Abstract
More than 500,000 scored words in 83 documents were used to conclude that it is possible to identify the source of documents (proto-orthodox Christian versus early Gnostic) on the basis of the emotions underlying the words. Twenty-seven New Testament works and seven Gnostic documents (including the gospels of Thomas, Judas, and Mary [Magdalene]) were scored with the Dictionary of Affect in Language. Patterns of emotional word use focusing on eight types of extreme emotional words were employed in a discriminant function analysis to predict source. Prediction was highly successful (canonical r = .81, 97% correct identification of source). When the discriminant function was tested with more than 30 additional Gnostic and Christian works including a variety of translations and some wisdom books, it correctly classified all of them. The majority of the predictive power of the function (97% of all correct categorizations, 70% of the canonical r2) was associated with the preferential presence of passive and passive/pleasant words in Gnostic documents.