Arabic has سلام‎ (salaam) and Hebrew has שָׁלוֹם‎ (shalom). The words have similar meanings of "peace". This seems like a case of an alveolar fricative shifting to post-alveolar fricative (or vice versa)

Similarly, Hebrew has "ani kotev" while Arabic has "ana aktub"; like a bi-labial fricative has gone to a labial-dental fricative (or vice versa).

This seems to indicate the possibility of some rules about how sounds have shifted. Is there some equivalent of a Grimm's law for semitic languages?

1 Answer 1


Quite a lot of them, in fact!

Grimm's Law is probably the most famous description of a regular sound change. But there are an enormous number of these in historical linguistics, some named, some not.

For example, the correspondence between Hebrew /v/ and Arabic /b/ stems from a change in Hebrew and Aramaic called begadkefat spirantization, where the short non-emphatic stops /b g d k p t/ (hence "begadkefat"; F instead of P because it got affected by this change) became fricatives after a vowel. Some parts of this change then got reversed by later mergers, depending on dialect, but I believe all modern Hebrew dialects still have fricative variants of /b k p/.

The correspondence between Hebrew /ʃ/ and Arabic /s/ doesn't have any nice name that I'm aware of. Proto-Semitic is reconstructed as having multiple different "S-like" sounds: *s₁ ("shin") became /ʃ/ in Hebrew and /s/ in Arabic, *s₂ ("sin") became /s/ in Hebrew and /ʃ/ in Arabic, and *s₃ ("samekh") became /s/ in both Hebrew and Arabic. Based on these correspondences (and other data like Ge'ez *s₂ > /ɬ/), many scholars think that *s₁ was a sibilant fricative of some sort, *s₂ was a lateral fricative, and *s₃ was a sibilant affricate.

The axiom that all phonological changes can be formulated as exceptionless laws like this was a core tenet of the Neogrammarian school of linguistics back in the 19th century. It's not generally accepted nowadays (because there are, it turns out, many changes that can't be formulated as universal laws, and the laws are never truly exceptionless), but the idea of regular, universal sound change laws like this is still a very important tool in historical linguistics.

  • 1
    Can you point me to any scholarly works on this topic? I'm particularly interested in learning how Akkadian might have diverged from Proto-semitic.
    – paceaux
    Commented Jul 8, 2021 at 16:46
  • 2
    @paceaux For Akkadian in particular, Huehnergard's grammar has a good overview in Appendix C.
    – Draconis
    Commented Jul 8, 2021 at 16:51
  • It’s not really clear from your answer, but there is the major difference that Grimm’s Law is not a single change, but the same as the First Germanic Sound Shift, which is an entire three-tier push-chain (or pull-chain) shift that sees voiced aspirates turn to voiced plosives, voiced plosives to unvoiced plosives and unvoiced plosives to unvoiced fricatives (not factoring in the voicing effect of Verner’s Law). So /b/ becoming /v/ in Hebrew is not parallel to Grimm’s Law (would have been /p/), but the /p/ > /f/ in begadkefat is. Commented Jul 8, 2021 at 18:21
  • Extraordinarily helpful, thank you. I'm on a bit of a quest for the etymology of נַעֲרָה (Narah) in Hebrew. I'm told that Arabic has "nuyarrat" for Sparrow, and mehri has "nəγγōr", so the question is, "could the Hebrew name "Naarah" (girl) have at one time been a cognate of some /n/ + /r/ word in another language which meant "sparrow" or "little bird". So my first step is to see how the consonants may have shifted....... suffice to say, if you have any other resources on Semitic phonetics you'd like to share... I'd appreciate it.
    – paceaux
    Commented Jul 8, 2021 at 18:24
  • 2
    @JanusBahsJacquet True, but given the mention of s~ʃ, I took "equivalent of Grimm's Law" as meaning a very consistent sound change across a language rather than specifically a chain shift of voiced stops > voiceless stops > fricatives.
    – Draconis
    Commented Jul 8, 2021 at 18:28

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.