Classifiers

This file defines the word-class record for classifiers and the vocabulary of Aikhenvald's typology of noun categorization devices. Classifier is the lexical entry Fragments store in their inventories: a form, a gloss, the semantic parameters motivating its choice, whether it is the general or a mensural classifier, and its dimensionality. A language's device is described by where its morphemes occur and what they characterize — the locus and the constituent — and its kind (noun class, numeral classifier, verbal classifier, …) is the classification of that pair by Table 15.1 (Classifier.kind), never a stored label; assignment principle, surface realizations, agreement, obligatoriness, general member, semantic parameters, and the language's number marking complete the description. Fragments record these values field by field beside the inventory (Mandarin.classifierLocus, …) or beside the gender data for noun-class languages; the record assembling them is study-local (Aikhenvald2000.Device).

Main definitions

• Classifier — the lexical entry; Classifier.Encodes, Classifier.parameters.
• Classifier.Parameter, Classifier.Dimension — the semantic parameters and dimensionality.
• Classifier.Scope, Classifier.Constituent, Classifier.Kind, Classifier.kind — the nine kinds of device as the classification of a locus by the constituent it characterizes.
• Classifier.Assignment, Classifier.Realization — assignment principles and realizations.
• Classifier.Strategy — the compositional strategies a framework may attribute to classifiers.

References

• [aikhenvald-2000], §1.5, §2.3, §4.4.1, §10.5, §11.1, Tables 15.1–15.3
• [Cor91]
• [downing-1996]
• [allan-1977]
• [little-moroney-royer-2022]

Semantic parameters

inductive Classifier.Parameter : Type

The semantic parameters noun categorization devices encode, in three large classes — animacy, physical properties, and function — with kind-specific preferences. Speech register is distinguished from the referent's social status, since honorific classifiers can index the style of speech rather than the rank of the referent.

• animacy : Parameter

  Animate versus inanimate.

• humanness : Parameter

  Human versus non-human.

• sex : Parameter

  Male versus female.

• socialStatus : Parameter

  The social status or rank of a human referent.

• kinship : Parameter

  The kinship relationship of a human referent.

• register : Parameter

  The speech register (honorific, common, humiliative) rather than the referent's status.

• shape : Parameter

  Shape and dimensionality.

• direction : Parameter

  Vertical versus horizontal orientation.

• interioricity : Parameter

  Differentiation of inside from outside, as between rings and holes.

• boundedness : Parameter

  Whether an outlined entity is delimited.

• size : Parameter

  Large versus small.

• consistency : Parameter

  Plasticity under manipulation: flexible versus rigid.

• constitution : Parameter

  Physical state, such as liquid or solid.

• material : Parameter

  The material an object is made of.

• nature : Parameter

  Other inherent, time-stable nature, often realized by classifiers specific to one noun.

• function : Parameter

  How an object is used or handled.

• arrangement : Parameter

  The configuration of objects, such as a coil or a row.

• quanta : Parameter

  A quantity of objects, such as a cluster or a flock.

• colour : Parameter

  Colour: perceptually salient but never a basis for noun categorization.

@[instance_reducible]

instance Classifier.instDecidableEqParameter : DecidableEq Parameter

Equations

• Classifier.instDecidableEqParameter x✝ y✝ = if h : x✝.ctorIdx = y✝.ctorIdx then isTrue ⋯ else isFalse ⋯

def Classifier.instReprParameter.repr : Parameter → Nat → Std.Format

Equations

• One or more equations did not get rendered due to their size.
• Classifier.instReprParameter.repr Classifier.Parameter.sex prec✝ = Repr.addAppParen (Std.Format.nest (if prec✝ ≥ 1024 then 1 else 2) (Std.Format.text "Classifier.Parameter.sex")).group prec✝
• Classifier.instReprParameter.repr Classifier.Parameter.size prec✝ = Repr.addAppParen (Std.Format.nest (if prec✝ ≥ 1024 then 1 else 2) (Std.Format.text "Classifier.Parameter.size")).group prec✝

@[instance_reducible]

instance Classifier.instReprParameter : Repr Parameter

Equations

• Classifier.instReprParameter = { reprPrec := Classifier.instReprParameter.repr }

inductive Classifier.Dimension : Type

The three values of dimensionality: one-dimensional (long), two-dimensional (flat), and three-dimensional (spherical).

• oneD : Dimension
• twoD : Dimension
• threeD : Dimension

@[instance_reducible]

instance Classifier.instDecidableEqDimension : DecidableEq Dimension

Equations

• Classifier.instDecidableEqDimension x✝ y✝ = if h : x✝.ctorIdx = y✝.ctorIdx then isTrue ⋯ else isFalse ⋯

@[instance_reducible]

instance Classifier.instReprDimension : Repr Dimension

Equations

• Classifier.instReprDimension = { reprPrec := Classifier.instReprDimension.repr }

def Classifier.instReprDimension.repr : Dimension → Nat → Std.Format

Equations

• One or more equations did not get rendered due to their size.
• Classifier.instReprDimension.repr Classifier.Dimension.oneD prec✝ = Repr.addAppParen (Std.Format.nest (if prec✝ ≥ 1024 then 1 else 2) (Std.Format.text "Classifier.Dimension.oneD")).group prec✝
• Classifier.instReprDimension.repr Classifier.Dimension.twoD prec✝ = Repr.addAppParen (Std.Format.nest (if prec✝ ≥ 1024 then 1 else 2) (Std.Format.text "Classifier.Dimension.twoD")).group prec✝

structure Classifier : Type

A classifier lexical entry: its form and gloss, the semantic parameters motivating its choice, whether it is the general classifier of its system, whether it is mensural rather than sortal, and its dimensionality when shape-based.

• form : String

  Surface form.

• gloss : String

  Gloss.

• semantics : List Parameter

  The semantic parameters motivating the choice of this classifier.

• isDefault : Bool

  Whether this is the general classifier that can replace the specific ones.

• isMensural : Bool

  Whether the classifier individuates by measure rather than by inherent properties.

• dimension : Option Dimension

  Dimensionality, when the classifier is shape-based.

def instReprClassifier.repr : Classifier → Nat → Std.Format

Equations

• One or more equations did not get rendered due to their size.

@[instance_reducible]

instance instReprClassifier : Repr Classifier

Equations

• instReprClassifier = { reprPrec := instReprClassifier.repr }

def instDecidableEqClassifier.decEq (x✝ x✝¹ : Classifier) : Decidable (x✝ = x✝¹)

Equations

• One or more equations did not get rendered due to their size.

@[instance_reducible]

instance instDecidableEqClassifier : DecidableEq Classifier

Equations

• instDecidableEqClassifier = instDecidableEqClassifier.decEq

def Classifier.Encodes (c : Classifier) (p : Parameter) : Prop

c.Encodes p when the classifier c is motivated by the parameter p.

Equations

• c.Encodes p = (p ∈ c.semantics)

@[instance_reducible]

instance Classifier.instDecidableEncodes (c : Classifier) (p : Parameter) : Decidable (c.Encodes p)

Equations

• c.instDecidableEncodes p = Classifier.instDecidableEncodes._aux_1 c p

def Classifier.parameters (cls : List Classifier) : List Parameter

The distinct semantic parameters attested across an inventory.

Equations

• Classifier.parameters cls = (List.flatMap (fun (x : Classifier) => x.semantics) cls).eraseDups

Kinds of device, by locus and constituent

inductive Classifier.Scope : Type

The morphosyntactic scope a noun categorization device operates in.

• headModifierNP : Scope

  Inside a head-modifier NP: head-modifier agreement.

• predicateArgument : Scope

  Outside the NP: predicate-argument agreement.

• noun : Scope

  The noun itself.

• numeralNP : Scope

  A numeral or quantifier NP.

• possessiveNP : Scope

  A possessive NP.

• clause : Scope

  The clause.

• adpositionalNP : Scope

  An adpositional NP.

• attributiveNP : Scope

  An attributive NP with a deictic.

@[instance_reducible]

instance Classifier.instDecidableEqScope : DecidableEq Scope

Equations

• Classifier.instDecidableEqScope x✝ y✝ = if h : x✝.ctorIdx = y✝.ctorIdx then isTrue ⋯ else isFalse ⋯

@[instance_reducible]

instance Classifier.instReprScope : Repr Scope

Equations

• Classifier.instReprScope = { reprPrec := Classifier.instReprScope.repr }

def Classifier.instReprScope.repr : Scope → Nat → Std.Format

Equations

• One or more equations did not get rendered due to their size.
• Classifier.instReprScope.repr Classifier.Scope.noun prec✝ = Repr.addAppParen (Std.Format.nest (if prec✝ ≥ 1024 then 1 else 2) (Std.Format.text "Classifier.Scope.noun")).group prec✝
• Classifier.instReprScope.repr Classifier.Scope.numeralNP prec✝ = Repr.addAppParen (Std.Format.nest (if prec✝ ≥ 1024 then 1 else 2) (Std.Format.text "Classifier.Scope.numeralNP")).group prec✝
• Classifier.instReprScope.repr Classifier.Scope.clause prec✝ = Repr.addAppParen (Std.Format.nest (if prec✝ ≥ 1024 then 1 else 2) (Std.Format.text "Classifier.Scope.clause")).group prec✝

inductive Classifier.Constituent : Type

The constituent a noun categorization device characterizes.

• headNoun : Constituent

  The head noun.

• argument : Constituent

  An A/S, S/O, or oblique argument.

• possessiveRelation : Constituent

  The possessive relation.

• possessedNoun : Constituent

  The possessed noun.

• possessor : Constituent

  The possessor.

• adpositionArgument : Constituent

  The argument of an adposition.

@[instance_reducible]

instance Classifier.instDecidableEqConstituent : DecidableEq Constituent

Equations

• Classifier.instDecidableEqConstituent x✝ y✝ = if h : x✝.ctorIdx = y✝.ctorIdx then isTrue ⋯ else isFalse ⋯

@[instance_reducible]

instance Classifier.instReprConstituent : Repr Constituent

Equations

• Classifier.instReprConstituent = { reprPrec := Classifier.instReprConstituent.repr }

def Classifier.instReprConstituent.repr : Constituent → Nat → Std.Format

Equations

• One or more equations did not get rendered due to their size.

inductive Classifier.Kind : Type

The nine kinds of noun categorization device, focal points on a continuum individuated by locus and the constituent characterized.

• nounClass : Kind

  Noun class or gender: a closed obligatory system realized by agreement inside and sometimes outside the noun phrase, with an inventory often of two to ten.

• nounClassifier : Kind

  Noun classifier: characterizes the head noun itself, independently of other NP elements, as a free form or an affix on the noun.

• numeralClassifier : Kind

  Numeral classifier: characterizes nouns in numeral and quantifier phrases, as a free form or an affix on the numeral, with an often large inventory.

• relationalClassifier : Kind

  Relational classifier: characterizes the possessive relation in a possessive NP.

• possessedClassifier : Kind

  Possessed classifier: characterizes the possessed noun in a possessive NP.

• possessorClassifier : Kind

  Possessor classifier: characterizes the possessor; very rare.

• verbalClassifier : Kind

  Verbal classifier: marks agreement on the verb with an S or O argument, as an incorporated classifier, an affix, or a suppletive classificatory stem.

• locativeClassifier : Kind

  Locative classifier: marks agreement with the head noun in an adpositional NP.

• deicticClassifier : Kind

  Deictic classifier: occurs with articles and demonstratives, marking spatial location or determination.

@[instance_reducible]

instance Classifier.instDecidableEqKind : DecidableEq Kind

Equations

• Classifier.instDecidableEqKind x✝ y✝ = if h : x✝.ctorIdx = y✝.ctorIdx then isTrue ⋯ else isFalse ⋯

def Classifier.instReprKind.repr : Kind → Nat → Std.Format

Equations

• One or more equations did not get rendered due to their size.
• Classifier.instReprKind.repr Classifier.Kind.nounClass prec✝ = Repr.addAppParen (Std.Format.nest (if prec✝ ≥ 1024 then 1 else 2) (Std.Format.text "Classifier.Kind.nounClass")).group prec✝

@[instance_reducible]

instance Classifier.instReprKind : Repr Kind

Equations

• Classifier.instReprKind = { reprPrec := Classifier.instReprKind.repr }

def Classifier.kind : Scope → Constituent → Option Kind

The kind of device determined by a locus and the constituent it characterizes: noun classes inside a head-modifier NP or agreeing with an argument outside it, noun classifiers on the noun, numeral classifiers in the numeral NP, the three possessive kinds by which part of the possessive NP they characterize, verbal classifiers agreeing with an argument in the clause, locative and deictic classifiers in adpositional and attributive NPs.

Equations

• Classifier.kind Classifier.Scope.headModifierNP Classifier.Constituent.headNoun = some Classifier.Kind.nounClass
• Classifier.kind Classifier.Scope.predicateArgument Classifier.Constituent.argument = some Classifier.Kind.nounClass
• Classifier.kind Classifier.Scope.noun Classifier.Constituent.headNoun = some Classifier.Kind.nounClassifier
• Classifier.kind Classifier.Scope.numeralNP Classifier.Constituent.headNoun = some Classifier.Kind.numeralClassifier
• Classifier.kind Classifier.Scope.possessiveNP Classifier.Constituent.possessiveRelation = some Classifier.Kind.relationalClassifier
• Classifier.kind Classifier.Scope.possessiveNP Classifier.Constituent.possessedNoun = some Classifier.Kind.possessedClassifier
• Classifier.kind Classifier.Scope.possessiveNP Classifier.Constituent.possessor = some Classifier.Kind.possessorClassifier
• Classifier.kind Classifier.Scope.clause Classifier.Constituent.argument = some Classifier.Kind.verbalClassifier
• Classifier.kind Classifier.Scope.adpositionalNP Classifier.Constituent.adpositionArgument = some Classifier.Kind.locativeClassifier
• Classifier.kind Classifier.Scope.attributiveNP Classifier.Constituent.headNoun = some Classifier.Kind.deicticClassifier
• Classifier.kind x✝¹ x✝ = none

Assignment and realization

inductive Classifier.Assignment : Type

The principle by which nouns are assigned to classes or classifiers.

• semantic : Assignment

  By the meaning of the referent.

• morphological : Assignment

  By morphological properties of the noun such as declension or derivational affix.

• phonological : Assignment

  By phonological properties of the noun such as its initial segment or final vowel.

• mixed : Assignment

  A semantic core with a morphological or phonological overlay.

@[instance_reducible]

instance Classifier.instDecidableEqAssignment : DecidableEq Assignment

Equations

• Classifier.instDecidableEqAssignment x✝ y✝ = if h : x✝.ctorIdx = y✝.ctorIdx then isTrue ⋯ else isFalse ⋯

@[instance_reducible]

instance Classifier.instReprAssignment : Repr Assignment

Equations

• Classifier.instReprAssignment = { reprPrec := Classifier.instReprAssignment.repr }

def Classifier.instReprAssignment.repr : Assignment → Nat → Std.Format

Equations

• One or more equations did not get rendered due to their size.

inductive Classifier.Realization : Type

The surface realization of a classifier morpheme.

• prefix : Realization

  Prefix or proclitic.

• suffix : Realization

  Suffix or enclitic.

• clitic : Realization

  Clitic.

• freeForm : Realization

  An independent lexeme.

• apophony : Realization

  Stem-internal vowel change.

• suppletion : Realization

  A suppletive stem.

• stress : Realization

  Stress.

• reduplication : Realization

  Reduplication.

• nounIncorporation : Realization

  Noun incorporation.

• repeater : Realization

  A repeater: the noun itself, or part of it, serving as its classifier.

@[instance_reducible]

instance Classifier.instDecidableEqRealization : DecidableEq Realization

Equations

• Classifier.instDecidableEqRealization x✝ y✝ = if h : x✝.ctorIdx = y✝.ctorIdx then isTrue ⋯ else isFalse ⋯

@[instance_reducible]

instance Classifier.instReprRealization : Repr Realization

Equations

• Classifier.instReprRealization = { reprPrec := Classifier.instReprRealization.repr }

def Classifier.instReprRealization.repr : Realization → Nat → Std.Format

Equations

• One or more equations did not get rendered due to their size.

Compositional strategies

inductive Classifier.Strategy : Type

The compositional strategy a framework attributes to classifier constructions. Strategy assignments to particular languages are made in the study files of the papers that make them (Chierchia1998, LittleMoroneyRoyer2022, Sudo2016), never in the Fragments.

• forNumeral : Strategy

  The classifier is a measure function required by the numeral, which takes it as its first argument ([krifka-1995b], [bale-coon-2014]); predicts classifier–plural co-occurrence and classifiers in counting contexts without a noun.

• forNoun : Strategy

  The classifier atomizes the noun denotation so that the numeral can count ([chierchia-1998]); predicts classifiers beyond numerals and complementary distribution with plural marking.

• sudoBlocking : Strategy

  Numerals are type-n singular terms shifted to predicates by a silent ∪-operator that overt classifiers in the lexicon block ([sudo-2016]); predicts classifiers with numerals only and no numeral or noun idiosyncrasies.

@[instance_reducible]

instance Classifier.instDecidableEqStrategy : DecidableEq Strategy

Equations

• Classifier.instDecidableEqStrategy x✝ y✝ = if h : x✝.ctorIdx = y✝.ctorIdx then isTrue ⋯ else isFalse ⋯

def Classifier.instReprStrategy.repr : Strategy → Nat → Std.Format

Equations

• One or more equations did not get rendered due to their size.

@[instance_reducible]

instance Classifier.instReprStrategy : Repr Strategy

Equations

• Classifier.instReprStrategy = { reprPrec := Classifier.instReprStrategy.repr }