Reading this and the comments on it I think I’m really starting to understand why folks find these sorts of models so appealing (and as a corollary why developmental models are unappealing). Also, my apologies in advance, because this comment contains an implicit and inherent element of status positioning that will place me, the commenter, above you, the reader who finds these sorts of models useful, but please know I too used to find these models very appealing so I know what it feels like from the inside, but now I know what it looks like from the outside.
If trying to directly predict the behavior of real people by modeling them and making predictions about what they will do doesn’t result in sufficiently accurate or precise predictions for your purposes, then you will be quite happy if there is a tool that will help you make better predictions. When there is only one category that contains all people your only method is to consider people fully as individuals, which is a lot of work and something you may not even know how to do (contrary to what the heavily Red and Green will tell you, understanding people on an individual level is a skill that must be developed just like any other if you want to do it well). If you add another category it gets a bit easier, though: now at least a person is either closer to one thing or another and you can make slightly better predictions without knowing anything more than 1 bit of information (think male/female, yin/yang, hot/cold, near/far). Adding more bits/categories means you can be more precise, so it’s tempting to build models that have more categories that let you more accurately predict what other people are like so long as you can learn enough about them to categorize them, which is generally much easier than understanding all of their individual nuance.
But the human brain has limits, and 5 things is about the max most people can handle in working memory, so models will generally max out around 5 categories. More categories might help but even if they did it would weaken the model because it would be hard to keep it all in your head at once and you’d need to look for ways to simplify things. Additionally if you had unbounded numbers of categories that’s just back to the same problem of as having only one category with everything in it only now you’re thinking in terms of an unbounded multi-dimensional feature space rather than an unboundedly complex category.
So being able to categorize people along 4 or 5 dimensions is basically the sweet spot of trying to understand other people (or complex systems in general, but we’ll stay on people here) without having to deal with the full complexity of treating them as individuals. It grants the most accuracy and precision possible without demanding potentially overwhelming effort. That these models consequently give us ways to understand ourselves and understand how others see us only adds to their appeal.
Developmental models, in contrast to these sorts of multi-dimensional models, provide no such usefulness because they are about one particular dimension (complexity) and say very little that will be directly useful to the object-level task of predicting what people will do except in those cases where you want to asses things like behaviors they are unlikely to be able to take. Yet developmental models also point at why I don’t find these multi-dimensional models very useful anymore.
The trouble is that greater complexity means more capacity means encompassing more of what is possible, and so these models only work so long as a person is not so complex as to encompass the whole thing and the user of the model is not capable of the complexity of considering others individually with sufficient fidelity. To take myself as a ready example, I’m sure there are those here who would accuse me of a lot of Greenness, but that’s mostly because I’m often leaning against the strong Blue vibe here and choose to emphasize the Green because my audience is already so Blue. And depending on what contexts you know of me in person, you may see me as more Black if you meet me in a White context and more White if you meet me in a Black context because I will be more of the other thing than the thing we are in. Red feels the most fake dimension to me, but if you think I am particularly Red or particular not-Red I suspect you it is likely because I am again closer to half-Red and so both farther away and closer to Red and not-Red than it looks. So I’m all the colors at once and what I look like depends more on the external lighting than any characteristic behaviors. And since I know this about myself I know it about others, so these models cease to seem very helpful because they can be washed out and I have other tools I can use in their place.
But lest I end this with you thinking I’m down on multi-dimensional models, I’m not. They are important tools and bridges that can enable deeper understanding. And in this sense everything is a bridge that can lead to deeper understanding. I just now understand for myself better why these models are so useful to folks even when there are “better” tools available.
I liked this comment for spelling out the ways this sort of model is useful and limiting. Some meta-ish commentary:
I felt like it could have been more succinct, and that the caveats about status positioning and apologizing for “seeming down on these models” felt mostly unnecessary.
(I think that sort of thing is sometimes necessary, but most often is more like a red-flag that you could probably refactor your comment/post until they no longer felt necessary)
Reading this and the comments on it I think I’m really starting to understand why folks find these sorts of models so appealing (and as a corollary why developmental models are unappealing). Also, my apologies in advance, because this comment contains an implicit and inherent element of status positioning that will place me, the commenter, above you, the reader who finds these sorts of models useful, but please know I too used to find these models very appealing so I know what it feels like from the inside, but now I know what it looks like from the outside.
If trying to directly predict the behavior of real people by modeling them and making predictions about what they will do doesn’t result in sufficiently accurate or precise predictions for your purposes, then you will be quite happy if there is a tool that will help you make better predictions. When there is only one category that contains all people your only method is to consider people fully as individuals, which is a lot of work and something you may not even know how to do (contrary to what the heavily Red and Green will tell you, understanding people on an individual level is a skill that must be developed just like any other if you want to do it well). If you add another category it gets a bit easier, though: now at least a person is either closer to one thing or another and you can make slightly better predictions without knowing anything more than 1 bit of information (think male/female, yin/yang, hot/cold, near/far). Adding more bits/categories means you can be more precise, so it’s tempting to build models that have more categories that let you more accurately predict what other people are like so long as you can learn enough about them to categorize them, which is generally much easier than understanding all of their individual nuance.
But the human brain has limits, and 5 things is about the max most people can handle in working memory, so models will generally max out around 5 categories. More categories might help but even if they did it would weaken the model because it would be hard to keep it all in your head at once and you’d need to look for ways to simplify things. Additionally if you had unbounded numbers of categories that’s just back to the same problem of as having only one category with everything in it only now you’re thinking in terms of an unbounded multi-dimensional feature space rather than an unboundedly complex category.
So being able to categorize people along 4 or 5 dimensions is basically the sweet spot of trying to understand other people (or complex systems in general, but we’ll stay on people here) without having to deal with the full complexity of treating them as individuals. It grants the most accuracy and precision possible without demanding potentially overwhelming effort. That these models consequently give us ways to understand ourselves and understand how others see us only adds to their appeal.
Developmental models, in contrast to these sorts of multi-dimensional models, provide no such usefulness because they are about one particular dimension (complexity) and say very little that will be directly useful to the object-level task of predicting what people will do except in those cases where you want to asses things like behaviors they are unlikely to be able to take. Yet developmental models also point at why I don’t find these multi-dimensional models very useful anymore.
The trouble is that greater complexity means more capacity means encompassing more of what is possible, and so these models only work so long as a person is not so complex as to encompass the whole thing and the user of the model is not capable of the complexity of considering others individually with sufficient fidelity. To take myself as a ready example, I’m sure there are those here who would accuse me of a lot of Greenness, but that’s mostly because I’m often leaning against the strong Blue vibe here and choose to emphasize the Green because my audience is already so Blue. And depending on what contexts you know of me in person, you may see me as more Black if you meet me in a White context and more White if you meet me in a Black context because I will be more of the other thing than the thing we are in. Red feels the most fake dimension to me, but if you think I am particularly Red or particular not-Red I suspect you it is likely because I am again closer to half-Red and so both farther away and closer to Red and not-Red than it looks. So I’m all the colors at once and what I look like depends more on the external lighting than any characteristic behaviors. And since I know this about myself I know it about others, so these models cease to seem very helpful because they can be washed out and I have other tools I can use in their place.
But lest I end this with you thinking I’m down on multi-dimensional models, I’m not. They are important tools and bridges that can enable deeper understanding. And in this sense everything is a bridge that can lead to deeper understanding. I just now understand for myself better why these models are so useful to folks even when there are “better” tools available.
I liked this comment for spelling out the ways this sort of model is useful and limiting. Some meta-ish commentary:
I felt like it could have been more succinct, and that the caveats about status positioning and apologizing for “seeming down on these models” felt mostly unnecessary.
(I think that sort of thing is sometimes necessary, but most often is more like a red-flag that you could probably refactor your comment/post until they no longer felt necessary)
^^^Parent comment is blue/green to a T