Grammaticalization processes in the genesis of Latin modal adverbs: rewriting the story of scilicet and videlicet

The grammaticalization path leading from lexical verbs to adverbs commonly takes place in verb-verb constructions where finite or non-finite verb forms tend to undergo formal erosion and semantic bleaching (Heine and Kuteva 2007: 73). A well-known example in Latin is the evolution of licet, which gives rise to the couple of synonymous adverbs scilicet and videlicet, traditionally explained as derived respectively from scire and videre + licet ‘one can know/see (that)’. On closer inspection, however, the process at stake is not as straightforward as it seems from the above description, and a detailed analysis of the data raises a number of issues. On the one hand, if we assume that the formations scire licet and videre licet were very early, we would expect any independent meaning of their component parts to have entirely faded away. In Plautus and Terence, however, both scilicet and videlicet can also govern indirect discourse just as would scio and video. Besides, licet combines with scire and videre precisely like any other infinitive, and nothing proves that such phrases are stereotyped, or frozen into adverbs. On the other hand, this seems to imply that the evolution was relatively late, but in that case probably too late for the phonetic development that must be posited if the first element is an infinitive. Moreover, a survey of Plautus’ uses of licet + infinitive (Hahn 1948: 316 ff.) shows that in most instances there is an individual agent, almost always either represented by a case-form (the dative, occasionally also the accusative), or else clearly supplied from the context. In point of fact, only 4 examples out of 99 display the generalizing sense of the locutions scire licet and videre licet. Also, both full constructions are indeed still frequent in classical age and later, and statistical data about the usage of licet, occasionally preposed and also inflected for tense/mood, point out that the form was neither fully desemanticized, nor decategorialized. In other words, this type of infinitival construction hardly accounts for the genesis of the adverbs at issue. A different explanation probably lies in the word used in funerals and when terminating judicial proceedings: ilicet ‘it’s all over’, a third, and often neglected adverb that in Classical Latin acquires the meaning ‘immediately’ by analogy with ilico. A detailed discussion of its origin and usage will highlight the role of analogy, frequency effects and new discourse functions (Bybee 2003) related to the autonomous use of licet in answers, whereas recent findings about the mechanisms of semantic change in modality (van der Auwera and Plungian 1998, Traugott and Dasher 2002, Magni 2005) will help to clarify why only the verbs ‘go’, ’know’ and ‘see’ are involved in this process.