‘Variant’ is originally an adjective, from Old French variant, from Latin variantem, participle of variare "to change".
A variant in a witness (manuscript, edition…) is a reading which is different from other readings in other witnesses to the same text on the same variant location.
Example: κύριον A D G : θεὸν B C F, where κύριον (in the reference text and in 3 manuscripts) and θεὸν (in 3 other manuscripts) are two variants.
The purpose of collation is to collect (all) variant readings in (all) witnesses. If virtually every word or even every character in a text can be subject to variation, in reality only a (smaller or larger) number of places in a text will: those places are called ‘variant locations’
The term ‘variant’ is neutral, it does not imply any decision about the direction of the variation. In the common errors method, however, only significant secondary readings (also called errors) can be used for the classification of the witnesses, often resulting into a stemma.
On the basis of that classification, the editor will choose the variants that will find their place in the reconstructed text (critical edition), the alternative variants will find their way to the critical apparatus. However, if all variants must be noted in the collation, not all will necessarily be recorded in the apparatus, depending on what the editor wants to be reflected in it: the critical decisions only (in that case only significant readings will be reported), the geographical / regional variations of the textual tradition (in that case even some orthographic variants may be recorded), etc. See also the distinction between analysis of (significant) variants and analysis of forms.
In many statistical methods, variants are used without inferring anything about the direction of variation (polarisation). Even if the result of that kind of classification may be displayed as a tree, the witnesses grouped on the basis of their similarities form groups rather than families.
– Colwell, Ernest C., and Ernest W. Tune. 1964. “Variant Readings: Classification and Use.” Journal of Biblical Literature 83 (3): 253–261.
– Irigoin, Jean. 1979. "Sélections et utilisations des variantes." in Jean Irigoin, and Gian Zarri, eds. 1979. La pratique des ordinateurs dans la critique des textes. Paris 29–31 mars 1978. Paris: Éditions du CNRS: 265-271.
– Love, Harold. 1984. “The Ranking of Variants in the Analysis of Moderately Contaminated Manuscript Traditions.” Studies in Bibliography 37: 39-57.
GE: alternative Lesart, Variante
FR: (leçon) variante
IT: lezione alternativa, variante