- people expect others to be close to the statistical average
- this expectation is often wrong
- people don't adjust their models for deviating individuals very quickly and sometimes never
- why?
- it's a tradeoff between model simplicity and ability to predict
model 1: "all people are exactly the same"
- the simplest model
- lots of wrong predictions
- misunderstandings from a lack of communication
model n: "there are no correlations between individuals whatsoever"
- the most complex model
- no predictions at all
- unnecessary communicative overhead
- no subscription of medicine possible
- one teacher per student
we perceive someone as a human being if they are close enough to the statistical average in a large enough number of aspects (dimensions) we consider central/essential.
we are able to recognize/build subgroups whose individuals share traits around an average that deviates from the general average. examples: men/women, people of different skin color, body shape, geographical origin, children/adults, gay/straight/bi, geeks/nerds/…, bankers/gardeners/criminals/… and so on (you can be a member of many groups, because there are so many groupable aspects).
the assumption that people can be categorized as belonging to a group of more than just a single person is economical. it enables us to save on storage and complexity and comes with the advantage of starting the exploration of new people with a fixed set of hypotheses to test — and that's why we are interested in statistical studies.
knowledge about individuals is stored in memory unless we have evidence that we should instead create a new category. for example, a person could come from a place we don't know and we would assume that what they look like is what people who live there in general look like. we don't create categories for every aspect that can be shared among people. one example is personal preferences. there is no point in having a category for people who like strawberry jam and one for those who don't, the reason being that you have no way of predicting which of those groups any given person would belong to before having asked them and stored this information in memory, specifically for this person. if liking or not liking strawberry jam is something important to you, you may also have a list of people who like/dislike it stored somewhere, so you don't forget about it, but it doesn't qualify as a group/category in the above sense, because of the lack of correlations between aspects of which you could know one (or a few) and infer others.
there is a point when memory and prediction overlap: the moment where we find out something about a person that we wouldn't have predicted. in this situation, we have to make a decision about what adjustments to make, depending on which case we're in:
- there could be a misunderstanding — we only need to adjust our understanding of the current conversation/situation
- the person is a member of a group we know of — we need to add them to it or add a new rule to our internal group definition
- the person is a member of a group unknown to us — create a new group, gather knowledge, build an internal definition
- in this specific aspect, the person is just different than other people — create a memory for this fact about them
the cases are ordered from simple (cheap) to complex (expensive). that's why we try to avoid going all the way down to person-specific memory.
the more person-specific knowledge is required to understand an individual, the harder it gets for us to deal with them. it depends on how important they are to us, personally, and how much time we spend with them, gathering more data, reinforcing existing memories, using memories to predict the person's behavior.
also, how important it is for us to be right most of the time and to prevent misunderstandings can be different from person to person. a way of dealing with general unpredictability would be to invest more time and energy in learning a more robust method of communicating with others. this approach might be especially fitting for the kind of person who likes to have a lot of friends rather than just a small number of close friends (and vice versa).