From grammar to dictionary. The early challenge of lemmatizing Tamil verbal forms, through categories used for Latin and Portuguese

This contribution,1 based on an examination of several Tamil dictionaries and Tamil grammars, composed in Portuguese and in Latin, by missionaries who were in Tamil Nadu during the 16th, 17th and 18th centuries, examines the lemmatization strategies which they followed, while dealing with Tamil verbal morphology. If nominal forms were not really a problem, verbal forms were more difficult to cope with. This is why for instance Proenca’s dictionary is very far from being completely lemmatized, and many of the forms which modern lexicographers would consider as falling under the same head, are listed as separate entries, and given separate translations. The complexity of the morphology was progressively mastered by grammarians, using labels taken from Portuguese or Latin terminology, although they did not always agree between themselves, concerning for instance what should be called infinitivus, some of them introducing new labels such as infinitivus substantivus and infinitivus absolutus. The most difficult nut to crack, however, was probably the existence of diathetic pairs, consisting of two paired verbs, which some modern linguists have referred to as ‘affective’ and ‘effective’, additionally accompanied by some causatives.