[New? Start here: https://meditationstuff.wordpress.com/articles/]
Johnson-Laird’s career has been spend on what he calls “mental models.” He has been an author on tens of research papers on topics such as negation, causality, etc.
Brainerd and Reyna have spent a portion of their career on what they call “Fuzzy-trace theory,” which is a theory about semantic memory.
I like mental models and fuzzy-trace theory because they are both phenomenologically provocative. That is, I can access referents in consciousness that seemingly correspond to moving parts in their theoretical machinery.
Further, I think these theories are a nice bridge between me and the rationality community, because both theories have dual-processing components and both theories empirically weigh in on cognitive biases.
According to Johnson-Laird, System 1 can manipulate and deploy mental models which are as explicit and iconic as possible, and System 2 is needed to mediate between models, search for counterexamples, and more. I’m not doing it justice, here, but he has a whole set of claims and falsifiable predictions. It’s amazing work, and he’s itching to prove himself wrong so he can come up with something even better.
My own spin is that these models can be “fully loaded” in consciousness (“are reality,” while they’re there), or not, and referents that correspond to alternative models are available as felt senses on the “periphery” of consciousness, if you know where to look. Emotion can make models “sticky,” and mind muscles can be strengthened to be able to deliberately call up, sort through, recurse through, and juxtapose mental models in an explicit and deliberate way. You can also learn to deliberately hang out in “no model” and “between model” territory which can sometimes afford unpredictable insights.
What I’ve found is that models can be simple or tremendously complex, but they don’t easily combine–it’s fascinating. They can be contradictory, and people ping-pong back and forth between them over seconds and minutes, based on priming, inner noise, and working memory limitations, and they don’t even realize they’re doing it.
I’m currently exploring automatic and deliberate fusing of mental models, the phenomenology of mental model inclusion and transcendence, ironing out inconsistencies, and getting a single, more intricate “chunk” instead of taking up two chunks of working memory. Of course, consciousness is the tip of the iceberg, so what are the attentional, intentional, and emotional levers that prioritize what’s happening while you’re daydreaming and sleeping?
Now, fuzzy-trace theory asserts that certain types of memory are composed of two components: verbal and gist. These components are stored separately and decay and evolve at different rates during different developmental periods of a person’s life. This non-coupling, in the literature, can experimentally account for all sorts of cognitive biases.
Fuzzy-trace theory’s “gist,” I assert, has its phenomenological component in Gendlin’s felt sense, Hurlburt’s nonsymbolic cognition, and so on.
Anyway, both mental model theory and fuzzy-trace theory have been used to predict and explain a wealth of cognitive biases. Furthermore, because these theories have phenomenological grounding, they may suggest discrete and deliberate “levers,” “moves in mental space,” that would be effective for debiasing. And, these debiasing strategies could be tested using pieces of the experimental protocols laid out in these two threads of literature.