The Quest for the Gist of Jesus: The Jesus Seminar, Dale Allison, and Improper Linear Models

Studies have found that while experts can be quite good at identifying criteria related to a particular phenomenon, they are typically outperformed by improper linear models (ilm), which assign equal weights to criteria. In this article, using widely-accepted criteria for assessing the authenticity of the sayings of Jesus, we generate a new ranking of Jesus’ sayings using an ilm. Then, drawing on recent advances in text mining—semantic network analysis—we first compare our ilm ranking to that of the Jesus Seminar’s and then to one based on Dale Allison’s recurrent attestation (RA) approach. We find that our ilm semantic network projects a more traditional understanding of Jesus than does the Jesus Seminar’s, but it is quite similar to the RA network. We conclude by suggesting that biblical scholars could benefit from various forms of computerized text mining in their quest for the historical Jesus.