For years, I've been stumped by the failure of most people, of society, to make simple and important inferences. They have the facts, it's important to them, and yet they do not put two and two together. How do they fail so?

Well, last week, a friend's comment suggested a trailhead for an explanation. This is part 1 of a ~4-part series where I share my current guess.

You know how to walk. When did you last think about contracting your quads?

You know how to talk. When did you last think about the placement of your tongue?

You know how to think. When did you last think about...the individual mental motions that make up the larger thought-acts you intentionally carry out?

Like walking, talking, and most actions, thinking is non-atomic. Any act of thinking is composed of sub-thoughts, micro-cognitions, and individual mental motions. And in comparison to walking and talking, the variety is staggering. There are only so many practical ways to walk, but the ways your mind could carry out a given thought-act are numerous.

Start with the example of answering something simple: where to get to lunch? You begin considering two options: the taqueria and the sushi place. Why those two? Your mind wordlessly queried "options" and that's what came back. The broader line of thought found that satisfactory and didn't push for more. On another day, your thinking would be querying what you felt like and hunting for justifications to pay the high sushi prices. Today that doesn't occur to you, instead the wordless considerations query came back with "what would your lunch companions like most?" (marked important) since you are choosing for them too.

My goal here isn't to suggest a definite or rigorous taxonomy of thought, more to gesture at the breakdown. Much of thinking seems to proceed in chains: trigger/response/trigger/response/trigger/response where both previous thought-responses and new events in the world can be new triggers. This makes you think of that, that makes you think of this, makes you think of that. Along the way there are wordless queries for information that some part of your mind provides. How and why? We don't usually think about that. You ask a question like what I should be doing right now and a list is produced as if from nowhere. Perhaps another part of your mind remembers that you're often forgetful and should consult your calendar. Thank you, helpful thought.

So there's information retrieval that happens in an opaque way and who knows how exactly the recall happens, but also there's a bevy of different reactions that a mind could throw up in response to any stimulus. Those reactions might look like a question. Somebody learns their boss is angry, their mind might variously: become afraid it is something they did and search for reasons, be gleeful because they enjoy their boss's displeasure, start considering how to take advantage, and so on. My guess is that in very rare cases do people pause to consider which of these responses is best, instead, much as the mind chooses our footfall for us, the mind decides which question one ought to be answering.

From another angle, there's the kind of reasoning a mind employs. Responding to a stimulus, a mind could call up past examples, it could simulate the reaction of certain other people, it could query the morality sub-module, it could execute a first-principles-simulation, it could make a query to the gut/heart about how this feels. The cognition done could be cerebral, it could be embodied, or some secret third thing.

To tie this to a concrete scenario, consider a young fellow on a date with a young woman at a nice restaurant when the time comes to pay the bill. For many, a part of the mind will recognize "pay the bill" as being a ritualized high-stakes moment and the optimal action being non-obvious. Having pattern-matched the scenario, another part of the mind might pump stress hormones^[1], priming the mind and body for action. The fellow's mind could go down any number of pathways here: (a) try to infer what this young woman is expecting, (b) mustering courage to take the option he believes is correct, (c) figuring out a joke to defuse any tension with humor, (d) figuring out how to be suave and deftly ferret out more information without committing to the wrong answer. My guess is that the conscious explicit thinking gets spent on the chosen part, e.g. thinking of good jokes, rather than choosing among the high-level options.

Like walking and speaking, these motions proceed automatically. And not just the intuitive, quick, rapid cognition System 1 mental acts. I contend that there's automatic mental sub-motion micro-cognitions happening incessantly and inseparably even with putatively explicit System 2 cognition, e.g. every time the System 2 cognition makes a query for lists of options; and at the meta-level too, the part of your mind that typically decides unexplicitly which things should be thought about deliberately vs not^[2].

All of this is to make two points: (1) thinking is not an atomic act. When you think "what shall I eat for lunch?", that mental act will be composed of a great many sub-pieces. And (2), almost all of these sub-pieces proceed automatically and without conscious thought – and all the conscious thought fragments can be decomposed into less conscious ones.

That's all for today. In tomorrow's piece, I'll muse on how the mind learns mental motions and when to use them.

1. ^{^}

   It's not crucial for my arguments, but I view emotions as part of cognition too, such that we should view generating a particular emotion (and corresponding physiological states) as part of the automatic constituent mental actions a mind conducts.

2. ^{^}

   Depending on the person and ideas they've absorbed, the mind eventually decides that one should think explicitly about whether or not to think explicitly.