Standard correspondence analysis (CA) is a visualisation method to show the relationship between two categorical variables on a simplified figure (usually of two-dimensions). This method represents each variable separately, in different spaces, but combines the results on the same figure for interpretation. Consequently, the distance between two categories of different variables is not interpretable and this complicates the interpretation. Additionally, the plausibility of the interpretation depends on the amount of information traded for the simplifications. In this work, we present Tensor CA, a method that circumvents these issues. Tensor CA is then used to analyse the occurrence of verbal forms in several ancient Egyptian texts in hieroglyphic script which were attested between ~400 BCE and ~100 BCE and pertain to the chronolect of Égyptien de Tradition. Our results highlight specific commonalities and differences in the verbal forms of the texts, some of which correlating with specific socio-historical changes occuring during Ptolemaic Egypt.