Bourdieu Vectors: A Framework for Quantifying Taste in Social Space

This paper introduces Bourdieu Vectors, a computational publicly accessible framework (https://bourdieuvectors.com) for estimating the social, cultural, economic, and symbolic capital of individuals based on the cultural practices they engage in. Each practice (e.g., football, opera, cooking) is represented as a multidimensional vector capturing likely levels of various forms of capital. Vectors are generated via prompt-based querying of the Gemini-2.5-Flash-Lite model. Bourdieu Vectors enable systematic analysis of social practices, measurement of similarity between practices, construction of individual or group taste profiles, and generation of personalized recommendations. The framework is intended for use in social sciences, digital humanities, market research, and cultural analytics, emphasizing transparency, reproducibility, and ethical responsibility.