In an onomasiological study that aims at comparing languages, a database can be created in the form of a corpus of texts which are translation equivalents of each other. The texts have the same designatum, and one can ascertain how it is coded in different languages. Texts that have been used for this purpose inlcude books that have been translated into many languages: the bible, Antoine de St. Exupéry, Le petit prince, Jane Rowling, Harry Potter.

A translation is an asymmetric relation between texts: one is the original, the other tries to render the former. The latter therefore makes use of the resources of its language guided by the use made in the former. Consequently, certain resources of the latter language usually lie fallow.

One can compare translations with each other and use the original as the tertium comparationis.¹ This does guarantee a symmetric relation between the texts compared. It does not, however, make up for the deficient use of the resources in each of the languages.

In contrastive studies that compare languages L1 and L2, one can make up for the drawback of the method by using one text of L1 and its translation into L2, and another text of L2 with its translation into L1.

¹ In principle, the New Testament would be an ideal object for that method. However, it is in fact not always easy to determine, and in fact often doubtful, whether available texts are actually translations from the Greek original.