INTRODUCTION
Relativization strategies in Amazonian languages
T. Givón
University of Oregon, Eugene, Oregon. White Cloud Ranch Ignacio, Colorado
This volume presents eight studies of relativization strategies in selected Amazonian languages. While it could not be comprehensive, it nonetheless ranges over several major themes in the cross-linguistic typology of relative clauses. First is the degree of grammaticalization of REL clauses; that is, the structural specificity of the marking system that makes such clauses distinct from either main clauses or other types of subordinate clauses. The languages treated here cover a considerable range between relatively little distinction between REL- and main-clause, to a clear contrast between main clauses and all subordinate clauses, to the rise of variable 'recoverability strategies' that help computing the case-role of the 'missing' argument inside the REL clause.
Second is the degree of 'classical' embedding of the REL clause within the main clause. Here one finds, often in the very same language, external heads modified by an adjacent REL clause, internal heads case-marked only for their function within the REL clause, and a variety of extrapositions, dislocations, repetitions, and headless or internally-headed constructions. And these familiar pragmatic-order variations interact, on occasion, with the grammar of case-role differentiation.
Third, one finds the grand theme of subordinate-clause nominalization. This major typological feature divides extremely-nominalizing languages – those that nominalized all subordinate clauses – from extremely-finite ones, and any shade and gradation in-between. The small sample of languages in the volume tends, at least historically, towards the extreme-nominalizing end – but to different degrees and at different time-depths from the prototype. And even within a single language one can see both extremes, with the diachronically-older nominalizing strategy alternating with the presumable-more-recent more finite REL-clause structure.
Eduardo Rivail Ribeiro's "Subordinate clauses in Karaja" describes a split strategy. For the subject and direct-object role, a minimal amount of specialized machinery is used, so that the REL-clause resembles the main clause – except for its clause-final (thus verb-final) stress. Whether this is a segmental rather than an intonational phenomenon remains to be seen, given that it is to some extent reminiscent of non-final raised intonation of REL-clauses in a V-final language. When the head occupies an oblique role in the main clause, the post-position appears at the NP-final – thus for this SOV language with N-REL order also a V-final – position in the REL clause. While this feature can become, potentially, part of the signal of oblique-roles relativization, it is not clear that it has so far. Rather, the various oblique types are marked with resumptive pronoun marked by post-position (a la modern Hebrew).
Ana Vilacy Galucio's "Relativização na língua Sakurabiat (Mekens)" describes nominalized REL-clauses. Subject relativization is zero-marked. The various other roles may e marked by two nominalizing V-suffixes, one for direct-object, the other for genitive. This system has the potential for gravitating towards the classical Philippine system. But it may also represent the final stages of collapse of such a system.
Carmen Lucia R. Rodrigues "Relativização em Xipaya (Tupi)" describes a subordinator V-suffix (or clitic) common to REL-clauses and other noun modifiers. No morphological provision is made for distinguishing SUBJ from OBJ REL-clauses. However, word-order devices such as extraposition or dislocation may play a role in marking such distinction. Word-order inside the REL-clause may also play a role here, distinguishing between 'the X that Y murdered' and 'The X that murdered Y'.
Sidi Facundes' "Estratégias de relativização em Apurinã (Aruak)" describes a 'basic' OSV language, one of the presumed 'impossible' word-order types. The REL-clause subordinator, perhaps a nominalizer historically, can differentiate between subject, direct-object and oblique roles. But the passive subordinator/nominalizer may also mark direct-object relativization, thus the alternative subordinate form for transitive clauses.
Hein van der Voort's "Construções atributivas em Kwaza" describes another nominalizing language, one where the nominalizer d oes not function to distinguish the various case-roles. An interesting interaction with the extensive nominal-classifier system may ultimately contribute to case-role recoverability strategy in relativization, though perhaps not yet.
Sergio Meira's "Orações relativas em línguas Karib" takes on the variety of relativization strategies within a single family. Nominalized REL-clauses are no doubt the diachronically-older pattern in this extreme-nominalizing family. And they may indeed e used to differentiate the case-roles involved. But the – probaly later – rise of invariant REL-clause subordinators is also observed in some languages, with various degrees of loss of the old nominalized structure, and thus the rise of new strategies for case-role differentiation in REL-clauses.
Pilar Valenzuela's "Aspectos morfosintacticos de la relativizacion en Shipibo-Konibo (Pano)" is a description of a rather typical Panoan language. Although subordinate clauses are historically nominalized (or take a 'participial' form), the nominalizing morphemes play no part in the strategy of case-role differentiation. Instead, the simple default strategy is employed, combining strong nominal case-marking of the surviving arguments with a relatively rigid SOV order, with the role of the missing argument inferred from what is missing, what is overtly there, and the verb-type (a la Japanese). Alternative sub-strategies of non-embedding are also possible, including 'internal head' constructions where the entire clause functions as a noun phrase (a la Bambara).
Denny Moore's "Cláusulas relativas em Gavião de Rondônia" reveals another nominalized RE-clause situation. This time the nominalizers may be, historically, old pronouns that differentiate, at most, between concrete and abstract arguments. Case-role differentiation – subject vs. object vs. oblique – is not well supported by such a system, and many REL-clauses remain, at least potentially, amiguous.
While obviously not intended to be comprehensive, given the massive spread of Amazonia, this collection nonetheless affords us an intriguing view of major typological themes as they reflect in a small sample of Amazonian languages. However restricted the sample, it succeeds in both instructing and wetting the appetite.
Publication Dates
-
Publication in this collection
28 Apr 2011 -
Date of issue
Apr 2006
