How could you explain or analyze these written Arabic from the non-concatenative morphology point of view? These verbs are derived from nouns.

bakkala (to buckle)
bukla (buckle)
tilifu:n (telephone)
talfana (to telephone)
tilifiziu:n (television)
talfaza (to televise)
zayt (oil)
zayyata (to oil)

thanks in advance!

  • I don't recognize the last two, but the rest are all English borrowings, in one way or another. Probly many more could be found with the same roots or the same meanings in other dialects of Arabic; there are a lot of strategies for what happens to borrowed words in Semitic languages, as the case of Modern Hebrew illustrates equally well.
    – jlawler
    May 16, 2014 at 18:25
  • How do you know that these are derived from nouns? What about veebs derived from nouns. How could we use prosodic morphology to treat them and what is our motivation for using prosodic mo?
    – user10590
    Oct 3, 2015 at 14:57
  • Actually the 'oil' derivation is incorrect. زَيَّتَ zayyata means (he oiled), to oil is تَزْييتْ (to oil)
    – WaelJ
    Jul 9, 2017 at 11:11

3 Answers 3


I don't know Arabic, but what I can extrapolate from Hebrew, which I know a bit, is that you treat the consonants of each expression in the same fashion as you do root consonants. Let's take the infinitives first. I use capitals for root sounds. The root of "buckle" is BKL. The infinitive pattern is CaCCaCa, where C stands for "consonant". Now you insert BKL for the Cs in the pattern, and you get BaKKaLa. The same would be true for "oil":

ZYT + CaCCaCa = ZaYYaTa

The interesting point here is the doubling of the middle root consonant, which seems to be required whenever the root is triliteral. The infinitive "to telephone" is a quadriliteral root:

TLFN + CaCCaCa = TaLFaNa

The root for "television" has five consonants, and it seems that the final /n/ simply gets dropped:

TLFZ + CaCCaCa = TaLFaZa

  • The interesting thing about such nonconcatenative morphology for me concerns how the segmentation of the morphs is construed. One has to assume morphs that consist of non-contiguous parts, right? In other words, a word like "bukkala" consists of two morphs, the one being b_kk_l_ and the other being _a__a_a. The one morph is then superimposed onto the other. Is this correct way to conceive of the morphology of such words? May 17, 2014 at 0:45
  • @TimOsborne: Except that gemination isn't part of the lexical morpheme but of the template: i.e. you have a root morpheme bkl and a template CaCCaCa, plus a rule telling you that the middle consonant of the root should get mapped onto both the middle C-slots of the template. That's one way of looking at it, anyway. But yes, in any case you have to assume non-contiguous morphs; a few analysts have tried to do without this assumption and use stem allomorphy instead (for Hebrew; don't know about Arabic), but this isn't very convincing (to me at least).
    – TKR
    May 17, 2014 at 3:04
  • Autosegmental Phonology (Goldsmith 1976/1990, McCarthy 1981, Lieber 1987) have a tier system that handles the consonant/vowel interaction in Semitic languages. The fundamental problem is that in order to distinguish radices (roots) and patterns in the way Tim indicates, one's approach must be unit-based. Of the most influential morphological theories, only Distributed Morphology is unit-based. The difficulties with establishing morphological units of form seems to have frustrated many a linguist. But see my angle this link May 17, 2014 at 5:53
  • @ThomasGross: What do you mean by "unit-based"?
    – TKR
    May 17, 2014 at 16:09
  • @TKR Morphological theories are - among other criteria - classified according to whether they are process-based, or unit-based. The classical distinction would be I(tem &)A(rrangement)-model vs. IP(process)-model. Unit-based morphologies must see non-lexical material, i.e. affixes of all sorts, as distinct units, while process-based theories don't acknowledge - at least - inflectional affixes as units. Process-based theories are thought to deal better with non-concatenative phenomena, such as the topic here, reduplication, etc. I believe there is a way to do all that with a unit-based system. May 17, 2014 at 18:17

bakkala (to buckle) bukla (buckle) a three consonontal root (b,k,l)

tilifu:n (telephone) talfana (to telephone) four consonantal root (t,l,f,n)

tilifiziu:n (television) talfaza (to televise)

four consonantal root (tlfz) zayt (oil) zayyata (to oil)

four consononta root z,y,y,t i wish this answers your question


Thank you very much guys! Are there other languages that apply the same rule as this written Arabic? Like whenever the root is triliteral the middle root consonant is doubling? I also want to research other words in wrtten arabic, if this "doubling rule" are applied in other words, but i don't know anything about arabic...

  • 1
    Could you please post additional questions as a new, separate question? Also, if you are satisfied with one of the answers, please log in to your original account and accept one of them by click the green checkmark next to the answer. :)
    – robert
    May 18, 2014 at 15:24

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.