Human Reasoning and Cognitive Science

by Keith Stenning, Michiel Van Lambalgen

In Human Reasoning and Cognitive Science, Keith Stenning and Michiel van Lambalgen -- a cognitive scientist and a logician -- argue for the indispensability of modern mathematical logic to the study of human reasoning. Logic and cognition were once closely connected, they write, but were "divorced" in the past century; the psychology of deduction went from being central to the cognitive revolution to being the subject of widespread skepticism about whether human reasoning really happens outside the academy. Stenning and van Lambalgen argue that logic and reasoning have been separated because of a series of unwarranted assumptions about logic. Stenning and van Lambalgen contend that psychology cannot ignore processes of interpretation in which people, wittingly or unwittingly, frame problems for subsequent reasoning. The authors employ a neurally implementable defeasible logic for modeling part of this framing process, and show how it can be used to guide the design of experiments and interpret results.

Genres: Philosophy

407 pages, Hardcover

First published August 1, 2008

Reviews

[Max · Rating 4 out of 5]

To what extend are the psychology of reasoning and logic relevant to each other. That's the question Stenning and van Lambalgen pose at the beginning of their book. It won't knock anybodies socks off if I spoiler that they make the case for their usefulness. Before this lecture, experiments like the infamous Wason's Selection Task had me believe something like this: Logic is not really natural to us, we have to train a lot to apply it correctly, it doesn't deal with all this uncertainty and isn't useful for complex data, so it's maybe a good model for thinking in highly familiar and unambiguous situations.

For example:
Assumption 1: If I cough dryly, I might be infected with this virus.
Assumption 2: I cough dryly. (at least kinda dryly)
Conclusion: I might be infected with this virus.

Also, apparantly like many others, I pretty much equated logic with propositional (and on good days predicate) logic. Stenning and van Lambalgen's TLDR would be something like:

"There are many more logical systems than that. When people are made to do reasoning tasks, two things happen:
(1) reasoning *to* an interpretation, where people try to understand the task and which encoding and reasoning approach (a.k.a. logic) is appropriate, and
(2) reasoning *from* an interpretation, where we start deducing according to the logic we have chosen."

They go on to show how many of the "irrational" choices of subjects might better be understood as applications of alternative logics. Examples I found convincing and where you can tell a story why it leads to "wrong" answers in the Wason Selection Task:
- some subjects seem to interpret the conditional (if A then B) as a normative law, a law that should be followed, instead of a law that can be true or false (as in propositional logic)
- people think differently between the course-of-events-conditional (if A, then B, and A usually happens) and the hypothetical conditional (if A, then B, but A won't happen)
- "[[strong falsity]]", where the negation of a conditional is taken to be the same as negating the consequent in the conditional
○ Example: If I tell you that the statement "If you go to Amsterdam, you will be stoned." is false, you conclude that "If you go to Amsterdam, you will not be stoned." (they say more than half of participants do that)

Other things to like:
- Cool discussion (and critique of the current discussion) of evolutionary origins of cognitive functions
- Many useful transcripts as concrete examples of how people think about their tasks
- An attempt to formalize human reasoning in a closed-world logic that they show to be implementable with neural networks (I skipped this)
- Two chapters discussing how their approach might explain some cognitive differences to people on the autism spectrum (also skipped)

Some complaints:
1) They discussed Bayesian approaches to model behavior here and there and were largely dismissive, e.g. complaining about the lack of need for the numerical precision that probabilities gives you. I didn't found this very convincing, and it was remarkable how often subjects used phrases like "probably", "likely", "there's a good chance" and even "I would bet on it" during solving puzzles that were supposed to unveil the used logic.
2) It was a bit lengthy here and there, though that probably says more about me not being able to skip content as much as I should.

Overall, I enjoyed it and would recommend skimming the second chapter.

[Marianne Lalancette · Rating 4 out of 5]

Réconcilier logique & cognition.

Propose la pertinence de l’utilisation de la logique comme outil de modélisation du raisonnement humain.

(ouvrage important pour mon mémoire de maîtrise)

[Trond · Rating 4 out of 5]

Pretty hard going some times, but enjoyable.