The typological diversity of morphomes #
Two paradigm fragments from [Her23] (The Typological Diversity of
Morphomes, OUP) instantiate Morphology.IsMorphome: the Spanish
L-morphome and a Darma agreement syncretism. Each is a systematic
syncretism whose cell set is captured by no feature-value conjunction — the
book's operationalization of "unnatural" (a natural class is one
"coextensive with a value or conjunction of values", §1.4). Naturalness is
supplied here as the value-or-conjunction predicate, the canonical
instantiation of Morphology's Natural parameter.
Systematicity — recurrence under more than one exponent — is what distinguishes a morphome from an accidental homophony: the Spanish set recurs across three lexemes with three different stem alternations, the Darma set across two voice allomorphs.
Source caveat #
The book's paradigm tables are images, so only cell sets and exponent labels recoverable from the surrounding prose are encoded (verbatim quotes cited below); full form-per-cell grids are not reproduced. Location cites are source-verified: Spanish §1.3 / Table 1.2, pp. 7-8; the definition §1.4, pp. 10-11; Darma §4.2.2.4, p. 149 (data from Willis 2007).
Main declarations #
Herce2023.spanishL_morphome_venir/_nacer/_caber— the L-morphome is a morphome under three distinct stem alternationsHerce2023.darma_morphome_intr/_tr— the Darma 1PL/2 syncretism is a morphome under both voice allomorphs
Morphome from a kernel-class Finset #
A reusable bridge: a nontrivial ker-class Finset that coincides with no
Natural class is a Morphology.IsMorphome.
The kernel class of a under f, as a Finset: the cells f sends to
f a. Its coercion is Morphology.syncretismClass.
Equations
- Herce2023.kerClass f a = {c : Cell | f c = f a}
Instances For
A nontrivial ker-class Finset that equals no natural class is a
morphome.
Spanish L-morphome (§1.3, Table 1.2, pp. 7-8) #
Verbatim (p. 7): the pattern "encompasses the 1SG present indicative and all the present subjunctive cells". The present-tense paradigm coordinates below are local slices (indicative/subjunctive × three persons × two numbers).
Equations
- Herce2023.instDecidableEqMood x✝ y✝ = if h : x✝.ctorIdx = y✝.ctorIdx then isTrue ⋯ else isFalse ⋯
Equations
- Herce2023.instFintypeMood = { elems := { val := ↑Herce2023.Mood.enumList, nodup := Herce2023.Mood.enumList_nodup }, complete := Herce2023.instFintypeMood._proof_1 }
Equations
- Herce2023.instReprMood.repr Herce2023.Mood.ind prec✝ = Repr.addAppParen (Std.Format.nest (if prec✝ ≥ 1024 then 1 else 2) (Std.Format.text "Herce2023.Mood.ind")).group prec✝
- Herce2023.instReprMood.repr Herce2023.Mood.sbjv prec✝ = Repr.addAppParen (Std.Format.nest (if prec✝ ≥ 1024 then 1 else 2) (Std.Format.text "Herce2023.Mood.sbjv")).group prec✝
Instances For
Equations
- Herce2023.instReprMood = { reprPrec := Herce2023.instReprMood.repr }
Equations
- Herce2023.instDecidableEqPer x✝ y✝ = if h : x✝.ctorIdx = y✝.ctorIdx then isTrue ⋯ else isFalse ⋯
Equations
- Herce2023.instFintypePer = { elems := { val := ↑Herce2023.Per.enumList, nodup := Herce2023.Per.enumList_nodup }, complete := Herce2023.instFintypePer._proof_1 }
Equations
- Herce2023.instReprPer.repr Herce2023.Per.first prec✝ = Repr.addAppParen (Std.Format.nest (if prec✝ ≥ 1024 then 1 else 2) (Std.Format.text "Herce2023.Per.first")).group prec✝
- Herce2023.instReprPer.repr Herce2023.Per.second prec✝ = Repr.addAppParen (Std.Format.nest (if prec✝ ≥ 1024 then 1 else 2) (Std.Format.text "Herce2023.Per.second")).group prec✝
- Herce2023.instReprPer.repr Herce2023.Per.third prec✝ = Repr.addAppParen (Std.Format.nest (if prec✝ ≥ 1024 then 1 else 2) (Std.Format.text "Herce2023.Per.third")).group prec✝
Instances For
Equations
- Herce2023.instReprPer = { reprPrec := Herce2023.instReprPer.repr }
Equations
- Herce2023.instDecidableEqNum x✝ y✝ = if h : x✝.ctorIdx = y✝.ctorIdx then isTrue ⋯ else isFalse ⋯
Equations
- Herce2023.instFintypeNum = { elems := { val := ↑Herce2023.Num.enumList, nodup := Herce2023.Num.enumList_nodup }, complete := Herce2023.instFintypeNum._proof_1 }
Equations
- Herce2023.instReprNum = { reprPrec := Herce2023.instReprNum.repr }
Equations
- Herce2023.instReprNum.repr Herce2023.Num.sg prec✝ = Repr.addAppParen (Std.Format.nest (if prec✝ ≥ 1024 then 1 else 2) (Std.Format.text "Herce2023.Num.sg")).group prec✝
- Herce2023.instReprNum.repr Herce2023.Num.pl prec✝ = Repr.addAppParen (Std.Format.nest (if prec✝ ≥ 1024 then 1 else 2) (Std.Format.text "Herce2023.Num.pl")).group prec✝
Instances For
A present-tense cell: mood, person, number.
Equations
Instances For
The L-morphome cell set: 1SG present indicative together with all six present subjunctive cells (7 cells).
Equations
- Herce2023.Lset = insert (Herce2023.Mood.ind, Herce2023.Per.first, Herce2023.Num.sg) {c : Herce2023.SpCell | c.1 = Herce2023.Mood.sbjv}
Instances For
venir 'come': velar /g/ stem in the L-cells (vengo, venga…), base
elsewhere. Defined by position, not by Lset — the syncretism is derived.
Equations
- Herce2023.venirStem (Herce2023.Mood.ind, Herce2023.Per.first, Herce2023.Num.sg) = "veng"
- Herce2023.venirStem (Herce2023.Mood.sbjv, fst, snd) = "veng"
- Herce2023.venirStem x✝ = "ven"
Instances For
nacer 'be born': velar /k/ stem in the L-cells (nazco, nazca…).
Equations
- Herce2023.nacerStem (Herce2023.Mood.ind, Herce2023.Per.first, Herce2023.Num.sg) = "naθk"
- Herce2023.nacerStem (Herce2023.Mood.sbjv, fst, snd) = "naθk"
- Herce2023.nacerStem x✝ = "naθ"
Instances For
caber 'fit': suppletive stem in the L-cells (quepo, quepa…).
Equations
- Herce2023.caberStem (Herce2023.Mood.ind, Herce2023.Per.first, Herce2023.Num.sg) = "kep"
- Herce2023.caberStem (Herce2023.Mood.sbjv, fst, snd) = "kep"
- Herce2023.caberStem x✝ = "kab"
Instances For
A feature specification: each feature is fixed to a value or left open. A natural class is the set of cells matching some spec (coextensive with a value or conjunction of values, [Her23] §1.4).
Equations
- Herce2023.SpSpec = (Option Herce2023.Mood × Option Herce2023.Per × Option Herce2023.Num)
Instances For
Does cell c match spec s (open features vacuously match)?
Equations
- One or more equations did not get rendered due to their size.
Instances For
The cells matching spec s.
Equations
- Herce2023.spNaturalClass s = {c : Herce2023.SpCell | Herce2023.spMatches s c = true}
Instances For
Naturalness for the Spanish space: coextensive with some feature-value conjunction.
Equations
- Herce2023.SpNatural X = ∃ (s : Herce2023.SpSpec), ↑(Herce2023.spNaturalClass s) = X
Instances For
No feature-value conjunction is coextensive with the L-set: it crosses the indicative/subjunctive divide, so it is unnatural.
The L-morphome under venir's velar /g/ alternation.
The same cell set under nacer's velar /k/ alternation — recurrence across a distinct exponent is what makes it systematic.
The same cell set under caber's stem suppletion.
Darma 1PL/2 syncretism (§4.2.2.4, p. 149; Willis 2007) #
Verbatim: "verbal agreement is characterized by a syncretism of 1PL and 2 … The formal affinity … is, therefore, morphomic." The shared n-based suffix is -he in intransitive verbs, -de in transitive verbs. Only the attested three-cell set and the two allomorph labels are encoded; the full grid is an image (see the source caveat).
The two voice allomorphs of the n-suffix (-he intr, -de tr).
Instances For
Equations
- Herce2023.instDecidableEqVoice x✝ y✝ = if h : x✝.ctorIdx = y✝.ctorIdx then isTrue ⋯ else isFalse ⋯
Equations
- Herce2023.instFintypeVoice = { elems := { val := ↑Herce2023.Voice.enumList, nodup := Herce2023.Voice.enumList_nodup }, complete := Herce2023.instFintypeVoice._proof_1 }
Equations
- Herce2023.instReprVoice.repr Herce2023.Voice.intr prec✝ = Repr.addAppParen (Std.Format.nest (if prec✝ ≥ 1024 then 1 else 2) (Std.Format.text "Herce2023.Voice.intr")).group prec✝
- Herce2023.instReprVoice.repr Herce2023.Voice.tr prec✝ = Repr.addAppParen (Std.Format.nest (if prec✝ ≥ 1024 then 1 else 2) (Std.Format.text "Herce2023.Voice.tr")).group prec✝
Instances For
Equations
- Herce2023.instReprVoice = { reprPrec := Herce2023.instReprVoice.repr }
A Darma agreement cell: person, number.
Equations
Instances For
The syncretic set: 1PL, 2SG, 2PL (three cells).
Equations
Instances For
The n-based agreement suffix, keyed by voice: the syncretic cells bear
it (value = the voice allomorph), other cells are left uncharacterized
(none), since the full grid is not recoverable. Defined by position.
Equations
- Herce2023.darmaAgr v (Herce2023.Per.first, Herce2023.Num.pl) = some v
- Herce2023.darmaAgr v (Herce2023.Per.second, snd) = some v
- Herce2023.darmaAgr v x✝ = none
Instances For
A Darma feature spec (person × number).
Equations
- Herce2023.DSpec = (Option Herce2023.Per × Option Herce2023.Num)
Instances For
Equations
- Herce2023.dMatches s c = ((match s.1 with | none => true | some p => decide (p = c.1)) && match s.2 with | none => true | some n => decide (n = c.2))
Instances For
Equations
- Herce2023.dNaturalClass s = {c : Herce2023.DCell | Herce2023.dMatches s c = true}
Instances For
Equations
- Herce2023.DNatural X = ∃ (s : Herce2023.DSpec), ↑(Herce2023.dNaturalClass s) = X
Instances For
{1PL, 2SG, 2PL} crosses persons 1 and 2 and fixes no number, so no feature-value conjunction is coextensive with it.
The 1PL/2 syncretism is a morphome under the intransitive allomorph -he.
…and under the transitive allomorph -de: recurrence across the two allomorphs establishes systematicity.