I'm going to try to engage with the following Encyclopedia Britannica article on "rationality".

Rationality, the use of knowledge to attain goals.

I have a bigger personal project to consider how definitions ought to be carried out, but let's say for the nonce we have to accept sometimes being able to provide a definition subjectively, even if imperfect. I already find the compact definition here strange. "Rationality" to me strongly calls to mind the word "reason" (the -ity form sometimes conveys a general phenomenon form, or noun form, what "perspicuity" is to "perspicuous"). I would find a safer definition to be, "the general tendency or phenomenon of being rational", then define "rational" as something along the lines of "having reason, having sound judgment, being accurate". Thus, I find the definition of rationality to encompass "attaining goals" unmotivated, even out of the blue.

Models of Rationality

Rationality has a normative dimension, namely how an agent ought to reason in order to attain some goal...

It sounds like Pinker is saying "rationality" partially is generalized prescriptions for people: what you should do (with regards to X, hence, deontic modality). I think this fails to separate two underlying aspects: one, what you should do; and two, why. The "why" part presumably depends on argumentation. If we think of reason as general validity of drawing inferences or describing the world, we are able to separate out this "normative" aspect: Pinker is saying that based on an analysis of a state of affairs, one is able to make prescriptions about what one should do, in light of them. This doesn't strike me as needing to claim that "rationality is inherently normative"; rather, normativity can be with regards to many things, for example, what one considers "rational" (or just, or good, or desirable, or smart, or proper, and so on).

and a descriptive or psychological dimension, namely how human beings do reason.

Having come this far in the article, it sounds like Pinker is using one word to mean many, many different things, and it lacks precision and clarity. Here, he discusses "reason" as a synonym for "think", as in, "a process by which people form judgments, make decisions, etc." The idea that "to reason" is an activity excludes the idea of "rationality" being about correct judgment, in my opinion. I find it dissatisfying that an attempt to define "rationality" would include, "including the way humans reason, including the profoundly erroneous ways they do." It would be better to discuss valid judgment, and contrast it with invalid judgment.

Normative models from logic, mathematics, and artificial intelligence set benchmarks against which psychologists and behavioral economists can compare human judgment and decision making.

It sounds like Pinker thinks of "rational decision-making" as a somewhat ad-hoc collection of techniques that have emerged in various fields in the course of history, but he doesn't claim that there are any that are inherently valid; just that they are "possible models" for reasoning processes. In other words, we could try to predict the weather based on the so-and-so statistical forecasting technique. Is that forecasting technique correct, though? Is it logically guaranteed to predict the future, correctly? It sounds like Pinker would claim "normatively" that we should base our decision on the results of this or that model, without a logical justification for why we should believe what that model models. Ironically, it sounds like Pinker doesn't think there is a capital-T Truth out there, but basically a grab-bag of theories to try out, with no unifying basis or underlying theory explaining how you can know if your conclusions are correct or not.

Formal logic, for example, consists of rules for deriving new true propositions (conclusions) from existing ones (premises). A common departure from formal logic is the fallacy of affirming the consequent, or leaping from “p implies q” to “q implies p,” for example, going from “If a person becomes a heroin addict, the person first smoked cannabis” to “If a person smokes cannabis, the person will become a heroin addict.”

Pinker gives a weak but decently acceptable definition for "formal logic", but then provides an example that is arguably the opposite - what is commonly called "informal logic", or, the study of common fallacious thinking patterns people exhibit. As I understand it, formal logic observes how deductive systems behave based on the precise rules established for those systems. It's not uncommon to see a huge variety of different logics, with different starting assumptions - for example, choosing to include or exclude the law of the excluded middle. I think there is nothing preventing me from provisioning a logic with the axiom, "if p implies q, then q implies p". It would be interesting to study that system of logic. Pinker seems to be saying that "formal logic" is a way of codifying valid reasoning, but arguably, I do not think modern logic is like that. Perhaps that is an open question I must learn more about, though.

The rest of the article goes through examples of these "reasoning heuristics" and "rules of drawing correct inferences" ("Bayes's rule", probability, statistical techniques, game theory), again, without any unifying theory of where all these "correct rules" come from (IMO). It seems like Pinker would advocate refuting any observed instance of reasoning which apparently violates these "techniques of reasoning", but this is a very patchwork understanding of correct reasoning, because we do not know how they all fit together into a complete theory; we don't know if there are other fallacies we aren't aware of; we haven't really proved that we know how to reason correctly, only that we know how to spot a "fallacy" now and then, but arguably, without even much of a proof that it really is a "fallacy", since, as I said, these claims do not descend from some fundamental principles of what valid judgment is and is not.

Is this Encyclopedia Britannica article profoundly flawed?

(There is a second part of the article which I suppose I could go into in a second post, about "why humans are irrational", but I think it's based on similar claims as the above.)

  • Britannica is written for general audience and geared towards colloquial use, "shortcomings" may be features when seen in that light. If you are into philosophical analysis of the concept, why not engage the more fine-grained and philosophically grounded definitions in SEP instead? Structural Rationality and Instrumental Rationality.
    – Conifold
    Jan 20 at 22:13
  • I don't know how the article could be profoundly flawed given that it is a superficial and rather random treatment of its subject. I agree that the headline definition 'the use of knowledge to attain goals' seems horribly imprecise. If I cite my knowledge of giant pink Martian rabbits invading Earth to attain the goal of providing the police with a plausible reason for my walking around town with a shotgun, I suspect their write-up of the incident will not include the line 'We took no further action, as the distinguished elderly male in possession of the firearm appeared to be acting rationall Jan 20 at 22:31
  • @Conifold it might be my memory playing tricks, but the content of the EB seems to have been dumbed down and popularised since I last had access to it in book form circa 1980. Jan 21 at 9:36

2 Answers 2


There was this enigmatic Ancient Greek word, logos. The concept it referred to has long been forgotten -- or, rather, people gradually lost the capacity to understand it.1 That's why logos does not have its equivalent in modern languages. Latin was already too modern, and Romans struggled to translate it. In the Gospels logos ended up as "the word". When working on Euclid's text about geometrical proportions, they managed to translate logos as "ratio" -- hence "rationality".

Now, with the above in mind, is it really surprising that we are still mighty confused about the meaning of logos/rationality?

1 "Even though logos always holds true, people prove unable to comprehend it, not even after they have been told about it." (Heraclitus, DK B1)
"In it was life, and the life was the light of men. And the light in the darkness shined; and the darkness comprehended it not." (John 1:5, KJV)


I share your aversion to Pinker's definition.

Its traditional in philosophy, at least in Europe from the 16th century on, to recognize the fundamental dichotomy of rationalism vs. empiricism, and it goes arguably all the way back to Plato vs. Aristotle.

To then give a manifestly empirical definition to the key ontic of the other — rationalist — party seems confused at best and frankly disingenuous.

It can be done intentionally of course as I do very briefly in the note in my profile. Or at greater length in first couple of sections of my paper on how programmers may understand recursion more deeply than the typical imperative presentation (outside paywall). However I would aver that there is a fundamental difference between accepting that there is a polar dichotomy followed by the attempt to find a resolution vs lazily and perhaps dishonestly using the opposing party's definition as though it were unproblematic.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .