What is the consequences to be a person who works in theory of science?

Question

this question rises into my head this day. First I read the comment by @mikeazo he says

Do you learn best by reading or doing? That will make a difference.

Also, today I had a class in cryptanalysis so we were doing a lot of computation in class during two hours (solving equations in order to find key of a cipher given ciphertext and plaintext). In the class I noticed that there are two types of students: faster and slower. I noticed that slower-people related to theory-people and faster-people to "non-theory people".

to define difference between theory and non-theory. Usually in science we have two types of people: theoretical people and experimental people. For example: in physics you can work in theory without doing experimental things or you can work in experimental without doing theory. Usually theory related to mathematics; since the language of science is mathematics. Now, in mathematics we have people who do "computation faster" like they solve puzzles quickly or they are very smart at games or multiply tens of digits without calculator, etc. While the other part, the theory-people, they analyze physical things to have a mathematical model and expect future results based on these models so if models doesn't work (such as no solution for the model under some conditions) then we say this result cannot happen in future under some conditions, etc.

From this point-of-view, I'm really interested to see whether is there any survey or research paper about "what is the consequences to be a theory-people or non-theory-people?" For example: I remember Michael Sipser says in his book of "Introduction to Theory of Computation", "theory is good because it expands your minds". Now, if someone is doing theory, then he expands his imagination and the way of thinking to things, but how about consequences? it seems to me for the example I had today in class that "theory people are slower to do normal jobs, even though they understand it". Is there any research papers about this issue! I would be happy to hear from this site anything related to this topic.

Answer 1

Quoting from "Cognitive Reflection and Decision Making"

Many researchers have emphasized the distinction between two types of cognitive processes: those executed quickly with little conscious deliberation and those that are slower and more reflective (Epstein, 1994; Sloman, 1996; Chaiken and Trope, 1999; Kahneman and Frederick, 2002). Stanovich and West (2000) called these "System 1" and "System 2" processes, respectively. System 1 processes occur spontaneously and do not require or consume much attention Recognizing that the face of the person entering the classroom belongs to your math teacher involves System 1 processes-it occurs instantly and effortlessly and is unaffected by intellect, alertness, motivation or the difficulty of the math problem being attempted at the time. Conversely, finding $\sqrt{19163}$ to two decimal places without a calculator involves System 2 processes-mental operations requiring effort, motivation, concentration, and the execution of learned rules.

This distinction is apparently also called dual process theory.

I for one like to call these "intuitionists" and "calculationists", but I also think the distinction is probably not too sharp. In my experience, the "intuitionists" exploit some patterns, partly "subconsciously", and tend to be better in games like Go or Backgammon (where the trees are very broad and humans cannot think too many moves ahead) and often have surprising idea that they themselves can't always explain how they came up with, whereas the "calculationists" tend to be better at Chess, long proofs and so forth.

Not coincidentally (in my intuition) these coincide with the best performing algorithms for such games: neural nets for Backgammon, but mostly deep tree exploration for chess. These are not entirely exclusive though, good Backgammon algorithm "think" 2-3 moves ahead but their performance comes mainly from a very well tuned neural net. I'm less familiar with Go, but from what I know of AlphaGo the story is similar. One should not discount the role of board evaluation in chess, but I think the tree exploration is main factor for success there. Anyhow, intuition is believed by some researched to work like a probabilistic decision, perhaps using pattern matching as discussed above.

As more anecdotal evidence: When I was in grad school, I had an adviser and co-adviser who had authored a large number of papers together. They described their roles in their collaboration as one being the idea generator and other one the tester.

More recently, unconscious thought theory posits that for complext decisions intution is essentially better, but after a highly-cited debut in Science, insofar it has failed replication particularly in large studies. I suspect a lot depends on the nature of the test.

On the other hand, dual-system theory seems to have fared better with a (simple test of distinction), introduced in the first paper I quote in this answer; Here's a quote from a more recent paper:

The present study replicated and extended earlier results from Campitelli and Gerrans (2014). In particular, Cognitive Reflection and Calculation behaved like distinct abilities, and Calculation was positively correlated with numeracy. However, Cognitive Reflection was positively correlated with numeracy as well; this correlation had not been tested in earlier studies.