Amharic Consonantal Roots

A minimal inventory of Amharic verbal roots used by @cite{faust-2026}'s re-analysis of @cite{broselow-1984}'s claim that Amharic admits OCP-violating biradical roots like √TT/√QQ.

@cite{faust-2026} argues that the seemingly-biconsonantal verbs (paradigm (5b), e.g. [wäddäd-ä] liked) are in fact triradical √wdd, satisfying the template by spreading; while the [t]-intruding paradigm (5c)/(12)/(13) is triradical with a [j] in the final position whose palatality merges with the preceding consonant — under this analysis Amharic has no OCP-violating roots after all.

def Fragments.Amharic.fdj : Core.Morphology.Root String

√fdj — base of [fädʤ-ä] scorch PFV.3MSG, [fädʤ-o] gerund. @cite{broselow-1984} analyzes this as a biradical √fd with /t/ as a default consonant inserted to satisfy the template; @cite{faust-2026} reanalyzes it as triradical with the third radical /j/, which palatalizes the preceding [d] to [dʒ] in the verbal paradigm and fails to surface as a separate segment. In nominal forms (gerund, INF), the feminine /t/ intrudes — not as a default consonant but as the n[+gen] exponent (@cite{faust-2026} (7)–(8), (11)–(12)).

Equations

• Fragments.Amharic.fdj = { segments := ["f", "d", "j"] }

def Fragments.Amharic.hid : Core.Morphology.Root String

√hid — base of [hed-ä] go PFV.3MSG, INF [mäh(i)d]. A "hollow" root in the standard analysis: the medial radical /i/ is non-consonantal and merges with the vocalization (@cite{faust-2026} (12e), (13c)).

Equations

• Fragments.Amharic.hid = { segments := ["h", "i", "d"] }

def Fragments.Amharic.sma : Core.Morphology.Root String

√sma — base of [sämm-a] hear PFV.3MSG, INF [mäsmat] (@cite{faust-2026} (12c), (13a)). The non-consonantal final radical /a/ merges with the vocalization.

Equations

• Fragments.Amharic.sma = { segments := ["s", "m", "a"] }

def Fragments.Amharic.sam : Core.Morphology.Root String

√sam — base of [sam-ä] kiss PFV.3MSG, INF [mäsam] (@cite{faust-2026} (12d), (13b)). The non-consonantal medial radical /a/ merges with the vocalization.

Equations

• Fragments.Amharic.sam = { segments := ["s", "a", "m"] }

def Fragments.Amharic.sbr : Core.Morphology.Root String

√sbr — base of [säbbär-ä] break PFV.3MSG, INF [mäsbär] (@cite{faust-2026} (5a), (12a)). A canonical type-A triradical with three distinct surface consonants.

Equations

• Fragments.Amharic.sbr = { segments := ["s", "b", "r"] }

def Fragments.Amharic.wd : Core.Morphology.Root String

√wd — base of [wäddäd-ä] liked PFV.3MSG (@cite{faust-2026} (5b), page 432). Both @cite{broselow-1984} and @cite{faust-2026} agree this is a biradical root. The two analysts diverge on √fdj (Broselow: biradical √fd; Faust: triradical √fdj) but agree on √wd. Crucially for @cite{faust-2026}: √wd does not violate the OCP at the root level, since /w/ ≠ /d/ — even though it surfaces with adjacent identical [d][d] in [wäddäd-ä]. The surface gemination is a template-spreading effect, not a root-level identity.

Equations

• Fragments.Amharic.wd = { segments := ["w", "d"] }

theorem Fragments.Amharic.fdj_triradical : fdj.triradical = true

theorem Fragments.Amharic.sbr_triradical : sbr.triradical = true

theorem Fragments.Amharic.wd_biradical : wd.biradical = true

theorem Fragments.Amharic.wd_no_adjacent_identical : Phonology.Constraints.adjacentIdentical wd.segments = 0

@cite{faust-2026}'s key claim about √wd (page 432): even though the surface form [wäddäd-ä] has adjacent identical [d][d], the root √wd has no adjacent identical segments — so the OCP is not violated at the root level. The biradical analysis (shared with @cite{broselow-1984}) is therefore maintained.