The Fisher-Yates exact test, that is implemented in the textometric Specificity measure, still proves to be relevant for textual data analysis because it is both meaningful and reliable. Its meaning comes from the understanding of the underlying mathematical model, that would potentially benefit from being better known by every user. In the textometric practice, the Specificity results are part of an interpretative path, along which the user gradually orients themself and combines several pieces of information. This importance given to the multidimensionality of analysis meets the tupleization approach proposed by Gries (2019). From a broader perspective, textual data analysis embraces a diverse range of practices with sometimes divergent expectations and objectives, concerning for instance the place given to text, the extent of automation, and the nature of evaluation, so that common elementary materials and tools can lead to substantiated yet completely different implementations and developments.