Tarifit Consonant Inventory #
@cite{afkir-zellou-2025}
Consonantal root data for Tarifit Berber (Amazigh, Afro-Asiatic; Glottocode: tari1263). Target words are CCəC triconsonantal verbs in the simple imperative form, from the production experiment in @cite{afkir-zellou-2025}.
Each word is specified by its three consonant natural classes on the Parker 8-level sonority scale (@cite{parker-2002}), enabling MaxEnt constraint evaluation based on cluster sonority profiles.
Prosodic template #
Tarifit has non-concatenative morphology with consonantal roots and prosodic templates. The CCəC template places a "prosodic template" schwa between C2 and C3. The C1–C2 cluster may receive an intrusive schwa (articulatory in origin) or the word may surface vowelless (no schwa at all).
Spirantization #
Singleton stops /b, d, t/ surface as fricatives [β, ð, θ] in all non-geminate environments (@cite{afkir-zellou-2025}). Natural class assignments below reflect surface realization.
Word list sourcing #
Words are from the paper's 38-item CCəC target word list. Glosses marked with (p. 20) are verified from the paper's main text; others are from the paper's word list / online appendix.
A CCəC target word from the Tarifit production experiment. The template schwa sits between C2 and C3.
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
Equations
- One or more equations did not get rendered due to their size.
- Fragments.Tarifit.Inventory.instBEqTriconWord.beq x✝¹ x✝ = false
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
C1–C2 sonority profile: rising (C1 less sonorous than C2).
Equations
- w.isRising = decide (w.c1.parkerSonority < w.c2.parkerSonority)
Instances For
C1–C2 sonority profile: falling (C1 more sonorous than C2).
Equations
- w.isFalling = decide (w.c1.parkerSonority > w.c2.parkerSonority)
Instances For
C1–C2 sonority profile: plateauing (equal sonority).
Equations
- w.isPlateauing = (w.c1.parkerSonority == w.c2.parkerSonority)
Instances For
/qrəβ/ : VLS(1)–liquid(6)–VDF(4), rise = 5. Table 9 "almost exclusively" C1ǎC2 (@cite{afkir-zellou-2025}). -- UNVERIFIED: gloss inferred from Berber root Q-R-B; paper does not -- gloss this word (distinct from /qrəʕ/ 'rip!' on p. 20).
Equations
- One or more equations did not get rendered due to their size.
Instances For
/qməʕ/ 'suppress!' (p. 20): VLS(1)–nasal(5)–VDF(4), rise = 4. Table 9 "variably" C1ǎC2 (@cite{afkir-zellou-2025}).
Equations
- One or more equations did not get rendered due to their size.
Instances For
/srəm/ : VLF(3)–liquid(6)–nasal(5), rise = 3. Table 9 "almost exclusively" C1ǎC2 (@cite{afkir-zellou-2025}). -- UNVERIFIED: gloss inferred from Berber root S-R-M; paper does not -- gloss this word explicitly.
Equations
- One or more equations did not get rendered due to their size.
Instances For
/ntəf/ 'pluck!' (p. 20): nasal(5)–VLS(1)–VLF(3), fall = 4. Table 7 "often vowelless" (>30%); Table 9 "never" C1ǎC2 (@cite{afkir-zellou-2025}).
Equations
- One or more equations did not get rendered due to their size.
Instances For
/nqəβ/ 'pick!' (p. 20): nasal(5)–VLS(1)–VDF(4), fall = 4. Table 9 "variably" C1ǎC2 (@cite{afkir-zellou-2025}).
Equations
- One or more equations did not get rendered due to their size.
Instances For
/ħkəm/ 'judge!' (p. 20): VLF(3)–VLS(1)–nasal(5), fall = 2. Table 9 "never" C1ǎC2 (@cite{afkir-zellou-2025}).
Equations
- One or more equations did not get rendered due to their size.
Instances For
/sχəf/ 'pass out!' (p. 20): VLF(3)–VLF(3)–VLF(3), plateau. Table 7 "often vowelless" (13–20%); Table 9 "never" C1ǎC2 (@cite{afkir-zellou-2025}).
Equations
- One or more equations did not get rendered due to their size.
Instances For
/sfən/ 'show!' (p. 20): VLF(3)–VLF(3)–nasal(5), plateau. Table 9 "never" C1ǎC2 (@cite{afkir-zellou-2025}).
Equations
- One or more equations did not get rendered due to their size.
Instances For
All target words from the production experiment used in our model.
Equations
- One or more equations did not get rendered due to their size.