Published

November 17, 2024

Poetry is a form of expression that uses language’s rhythmic and aesthetic qualities to evoke emotions and convey ideas. Though poetry may not always offer direct solutions or translations for our daily lives, it broadens our perspectives by introducing us to diverse emotions, thoughts, and experiences.

Quantitative social science is the study of human behavior and societal patterns using statistical analysis and empirical data. Quantitative social science does not yield findings immediately applicable to individual experiences, it simply provides invaluable insights into societal trends and structures, broadening our understanding of the world around us.

The two are much closer than is usually assumed, we generally assume little in common.

Quantitative social science primarily describes, and even when addressing causality, it’s still one possible description of a phenomenon. Our social scientific descriptions aren’t real like a rock is real, they tell one possible story, other possible stories are possible, one perspective out of many.

The usual reply is that quantitative social science is anchored in empirical evidence, its conclusions undergo formal tests, it seeks objectivity in contrast to the subjectivity of poetry, and while poetry stirs emotions, quantitative social science aims to explain and predict. While I can’t outright dismiss these points, I urge you to approach what I’m saying with an open mind.

Much of quantitative social science veers towards edutainment: the observed effects are small, it rarely offers accurate predictions and while it might show some patterns, it almost never exerts tangible control on anything, its real-world impact is minimal. True control of something, like in the non-social and non-human sciences, we only get in randomized controlled trials, but even there, the effects are small, the large RCTs are suspect (I’m talking about social science, not vaccines; vaccines have a physical causal mechanism and the effect can be large).

The absence of universal laws in social science means that your regression analysis on 36 year-states isn’t a law but a stylized fact; akin to a narrative. Researchers publish these edutainment stylized facts with little real world impact to get tenure. We got calls, recently, from leading economists, for humility. Quantitative social science is not useful in the sense that it helps us predict, understand and control the world, it doesn’t. It does describe: rich people are rich, have more opportunities, some people disagree, others agree. This kind of description is very limited in what it ways about the world. It does put certain narratives out there but these narratives are in competition with other narratives and it’s unclear to me whether the narratives presented in peer-reviewed journals genuinely provide clarity in our comprehension of the world.

The web of belief is this metaphor describing knowledge as an interconnected network of beliefs, where no belief is immune to revision, and empirical observations influence the web’s edges, while central, foundational beliefs are more resistant to change. Poetry is much closer to empirical social science than non-social non-human science. One’s about saying things that matter to humans, the other is to describe and control the non-human world. Very different.