This popped into my mind the other day, and I’ve been distracted by it since… You know when you’re trying to recall something, and a wrong answer pops into your head, but you know it’s wrong. Like how does that work? E.g. if you’re trying to remember who made a song, and your brain can almost simultaneously go - oh it’s that band, and then oh no not them. It feels like there has to be two (at least) parts of the brain working on it at the same time.
Maybe I’ll be lucky and a neuroscientist will drop in and link me to a paper. More likely it’s something to discuss with wild speculation. Either way, I’m hoping writing it down will stop it distracting me.
Really, what you are asking about is called metamemory. Knowing the jargon in the domain might help you find more useful information. Neuroscientists have examined how the brain monitors and corrects error. For example, here is a paper that examines what regions of the brain appear to be responsible for error correction in a semantic recall task. In some sense, you are right, there are multiple parts of the brain working on recall and error correction at once, but you should really think of the brain as a larger system whose components work together in the same way the fuel injectors and pistons of an engine are part of a larger whole.
That’s very helpful thank you. I read the abstract of the paper, I think it might take me a couple of goes to really grok it. I think it’s testing why are more likely to correct a wrong answer given on a test (in a subsequent test), if they are enthusiastically told it’s right the first time. This is compared to if they are told that they might be wrong!
Given it’s the first time I’ve heard of this, I’m finding even the premise a challenge! ‘Hypercorrection’ apparently, for anyone not going to the paper.
What I’ve read of the article, meta memory seems to be more about our ability to judge how well we know something, rather than evaluate if our recall is correct.
I say ‘rather’… The concepts are obviously (or maybe not obviously!) related, but that sounds like assigning a score to the information we possess. While my original question was around evaluating knowledge as incorrect after recall.
That’s why the engine analogy doesn’t quite work for me. It’s not one answer, it’s two! So if it is an engine, it’s one that drives the car both forwards and backwards initially, and then switches off the one it doesn’t need.
I’m definitely going to read more into these concepts though. Thanks again for the links!
Heuristic processing, determination based on induction rather than certainty.
let’s use “on the roof again” by as an example, a 90s pop punk song.
ask your brain “who made this song?”
very rapidly it goes
music people!
not all music people.
and discards bands who definitely don’t sound like it: instrumental bands, classical music, slower love ballads, female singers (since the lead singer is a male in this case).
and now there’s a way smaller pool to choose from, so your brain keeps going, matching broad strokes to the broad strokes you remember from the song you’re thinking of, until one band sticks out as the most likely candidate, and usually at the end you can stick that determination to a specific memory you have of looking at the album cover while you listen to the song, or some concrete moment in which a line struck you and that more concrete memory helps confirm your final answer.
this is also why people can remember things “for sure” that turn out to be incorrect, because they’ve gone through the heuristic process and determined a most likely answer that may not be correct, because it takes much less processing power and time to heuristically determine an answer rather than specifically determining it answer, and heuristic processing does work well most of the time.
I don’t know the exact neuroscience behind it, but suspect this relates to the fallibility of memory, and whatever goes in in the brain during learning and reasoning.
So you know of multiple bands and songs, you attempt to relate a specific band and song, but because of some murky memories and rough recall, you briefly relate the wrong pieces of information, recognize the mistake, and correct for it. How that process works precisely may vary across people and their methods of recall and knowing.
In a way I suspect that the false-positive you mention may be a sort of synaptic shortcut to the correct information through whatever systems are at play in the brain for this reasoning. For some it may be that this process of rapid error correction is sometimes faster than immediate, accurate recall, or may be more of an unconscious aspect of accurate recall.
I’m certain (indeed more certain than I likely should be, which may be meta-meta memory?!) that what you say that the end is the case. There’s almost certainly a bias towards error correction over direct recall. Certainly my experience is of testing wrong answers in my head before alighting on the right one.
That implies a set up more like an adversarial neural network (I’m not saying this is actually how it is, just trying to draw an analogy from something I understand), as opposed to a function in code. But that seems like a bit of a waste, but also means that two (or more) distinct processes could be working on the same task?