Insights Over Ontologies
  • When you have a set of ideas, one natural human thing to do is to try and find an overarching theme that can hold them together
    • e.g. a model about how habits work, so you can explain interventions and experimental results in terms of this model
  • In many self-help cases, it may actually be better to offer insights rather than completely packaged ontologies
    • by insight, I refer to an observation about how the world works, e.g. something less overarching than an ontology
    • an ontology about habits might focus on their being a cue-response pattern established by repetition and certain sensory timings, and everything is about how it relates to this model
    • an insight might just be "if you do X a whole bunch, it's much more likely to become habitual"
  • Insights can be better because they don't require the user to discard their previous ontology or their existing frameworks
  • Something-something, it seems good to respect people's existing ideas and frameworks. To that end, it seems more morally good to offer them ways to locally change rather than giving them an entire new set of views.
    • This seems true even if the other person is a Kegan 5 and totally understands the relative merits between worldviews. Although maybe not. This is up for debate.
  • Insights are easier to incorporate and easier to remember. If you frame your model as a set of insights, it makes it easier for the user to pick and choose which parts are useful to them. I suspect we already do this when people give us their models in the form of an ontology, so just giving the user more modular pieces seems useful to this end.
  • Moreover, it seems like people usually don't remember anything anyway. EX: for books, people might remember the central thesis, but not likely all of the supporting points.
  • So instead of making all the cool things you want to share seem like just corollaries of some grand theory, it seems good to just focus on the cool things themselves. If the user buys that they're all useful, they might end up reconstructing a similar ontology to explain them all. If not, then at least they might take away a few things, which seems better than just leaving them with the vague shape of the connective tissue you came up with post-hoc to explain your cool observations.

Last Updated: 2020-06-19 23:41
First Published: 2020-06-19 23:41