I am looking for the most ancient proto-world lemmas and it seems, the word for tongue is shared by many families from over the world. Here are some selected examples:

* Proto-Heiban:       d̪ingәla / ringәla
* Proto-Talodi:       t̪ulenge / ləlenge
* Proto-Katloid:      -langed
* Rashad:             tangela-k / yangela-k

* PIE:                dn̥ǵʰwéh₂s
* Manchu:             ilenggu

* Burushaski:         yuŋus

* Mapucho:            ðungu (language)

* Haida (Alaska):     tʼāngal

Sahul & Indo-Pacific:
* Proto-Nyulnyulan:   d-yangala
* Proto-Pama-Nyungan: dyalang
* Bunuban:            djälän
* Tiwi:               tingawa (lip), yunguk (language)
* Proto-Gunwinyguan:  thelng

These examples suggest possible reconstructions as dyungala/lingala.

But what we know is that in Proto-Heiban and Proto-Talodi, d-/r- and t-/l- are prefixes, meaning "long objects or domestic animals". In Proto-Nyulnyulan d-yangala, d is also a prefix, giving reflexes with and without it. In some other languages presence of the initial stop changes the meaning, such as in Tiwi tingawa (lip) versus yunguk (language).

So, I wonder, whether these two forms could persist into PIE?

We know, that the split exists in PIE as well: Germanic tongue, Latin lingua and dingua, Baltic liežùvis and insuwis, Armenian lezu. The change d->l has been explained by influence from the word for "lick", but could it be the case that PIE inherited in fact the both forms, with initial d- and with initial l-?

  • Wouldn't we expect to see initial *l- in some Indo-European forms, then, outside of Latin? (Some people attribute the alternation in Latin between initial d and l to Etruscan; there's not a ton of evidence but it's a nice theory. Either way we don't really see it outside of Latin.)
    – Draconis
    Commented Jan 20 at 5:22
  • @Draconis we do have. As you can see from the question, we have Lithuanian liežùvis, Armenian lezu, both inherited from PIE.
    – Anixx
    Commented Jan 20 at 5:33
  • Yeah, I realized addressing that would take more than a comment, so I turned it into an answer; see below.
    – Draconis
    Commented Jan 20 at 5:34

1 Answer 1


Unlikely. Almost all of the descendants of *dn̥ǵʰwéh₂s (a truly wonderful mess of diacritics, that) show a dental consonant at the beginning; the main ones that don't are Latin and Armenian, and both of those conveniently have a word for "lick" that has a very similar shape (Latin lingua and lingere, Armenian lezu and lizem). In fact, while Balto-Slavic lost the initial dental completely, we later see an l get inserted there in dialectal forms (e.g. apparently some Russian dialects use l'azyk for standard jazyk). Since it would be extremely difficult for this *l to survive all the way into Russian without leaving any traces along the way, it's almost certain that it's an innovation within Russian.

Some linguists have also attributed this Latin d~l alternation to Etruscan, which shows confusion between initial d and l in various words; since our knowledge of Etruscan is limited, there's not a ton of evidence for it (and contamination from lingere is more likely), but it's a fun theory. And if it's correct, it would explain littera and laurus from Greek diphthera and daphnē.

  • Also, Lithuanian liežùvis...
    – Anixx
    Commented Jan 20 at 5:37
  • 1
    @Anixx have you considered sound symbolism to explain the similarities you observe?
    – Keelan
    Commented Jan 20 at 7:28
  • 3
    @Anixx onomatopoeia are a type of sound symbolism. I'm not suggesting tongues make l and ng-sounds. However, ng is a consonant where you can show someone your tongue quite well. And l is used for flexible objects and motions in many words, so there is some relation there as well. I'm not saying that either of these two is necessarily correct. I only want to show that there are reasonable explanations for these parallels other than shared ancestry or language contact. And as long as the parallels are confined to one lexeme, such an alternative may be the simpler explanation.
    – Keelan
    Commented Jan 20 at 14:44
  • 2
    @Anixx do you have regular sound correspondences for these? You say “often reconstructed for Nostratic”, but I’d say most people think there is not enough evidence for Nostratic.
    – Keelan
    Commented Jan 20 at 19:08
  • 1
    @Anixx I’m not saying it’s a coincidence, I’m saying that sound symbolism is another option that should not be discounted.
    – Keelan
    Commented Jan 20 at 19:50

Your Answer

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

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