2 How to think like an ipseologist
2.1 Personally expressed identity is the data.
One measures ipseity with language. Specifically, the data sought is personally expressed identity text. This text must satisfy all three conditions:
It should be personal - the authors are describing themselves.
It should be expressed - the authors’ text is “published,” i.e. the words are available where others might see them.
It should describe identity - the explicit purpose of the text is description of the author.
Currently, personally expressed identity text data is best sourced from social media profiles. For the years 2012-2023, Twitter profile biographies (bios) were the best source because of the scale at which the platform was used and the open availability of public profile data.
2.2 Individuals are bags of signifiers.
Individuals use words to describe themselves. Ipseologists presume these words are a strong signal of self-perceived identity. Others worry whether these words should be considered veridical, aspirational, performative or some other category. Ipseologists don’t worry. Or more accurately, we don’t pretend that there is some gauge of the ‘true self’ that a bio fails to comport with.
When we want to know who someone is, we ask them. Then we analyze the language.
Specifically, we look at the signifiers (words) they choose. Consider @BarackObama’s bio: Dad, husband, President, citizen. Here is a model subject - no function words, just a comma-delimited list of identity signifiers.
Everyone is a bag of signifiers. Presuming this vastly simplifies analysis. Each individual, at each time point is a small set of self-ascribed words. Analysis can begin with mere counting. How many individuals consider themselves dad and how many mom? Small increases in analysis complexity reap big rewards. One can proceed to measure which signifiers co-occur with mom but not dad. If an individual removes the word dad from his bio, which signifier is he most likely to replace it with?
We will look into these more sophisticated questions soon, but let’s first build a foundation with simpler analyses. Chapter 3