I'm trying to figure out which phenomena may be involved in the development of Italian "dito" from Latin DĬGĬTU(M).

I think one of them may be a loss of intervocalic -G-, as explained in this article for other words. This is also explained in the book Lineamenti di grammatica storica dell'italiano by Giuseppe Patota (Bologna, il Mulino, 2002):

In posizione interna la velare sonora, dopo essersi palatalizzata, ha subito un'ulteriore trasformazione, e in alcuni casi si è intensificata (come in LĔGIT > legge), in altri si è dileguata perché assorbita da una I successiva, detta omorganica perché pronunciata con gli stessi organi articolatori della consonante precedente (in entrambi i casi si tratta di foni palatali). Per esempio, dalla base latina SAGĬTTA(M), alla palatalizzazione della velare ([sa'ʤitta]) è seguito il suo dileguo, che ha prodotto SAĬTTA e poi saetta, con regolare trasformazione della Ĭ tonica in e chiusa [e].

My translation:

In internal position the voiced velar, after being palatalized, has undergone further transformation, and in some cases has been intensified (as in LĔGIT > legge), in other ones it was lost because it has been absorbed by a subsequent I, said homorganic because it's pronounced with the same articulatory organs as the previous consonant (in both cases they are palatal phones). For instance, from the Latin base SAGĬTTA(M), to the palatalization of the velar ([sa'ʤitta]) followed its lost, which has produced SAĬTTA and then saetta, with regular transformation of the tonic Ĭ into closed e, that is, [e].

But then, these two Ĭ in contact in *DĬĬTU, could they have behaved like a sort of long vowel, in such a way that the development Ī > [i] was involved? Has this any sense?

My problem is that I don't know if [ii] > [i:] is an usual development in Vulgar Latin or in early Romance languages (in particular, in early Florentine), although it seems reasonable to me.

I imagine that something similar to Italian happened in Catalan (one of my native languages), in which DĬGĬTU(M) > dit.

I've seen that, according to Daniel Recasens i Vives, what it happened in this case both in Tuscan Italian and in Catalan, is the evolution GI > [j] followed by the loss of this [j].

In RECASENS, D., "Velar and dental stop softening in Romance", published in Diachronica, 28:2, (2011), 186-224, there is a section in which the author is analyzing the phonetic outcomes for intervocalic GI/E, GY, DY and I in different Romance languages showing the differences on the place of articulation. In this section, there is a note which states:

Minor phonetic solutions deviating from the general pattern have not been included in the tables, e.g., [j] or zero for GI/E and an alveolar affricate for DY in Tuscan Italian words ([ˈdito] DIGITU “finger”, [sarˈtana] SARTAGINE “pan”, [ˈmeddzo] MEDIU “half (masc. sing.)”).

The book Fonètica històrica del català by the same author (Barcelona, Institut d'Estudis Catalans, 2017) explains something similar for Catalan in a section in which the author is analyzing the different outcomes of the sequences J, B[j]-V[j], D[j], GE and GI:

En alguns vocables del català general la solució fonètica dels grups consonàntics que ens ocupen ha estat [j] o l'elisió de la consonant.

Alguns mots experimentaren elisió d'una realització afeblida [j] de [ɟ] provinent de Ge,i, ja en llatí vulgar, per superposició gestual i/o confusió perceptiva entre aquest gradual i una vocal anterior immediatament precedent o següent. Exemples: mestre MAGISTRU, vint VIGINTI, ficar *FICCARE < *FIGICARE, dit *DITU < DIGITU, fred FRIGIDU, mai, més MAIS < MAGIS; cuidar COGITARE (cat. ant. cug, cuyg amb africada; DECat, vol. 2, p. 1091).

My translation:

In some general Catalan words the phonetic solution of the consonantic sequences that concern us has been [j] or the loss of the consonant.

Some words experienced the loss of a weak realization [j] of [ɟ] coming from GE or GI already in Vulgar Latin, through gestural blending and/or perceptive confusion of this gradual release and an immediately preceding or following front vowel. Examples: mestre MAGISTRU, vint VIGINTI, ficar *FICCARE < *FIGICARE, dit *DITU < DIGITU, fred FRIGIDU, mai, més MAIS < MAGIS; cuidar COGITARE (old Catalan cug, cuyg with affricate; Diccionari etimològic i complementari de la llengua catalana, vol. 2, p. 1091).

So, according to Recasens, for both Italian dito and Catalan dit


However, this doesn't explain how one gets [i] from Ĭ.

  • 1
    Since Italian doesn't have long vowel phonemes, what's not clear about Ī > [i] development?
    – Yellow Sky
    Commented Jul 18, 2022 at 17:14
  • 1
    @YellowSky: The point is that there are two Ĭ in the Latin word, not an Ī.
    – Charo
    Commented Jul 18, 2022 at 17:22
  • 2
    @Charo That’s a pretty trivial change, though. I’d say it’d be harder to find a language where /i.i/ doesn’t eventually become /i:/ over time than one where it does. You do sometimes see dissimilations like /i.i/ > /ei/, especially if the sounds belong to separate morphemes (and in fact I think something like that must have happened in French/Spanish/Portuguese, which all have reflexes of earlier /e:/), but I don’t think that applies to Italian. Commented Jul 18, 2022 at 18:22
  • 1
    I'm sorry if the answer of my question is so simple. I'm newbie in historical Romance linguistics. Can anyone please write an answer?
    – Charo
    Commented Jul 26, 2022 at 12:33
  • @JanusBahsJacquet: After reading the works of Recasens, this doesn't seem so trivial to me.
    – Charo
    Commented Jul 27, 2022 at 18:21


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.