I have created a community wiki question to encourage/enable users to share the books or paper from cognitive science that you think is most important for others to read and know about. Please just give one suggestion per answer and please make sure that you explain why it is important that people understand this reading and what it explains.

Note that you can suggest more several pieces of content if you think that they are all equally essential but this is generally discouraged.

I note that a previous question, which I have voted to reopen, asked for the must know papers of cognitive science and was closed for being too broad. If that policy of closing these sorts of "polls" remains in place then I am happy to remove or modify my question.

However, I would suggest that the question be left open as I have seen similar questions (1, 2) being very well received on other stackexchange forums and attracting considerable interest and engagement.

While some of these question have since been closed, I oppose that policy as I accord with the view from here, that "'must read books' questions are pretty much my favourite part of every stackexchange site – Damon Mar 9 '11 at 17:48"

  • $\begingroup$ Moderators - I wanted this to be a wiki - but, perhaps as I did not answer my own question (I planned to wait), I cannot find this option anymore. Can you please do this for me, or let me know if it is impossible so that I can post a new question as a community wiki. $\endgroup$ Commented Jan 20, 2016 at 5:29
  • $\begingroup$ why don't you phrase it such that it is 1 reading per answer and let the votes tell us what is popular? $\endgroup$
    – RJ-
    Commented Jan 21, 2016 at 8:25
  • $\begingroup$ Users cannot make questions CW, it is a mod only tool. You should flag the question for moderation attention. Note, however, that CW questions are often not good fits. This one seems like a shopping opinion based question. $\endgroup$
    – StrongBad
    Commented Jan 27, 2016 at 18:59
  • 1
    $\begingroup$ Relevant Meta question on whether or not we welcome this type of questions, and if so, how to structure it: meta.cogsci.stackexchange.com/q/36/21. Given this, I personally feel the scope of the question is too broad. I welcome list-type questions myself, but only when well defined (as you can find examples in the meta post). $\endgroup$
    – Steven Jeuris
    Commented Jan 28, 2016 at 10:12
  • 2
    $\begingroup$ I think this question would be better if we listed one book/paper per answer, not three. And provided an explanation of why that is a good choice. Since I've made this CW, you can always feel free to provide more than one answer. Of course, the community can still vote to close this question, if they think it is too broad like @StevenJeuris suggests. $\endgroup$ Commented Jan 28, 2016 at 19:04

3 Answers 3


Ok, I will start. I don't necessarily know if these are definitely the best three books I have read, but they are three of the best. If I think of better ones then I will add them.

  1. Thinking, Fast and Slow
  2. Predictably Irrational
  3. Influence: Science and practice

My interest in cognitive science is about i) understanding my own irrationality and ii) understanding how to change behaviour (e.g., persuasion, influence etc) to help people to have better lives. My three choices reflect this.

As related to this, I think that these books would be valuable for anyone to read as they could help you to better understand how to i) make better decisions, ii) help other people to make better decisions.


Chomsky's review of Skinner's book Verbal Behavior. This review helped kick off both modern linguistics and the cognitive revolution.

Marr and Poggio (1976). Here is where Marr's famous levels of analysis of information processing systems are detailed. These levels provide a framework for organizing all of cognitive science research, in my and many others' opinion.

EDIT: My girlfriend (another psych PhD student) suggested a book that I agree is pretty foundational in cogsci: Jerry Fodor's 1983 monograph Modularity of Mind. Here's the precis. This is basically a thesis about how the mind is organized: into functionally specialized 'modules'. The modularity thesis makes a number of other claims about the qualities of these modules, including informational encapsulation (modules are not sensitive to all potentially relevant information) and domain-specificity (they work only on certain inputs).

EDIT 2: Okay, once I got started on Fodor, I couldn't stop. Just as important as knowing about modularity is knowing about Fodor's language of thought hypothesis. I haven't read his 1975 book in which he lays out these ideas, but the gist is that thought takes place via a mental language with its own syntax and semantics. Check out the Stanford Encyclopedia of Philosophy entry on the LoT.

  • $\begingroup$ Are you intentionally going for the most one-sided answer possible? $\endgroup$
    – jona
    Commented Jan 27, 2016 at 1:15
  • $\begingroup$ Whether or not one agrees with these works (I happen to find them all compelling), I don't think you can deny that they are among the most influential. It's fine if you want to try to tear them down, but you need to understand them first. $\endgroup$ Commented Jan 27, 2016 at 1:52
  • $\begingroup$ I didn't say any of these texts is bad. I pointed out this answer is one-sided - it lacks perspective. It offers a narrow-minded selection. $\endgroup$
    – jona
    Commented Jan 27, 2016 at 2:04
  • $\begingroup$ You're being a little undiplomatic, if I may be frank with you. If a connectionist, embodied cognition person, DST proponent, etc. wants to post, they're welcome to. $\endgroup$ Commented Jan 27, 2016 at 2:24

My answer is going to be super skewed, I think, toward my interests. Nonetheless, I'd consider these papers "revolutionary," and I think their legacy will be long-lasting and their impact expansive (across all domains of psychology).

  1. Are emotions natural kinds? (Barrett, 2006). This paper upends many decades of thinking about what emotions are and has sparked a huge debate and a paradigm shift.
  2. Whatever next? Predictive brains, situated agents, and the future of cognitive science. (Clark, 2013). If this paper doesn't blow your mind, I don't know what will. Predictive coding is gradually taking over psychology…
  3. Oh gosh…I'll have to come up with a third one later. :)

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.