Universal #107
Legacy number: 107
| Number | 107 |
| Original | Nominal modifiers (such as relative, adjectival, and attributive expressions) follow nouns in VO languages and precede nouns in OV languages |
| Standardized | IF basic order is VO , THEN relative clauses, attributive adjectives, and attributive nouns follow their head nouns. IF basic order is OV, THEN relative clauses, attributive adjectives, and attributive nouns precede their head nouns. |
| Formula | VO ⇒ N Rel & N Adj & N Gen; OV ⇒ Rel N & Adj N & Gen N |
| Keywords | order, OV, VO, verb, object, noun, modifier, relative clause, attributive, genitive, adjective |
| Domain | syntax |
| Type | implication |
| Status | achronic |
| Quality | statistical |
| Basis | sample of 30 languages in Greenberg 1963; languages in Lehmann 1973 |
| Source | W. P. Lehmann 1971: 19, W. P. Lehmann 1972: 276-7, W. P. Lehmann 1973: 48; claimed also in Pinnow 1960: 96-97, Greenberg 1963: 85-90, Keenan 1978a: 115, Downing 1978: 383 |
Discussion
Counterexamples
Chinese (Sinitic, Sino-Tibetan): SVO & Rel N; Persian (Iranian, IE), Bengali (Indic, IE); Georgian (S. Caucasian): SOV & N Rel; Classical Tibetan (Tibeto-Karen, Sino-Tibetan); Nama (Khoisan); Quechua (Andean), Papago (Tepiman, Uto-Aztecan); Turkish (Turkic, Altaic): all have both preposed and postposed relative clauses (Downing 1978).Tagalog (W. Malayo-Polynesian, Austronesian) and possibly other Philippine languages are verb-initial and have both prenominal and postnominal Rel (Keenan 1985: 144).Woods Cree (Algonquian): VO & Gen N (Starks 1987: 225).Guajajara (Tupi): VO & N Rel & Gen N & N A (Harrison 1983, 1986).Tigre (Semitic, Afroasiatic): SOV & A N & N Gen (Dryer 2000). Burmese (Burman): SOV & N A (Foster & Hofling 1987), cf. #1616.Mende (Mande, Niger-Congo): SOV & N Rel (Foster & Hofling 1987), cf. #1616.see also counterexamples to #56.
Notes
1. According to #15 (also due to W. P. Lehmann), verbal modifiers such as negation, causation, and reflexive or reciprocal exhibit conversely ordered placement.2. Concerning W. P. Lehmann’s 1973 principle that there tend to be no intervening elements between object and verb (see ##15, 107); hence Obj modifiers - Obj - V - Sentence mood markers, Sentence mood markers - V - Obj - Obj modifiers Krifka (1985: 87) suggested a more general principle: “Whenever a complex constituent consisting of a head h and a modifier m is itself a modifier to a head H, then there is a tendency for m not to intervene between h and H. Thus preferred orders: (mh)H, H(hm); dispreferred orders: (hm)H, H(mh).” 3. Dryer 1986: 98, 1988: 191: “There is no evidence of any relationship between the order of Verb and Object and the order of Adjective and Noun.” See also #1021.4. Dryer 1986: 101-103: “Although there is a clear preference for Gen N order among OV languages, the preference for N Gen order among VO languages is considerably weaker. Only in Africa and North America the VO languages are predominantly N Gen. (...) Furthermore, there is a significant difference between SVO and verb-initial languages in the position of genitives.”5. Dryer 1991 shows that while SVO languages generally exhibit properties associated with verb-initial languages rather than properties associated with verb-final languages, SVO languages are intermediate between verb-initial and verb-final languages as far as the order of genitive and noun is concerned, with G N and N G order approximately equal in frequency.
Log in to join the discussion.
2. Concerning W. P. Lehmann's 1973 principle that there tend to be no intervening elements between object and verb (see <a class="innerLink" href="browse.php?entry_id=15"><img src="../img/entry.gif" border="0">, <a class="innerLink" href="browse.php?entry_id=107"><img src="../img/entry.gif" border="0">); hence
Obj modifiers - Obj - V - Sentence mood markers,
Sentence mood markers - V - Obj - Obj modifiers
Krifka (Krifka 1985: 87) suggested a more general principle: "Whenever a complex constituent consisting of a head h and a modifier m is itself a modifier to a head H, then there is a tendency for m not to intervene between h and H. Thus preferred orders: (mh)H, H(hm); dispreferred orders: (hm)H, H(mh)."
3. Dryer 1986: 98, Dryer 1988: 191: "There is no evidence of any relationship between the order of Verb and Object and the order of Adjective and Noun." See also <a class="innerLink" href="browse.php?entry_id=1021"><img src="../img/entry.gif" border="0">.
4. Dryer 1986: 101-103: "Although there is a clear preference for Gen N order among OV languages, the preference for N Gen order among VO languages is considerably weaker. Only in Africa and North America the VO languages are predominantly N Gen. (...) Furthermore, there is a significant difference between SVO and verb-initial languages in the position of genitives."
5. Dryer 1991 shows that while SVO languages generally exhibit properties associated with verb-initial languages rather than properties associated with verb-final languages, SVO languages are intermediate between verb-initial and verb-final languages as far as the order of genitive and noun is concerned, with G N and N G order approximately equal in frequency.