0. Many discourse markers (DMs) could have also another status in the language: in Italian, f.ex., it could be a verb (figuriamoci, diciamo, guarda, dai, senti), a noun (in realta', in pratica), an adverb (veramente, praticamente) etc. A large group of italian DMs corresponds to a form of the verb, that loses partially not only its meaning but also the typical syntactic and morphological properties, and acquires functions and features of DMs (such as positional variability, intonation patterning, invariable form etc.).The purpose of this presentation is to demonstrate the relation between the semantics and the morphological form of the verb and the conditions of use of the verb-formatted DM.

1. In different languages, the verb-formatted DMs often derive from the equivalent verbs that describe, as a rule, the action accomplished by the participants of the communication (the speaker (S0) and the addressee (S1)). Ex. Dire (to say) - S0 says p to S1, sentire (to listen) - S1 listens to what S0 says (=p), guardare (to look) - S1 imagines p (= what S0 says) etc. The relation between the speaker (S0) // the addressee (S1) - the scope (what is said = p) and the other context (q), indicated by the DM, is presupposed by the semantics of the verb transformed into the DM. But the link between this three components of communication is established by each DM in a different way. It meens that to characterize a DM and to find an equivalent translation, we need to describe three types of relation: - the relation between S0-S1, - the relation between S0/S1-p, and - the relation between p-q.

Ex. The Italian DM 'guarda' and the DMs from the equivalent verb in French and in English (regarde, look) are interpreted in the situation of the communication as: ‘the speaker draws the attention of the addressee to p’. But this definition could be specified for each DM according to three types of relation (indicated above). Guarda: 'S1, to accomplish a correct action, need to have the information p’, p has a status of a warning. Ma, guardi, c’è un equivoco, - cerchi di avvertirlo, - questo non è un testo (Calvino) In French, to indicate the same type of relation we need to use another expression: Attention, écoutez, il y a un malentendu /.../

> The DMs that derive from the equivalent verb often mark different character of relations and cannot be translated one by another. The description of three types of relation allows to specify the conditions of use of every DM.

2. The character of relation is often specified by the remained morphological form (that has a deictic character) of the DM: - the tense is suggested to indicate the relation p-q, - the mood specifies the relation S0/S1-p, and - the person - the relation S0-S1. The contextual conditions of use reflect the relation between the semantical features of a DM and the morphological form.

Ex. The 1 pers.pl. indicates that S0 and S1 accomplish an action together. The DMs of this form are often used by S0 as a mark of his union with S1. S0 and S1 accept the role of a collective author of p. It presupposes the presence of the shared knowledge or opinion (ex. Figuriamoci) and the assumption of the responsability for p by both S0 and S1 (ex. Diciamo).

Noi pero’ avevamo torto di ritenerci membri dell’unica minoranza che in Italia fosse perseguitata. Figuriamoci! (Bassani) (=We were wrong to consider ourselves the member of the only minority that was persecuted in Italy.Figuriamoci!). Figuriamoci! is translated explicitly in French: ... sans nous rendre compte qu’il y avait plusieurs autres , de ces minorités, qui souffraient comme nous et plus que nous.(= ... without realize that there was several other minority that suffered as we did and even more.)

> The DMs, derived from the different form of the same verb, are transformed into different markers with different conditions of use and cannot be substituted by the other form of the same verb (ex. Figuriamoci - figurati). It suggests that the morphological form of a DM of verbal origin has a sense-formatting function.