Semantic Alchemy: How We Used Algorithmic Rituals to Study Ritual

This paper ambiguates the line between algorithm and ritual, demonstrates a novel natural language processing (NLP) technique, and shares some sociological insights from our data analysis of non-religious rituals. In one sense, this paper is a description of an NLP methodology and the insights it offered to the study of the rituals that non-religious people perform. In another sense, this paper describes an esoteric ritual of translation: a procedure to transpose semantic objects into and out of latent space to reveal patterns in a dataset. This method was invented for the interpretation of unstructured data collected in a study of rituals of the non-religious. It has assisted us with the study of interview data (n=40) and survey data (n=3,175). I show how the subject of study (ritual) merged with the methodology of study (NLP) to describe how NLP is something of a ritual itself.