This post is written with the goal of finding patterns that are predicted by formal analyses, but possibly do not exist. If they really do not exist, one might ask why. The issue is important to me in light of an upcoming invitation to speak about “prosodic templates” in May 2025.
Consider the three following denominal verbs from Modern Hebrew (stress is final unless marked by an accute accent):
(1) |
a. |
[tilfen] |
‘he phoned’ |
< [télefon] |
|
b. |
[fikes] |
‘he focused’ |
< [fókus] |
|
c. |
[mikses] |
‘he mixed (music)’ |
< [miks] |
One “traditional” analysis of such facts uses a template CiCeC, specifying consonantal positions and a melody. This template is imposed on the base. This analysis is problematic for (1a,c), because the verbs have four consonants, not three. The analyst has to use the metaphor of two consonants being “squeezed into” one slot (in the case of 1c, because the original cluster is preserved).
Bat-El (1994) proposed another analysis, whereby the template is only its vowels <i,e>. These vowels replace the first and last vowel of the base through a process called “melodic overwriting’. This analysis also requires “plug-ins”. First, for [télefon], it amounts to the wrong output *[tilefen]. Bat-El assumes a process of “stray erasure” of the non-overwritten vowel, possibly motivated by word-size constraints. Second, for [fikses], the analysis requires reduplication to be triggered. Following McCarthy & Prince (1996), Bat-El assumes a constraint “Final C”, requiring stems to end in a consonant. However, many stems do not end in a consonant in Modern Hebrew (speakers have no problem with verbs like [kine] ‘he envied’).
If Bat-El’s analysis is on the right track, one might imagine a language that is like Modern Hebrew except for the size constraint. In such a language, one expects to find a morphological process that replaces the first and last vowel, but does not touch the middle vowel: [télefon]+/u,e/ => [tulefen]. This would be a sort of melodically overwriting circumfix. I am unaware of such a morphological process. Is anybody aware of one?
Setting aside the circumfix quality, one might expect a melody to be introduced from one edge of the base. For instance [télefon]+/u,e/ => [télufen]. Again, I do not know of such a case; though Tamashek Berber (Heath 2005) plurals come close, but do not leave the residue vowels untouched.
One might ask the same thing about the syllabic structure. Suppose we have a template CVCC (the vowel would be unspecified). We might expect a process to be expressed only by changing the prosody of the base. For instance (2). Yawelmani Yokuts (Archangeli 1983) has something like this, but templates are always imposed by outside suffixes. As in the case of “melodic overwriting”, a template can also be imagined to be imposed on one side of the base, as opposed to on the whole base. For instance, a process that imposes CVCC on the right extremity would look like (3).
(2) |
|
singular |
plural |
|
|
(3) |
|
singular |
plural |
|
a. |
[safir] |
[safr] |
|
|
|
a. |
[baronik] |
[baronk] |
|
b. |
[sfur] |
[sufr] |
|
|
|
b. |
[farka] |
[farakk] |
|
c. |
[sfri] |
[sifr] |
|
|
|
c. |
[normal] |
[noraml] |
|
d. |
[sefr] |
[sefr] |
|
|
|
d. |
[onest] |
[onest] |
For (3), the broken plurals of Modern standard Arabic as analyzed in McCarthy & Prince (1990) are a possible candidate: a prosody CvCvv is imposed on the left edge, and the prosody of the residue is retained. /maktab/ + /CaCaa, i/ => [makaatib] ‘desk’ and /sultˁaan/ => + /CaCaa, i/ => [salaatˁiin]. But in this case, the melody is also changed in both parts of the output. Moreover, there is a templatic limit on the residue, in that it cannot exceed one syllable.
Noam Faust, noam.faust@univ-paris8.fr
References
Archangeli, Diana. 1983. “The Root CV-Template as a Property of the Affix: Evidence from Yawelmani.” Natural Language and Linguistic Theory 1: 347-384.
Bat-El, Outi. 1994. “Stem Modification and Cluster Transfer in Modern Hebrew.” Natural Language & Linguistic Theory 12, 4: 571–96.
Heath, Jeffrey. 2005. A grammar of Tamashek (Tuareg of Mali). Mouton de Gruyter (Mouton Grammar Series).
McCarthy, John J. and Alan Prince. 1990. “Foot and word in Prosodic Morphology: The Arabic broken plural”. Natural Language & Linguistic Theory 8 (2): 209-283
McCarthy, John J. and Alan Prince. 1996. Prosodic Morphology 1986. Ms, retrieed online 26/01/2025.
OpenEdition suggests that you cite this post as follows:
noamfaust (February 26, 2025). Is partial templaticity an attested process? Diversity Linguistics Comment. Retrieved March 16, 2025 from https://doi.org/10.58079/13dtt
“Setting aside the circumfix quality, one might expect a melody to be introduced from one edge of the base. For instance [télefon]+/u,e/ => [télufen]. Again, I do not know of such a case; though Tamashek Berber (Heath 2005) plurals come close, but do not leave the residue vowels untouched.”
How about Hausa -a-e plurals? Examples from Newman’s grammar:
gishirī > gishā̀rē
guzumā > guzā̀mē
‘askā > ‘asā̀kē
But I don’t think Hausa applies this plural type to longer stems.
Compare also plurals in Northern Guera languages, e.g. Mogum:
mírtele > mírtal
colkom > colkam
dúŋgùm > duŋgam
(For Northern Guera plurals, see my article Pattern borrowing and hybridization in Mubi (East Chadic): The importance of congruence, section 3.3, and associated references.)