Cross-Linguistic Typology of Tense and Aspect (WALS Chapters 65--69) #
@cite{comrie-1985} @cite{dryer-haspelmath-2013} @cite{dahl-velupillai-2013} @cite{de-haan-2013}
Cross-linguistic data on tense-aspect systems from the World Atlas of Language Structures (WALS), covering five parameters:
Ch 65 (Perfective/Imperfective Aspect): Whether a language grammaticalizes the perfective/imperfective distinction. 222 languages, two values.
Ch 66 (The Past Tense): Whether a language has grammatical marking of past tense, and if so how many remoteness distinctions it makes. 222 languages, four values.
Ch 67 (The Future Tense): Whether a language has inflectional marking of future tense. 222 languages, two values.
Ch 68 (The Perfect): Whether a language has a distinct perfect category (resultative + experiential uses), and its diachronic source. 222 languages, four values.
Ch 69 (Position of Tense-Aspect Affixes): How tense-aspect is morphologically expressed. 1062 languages, five values.
Key findings #
@cite{dahl-velupillai-2013} @cite{bybee-perkins-pagliuca-1994} note that there is no evidence for a typological division into "tense languages" vs "aspect languages" -- languages that have aspectual marking tend also to have tense marking. Suffixing is overwhelmingly the dominant strategy for tense-aspect morphology. South-East Asian languages consistently lack morphological tense-aspect marking.
WALS Ch 65: Whether the perfective/imperfective distinction is grammaticalized in the language. "Grammatical marking" includes both morphological and periphrastic constructions.
- grammatical : AspectMarking
Language has grammatical marking of perfective/imperfective
- none : AspectMarking
Language has no grammatical marking of perfective/imperfective
Instances For
Equations
- Phenomena.TenseAspect.Typology.instBEqAspectMarking.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 66: Past tense marking and remoteness distinctions.
Past marking includes both inflectional and periphrastic constructions. Past imperfectives (restricted to imperfective contexts, as in the Armenian Imperfect) count as past marking.
- marked : PastMarking
Past/non-past distinction marked, no remoteness distinctions
- markedRemoteness2_3 : PastMarking
Past/non-past distinction marked, 2-3 degrees of remoteness
- markedRemoteness4plus : PastMarking
Past/non-past distinction marked, 4+ degrees of remoteness
- none : PastMarking
No grammatical marking of past/non-past distinction
Instances For
Equations
- Phenomena.TenseAspect.Typology.instBEqPastMarking.beq x✝ y✝ = (x✝.ctorIdx == y✝.ctorIdx)
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
Whether a language has any past tense marking (any of the three marked categories).
Equations
Instances For
WALS Ch 67: Whether the language has inflectional future marking.
Only inflectional marking is counted here (not periphrastic constructions like English "will" + V). Irrealis markers that obligatorily encode future reference are included.
- inflectional : FutureMarking
Inflectional marking of future/nonfuture distinction
- none : FutureMarking
No inflectional marking of future/nonfuture distinction
Instances For
Equations
- Phenomena.TenseAspect.Typology.instBEqFutureMarking.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 68: Whether the language has a distinct perfect category, and its diachronic source.
A form counts as a perfect only if it has both resultative and experiential uses (not a general past, not a dedicated resultative).
- fromPossessive : PerfectType
Perfect derived from possessive construction ('have'-perfect). Restricted almost exclusively to western Europe.
- fromFinishAlready : PerfectType
Perfect derived from word meaning 'finish' or 'already'. Concentrated in South-East Asia and West Africa.
- other : PerfectType
Other perfect (including from dedicated resultatives, or diachronic source undetermined).
- none : PerfectType
No perfect category.
Instances For
Equations
- Phenomena.TenseAspect.Typology.instBEqPerfectType.beq x✝ y✝ = (x✝.ctorIdx == y✝.ctorIdx)
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
Whether a language has any perfect category.
Equations
Instances For
WALS Ch 69: Primary morphological strategy for tense-aspect marking.
Languages with heterogeneous strategies where no single type is
dominant are classified as mixed. Infixes and stem changes are
treated as subtypes of prefixes/suffixes when localized at edges.
- prefixing : TAAffixPosition
Tense-aspect prefixes (primary strategy)
- suffixing : TAAffixPosition
Tense-aspect suffixes (primary strategy)
- tonal : TAAffixPosition
Tense-aspect tone (primary strategy). Almost exclusively African.
- mixed : TAAffixPosition
Combination of strategies with none primary
- noInflection : TAAffixPosition
No tense-aspect inflection
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
Whether a language has any tense-aspect affixation.
Equations
- Phenomena.TenseAspect.Typology.TAAffixPosition.prefixing.hasAffixes = true
- Phenomena.TenseAspect.Typology.TAAffixPosition.suffixing.hasAffixes = true
- Phenomena.TenseAspect.Typology.TAAffixPosition.tonal.hasAffixes = true
- Phenomena.TenseAspect.Typology.TAAffixPosition.mixed.hasAffixes = true
- Phenomena.TenseAspect.Typology.TAAffixPosition.noInflection.hasAffixes = false
Instances For
A language's tense-aspect typological profile across WALS chapters 65-69.
- language : String
Language name
- iso : String
ISO 639-3 code
- family : String
Language family
- aspect : AspectMarking
WALS Ch 65: perfective/imperfective aspect
- past : PastMarking
WALS Ch 66: past tense marking
- future : FutureMarking
WALS Ch 67: inflectional future
- perfect : PerfectType
WALS Ch 68: perfect category
- affixPosition : TAAffixPosition
WALS Ch 69: tense-aspect affix position
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.
Instances For
Equations
- One or more equations did not get rendered due to their size.
- Phenomena.TenseAspect.Typology.instBEqTAProfile.beq x✝¹ x✝ = false
Instances For
WALS Ch 65 distribution: perfective/imperfective aspect.
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.
Instances For
Equations
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.
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.
- Phenomena.TenseAspect.Typology.instBEqCh66Counts.beq x✝¹ x✝ = false
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
WALS Ch 67 distribution: inflectional future.
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.
Instances For
Equations
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.
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.
- Phenomena.TenseAspect.Typology.instBEqCh68Counts.beq x✝¹ x✝ = false
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.
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.
- Phenomena.TenseAspect.Typology.instBEqCh69Counts.beq x✝¹ x✝ = false
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
Ch 65 sample size: 222 languages.
Ch 66 sample size: 222 languages.
Ch 67 sample size: 222 languages.
Ch 68 sample size: 222 languages.
English (Indo-European, Germanic). Periphrastic perfective (simple past vs progressive), inflectional past, no inflectional future (uses "will" — periphrastic), 'have'-perfect, suffixing (-ed, -ing).
Equations
- One or more equations did not get rendered due to their size.
Instances For
Russian (Indo-European, Slavic). Robust perfective/imperfective (typically via prefixation), inflectional past, no inflectional future (periphrastic "budet" + infinitive for imperfective future, perfective present = future), no perfect, suffixing (-l, -la for past).
Equations
- One or more equations did not get rendered due to their size.
Instances For
French (Indo-European, Romance). Imparfait/passé composé aspect distinction, inflectional past (passé simple), inflectional future (-ai, -as, -a...), 'have'-perfect (avoir + past participle), suffixing.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Spanish (Indo-European, Romance). Imperfecto/preterito (imperfective/perfective), inflectional past and future, 'have'-perfect (haber + participle), suffixing.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Finnish (Uralic). No grammatical perfective/imperfective, inflectional past (-i), no inflectional future (present tense used for future reference), perfect (on + past participle, 'be'-type), suffixing.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Turkish (Turkic). Perfective/imperfective (-di vs -iyor), inflectional past, inflectional future (-ecek), other perfect (-miş, from evidential/inferential), suffixing.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Japanese (Japonic). Perfective/imperfective (ta-form vs te-iru), inflectional past (-ta, -da), no inflectional future, no distinct perfect, suffixing.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Korean (Koreanic). Perfective/imperfective (-었- vs -고 있-), inflectional past (-었-), inflectional future (-겠-), no distinct perfect, suffixing.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Mandarin Chinese (Sino-Tibetan). Perfective le/guo, but these are particles (not inflectional), no inflectional past, no inflectional future, no distinct perfect, no tense-aspect inflection. Quintessential isolating language.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Vietnamese (Austroasiatic). No grammatical perfective/imperfective, no inflectional past, no inflectional future, no perfect. Tense-aspect expressed by separate particles (da, se). Part of the South-East Asian isolating area.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Thai (Tai-Kadai). No grammatical perfective/imperfective, no inflectional past, no inflectional future, no perfect. South-East Asian isolating type.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Indonesian (Austronesian, Malayo-Polynesian). No grammatical aspect, no past marking, no inflectional future, no perfect. Classic example of tenselessness (air itu dingin = 'the water is/was cold'). Minor tense-aspect suffix exists (-i repetitive), but marginal.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Swahili (Niger-Congo, Bantu). Perfective/imperfective (-li-, -na-), past marking, no inflectional future, other perfect (-me-), prefixing (T/A markers are verbal prefixes).
Equations
- One or more equations did not get rendered due to their size.
Instances For
Yoruba (Niger-Congo, Atlantic-Congo). Perfective/imperfective distinction, no past tense marking, no inflectional future, perfect from 'already' (ti), no tense-aspect inflection (preverbal particles).
Equations
- One or more equations did not get rendered due to their size.
Instances For
Hindi (Indo-European, Indo-Aryan). Perfective/imperfective (-aa, -taa), inflectional past, inflectional future (-egaa), other perfect, suffixing.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Arabic (Egyptian) (Afro-Asiatic, Semitic). Perfective/imperfective (katab/yiktib), inflectional past (perfective form encodes past), no inflectional future (uses preverbal particles), no distinct perfect, suffixing.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Quechua (Quechuan). No grammatical perfective/imperfective, inflectional past with 2-3 remoteness distinctions (direct vs indirect past), inflectional future (-saq), other perfect (-sqa resultative), suffixing.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Tagalog (Austronesian, Malayo-Polynesian). Perfective/imperfective (completed vs contemplated aspect), no inflectional past or future (aspect-based system), no distinct perfect, prefixing (mag-, nag-, -um- etc.).
Equations
- One or more equations did not get rendered due to their size.
Instances For
Lango (Nilotic, Eastern Sudanic). Perfective/imperfective/progressive marked primarily by tone, past tense marking, no inflectional future, other perfect, tonal T/A marking.
Equations
- One or more equations did not get rendered due to their size.
Instances For
All languages in the sample.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Total languages with past marking in Ch 66.
Equations
- One or more equations did not get rendered due to their size.
Instances For
A majority (134/222 = 60%) of WALS Ch 66 languages have past marking.
In the sample, languages with BOTH aspect and past marking outnumber languages with ONLY one of the two.
@cite{dahl-velupillai-2013}: "there are considerably more languages in the sample that have both the aspectual and the temporal categories, or neither of the alternatives, than have one only."
Equations
Instances For
Equations
Instances For
Equations
Instances For
In our sample: more languages have both aspect+past than aspect-only.
In our sample: more languages have both aspect+past than past-only.
Whether a language lacks all three major T/A gram types from Chs 65-67 (no aspect, no past, no future). This is the hallmark of the South-East Asian isolating cluster.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Vietnamese, Thai, and Indonesian all lack the three major T/A gram types: no grammaticalized perfective/imperfective, no past marking, no inflectional future.
@cite{dahl-velupillai-2013}: "The languages in this area lack not only past tenses but also marking of the imperfective/perfective distinction and inflectional futures."
Note: Indonesian has a minor repetitive suffix (-i) and so WALS classifies it as having T/A suffixes, but it lacks the major T/A categories.
Inflectional future is approximately 50/50 in Ch 67 (110 vs 112). Neither value constitutes a strong majority.
'Have'-perfects (from possessive constructions) are extremely rare cross-linguistically — only 7 out of 222 languages. They are restricted almost exclusively to western Europe.
Most languages lack a distinct perfect category entirely.
Among languages in our sample with grammatical perfective/imperfective aspect, the majority also have past tense marking.
This supports Dahl & Velupillai's observation that aspect and tense tend to co-occur rather than being alternatives.
Tone as the primary tense-aspect strategy is extremely rare (13/1131).
In our sample, only Lango uses tone as the primary T/A strategy.
Most past-marking languages make no remoteness distinctions (94 out of 134 with past marking). Languages with 4+ degrees of remoteness are extremely rare (only 2: Yagua being the richest with 5 degrees).
Perfects from 'finish'/'already' are more common than 'have'-perfects (21 vs 7), concentrated in South-East Asia and West Africa.
Languages in our sample that lack ALL four T/A categories (no aspect, no past, no future, no perfect).
Equations
- One or more equations did not get rendered due to their size.
Instances For
There exist at least 2 languages in our sample with no T/A categories.
Languages that lack all four T/A categories AND have no T/A inflection form the core of the South-East Asian isolating cluster.
This matches the pattern described by @cite{dahl-velupillai-2013}: the same languages that lack tense and aspect grams are the isolating languages that lack inflectional morphology.
Note: Indonesian has a minor repetitive suffix (-i) in WALS, so it
has lacksAllTA but not noInflection. Vietnamese and Thai have both.
WALS Ch 78: How evidentiality is coded in the language.
"Evidentiality" here means grammaticalized marking of information source. Languages differ in whether evidentiality is part of the tense system, expressed by verbal affixes, clitics, particles, or not grammaticalized.
- none : EvidentialityCoding
No grammatical evidentiality
- partOfTense : EvidentialityCoding
Evidentiality is part of the tense paradigm (e.g., Turkish -miş)
- verbalAffix : EvidentialityCoding
Evidentiality marked by verbal affix
- clitic : EvidentialityCoding
Evidentiality marked by clitic
- particle : EvidentialityCoding
Evidentiality marked by particle
- other : EvidentialityCoding
Other strategy
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
Whether a language has any grammatical evidentiality.
Equations
Instances For
A language's typological profile across the full TAME space: tense (Ch 66–67), aspect (Ch 65), perfect (Ch 68), morphological position (Ch 69), and evidentiality (Ch 78).
Extends TAProfile with the evidentiality dimension.
- language : String
Language name
- iso : String
ISO 639-3 code
- family : String
Language family
- aspect : AspectMarking
WALS Ch 65: perfective/imperfective aspect
- past : PastMarking
WALS Ch 66: past tense marking
- future : FutureMarking
WALS Ch 67: inflectional future
- perfect : PerfectType
WALS Ch 68: perfect category
- affixPosition : TAAffixPosition
WALS Ch 69: tense-aspect affix position
- evidentiality : EvidentialityCoding
WALS Ch 78: coding of evidentiality
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.
Instances For
Equations
- One or more equations did not get rendered due to their size.
- Phenomena.TenseAspect.Typology.instBEqTAMEProfile.beq x✝¹ x✝ = false
Instances For
Turkish (Turkic). Evidentiality is part of the tense paradigm: -miş marks indirect evidence (hearsay/inference) vs -di (direct).
Equations
- One or more equations did not get rendered due to their size.
Instances For
Quechua (Quechuan). Evidentiality via verbal suffixes: -mi (direct), -si (hearsay), -chá (conjecture).
Equations
- One or more equations did not get rendered due to their size.
Instances For
Korean (Koreanic). Evidentiality via verbal suffixes: -te (retrospective evidential), -ney (contemporaneous evidential). @cite{cumming-2026} analyzes these as tense-evidential interactions.
Equations
- One or more equations did not get rendered due to their size.
Instances For
English (Indo-European). No grammatical evidentiality. Evidential distinctions are expressed lexically ("apparently", "I heard that..."), not grammatically.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Mandarin Chinese (Sino-Tibetan). No grammatical evidentiality, no tense-aspect inflection. Isolating type.
Equations
- One or more equations did not get rendered due to their size.
Instances For
All TAME-profiled languages.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Whether a TAMEProfile has any tense or aspect marking
(aspect, past, or future).
Equations
Instances For
In our TAME sample, every language with grammatical evidentiality also has tense or aspect marking.
This reinforces the "no tense-vs-aspect divide" finding (Generalization 3) by showing that evidentiality also clusters with tense/aspect marking rather than replacing it.