Universal #619
Legacy number: 619
| Number | 616 |
| Original | If a given language presents NPrel as a pronoun for any position in the hierarchy , then it presents NPrel as a pronoun for all lower positions on the hierarchy: subject > direct object > non-direct object > possessor. |
| Standardized | If an NPrel is a pronoun for subjects, THEN an NPrel can be a pronoun for non-direct objects, direct objects, and possessors as well. If NPrel is a pronoun for direct objects, THEN NPrel can be a pronoun for non-direct objects, and possessors as well. If NPrel is a pronoun for non-direct objects, THEN NPrel can be a pronoun for possessors as well. |
| Formula | NPrel as a pronoun for subject ⇒ NPrel as a pronoun for direct object ⇒ NPrel as a pronoun for non-direct object ⇒ NPrel as a pronoun for possessor |
| Keywords | accessibility hierarchy, pronoun, relative clause, attributive, non-direct object, direct object, subject |
| Domain | syntax |
| Type | implicational hierarchy |
| Status | achronic |
| Quality | absolute |
| Basis | languages in Keenan 1985b |
| Source | Keenan 1985b: 147ff. |
Discussion
Counterexamples
Vata, a Kru language (Niger-Congo) described by Hilda Koopman, in which resumptive pronouns are obligatory when the subject is relativized and impossible in all otherpositions. (P. Kroeger, Discussion List for ALT, March 8, 2001)
Notes
1. The lower NPrel is on the Hierarchy the more common it is to find it expressed by a personal pronoun.2. Note that the two implicational patterns #619 and #622 go in opposite directions in the relative clause data: gaps from low to high, copy pronouns from high to low. The gaps cut off in more complex environments, while the pronouns cut off in simpler environments.3. RCs which express NPrel as a personal pronoun typically allow a greater range of positions to be relativized compared with RCs which do not use such pronouns (though they may use specifically relative pronoun). For example, some languages (Basque, N.Frisian) do not easily relativize possessor NPs at all, but possessors are always relativizable in ‘pronoun retaining’ RCs.4. Hawkins 1999: 258 reformulates this universal in terms of complexity hierarchy. Relative Clause Copy Pronoun Hierarchy:If a copy pronoun (also called resumptive pronoun) in a relative clause is grammatical in position on a complexity hierarchy H, then copy pronouns will be grammatical in all lower positions that can be relativized at all.
Log in to join the discussion.
2. Note that the two implicational patterns <a class="innerLink" href="browse.php?entry_id=619"><img src="../img/entry.gif" border="0"> and <a class="innerLink" href="browse.php?entry_id=622"><img src="../img/entry.gif" border="0"> go in opposite directions in the relative clause data: gaps from low to high, copy pronouns from high to low. The gaps cut off in more complex environments, while the pronouns cut off in simpler environments.
3. RCs which express NPrel as a personal pronoun typically allow a greater range of positions to be relativized compared with RCs which do not use such pronouns (though they may use specifically relative pronoun). For example, some languages (Basque, N.Frisian) do not easily relativize possessor NPs at all, but possessors are always relativizable in 'pronoun retaining' RCs.
4. Hawkins 1999: 258 reformulates this universal in terms of complexity hierarchy. Relative Clause Copy Pronoun Hierarchy: If a copy pronoun (also called resumptive pronoun) in a relative clause is grammatical in position on a complexity hierarchy H, then copy pronouns will be grammatical in all lower positions that can be relativized at all.