Will Barratt, PhD
Coffman Distinguished Professor of Higher Education Leadership Emeritus
Indiana State University
Categorical thinking is a bad idea. The classic western model of data types, (Stanley Smith Stevens, "On the Theory of Scales of Measurement", 1946) proposed Nominal, Ordinal, Interval, and Ration types of data a model. This proposed scheme / typology is often taught as gospel / dogma in statistics and measurement classes. However, there are a lot of alternative schemata for typing data. Nearly all of these data schemata use some form of categorical data (Nominal in Steven's case, Discrete in others) and this is a problem. Simplifying reality, which is always a part of measurement, leads to simple thinking. We all simplify reality in order to make an orderly world, and we all try to ignore the consequences of categorical thinking.
Example: How big is a 2 x 4, a standard piece of wood in the USA used for a lot of construction? The answer depends on how closely you want to measure the wood the day you measure it (tape measure or micrometer), on the humidity that day (moist wood swells, dry wood shrinks), on the temperature that day, and on many more things that might affect the real size of the 2 x 4 (which are supposed to be 1 1/2 x 3 1/2 inches (38 x 89 mm)).
Other convenient, and false, categories are gender, age, race, ethnicity, religion, and social class. It is a psychological and statistical convenience to treat these ideas as nominal or categorical data, and this leads to simple thinking about complex things. As an example think of how many levels of categories are there for Christian/Muslim/Hindu/Buddhist to get clarity on the category of Christian/Muslim/Hindu/Buddhist for Christians/Muslims/Hindus/Buddhists who you know?
As I have said to the amusement of many friends "2 is a stupid number". Even the idea of trans-gender reflects a categorical data world view. Self identification of race on the US decennial census has only been used since 1960. Again, classifying people into groups simplifies analysis. Anyone who has gotten the results of their genetic analysis knows that categories fail when faced with percentages of ancestry associated with current nation states. How much sub-Saharan African do you need to be to be "Black"? How much European do you need to be "White"? One drop? How should we assign borders to these categories?
The categories of social class, working class, middle class and so on, are a convenience. I developed the Barratt Simple Measure of Social Status (this is a measure of status and not class) to help researchers work with status, and by inference class, as a variable. When queried, usually by young researchers, about how to turn the BSMSS score into a social class, I note, politely, that class is not a category but a fuzzy and fluid idea, so please use correlation in your analysis and not ANOVA (which is based on categorical membership).
Bourdieu, in Forms of Capital, provided a post-Marxist view of class by including social capital and cultural capital to add to the Marxist notion of economic capital. Bourdieu's model is better, in that it is a more comprehensive model of class, and looks at social, cultural, and economic wealth as continuous (not categorical) variables. This is a much more fluid and fuzzy model, departing radically from class categories.
Our minds confuse the map (categories) with the territory (real life). Our brains have been colonized with a kind of thinking that leads us to identify us and them, with all the consequences of that world view.
Fluid and Fuzzy
When I read or hear people use phrases like "White People", "CIS", or "Middle class" I am tempted to stop reading or listening because of their acceptance of categorical models of reality demonstrates ignorance of the real world. I am learning patience with categorical thinking, but it is a challenge.
Fluid and fuzzy is a much more difficult way of being in the world. What used to be nouns become, well, nouns with permeable and flexible boundaries. At what point does yellow become orange? Fluid and fuzzy makes some people very uncomfortable. It takes a lot more energy to be in a fluid and fuzzy world than in a categorical world.
Google social class categories and you will get an array of answers - all categorical. An old answer of mine would be that social class groups are arrayed like pearls on a string from low to high prestige. Even pearls imply discrete groups. Even the 1% is categorical, requiring as of this writing an annual income above USD 540,000. Or do you mean the 1% in wealth in the world, which is about USD 4,400,000. Again with the categories. The clear bright line between 1.01% and 1.00%. Yeah, that boundary line to include and exclude is arbitrary. Remember the 1 drop rule for being Black in the USA? Not 1/2 drop, but 1 drop. This is an equally stupid way to categorize what is not inherently categorizable - better check your genetic profile again to make sure of your racial category.
tl;dr fluid and fuzzy is better than categories
No comments:
Post a Comment