This is more philosophy than psychology but it has a psychological component. I think that we humans, along with the higher animals, see new patterns and recognize old patterns and jump to conclusions when we see partial patterns. We do this in place of logic, largely even when we do mathematics.
One of the patterns we have collectively discovered is logic itself, formal or otherwise. The careful mathematician will attempt to put his proofs into a form that matches these logical patterns, even occasionally so that a computer may verify conformity to patterns it has been taught. Thus do patterns bear on logic.
AI people generally agree that patterns are a vital tool. I agree that a logic based AI can harness pattern hardware and could in principle perform algorithms to find patterns. As a practical matter, however, I think that special pattern hardware is indicated. I am yet ambivalent as to whether AI should itself be based on patterns, like us, or on logic.
A problem that I recently noticed (when someone pointed it our to me) is that special pattern hardware is unlikely to be able to describe the patterns it knows; it can merely say whether a particular pattern is present. We may share this weakness with future AIs. Curiously logic seems to be the language of patterns, and thus does logic bear on patterns.
This leaves open the question of how the Greeks came to note that ‘all men are mortal. ...’. They were not talking about men but about the logical syllogism which was a familiar but unobserved pattern to some. If logic is embedded to some degree in our language, it is presumably likewise embedded in our ways of thinking, or at least convincing others with ‘proof’ like utterances. “All men are mortal” is more a pattern than a proposition but there is no clear line. In order to predict the consequences of our actions, we need to pattern match against imagined situations, as well as real situations.
This needs much more development.
Patterns are seldom without exception. That some dogs have lost a leg does not make “Dogs have 4 legs.” much less useful. Mathematicians specialize in patterns without exceptions. (Of course there are oddball constructs such as ‘odd primes’.) Here I accuse Arthur Conan Doyle of confusing the logic of math with the logic of real things. Even vague tendencies are useful to observe, and the language of logic provides a means conveying such observations.
Metaphors recognize patters that appear in different realms. The mathematician’s pattern is caught in the notion of isomorphism which is very central to math. Even the concept of 7 is like the equivalence class of sets with 7 elements. There is a one-to-one correspondence, or isomorphism, between two members of that class. See this.
I use the term rhetoric for that use of logic to persuade others to believe as you would have them believe. This has a negative connotation, but may sometimes serve mutually beneficial ends. This is thus an aberration of logic. This line tips dangerously into ethics, a topic which I avoid. I think that rhetoric developed along with language and has been used as an organizational tool among humans. A cynical term has been suggested recently: the ‘domestication of humans’. There are some advantages in hierarchies in any sort of information systems and large groups of people are in some sense an information system. I thus agree with Hugo Mercier to a degree that rhetoric has been used more to influence others than to find the truth. It is logic’s earlier and current ability to find the truth, however, that gives rhetoric its effectiveness. Rhetoric is thus parasitic on logic. This is not a subtle point—it is highly analogous to saying that the telephone is used to convey facts and is consequently used to convey lies. Logic augments speech to spread belief and is thus the natural tool for control. That it is abused cannot be used as a reason to abandon logic. In so far as logic is part of language, then perhaps rhetoric, in my sense, has been the main use of logic. The processing of patterns preceded logic as part of speech, however, and I use ‘logic’ to include that prelinguistic phase.
Said otherwise, inter or intra species signals are invariably corrupted to the benefit of the sender. Logic is a signal to cause a recipient to believe. Somehow we also manage to deceive ourselves and this simple idea does not account for that. Perhaps we deceive ourselves as practice, better to deceive others.