We seem to be hardwired to seek patterns, to make simplifications, to reduce things until they are small enough to fit in our existing mental frameworks. An important part of this process involves the usage of assumptions, pre-conceived notions about how things are that are taken as truths. Assumptions are incredibly powerful, as they often distill immeasurably complex things down to a single concept, and allow us to carry on with our thoughts. But we should beware that most of our assumptions are probably wrong.

For example, when you encounter a distasteful piece of content on a social media platform, an assumption you may make is that the content is allowed to stay due to the ad revenue that it generates, because all companies want to maximize profits. The reality could be one of a large number of possibilities, such as you just managed to get unlucky and caught one of the small number of objectionable content that the otherwise extremely effective content filtering system let through by mistake. The assumption allowed you to ignore all the complexities that you cannot see (how the social media platform actually functions), and gave you an easy answer based on a simple heuristic (all companies want to maximize profits).

One thing I have noticed is that having a good principle or heuristic does not make the assumptions based on it any more accurate when applied in individual circumstances. In fact, most assumptions are typically based on good heuristics. The one above for example, that companies seek to maximize profits, is actually a pretty accurate high level statement. But still, assumptions based on it can lead us astray. I think this has to do with reductionism in general: high level heuristics used to make assumptions are simply not very good at analyzing specific situations.

Individual events usually have more direct and more nuanced explanations. The higher level principle may still be, at least in part, ultimately responsible, but it is so far and detached from the specific matter at hand that using it to draw conclusions can take us down the wrong path. This is similar to applying population level statistical data to an individual person, an equally egregious misuse of a powerful tool in my opinion.

Despite the potential problems it brings, we cannot just abandon the usage of assumptions. Without assumptions we will never be able to do anything, as we’ll be stuck analyzing every little detail on the way to figure out the real question at hand. But we should always be consciously aware of the assumptions that we make along the way in our thoughts and arguments. We should always try to put some thoughts into the alternative realities where our assumptions are wrong, and see how that changes our thinking. It will make us less certain of our conclusions, but that is generally (but perhaps not always) a good thing.