it as we do in physics class. , where we know that every action has an equal and opposite reaction. The conclusion we’ve drawn is not about what DID happen or even what WILL happen but rather that one makes the other MORE LIKELY to happen. Think about it in terms of smoking and cancer: smoking doesn’t ensure you will get cancer; it just makes it more likely.

But if that’s true, how can we establish that A does in fact cause B? What does it mean to “cause”? This is a question that is the central issue in so many academic and professional pursuits, including the law (just wait until you get to Torts 1…)

Ultimately where we land in evaluating these ideas is that we can’t know for sure that A is the sole cause of B but rather that it is a significant and probable cause. In civil cases, we call that “a preponderance of the evidence,” enough to make it more likely than not.

How do we demonstrate that? Well, we’re not going to reinvent the wheel here. Let’s instead be led by the professionals who have spent the most time considering data as a means of demonstrating causation: epidemiologists, statisticians, and public health professionals.

The Three Pillars of Causation

How to quickly examine a causal claim

To examine a causal claim, scientists need three basic forms of evidence: DATA, TIMING, and MECHANISM.

Having these pieces of evidence will not GUARANTEE that the proposed factor causes the effect (remember, causation is about probabilities, not deductive certainties) but it will certain make it more likely.

On the flip side, if one can demonstrate that evidence for one of these is not present, that does not prove that there is NOT a causal relationship; it only indicates that it is less likely.

Data

It's all fine and good to have a theory about a causal connection, but if you don't have any observable data to back you up you're going to have a tough time getting anyone to believe you.

For instance, "smoking causes cancer" makes a lot of sense, but the first thing you're going to want to know is how often do people that smoke actually get cancer as compared to those who do not smoke. This is a simplified combination of a number of statistical concepts but primarily that of correlation, the statistical relationship between two random variables.

In simplest terms, to have good evidence of Data, we need

the presence of the cause makes the presence of the effect more likely, and
the absence of the cause makes the presence of the effect less likely

Timing

Causal arguments rely on a timeline. Which thing happens first, followed by what, in order to end up at the result. Without this information, making a causal claim seems a bit silly.

Have you ever stopped to wonder whether cancer causes smoking? I mean, it flies in the face of years of research, but let's put that aside for a second. Let's pretend it's 1900, and you are presented with a mostly unknown and entirely unproven hypothesis that smoking tobacco causes lung cancers. It would be a really good idea to confirm that people starting smoking early in life and develop cancer later in life.

This is true of all causal claims: we need evidence that we have the correct direction, the Timing, that factor A occurs BEFORE effect B. Otherwise, we may just have things flipped around.

Mechanism

Sticking with our causal timeline, we now have the endpoints of the story, but what happened? How did we get there? That physical process that starts with the input of the cause and ends with the effect needs to be filled in with all the steps between. This is the Mechanism.

Once researchers had sufficient data to convince them of the association between smoking and cancer, they still had a long road ahead of them: HOW does smoking cause cancer? What are the steps, the biological, chemical, and/or physical steps that make it happen?

After many years, many experiments, many labs, we have examined that process and generally agree that smoke from burning cigarettes is inhaled, carcinogenic particulate matter from that smoke is absorbed by lung tissue, the cells in the lungs then mutate due to the carcinogens, and those mutations are cancerous.

Wow, that's a mouthful. But now we know that this isn't just supposition; the scientists have traced this Mechanism and confirmed that these steps actually happen.

So, where have we gotten? Did we prove that smoking causes cancer?

Yes and No.

No, we did not prove that every smoker gets lung cancer, or even that every smoker of 2 packs or more per day for at least 40 years gets lung cancer. There are exceptions, there is randomness to the data, and we cannot either descriptively of the past or 100% predictively about the future say what did or will happen.

But on the other hand...

Yes, we have demonstrated that smoking is significant factor in causing lung cancers, and doing so makes it more likely to develop cancer in the future.

This is the crux of deductive vs inductive reasoning. Deductive arguments take facts, apply them to known rules, and make a specific conclusion. Inductive arguments take observed data and attempt to use them to make a probable generalization about that observation. When they do, you cannot prove anything; the best you can do is make that generalization more reasonable.