Universal 565: Du (3 person) ⇔ Du (2 person)

Original: If 3rd Person differentiates a Dual, so does 2nd, and vice versa.
Standardized: IF 3rd person differentiates a dual, THEN so does 2nd.
IF 2nd person differentiates a dual, THEN so does 3rd.
Keywords: pronoun, personal pronoun, number, dual, 3rd person, 2nd person
Domain: inflection
Type: mutual implication
Status: achronic
Quality: absolute
Basis: languages in Humboldt 1830, including Basque (isolate), Greenlandic (Eskimo-Aleut), Saami (Uralic), Tahitian, Malay (both Malayo-Polynesian), Indo-European (e.g. Sanskrit, Ancient Greek), Semitic languages, American languages (e.g. Quechua, Totonaca, Huasteca, Mapuche, Tamanaca, Chayma)
Source: Humboldt 1830, as interpreted in Plank 1989: 302; Plank 1994: 232

Counterexamples: Languages with pronominal dual in 3rd but not 2nd person (thus violating the first sub-claim):1. In Hupa (Athabaskan), the indirect way of expressing duality, by means of simultaneously marking verbs as plural and non-plural, is restricted to 3rd person subjects (Plank 1989: 303-4).2. In South Arabian (Semitic, Afro-Asiatic), dual with nouns and verb agreement is restricted to 3rd person.3. In Tunica (isolate remotely related to Algonquian), the pronominal dual is limited to the masculine gender of the 3rd person.4. Certain dialects of Ostyak (Ob-Ugric, Uralic) (Kazym: Rédei 1968).5. Sedang (Mon-Khmer, Austroasiatic) has a dual in both inclusive and exclusive 1st persons and in 3rd person but not in 2nd (Smith 1979). Language with pronominal dual in 2nd person but not 3rd person (thus violating the second sub-claim):1. In Dizi (Omotic, Afro-Asiatic), independent pronouns and prefixal possessives have been claimed to have no dual except for 2nd person (Allan 1976) . This, however, might be a misanalysis as suggested by Hayward (pc to Plank, see Plank 1989: 326, note 8). No other Omotic language has a dual.

FP