Cross-Linguistic Typology of Negation (WALS Chapters 112--115, 143--144) #
@cite{dryer-haspelmath-2013} @cite{haspelmath-2013} @cite{miestamo-2005} @cite{miestamo-2013} @cite{dryer-2013-wals}
Cross-linguistic data on clausal negation from WALS chapters 112--115, 143, and 144.
Ch 112: Negative Morphemes #
How standard (clausal) negation is expressed. Six categories based on morpheme type: negative affix, negative particle, negative auxiliary verb, negative word of unclear status, variation between word and affix, and double (bipartite) negation requiring two co-occurring markers.
Ch 113: Symmetric and Asymmetric Standard Negation #
Whether negation changes clause structure beyond adding a negative marker. Symmetric negation adds only the negator; asymmetric negation introduces further structural changes (e.g., changes in finiteness, verb paradigm, or tense-aspect marking). Three types: Sym only, Asy only, or both.
Ch 114: Subtypes of Asymmetric Standard Negation #
For languages with asymmetric negation, what structural domain is affected: finiteness (A/Fin), reality status (A/NonReal), or other grammatical categories (A/Cat). Languages may combine subtypes.
Ch 115: Negative Indefinite Pronouns and Predicate Negation #
How negative indefinites ('nobody', 'nothing') interact with clausal negation. Whether they co-occur with predicate negation (negative concord, the dominant pattern worldwide) or preclude it.
Ch 143: Order of Negative Morpheme and Verb #
Seven sub-features (143A--143G) covering the position of the negative morpheme relative to the verb. 143A gives the overall classification (NegV, VNeg, [Neg-V], [V-Neg], double/triple negation, etc.). 143B--143D detail obligatory/optional double and triple negation patterns. 143E--143G decompose into preverbal morphemes, postverbal morphemes, and minor morphological means (negative tone, infix, stem change). All seven features cover the same 1325-language sample except 143B (119), 143C (81), 143D (6).
Ch 144: Position of Negative Morphemes by Word Order Type #
Twenty-five sub-features (144A--144Y) cross-tabulating negation position with basic word order type. 144A gives the overall position of the negative word relative to S, O, and V (1190 languages). 144B gives clause-edge and verb-adjacency position (609 languages). 144C covers languages where negation changes word order (28 languages). 144D--144K break down SVO languages by negation position. 144L--144S break down SOV languages. 144T--144X break down verb-initial languages. 144Y covers object-initial languages (16 languages).
WALS Ch 112: How standard (clausal) negation is expressed.
Six categories based on the morphological status of the negative marker: (1) affix on the verb, (2) free particle, (3) auxiliary verb inflecting for verbal categories, (4) negative word whose status is unclear, (5) variation between word and affix constructions in the same language, (6) bipartite ("double") negation requiring two co-occurring markers.
- affix : NegMorphemeType
Negative affix on the verb (e.g., Kolyma Yukaghir
el-jaqa-te-je'NEG-achieve-FUT-1SG'). - particle : NegMorphemeType
Negative particle: free word, no verbal inflection (e.g., English
not, Musgupay). - auxVerb : NegMorphemeType
Negative auxiliary verb: inflects for person, number, or TAM like verbs in the language (e.g., Finnish
e-n'NEG-1SG'). - wordUnclear : NegMorphemeType
Negative word whose status as verb or particle is unclear, typically in isolating languages with little verbal morphology (e.g., Maori
kaahore). - variation : NegMorphemeType
Language uses both a negative word and a negative affix in different constructions (e.g., Rama: preverbal particle in one construction, verbal suffix in another).
- doubleNeg : NegMorphemeType
Bipartite negation: two co-occurring negative morphemes, one preceding and one following the verb (e.g., French
ne...pas, Izito-ome-du'NEG-do-NEG').
Instances For
Equations
- Phenomena.Negation.Typology.instBEqNegMorphemeType.beq x✝ y✝ = (x✝.ctorIdx == y✝.ctorIdx)
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
WALS Ch 113: Whether negation changes clause structure.
Symmetric negation: the negative differs from the affirmative only by adding the negative marker(s) -- no structural changes to verb form, paradigm, or clause type.
Asymmetric negation: the negative construction differs structurally from the affirmative in additional ways (changed finiteness, different verb paradigm, different TAM marking, etc.).
- symmetric : NegSymmetry
Symmetric only (Type Sym): negation never changes clause structure. (e.g., German
ich singe/ich singe nicht). - asymmetric : NegSymmetry
Asymmetric only (Type Asy): negation always introduces structural differences (e.g., Finnish: negative verb + connegative).
- both : NegSymmetry
Both symmetric and asymmetric (Type SymAsy): some constructions are symmetric, others asymmetric (e.g., Lezgian: present symmetric, past imperfective asymmetric).
Instances For
Equations
- Phenomena.Negation.Typology.instBEqNegSymmetry.beq x✝ y✝ = (x✝.ctorIdx == y✝.ctorIdx)
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
WALS Ch 114: Which grammatical domain is affected by asymmetric negation.
Four primary subtypes (@cite{miestamo-2005} Table 2, p. 60):
- A/Fin: negation changes finiteness (adds negative verb, lexical verb becomes nonfinite / subordinate)
- A/NonReal: negation introduces irrealis/nonrealized marking
- A/Emph: negative contains marking that expresses emphasis in non-negatives (rare; 4 languages in RS)
- A/Cat: negation changes marking of TAM, person, number, etc.
Note: WALS Ch 114 does not distinguish A/Emph as a separate value, collapsing it into other categories. The four-way distinction is from @cite{miestamo-2005} only.
- finiteness : AsymmetrySubtype
A/Fin: asymmetry in finiteness. Typically a negative auxiliary becomes the finite verb, and the lexical verb appears in a nonfinite form (e.g., Finnish:
e-n tule'NEG-1SG come.CONNEG'). - realityStatus : AsymmetrySubtype
A/NonReal: asymmetry in reality status. The negative is obligatorily marked with an irrealis/nonrealized category that the affirmative lacks (e.g., Imbabura Quechua: negative requires
-chuirrealis). - emphasis : AsymmetrySubtype
A/Emph: the negative contains marking that expresses emphasis in non-negative contexts. Rare (4 languages in the RS). @cite{miestamo-2005} §3.3.3, Table 2 (p. 60).
- otherCategories : AsymmetrySubtype
A/Cat: asymmetry in other grammatical categories (TAM, person-number affixes, etc.). The negative uses different category markers than the affirmative (e.g., Karok: different person-number affixes under negation).
- finAndNonReal : AsymmetrySubtype
Combined: A/Fin and A/NonReal (e.g., Copainalá Zoque, Squamish).
- finAndEmph : AsymmetrySubtype
Combined: A/Fin and A/Emph (e.g., Meithei).
- finAndCat : AsymmetrySubtype
Combined: A/Fin and A/Cat (e.g., Kolokuma Ijo).
- nonRealAndCat : AsymmetrySubtype
Combined: A/NonReal and A/Cat.
- emphAndCat : AsymmetrySubtype
Combined: A/Emph and A/Cat (e.g., Cantonese, Meithei).
- nonAssignable : AsymmetrySubtype
Non-assignable: language has only symmetric negation (Type Sym in Ch 113), so no asymmetry subtype applies.
Instances For
Equations
- Phenomena.Negation.Typology.instBEqAsymmetrySubtype.beq x✝ y✝ = (x✝.ctorIdx == y✝.ctorIdx)
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
@cite{miestamo-2005}'s two dimensions of asymmetry. WALS Ch 113 collapses these into a single symmetric/asymmetric distinction; Miestamo decomposes asymmetry into two independent dimensions.
- constructional : AsymmetryDimension
The negative clause has a different syntactic structure than the affirmative, beyond just the negation marker. E.g., Finnish neg aux restructures the clause; Japanese -nai changes verb to i-adjective.
- paradigmatic : AsymmetryDimension
The negative paradigm has fewer formal distinctions than the affirmative. E.g., Burmese -bu neutralizes TAM; Turkish aorist uses a different marker.
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
Whether the asymmetry is derived from the negation marker type or independent of it (@cite{miestamo-2005}).
- derived : AsymmetrySource
The asymmetry follows structurally from the negation marker's properties. A negative verb necessarily creates A/Fin.
- independent : AsymmetrySource
The asymmetry is not predictable from the marker type alone. E.g., TAM neutralization in Burmese is independent of circumfixing.
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
Equations
- Phenomena.Negation.Typology.instBEqAsymmetrySource.beq x✝ y✝ = (x✝.ctorIdx == y✝.ctorIdx)
Instances For
WALS Ch 115: Interaction of negative indefinites ('nobody', 'nothing') with clausal negation.
Cross-linguistically, negative concord (co-occurrence) is overwhelmingly dominant. The preclusion pattern is concentrated in Western Europe and Mesoamerica; the normative criticism of "double negation" as "illogical" is a prescriptive artifact rooted in Latin prestige (@cite{haspelmath-1997}, sec. 8.2).
- cooccur : NegIndefiniteStrategy
Negative indefinites co-occur with predicate negation (negative concord). 'Nobody NEG came' = 'Nobody came'. The dominant pattern worldwide. (e.g., Russian
nikto ne prisel'nobody NEG came'). - preclude : NegIndefiniteStrategy
Negative indefinites preclude predicate negation. The indefinite alone carries the negation. (e.g., German
Niemand kam'Nobody came', *Niemand kam nicht). - mixed : NegIndefiniteStrategy
Mixed behavior: some negative indefinites co-occur with negation, others preclude it (e.g., position-dependent as in Spanish:
Nadie vinobutNo vi nada; or different indefinite series as in Swedish). - negExistential : NegIndefiniteStrategy
Negative existential construction: a negative/negated existential verb serves as the main predicate (e.g., Nelemwa
kia agu i uya'not.exist person 3SG arrive' = 'Nobody came').
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
WALS Ch 143A: Position of the negative morpheme relative to the verb.
Covers 1325 languages. Single-negation types distinguish NegV (preverbal particle), VNeg (postverbal particle), [Neg-V] (preverbal affix), and [V-Neg] (postverbal affix). Multi-negation types cover obligatory double negation, optional double negation, and optional triple negation with obligatory or optional double negation.
- preverbalParticle : NegVerbPosition
Preverbal negative particle:
NegV. - postverbalParticle : NegVerbPosition
Postverbal negative particle:
VNeg. - preverbalAffix : NegVerbPosition
Preverbal negative affix:
[Neg-V]. - postverbalAffix : NegVerbPosition
Postverbal negative affix:
[V-Neg]. - negativeTone : NegVerbPosition
Negative tone (suprasegmental).
- mixedSingle : NegVerbPosition
Mixed: two single-negation types coexist.
- obligDoublNeg : NegVerbPosition
Obligatory double negation.
- optDoubleNeg : NegVerbPosition
Optional double negation.
- tripleNeg : NegVerbPosition
Optional triple negation (with obligatory or optional double).
- optSingleNeg : NegVerbPosition
Optional single negation.
Instances For
Equations
- Phenomena.Negation.Typology.instBEqNegVerbPosition.beq x✝ y✝ = (x✝.ctorIdx == y✝.ctorIdx)
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
WALS Ch 143E/F: Whether a language has preverbal and/or postverbal negative morphemes.
- preverbalOnly : NegMorphemePosition
Preverbal particle only.
- postverbalOnly : NegMorphemePosition
Postverbal particle only.
- preverbalAffixOnly : NegMorphemePosition
Preverbal affix only.
- postverbalAffixOnly : NegMorphemePosition
Postverbal affix only.
- both : NegMorphemePosition
Both preverbal and postverbal.
- none : NegMorphemePosition
None (language uses minor means or double negation).
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
Ch 113 and Ch 114 use the same sample.
Ch 143A, 143E, 143F, 143G all cover the same 1325-language sample.
A language's negation profile across WALS Chapters 112--115.
- language : String
Language name.
- iso : String
ISO 639-3 code.
- morphemeType : NegMorphemeType
Ch 112: How standard negation is expressed.
- symmetry : NegSymmetry
Ch 113: Symmetric, asymmetric, or both.
- asymmetrySubtype : AsymmetrySubtype
Ch 114: Asymmetry subtype (nonAssignable if symmetric only).
- negIndefinite : Option NegIndefiniteStrategy
Ch 115: Strategy for negative indefinites, if attested.
Illustrative negative marker form(s).
Is the negation marker a syntactic head (X°) rather than a phrase (XP)? Relevant for @cite{greco-2020}: only head-status markers can merge in CP to produce surprise negation.
- notes : String
Notes on the negation system.
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
English: negative particle not; WALS classifies as both symmetric and
asymmetric (do-support is an asymmetric structural change). A/Cat: the
category-level change is the introduction of auxiliary do in negation.
Negative indefinites show mixed behavior: nobody precludes
predicate negation (*Nobody didn't come), but anything requires it
(I didn't see anything).
Equations
- One or more equations did not get rendered due to their size.
Instances For
German: negative particle nicht; symmetric negation — adding nicht
causes no structural change to the clause. Negative indefinites preclude
predicate negation: Niemand kam (*Niemand kam nicht).
(But note: substandard / Bavarian German has negative concord.)
Equations
- One or more equations did not get rendered due to their size.
Instances For
French: bipartite negation ne...pas (WALS codes as particle since
colloquial French drops ne). In colloquial register, only pas is used.
WALS Ch 115 classifies as mixed: some negative indefinites co-occur with
ne (Je n'ai rien vu), while personne can appear without ne in
some registers.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Russian: negative particle ne; symmetric negation. Negative indefinites
obligatorily co-occur with predicate negation (negative concord):
Nikto ne prisel 'Nobody NEG came' = 'Nobody came'.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Finnish: negative auxiliary verb e- inflects for person-number; the lexical
verb appears as a connegative (present) or past participle (past). Always
asymmetric (A/Fin): negation changes finiteness structure.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Japanese: negative suffix -nai (affix on verb). WALS classifies as
asymmetric (A/Fin + A/Cat): the negative form involves both finiteness
changes and different category markings.
Negative indefinites co-occur with predicate negation:
dare-mo ko-nakat-ta 'who-MO come-NEG-PST' = 'Nobody came'.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Mandarin Chinese: negative particles bu (general) and mei(you) (perfective).
WALS classifies as both symmetric and asymmetric: the bu/mei distinction
introduces an asymmetry of type A/Fin (finiteness-like distinction).
Equations
- One or more equations did not get rendered due to their size.
Instances For
Turkish: negative suffix -mA- on the verb. WALS classifies as both
symmetric and asymmetric: some constructions (aorist) are symmetric while
others show category changes (A/Cat).
Negative indefinites co-occur with predicate negation:
Hic kimse gel-me-di 'at.all person come-NEG-PST' = 'Nobody came'.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Czech: negative prefix ne- on the verb; symmetric negation. Negative
indefinites obligatorily co-occur with predicate negation (negative concord):
Nikdo neprisel 'Nobody NEG.came' = 'Nobody came'.
Note: Czech is not in the WALS Ch 113--115 sample; Ch 112 classification
is grounded.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Spanish: negative particle no; symmetric negation. Mixed behavior for
negative indefinites: preverbal indefinites preclude negation
(Nadie vino 'Nobody came'), but postverbal indefinites require it
(No vi nada 'NEG I.saw nothing').
Equations
- One or more equations did not get rendered due to their size.
Instances For
Italian: negative particle non; symmetric negation. Mixed behavior for
negative indefinites (paralleling Spanish): preverbal n-words stand alone
(Nessuno è venuto 'Nobody came'), but postverbal n-words require non
(Non ho visto nessuno 'NEG have seen nobody').
Equations
- One or more equations did not get rendered due to their size.
Instances For
Burmese: bipartite negation with prefix ma- and suffix -bu; the negative
suffix -bu replaces the TAM markers used in the affirmative. Always
asymmetric: the negative neutralizes TAM distinctions.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Maori: negative word kaahore; isolating language makes it unclear whether
the negator is a verb or particle. Classified as 'wordUnclear' per WALS.
WALS Ch 113 classifies Maori as asymmetric with A/Fin: the negator
functions as a (quasi-)auxiliary that changes the finiteness structure.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Izi (Igboid, Niger-Congo): bipartite negation with prefix and suffix on the
verb: to-ome-du 'NEG-do-NEG'. Always asymmetric.
Note: Izi is not in the WALS Ch 113--115 sample; Ch 112 is grounded.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Kolyma Yukaghir: negative prefix el- on the verb. WALS classifies as
both symmetric and asymmetric, with A/Cat subtype: tense marking may
differ under negation but not in all constructions.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Rama (Chibchan; Nicaragua): WALS Ch 112 classifies as negative particle. Has both symmetric and asymmetric negation (Ch 113), with A/Fin + A/Cat asymmetry subtype (Ch 114).
Equations
- One or more equations did not get rendered due to their size.
Instances For
Hixkaryana (Carib; Brazil): asymmetric negation of subtype A/Fin. A
non-negative copula functions as the finite element, and the lexical verb
is deverbalized by the negative suffix -hira:
amryeki-hira w-ah-ko 'hunt-NEG 1SUBJ-be-IMM.PST'.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Nelemwa (Oceanic; New Caledonia): negative indefinites use a negative
existential construction: kia agu i uya 'not.exist person 3SG arrive'
= 'Nobody came'. Classified as negExistential for Ch 115.
Note: Nelemwa is only in the WALS Ch 115 sample; Ch 112-114 values
are based on descriptive sources.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Imbabura Quechua: negative particle mana with optional irrealis suffix
-chu. WALS classifies as SymAsy with A/NonReal: some constructions are
symmetric (particle alone), others require -chu irrealis marking on the
verb. The A/NonReal asymmetry is paradigmatic — the negative obligatorily
includes an irrealis category absent from the affirmative.
Equations
- One or more equations did not get rendered due to their size.
Instances For
All language profiles in the sample.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Does a language use a given morpheme type?
Equations
- p.hasMorphemeType t = (p.morphemeType == t)
Instances For
Does a language have symmetric negation (either symmetric only or both)?
Equations
Instances For
Does a language have asymmetric negation (either asymmetric only or both)?
Equations
Instances For
Does a language show negative concord (co-occurrence of negative indefinites with predicate negation)?
Equations
Instances For
Count of languages in the sample with a given morpheme type.
Equations
- Phenomena.Negation.Typology.countByMorphemeType langs t = (List.filter (fun (x : Phenomena.Negation.Typology.NegationProfile) => x.hasMorphemeType t) langs).length
Instances For
Count of languages in the sample with a given symmetry type.
Equations
- Phenomena.Negation.Typology.countBySymmetry langs s = (List.filter (fun (x : Phenomena.Negation.Typology.NegationProfile) => x.symmetry == s) langs).length
Instances For
Ch 112: Negative particles outnumber negative affixes.
Ch 112: Negative auxiliary verbs are rare (< 5% of sample).
Ch 112: Particles + affixes together account for the vast majority.
Ch 113: SymAsy (both) is the most common type, followed by Sym. Asy-only is the least common.
Ch 113: Languages with at least some symmetric negation (Sym + Both) far outnumber purely asymmetric languages.
Ch 114: Among languages with a single asymmetry subtype, A/Cat is the most common, followed by A/Fin and A/NonReal.
Ch 115: Co-occurrence with predicate negation (negative concord) is by far the most common pattern worldwide. Co-occurrence outnumbers preclusion by more than 15x.
Ch 115: Preclusion of predicate negation by negative indefinites is the rarest of the four strategies.
In our sample, every language with bipartite ("double") negation morphemes (Ch 112) has asymmetric negation (Ch 113). This makes sense: if negation requires two markers whose placement changes the clause structure, the negative clause structurally differs from the affirmative.
In our sample, all Slavic languages (Russian, Czech) show negative concord: negative indefinites obligatorily co-occur with predicate negation.
Equations
Instances For
Languages with negative words of unclear status (Ch 112 type 4) are common in isolating languages where verbal morphology is minimal. Maori (isolating, Polynesian) illustrates this: without verbal inflection, there is no morphological basis for deciding if the negator is a verb.
Italian and Spanish share the same mixed n-word strategy.
In our sample, every language classified as symmetric-only (Ch 113) has a non-assignable asymmetry subtype (Ch 114).
In our sample, no language classified as asymmetric or both has a non-assignable subtype.
In the WALS data, the count of non-assignable languages in Ch 114 exactly equals the count of symmetric-only languages in Ch 113. This is a consistency check: the same set of languages.
Negative auxiliary verbs (Ch 112) are always associated with asymmetric
negation of subtype A/Fin: the auxiliary becomes the finite element, and
the lexical verb is defiticized. Finnish illustrates this perfectly:
e-n tule [NEG-1SG come.CONNEG]. In our sample, Finnish is the only
negative auxiliary verb language, and it has A/Fin asymmetry.
Areal pattern: the negative auxiliary verb type is concentrated in northern Eurasia, stretching from Finland to western Siberia (@cite{dryer-haspelmath-2013}, sec. 2). Our sample contains Finnish as the representative; other languages in this belt include Estonian, Nenets, Evenki, Khanty.
Number of languages in our sample.
Morpheme type distribution in our sample.
Symmetry distribution in our sample.
Ch 143A: Preverbal particle (NegV) is the most common single-negation type, accounting for 525 of 1325 languages.
Ch 143A: Postverbal affix ([V-Neg]) is the second most common type.
Ch 143A: Postverbal particle (VNeg) count.
Ch 143A: Preverbal affix ([Neg-V]) count.
Ch 143A: Obligatory double negation count.
Ch 143A: Optional double negation count.
Ch 143A: Preverbal negation (particle or affix) is far more common than postverbal negation (particle or affix).
Ch 143A: Particles (free words) are more common than affixes (bound morphemes) for both preverbal and postverbal positions.
Ch 143B: NegVNeg (discontinuous double negation) is the most common obligatory double negation pattern (35 of 119 languages).
Ch 143B: Neg[V-Neg] (28 languages) and [Neg-V-Neg] (27 languages) are close behind NegVNeg (35) as the most common double negation patterns.
Ch 143E: Most languages have a preverbal negative particle.
Ch 143E: Preverbal affix count.
Ch 143E: Languages with no preverbal negative morpheme.
Ch 143F: Postverbal affix is more common than postverbal particle.
Ch 143F: Postverbal affix count.
Ch 143F: Postverbal particle count.
Ch 143F: Most languages have no postverbal negative morpheme.
Ch 143G: The vast majority of languages use no minor morphological means (negative tone, infix, stem change) for negation.
Ch 143G: Negative tone is the most common minor means.
Ch 143E + 143F: Most languages have at least one preverbal or postverbal negative morpheme. The "none" counts for preverbal (390) and postverbal (712) do not sum to more than the sample size, meaning some languages lack both and rely on double negation or minor means.
Ch 144A: MorphNeg (morphological negation, no negative word) is the largest single category.
Ch 144A: SNegVO (subject, then negative word, then verb-object) is the most common word-order-specific position.
Ch 144A: ObligDoubleNeg count.
Ch 144A: OptDoubleNeg count.
Ch 144A: SVONeg count.
Ch 144A: SONegV count.
Ch 144A: SOVNeg count.
Ch 144A: NegVSO count.
Ch 144A: NegVOS count.
Ch 144D: In SVO languages, SNegVO is the most common negation position.
Ch 144D: SNegVO count.
Ch 144D: SVONeg count.
Ch 144D: S[Neg-V]O (preverbal affix) count.
Ch 144L: In SOV languages, postverbal affix SO[V-Neg] is the most common single-negation position.
Ch 144L: SO[V-Neg] count.
Ch 144L: SONegV count.
Ch 144L: SOVNeg count.
Ch 144L: SO[Neg-V] count.
Ch 144T: In verb-initial languages, NegVSO is the dominant pattern.
Ch 144T: NegVOS count.
Ch 144Y: Object-initial languages are rare (16 in sample).
Ch 144B: Immediately preverbal is the most common position for negative words.
Ch 144B: Immediately preverbal outnumbers all other positions.
Ch 143A: Double negation (obligatory + optional) accounts for about 15% of all languages in the sample.
Ch 144A: Languages with morphological negation (no negative word) are the single largest category, but languages with a negative word of some kind outnumber them.
Ch 143A vs Ch 144A: The two chapters cover overlapping but different samples (1325 vs 1190 languages).
Ch 144D + 144L + 144T: SVO, SOV, and verb-initial languages together account for the entire Ch 144A sample (with overlap, since some languages appear in multiple word-order-specific sub-features).
Expletive negation survey #
@cite{jin-koenig-2021}
Expletive negation (EN) — semantically vacuous negation triggered by the lexical meaning of an embedding predicate or operator — was surveyed across 722 languages. EN was attested in 74 languages across 37 genera (every continental area except South America).
The two most widespread EN triggers:
- BEFORE (UNTIL): 50 of the 74 EN-attesting languages
- FEAR (AFRAID): 39 of the 74 EN-attesting languages
Cross-linguistic EN survey results.
- totalSurveyed : Nat
Total languages surveyed
- languagesWithEN : Nat
Languages where EN was attested
- generaWithEN : Nat
Genera where EN was attested
- beforeTriggerCount : Nat
Languages with EN in BEFORE-clauses specifically
- fearTriggerCount : Nat
Languages with EN in FEAR-clauses specifically
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
The overall EN survey from @cite{jin-koenig-2021}.
Equations
- Phenomena.Negation.Typology.enSurvey = { totalSurveyed := 722, languagesWithEN := 74, generaWithEN := 37, beforeTriggerCount := 50, fearTriggerCount := 39 }
Instances For
EN is attested in a substantial minority of surveyed languages.
EN is found across many genera (not an areal phenomenon).
BEFORE is the most common EN trigger.
BEFORE triggers occur in the majority of EN-attesting languages.
Continental areas from @cite{jin-koenig-2021} Table 2.
- africa : ContinentalArea
- australiaNewGuinea : ContinentalArea
- eurasia : ContinentalArea
- northAmerica : ContinentalArea
- southAmerica : ContinentalArea
- pidginsCreoles : ContinentalArea
Instances For
Equations
- Phenomena.Negation.Typology.instBEqContinentalArea.beq x✝ y✝ = (x✝.ctorIdx == y✝.ctorIdx)
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
Per-area EN survey data (Table 2).
- area : ContinentalArea
- languagesLookedAt : Nat
- languagesWithEN : Nat
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
Table 2: Distribution of languages with and without EN by continental area (@cite{jin-koenig-2021}).
Equations
- One or more equations did not get rendered due to their size.
Instances For
The per-area EN counts sum to 74. (The per-area language-looked-at counts sum to 728, not 722 — the paper's total is 722, suggesting 6 Pidgin/Creole languages are also counted in geographic areas.)
EN is not attested in South America in this sample.
Eurasia has the highest concentration of EN-attesting languages.
Per-language EN attestation #
Table 3 lists all 74 languages where EN was attested, grouped by
continental area and genus. Each entry records the EN trigger concepts
attested in that language (using the concept labels from the paper,
in small capitals). Lexical forms are recorded in the forms field
where available in the paper.
A language with attested EN and its trigger concepts (Table 3).
- name : String
Language name
- iso : String
ISO 639-3 code
- genus : String
Genus (following the paper's classification)
- area : ContinentalArea
Continental area
EN trigger concepts attested (concept labels from Table 3)
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
Table 3: All 74 languages where EN was attested, with their trigger concepts (@cite{jin-koenig-2021}, pp. 45–48).
Equations
- One or more equations did not get rendered due to their size.
Instances For
Table 3 has exactly 74 languages.
Per-area counts match Table 2.
No South American languages have EN in this sample.
Every language has at least one trigger concept.
The number of distinct genera represented (counting unique genus strings).
Cross-validating Table 3 and Table 4 #
Table 4 reports that BEFORE (UNTIL) triggers EN in 50 languages and FEAR (AFRAID) in 39. We verify lower bounds from the per-language trigger lists in Table 3. (Some languages in the paper's broader survey data may have BEFORE/FEAR triggers listed in the Appendix but summarized under different concept labels in Table 3, so Table 3 counts are conservative.)
At least 42 languages in Table 3 have BEFORE/UNTIL as an EN trigger. (Table 4 reports 50 — the difference is likely due to additional triggers identified through detailed study of the five core languages and listed in the Appendix but not in Table 3's compact listing.)
At least 38 languages in Table 3 have FEAR/AFRAID as an EN trigger. (Table 4 reports 39.)
BEFORE/UNTIL is more widely attested than FEAR/AFRAID in Table 3, consistent with Table 4's ranking (50 > 39).