I don't speak Italian at all, but I was a bit surprised that the word "flame" in Italian is "fiamma" (IPA: /ˈfjam.ma/) (to compare with flamme in French, flamma in Latin and llama in Spanish). My superficial idea is that it went through the following evolution process:

flamma -> fllamma -> fiamma

I guess that Spanish has influenced the ll to be pronounced as j.

Question: Does this guess make sense? If not, how did fiamma evolve from flamma?

  • 6
    I don't think there is any need to posit Spanish influence here. Spanish pronouncing <ll> as [j] is itself likely a recent-ish phenomenon, with certain regions still pronouncing it as [ʎ], while this sound change in Italian was completed by the time Dante wrote (for some reason, his best-known work does involve the word fiamme a lot!).
    – LjL
    Nov 1, 2020 at 0:39
  • 1
    @VladimirF Absolutely agree about the Code of Conduct. I was not saying it's against any rule but was mostly referring to the "If you’re here to help others, be patient and welcoming" part. I consider it as a guideline and believe that downvoting is not the best way to address neophytes. They may also hover on the "New contributor" indicator and meditate one second or so. Obviously, each site has its own culture, and personally, I prefer a harsh comment than nothing. Nov 1, 2020 at 15:51
  • 1
    I don't think the downvotes are justified here, especially with no explanation, or on a question by a new contributor. Whilst knowledge of how these consonants evolved in the Romance languages may be considered basic by those with some expertise in historical linguistics, there is no sufficiently detailed account somewhere an interested layman would find it (e.g. wikipedia). As such I don't think the question does reflect a lack of research effort. The question is also perfectly clear
    – Tristan
    Nov 2, 2020 at 14:10
  • 1
    the close votes (justified as "Language-specific grammar and usage questions are off-topic unless primarily concerned with linguistics rather than usage") make no sense whatsoever. This question is not about the grammar or usage of any particular language, but about comparing the evolution of two related languages. Such questions are on topic. Even questions about the evolution of a single language are consistently allowed as on topic. This should be no different
    – Tristan
    Nov 2, 2020 at 14:17
  • 1
    @Tristan Indeed, if this site was reserved for research-level Q&A (consider e.g. cstheory.SE vs. cs.SE, or mathoverflow vs. math.SE), then this question would be instantly closed, and I would move on. Nov 2, 2020 at 14:46

2 Answers 2


Syllable-initial Latin "Xl" clusters, where X is a consonant, regularly become "Xi" in Italian.


platea -> piazza ('square')

clamare -> chiamare ('call')

flumen -> fiume ('river')

glacia -> ghiaccio ('ice')

blancus -> bianco ('white')

As you surmise, these went through a stage of /ʎ/ (like Spanish <ll>).

Edit: in some contexts, Spanish goes even further, and loses the preceding consonent:

clamare -> llamar

plenus -> lleno ('full')

but this is not universal in Spanish the way it is in Italian. Compare

platea -> plaza

florem -> flor ('flower' - cf Italian 'fiore').

  • 1
    Thank you for pointing out the pattern and giving several examples. Nov 1, 2020 at 16:04

See Colin Fine's answer for the Spanish story. There exist some similar consonant-vowel pairs in West Germanic:

Standard German l : Austro-Bavarian ɪ
/fal/ : /foɪ/ ("fall" means fall)
/fɔl/ : /fuɪ/ ("voll" = full)
/fyl/ : /fyɪ/ ("füll" = fill)
/va:l/ : /vɔ:ɪ/ ("wahl" = choice)
/vo:l/ : /vo:ɪ/ ("wohl" = well)

Standard German l : Alemannic ʊ
/fi:l/ : /fɪ:ʊ/ ("viel" = much)
/ni:l/ : /nɪ:ʊ/ ("Nil" = Nile)
/mɪlç/ : /mɪʊχ/ ("Milch" = milk)
  • 1
    The long /l:/s don't exist, though, right? Nov 1, 2020 at 9:35
  • 5
    I don't see this answer is very relevant. The Romance phenomenon (in Italian, not Spanish, though it happens to a limited degree in Spanish too) is after a consonant.
    – Colin Fine
    Nov 1, 2020 at 11:43
  • phipsgabler, fair point, I wanted to indicate syllable-length consonants here. Speaking for the variety of Standard German spoken where Austro-Bavarian is spoken, this transcription is adequate though. ColinFine fair enough, I'll see if someone likes the answer and eventually remove it if the users decide to find it irrelevant. I admit it's intended as a background information, not as a direct answer. (addressing the asker's question "how else would such a process happen if not analogous to this Spanish")
    – purlupar
    Nov 1, 2020 at 15:04
  • 1
    What's a syllable-length consonant? Do you mean the ambisyllabicity it develops in /fɛl.lɛ/? (I'm from Graz, but I think the only time I use long consonants is when I speak Finnish...) Nov 1, 2020 at 15:14
  • 1
    According to traditional anaysis, no, AFAIK -- at least there are no minimal pairs (except maybe for some nasals like /hæŋ hæŋː/ häng! - hängen in some varieties). Maybe it's just confusion arising from the spelling. Altough there may be measureable phonetic differences, of course. Nov 2, 2020 at 7:51

Not the answer you're looking for? Browse other questions tagged or ask your own question.