Acessibilidade / Reportar erro

The particle ‘ke’ as a differential object and subject marker in Ka’apor

A partícula ‘ke’ como estratégia de marcação diferencial de sujeito e objeto em Ka’apor

Abstract

The main purpose of this article is to show that the Ka’apor language exhibits both direct object marking (DOM) and differential subject marking (DSM). This research demonstrates that the particle ke is responsible for triggering these systems, since it is used when agent subjects are semantically affected by the event described by the verb and when objects are high on the animacy scale. In this sense, the DOM mechanism in Ka’apor is regulated by both the animacy and definiteness scales. With regard to DSM, I hypothesize that it emerges as an example of a markedness reversal, since affectedness is not a typical property of subjects, but only of objects. As a result, DSM in Ka’apor is characterized by the fact that only subjects which resemble typical patient objects are overtly case-marked by the particle ke.

Keywords
Dative; Ergative; Nominative; Differential marking; Ka’apor

Resumo

O objetivo deste artigo é demonstrar que a língua Ka’apor exibe tanto marcação diferencial de objeto, doravante DOM, como marcação diferencial de sujeito, doravante DSM. O próposito é demonstrar que a partícula ke é a responsável por acionar esses dois sistemas, visto que é utilizada quando sujeitos são semanticamente afetados pelo evento e quando objetos ocupam uma posição alta na escala de animacidade. Nesse sentido, o mecanismo de marcação diferencial de objeto em Ka’apor é regulado tanto pela escala de animacidade como pela escala de definitude. Já em relação à marcação diferencial de sujeito, a hipótese assumida é a de que DSM emerge como o reflexo de uma marcação reversa, no sentido de que afetação não é uma propriedade semântica prototípica de sujeitos, mas somente de objetos. Em suma, a marcação diferencial de sujeito é acionada quando sujeitos apresentam as propriedades semânticas que são típicas de pacientes.

Palavras-chave
Dativo; Ergativo; Nominativo; Marcação diferencial; Ka’apor

INTRODUCTION

According to Duarte (2014, p. 100)DUARTE, Fábio Bonfim. On the semantics of affectedness and its implications for argument structure in the Ka’apor language. Revista Linguística, Rio de Janeiro, v. 10, n. 1, p. 99-122, jun. 2014., “Ka’apor is spoken by about 1,000 people who live in the state of Maranhão, in northern Brazil. This language belongs to the Tupí-Guaraní family, Tupí Stock”. The main purpose of this paper is to show that Ka’apor exhibits both direct object marking, henceforth DOM, and differential subject marking, hereafter DSM. This paper will also show that the particle ke is responsible for triggering these systems, since it signals that subjects are semantically affected by the event and that objects are high in the animacy scale. Another aim is to identify what abstract Case1 1 In this paper, abstract Case will be spelled out with capital letter, while morphological case will remain with lowercase letter. See details on the difference between abstract Case and morphological case in Ura (2001). this particle encodes both in DSM and DOM systems. In line with these assumptions, the theoretical hypothesis to be explored in this paper is that this particle is used to convey animacy and definiteness on the one hand, and the semantics of affectedness on the other hand. In line with this proposal, the theory to be developed is that this particle plays a strong role in determining when the object is high in the animacy and definiteness scale, whereas lack of the association with agenthood is what determines the differential subject marking for both transitive and unergative subjects. The analysis is based on theories about subject and object differential marking, such as the ones developed by Aissen (2003)AISSEN, Judith. Differential object marking: iconicity vs. economy. Natural Language and Linguistic Theory, [S.l.], v. 21, n. 3, p. 435-483, 2003., Butt (2003BUTT, Miriam. Case systems: beyond structural distinctions. In: BRANDNER, Ellen; ZINSMEISTER, Heike (ed.). New perspectives on case theory. Stanford: CSLI Publications, 2003. p. 53-87., 2006)BUTT, Miriam. The dative-ergative connection. In: HOFHERR, Patricia Cabredo; BONAMI, Olivier (ed.). Empirical issues in syntax and semantics 6. The Hague: Thesus, 2006. p. 69-92., Butt and King (1991BUTT, Miriam; KING, Tracy Holloway. Semantic case in Urdu. In: DOBRIN, L.; NICHOLS, L.; RODRIGUEZ, R. M. (ed.). Papers from the 27th Regional Meeting of the Chicago Linguistic Society. Chicago: CLS, 1991. p. 31-45., 2004)BUTT, Miriam; KING, Tracy Holloway. The status of case. In: DAYAL, Veneeta; MAHAJAN, Anoop (ed.). Clause structure in south Asian languages. Dordrecht: Springer, 2004. v. 61, p. 153-198., Comrie (1989)COMRIE, Bernard. Language universals and linguistic typology. 2nd ed. Oxford: Basil Blackwell, 1989., Croft (1988CROFT, William. Agreement vs. case marking and direct objects. In: BARLOW, Michael; FERGUSON, Charles. (ed.). Agreement in natural language: approaches, theories, descriptions. Chicago: University of Chicago Press, 1988. p. 159-179., 1990)CROFT, William. Typology and universals. 4th ed. Cambridge: Cambridge University Press, 1990., Givón (1976)GIVÓN, Talmy. Topic, pronoun and grammatical agreement. In: LI, Charles (ed.). Subject and Topic. New York: Academic Press, 1976. p. 149-188. and Woolford (2000)WOOLFORD, Ellen. Object agreement in Palauan: specificity, humanness, economy, and optimality. In: PAUL, Ileana; PHILLIPS, Vivianne; TRAVIS, Lisa. Formal Issues in Austronesian Linguistics. Dordrecht: Kluwer, 2000. p. 215-245., among others.

The article is subdivided into nine sections. Section “Theoretical assumptions” presents the theoretical backgrounds that will support the analysis. Section “Some relevant grammatical properties of Ka’apor” provides an overview of some grammatical properties of Ka’apor language. Section “What kind of case does ke convey?” demonstrates that ke can in fact be interpreted as a morphological instantiation of inherent dative Case. Sections “The animacy dimension” and “The definiteness dimension” outline the way DOM is triggered when the object is high both in the animacy hierarchy and in the definiteness scale. Section “Differential subject marking” explores the way that DSM is expressed when the subject corresponds to an affected agent in the argument structure of unergative and transitive verbs. Section “What happens when both DSM and DOM compete?” addresses the three grammatical patterns in which both DSM and DOM may cooccur. Section “On the grammatical status of ke” presents an analysis on the grammatical category to which this particle belongs. The final section concludes the paper.

THEORETICAL ASSUMPTIONS

In this article I will be adopting the Differential Case Marking Theory, henceforth DCMT, as is originally proposed by Butt and King (1991BUTT, Miriam; KING, Tracy Holloway. Semantic case in Urdu. In: DOBRIN, L.; NICHOLS, L.; RODRIGUEZ, R. M. (ed.). Papers from the 27th Regional Meeting of the Chicago Linguistic Society. Chicago: CLS, 1991. p. 31-45., 2004)BUTT, Miriam; KING, Tracy Holloway. The status of case. In: DAYAL, Veneeta; MAHAJAN, Anoop (ed.). Clause structure in south Asian languages. Dordrecht: Springer, 2004. v. 61, p. 153-198. and Butt (2003BUTT, Miriam. Case systems: beyond structural distinctions. In: BRANDNER, Ellen; ZINSMEISTER, Heike (ed.). New perspectives on case theory. Stanford: CSLI Publications, 2003. p. 53-87., 2006). The core of this theoretical assumption is that semantic factors seem to be at the root of most case alternations among languages. Butt (2006)BUTT, Miriam. The dative-ergative connection. In: HOFHERR, Patricia Cabredo; BONAMI, Olivier (ed.). Empirical issues in syntax and semantics 6. The Hague: Thesus, 2006. p. 69-92., for instance, assumes that case systems are better understood if one takes semantic parameters into account. “One piece of evidence in favor of this analysis is the fact that there is a tendency among languages to use case alternations, both in subjects and in objects, in order to express semantic contrasts” (Duarte, 2014DUARTE, Fábio Bonfim. On the semantics of affectedness and its implications for argument structure in the Ka’apor language. Revista Linguística, Rio de Janeiro, v. 10, n. 1, p. 99-122, jun. 2014., p. 104). In Urdu, for instance, the semantic interpretation of the subject leads to a differential subject marking, henceforth DSM. For this reason, the ergative Case {=ne} marks that the subject is an agent with control and that it acts on purpose, whereas the dative Case {=ko} indicates that it does not have control over the action. In this sense, Butt (2006, p. 71)BUTT, Miriam. The dative-ergative connection. In: HOFHERR, Patricia Cabredo; BONAMI, Olivier (ed.). Empirical issues in syntax and semantics 6. The Hague: Thesus, 2006. p. 69-92. states that “[…] only the want modality is expressed with an ergative subject, whereas the dative can express both necessity and desire”. Compare the examples below:

In (1), the ergative Case serves to indicate that the subject performs a greater control over the action, whereas the dative Case in (2) denotes that the subject has no control. In this sense, the meaning of volition/wanting is directly obtained when the subject is marked with the ergative Case, whereas the meaning of necessity/desire is achieved by marking the subject with dative. In sum, these examples point out that the ergative is associated with control over an action, whereas the dative is typically associated with goals and experiencers. According to this theory, if an ergative language exhibits a differential subject marking, the ergative Case tends to be used to mark agents, whereas the dative Case is triggered to mark the affected agent.

Ka’apor exhibits a similar pattern as the one above, since the particle ke marks subjects with low control, whereas prototypical agents remain unmarked. However, Ka’apor differs from Urdu and other ergative languages in that it is not an ergative-absolutive language, since it follows a nominative-accusative system. In such a Case system, it is the nominative Case that marks prototypical agents, whereas the dative Case is used on affected subjects to indicate the DSM. Details of this Case alternation system in Ka’apor will be addressed in sections “The definiteness dimension” and “Differential subject marking”.

As to DOM, I will be assuming the proposal advocated by Aissen (2003, p. 435)AISSEN, Judith. Differential object marking: iconicity vs. economy. Natural Language and Linguistic Theory, [S.l.], v. 21, n. 3, p. 435-483, 2003. and by Butt and King (2004)BUTT, Miriam; KING, Tracy Holloway. The status of case. In: DAYAL, Veneeta; MAHAJAN, Anoop (ed.). Clause structure in south Asian languages. Dordrecht: Springer, 2004. v. 61, p. 153-198., according to which, in a differential object Case marking system, the higher in prominence a direct object is, “[…] the more likely it is to be overtly case-marked […]”. In line with this view, there are two dimensions along which prominence is assessed cross-linguistically: the animacy and definiteness dimensions. These dimensions can be expressed by two distinct scales, as follows:

According to Aissen (2003)AISSEN, Judith. Differential object marking: iconicity vs. economy. Natural Language and Linguistic Theory, [S.l.], v. 21, n. 3, p. 435-483, 2003., the scales2 2 Aissen (2003, p. 436-439) posits that higher ranked direct objects tend to be more marked than the lower ranked ones. Also, “If any inanimate objects are case-marked in a language with DOM, then at least some animate objects will be case-marked; […] then at least some definite objects will be marked, etc. […] The intuition is that high prominence which motivates DOM for objects is exactly the prominence which is unmarked for subjects. Thus, it is those direct objects which most resemble typical subjects that get overtly case-marked. Because of the association of subjects with agenthood, on the one hand, and topicality on the other hand, animacy and definiteness are unmarked properties for subjects (Keenan, 1976). But they are not unmarked properties for objects. In fact, they seem to be marked properties for objects, in part perhaps because of the pressure to maximally differentiate subject and object. Thus, exactly what is marked for objects is unmarked for subjects, and vice versa - an instance of what has been termed markedness reversal”. above make clear cross-linguistic predictions. They entail, for instance, that the high position objects tend to be more marked than the ones that occupy the lower positions in the two scales. In line with this, animate, definite or specific objects tend to be more often marked by means of a case marker than the ones that are inanimate, indefinite or non-specific. I will also follow the assumptions of Comrie (1989)COMRIE, Bernard. Language universals and linguistic typology. 2nd ed. Oxford: Basil Blackwell, 1989. and Croft (1988CROFT, William. Agreement vs. case marking and direct objects. In: BARLOW, Michael; FERGUSON, Charles. (ed.). Agreement in natural language: approaches, theories, descriptions. Chicago: University of Chicago Press, 1988. p. 159-179., 1990)CROFT, William. Typology and universals. 4th ed. Cambridge: Cambridge University Press, 1990. and Enç (1991)ENÇ, Mürvet. The semantics of specificity. Linguistic Inquiry, Canada, v. 22, n. 1, p. 1-25, 1991. and Duarte (2014)DUARTE, Fábio Bonfim. On the semantics of affectedness and its implications for argument structure in the Ka’apor language. Revista Linguística, Rio de Janeiro, v. 10, n. 1, p. 99-122, jun. 2014. and Ngunga et al. (2016)NGUNGA, Armindo Saúl Atelela; DUARTE, Fábio Bonfim; CAMARGOS, Quesler Fagundes. Differential object marking in Mozambican languages. In: PAYNE, Doris L.; PACCHIAROTTI, Sara; BOSIRE, Mokaya (ed.). Diversity in African languages. Berlin: Language Science Press, 2016. p. 333-354. DOI: http://dx.doi.org/10.17169/langsci.b121.489.
https://doi.org/10.17169/langsci.b121.48...
that animacy, definiteness and specificity features play a major role regarding the activation of differential object marking across languages. One of the classical examples cited in the literature is the Case alternation in Urdu in the sense that the accusative alternates with nominative in order to encode a semantic distinction related to specificity/animacy. As such, when the accusative morpheme {=ko} appears on the object, it denotes specificity/animacy. This is the situation in the contexts below, where the morpheme {=ko} is a marker of specificity. Note that the occurrence of the ko-marked NP ‘the car’ presupposes that it is known to exist previously in the discourse. Compare the examples (a) and (b) below.

It is important to call to the reader’s attention that the concept of specificity and definiteness does not coincide semantically. Based on this, I will assume, hereafter, that specific DPs may be interpreted as familiar to the speaker, but they are not necessarily definite. In line with this view, the referents of specific D/NPs are ‘weakly linked’ in a previously established conversation, whereas definite D/NPs must be tied to a ‘strongly linked’ discourse.

With this theoretical background in mind, I will propose that the DOM mechanism in Ka’apor is quite similar to the one exhibited by Urdu above, insofar as it is regulated both by the animacy scale and by the definiteness scale. Ka’apor DSM mechanism is also similar to Urdu in that it is constrained by volition and agentivity, on the one hand, and by lack of control, on the other hand. Based on these assumptions, I will be assuming that DOM in Ka’apor is better explained if we adopt a two-dimensional approach in which animacy and definiteness play an important role in determining whether or not objects will be marked by the particle ke within the clause. Before examining the two subsystems, the aim in the next section is to give a general overview of the grammatical aspects of the Ka’apor language that will be relevant to the theoretical analysis to be developed in the subsequent sections.

SOME RELEVANT GRAMMATICAL PROPERTIES OF KA’APOR3 3 I refer the reader to Lopes (2009) and Cabana (2015) for a more comprehensive description on the Ka’apor Language.

In an unmarked situation, one may argue that prototypical agent subjects of transitive and intransitive action verbs do not receive the Case marker ke, as opposed to themes and affected objects that may be marked with this particle. Compare the examples below.

Transitive verb

Inergative verbs

As the data above indicate, agents under control of the action are not marked with the particle ke. However, theme and affected objects, whose referents are humans, specific and definite, may be marked with this particle. Additionally, affected agent subjects that are not a prototypical agent, but an external argument whose θ-role is hybrid in nature, are marked by the particle ke. More to the point, although this argument is selected by an action verb, it is affected by the event. This led Duarte (2014, p. 109) to posit that “[…] this external argument corresponds to what Saksena (1980)SAKSENA, Anuradha. The affected agent. Language, [S.l.], v. 56, n. 4, p. 812-826, Dec. 1980. describes as being the affected-agent in languages such as Hindi […]”.4 4 Saksena (1980, p. 821 apud Duarte, 2014, p. 109) assumes that affected agents “[…] undergo a change of state physically (as in the activity expressed by running) or psychologically (as in the activity of studying). In other words, these agents have some of the properties that one typically expects of patients. These agents are not only doers (performers of their activities) but also doees (recipients of these same activities)”. In the examples below, the presence or absence of [ke] encodes contrasts such as volition/purpose versus necessity/obligation/affectedness.

As to the agreement system, it is observed that only transitive and intransitive subjects trigger agreement on the verb, since objects are not cross-referenced on the verb stem by personal prefixes. In order to facilitate the understanding of the agreement pattern that is established between the subject and the verb in the predicates, Table 1 shows the complete set of the personal pronouns and the subject prefixes that appear on the verb stem. Note that there is no set of agreement affixes for cross-referencing objects.

Table 1
Personal Markers. Source: Duarte (2014, p. 105)

The fact that the set of agreement prefixes above encodes the intransitive and transitive subjects thus signals that verb agreement follows the nominative-accusative alignment. In conclusion, one may argue that transitive and intransitive subjects are cross-referenced by means of the same set of personal markers. Furthermore, observe that the third person prefix is realized by two allomorphs: (u- ∞ Ø-). The allomorph (Ø-) is restricted to contexts where the verb stem is composed of more than one syllable. Compare the examples below:

intransitive

As to the verbal morphological template, it is notable that the subject agreement, the reflexive voice and the causative morpheme precede the verb stem, whereas the negation suffix and the tense suffix follow it. Additionally, no other morpheme may intervene between the root and causative prefix {-mu-}, since they must be adjacent to each other, as the Table 2 below indicates.

Table 2
Linear order of verbal morphemes. Source: Duarte (2014, p. 119), added with the column named suffix 2.

In root clauses, subjects and objects tend to occur before the verb, whereas tense/aspectual particles and auxiliaries obligatorily follow the predicate, thereby producing the [SOV-tense/aspect] and [SOV-Aux] word order possibilities, as is indicated by the following examples.

SOV-aspect

Although the SOV is the more general word order in narratives and spontaneous speech, we also find other linear order possibilities, such as VSO-Aux and SVO-Aux. The reason for this syntactic variation is a topic that needs to be addressed in future research. Compare the examples below.

In temporal/conditional subordinate clauses, the complementizer rahã ‘when, if’ is systematically placed at sentence-final position. As a consequence, it appears after the predicate, thereby producing (at least) two word order possibilities, such as the [SOV-C] and [OSV-C]. In these clauses, the verb and its core arguments precede the complementizer particle, as follows:

After presenting this short overview of some aspects of the Ka’apor grammar, the next sections aim to investigate the intricate system of DOM and DSM. Let us then start with the grammatical status of the particle ke, since its syntactic-semantic distribution will be crucial to determining the differential marking system in root and subordinate clauses.

WHAT KIND OF CASE DOES ke CONVEY?

Taking into consideration the fact that objects and subjects can be identically case-marked with ke, I will follow the analysis presented in Duarte (2014)DUARTE, Fábio Bonfim. On the semantics of affectedness and its implications for argument structure in the Ka’apor language. Revista Linguística, Rio de Janeiro, v. 10, n. 1, p. 99-122, jun. 2014., according to which this particle is a morphological spell-out of an abstract Case that is used to mark animate and definite objects and affected transitive subjects with reduced control. Let us further assert that it corresponds to a dative Case in the sense of Woolford (1997WOOLFORD, Ellen. Four-way case systems: ergative, nominative, objective and accusative. Natural Language and Linguistic Theory, [S.l.], v. 15, n. 1, p. 181-227, Feb. 1997., 2006)WOOLFORD, Ellen. Lexical case, inherent case, and argument structure. Linguistic Inquiry, Canada, v. 37, n. 1, p. 111-130, 2006. DOI: https://doi.org/10.1162/002438906775321175.
https://doi.org/10.1162/0024389067753211...
, as it is semantically oriented, rather than a structural Case. Strong evidence in favor of this proposal comes from the fact that the occurrence of ke is highly predictable, inasmuch as it can be associated with a fixed semantic interpretation. It is typically related to the semantics of animacy and definiteness, when in object position, and to the semantics of affectedness when it appears enclitic to unergative and transitive subjects. Another piece of evidence is that, in addition to marking objects and affected agent subjects, it is also possible to find ke marking goals in ditransitive verbs, as follows:

Based on the data above, one may be tempted to postulate that ke originates as a dative Case, initially marking goal arguments, and that it then extends further to mark theme/patient arguments as well as agents with low control.

Another piece of evidence is the tendency among languages to use case alternations, both with subjects and with internal arguments, in order to express semantic contrasts. This is, for instance, the situation in Urdu, where the dative alternates with the ergative, and in Bengali, where the genitive alternates with the nominative.5 5 Butt (2006, p. 84) proposes that the dative may be interpreted both “[…] as a goal (place) and, in contrast to another case marker, as an agent with reduced control over the action. […]. In Urdu, the dative contrasts with the ergative. In Bengali, the genitive contrasts with the nominative. Given that Bengali has no ergative case, the nominative acts as the default marker for agents; and the genitive in contrast with the nominative indicates reduced control over the action”. For instance, since there is no dative available in Bengali grammar, the genitive is used to cover the meaning of affectedness. Butt (2006, p. 74)BUTT, Miriam. The dative-ergative connection. In: HOFHERR, Patricia Cabredo; BONAMI, Olivier (ed.). Empirical issues in syntax and semantics 6. The Hague: Thesus, 2006. p. 69-92. calls our attention to the fact that “[…] Bengali uses genitive case where other languages tend to employ the dative […]”. Thus, in Bengali, the nominative acts as the default marker for agents, whereas the genitive is used to express an argument that has reduced control over the action, as follows:

Interestingly, the same case alternation appears in Ka’apor, since the dative Case, instantiated by the particle ke, alternates with the unmarked nominative subject, both in unergative and transitive sentences, as repeated below:

Based on these data, it thus seems quite plausible to postulate that Ka’apor exhibits the same case alternation as Hindi, Urdu and Bengali. However, as Ka’apor is not an ergative language like Urdu and Hindi, it will be the unmarked nominative that alternates with the dative to indicate prototypical agents with a high degree of control over the action. Based on these assumptions, Duarte (2014, p. 113)DUARTE, Fábio Bonfim. On the semantics of affectedness and its implications for argument structure in the Ka’apor language. Revista Linguística, Rio de Janeiro, v. 10, n. 1, p. 99-122, jun. 2014. assumes “[…] that the inherent dative Case, which is expressed by the enclitic particle ke in Ka’apor, bears the following semantic interpretations […]”:

Based on the above assumptions, Duarte (2014)DUARTE, Fábio Bonfim. On the semantics of affectedness and its implications for argument structure in the Ka’apor language. Revista Linguística, Rio de Janeiro, v. 10, n. 1, p. 99-122, jun. 2014. points out that, in many languages, distinct syntactic functions are often expressed by homophonous case markers. This occurs in Urdu for example, where the case marker ko is used for marking the dative.6 6 See Butt (2006) for a detailed analysis on Urdu Case system. In other languages, the markers of instrumentals and ergatives or of instrumental and genitives also tend to be form-identical. Based on the empirical data examined thus far and on the differential marking theories, we may postulate that the same situation also holds in Ka’apor, since ke covers different syntactic slots, resulting in a complex system of DOM and DSM. Therefore, it seems quite reasonable to assume that ke is in fact an instantiation of the inherent dative Case that engages in competition with the unmarked nominative and accusative Case. This encoding of a system of semantic contrasts is thus consistent with that in languages such as Hindi, Urdu and Bengali,7 7 Notice that under Butt’s (2006, p. 84-85) proposal, “This is expected as part of language change when new case markers enter the language or engage in competition in a system of semantic contrasts […]”. According to Butt’s theory, “[…] if a Case marker can express both low control (affectedness) and the dimension of place and path, then this Case marker can take over the semantic space of the accusative as well as the dative, thus resulting in homophony of the accusative and dative […]”. among others’.

THE ANIMACY DIMENSION

As to the animacy dimension, recall that subjects tend to be more associated with animate referents than objects. Because of this, animate objects tend to be more marked than animate subjects crosslinguistically, since this semantic property is not an unmarked property for objects. This explains the reason why Ka’apor grammar requires extensive case-marking for human personal pronouns when they are in the syntactic position of direct object. For this reason, I will assume henceforth that one of the roles of DOM in Ka’apor is to disambiguate subject from object in transitive action predicates, especially when they compete for the slots of the core syntactic functions. In such contexts, when the object corresponds to a (human) animate NP, it must be obligatorily case-marked with the particle ke, as follows:

In addition to the contexts shown above, first and second person human pronouns obligatorily have overt case-marking whenever they occur in the object position. Based on this, the generalization seems to be that personal pronouns will be obligatorily case-marked with the particle ke, inasmuch as they are intrinsically definite. This in turn signals that Ka’apor is a language that always case-marks personal pronouns, thereby denoting an intricate DOM system. This grammatical pattern is in line with the theoretical assumption that human NPs and first and second personal pronouns tend to be grammatically marked cross-linguistically, since they outrank all other elements in the animacy and definiteness scale. This in part explains why they are the most susceptible arguments to trigger DOM in Ka’apor. Compare the following examples.

In sum, we can conclude that the two subsystems shown above are directly determined by the animacy scale. The fact that human NPs and personal pronouns require overt marking is, after all, a reflection of the pervasive pressure to maximally differentiate subject and object in Ka’apor, especially when they are animate or first- or second-person pronouns.8 8 I refer the reader to Cabana (2015) for a similar analysis regarding the animacy scale.

THE DEFINITENESS DIMENSION

In contrast to the contexts shown in the previous section, there are cases in which DOM is required in Ka’apor, even when the absence of case marking would not lead to any ambiguity. In such contexts, the object is not high on the animacy scale, but it is highly prominent on the definiteness scale, since its referent is clearly definite in the immediate pragmatic context. Compare the examples (a) and (b) below:

Even human NPs in object position may remain morphologically unmarked, if they are indefinite and non-specific, as the example below illustrates.

Based on the contrast shown above, the generalization we can make is that the particle ke marks only definite objects, regardless of whether they are human, animate or inanimate. Hence, it seems plausible to state that case-marking of definite NPs in Ka’apor is not optional, but obligatory. This in turn explains why the pronouns for first and second persons, as well as [+definite] human, animate and inanimate NPs are all obligatorily case-marked, whereas [-definite] NPs cannot be case-marked, regardless of whether they correspond to animate or inanimate objects. In this sense, Ka’apor differs from Hindi, for example, where case-marking of animates is also possible for indefinites. As such, the presence of the case marker {-ko} is obligatory with human-referring definite objects as well as with human-referring indefinite ones, as follows:

In sum, Ka’apor DOM is largely restricted to contexts in which objects are either first and second person or definite NPs. Therefore, we could make the following generalization about Ka’apor DOM:

DIFFERENTIAL SUBJECT MARKING

In addition to marking objects, it is also possible to find contexts in which the particle ke marks the subject of agentive verbs (in principle, a situation the reader might have thought to be impossible). The examples below illustrate clear cases of differential subject marking. Interestingly, in these examples, the particle ke can become enclitic to the subject of unergative verbs. According to Duarte (2014, p. 109), “In such contexts, the subject does not correspond to a prototypical agent, but to an argument whose θ-role is hybrid in nature. In other words, although the subject is an argument of an action verb, it does display some degree of affectedness […]”. As such, this argument corresponds to what Saksena (1980)SAKSENA, Anuradha. The affected agent. Language, [S.l.], v. 56, n. 4, p. 812-826, Dec. 1980. describes as being the affected agent in languages such as Hindi. Notice that the presence or absence of ke in the examples below serves to encode contrasts such as volition/purpose versus necessity/obligation. Thus, in the (a) example, the meaning is that the subject performed the action with some affectedness; in (40a), it is inferred that something (a stone, a knife, a chair, etc.) might have fallen on Purutu’s foot, so that he did not have a chance to avoid it. However, the agentive meaning is obtained when the subject does not co-occur with the particle ke, as in (40b). In such contexts, since ke is omitted, the affectedness interpretation cannot be inferred.

Similar semantic alternation is also found in transitive constructions. For example, the verb -’u ‘eat’ can select an affected agent or an agent. Then, below in (41a), the subject has control over the action of eating, and as a consequence, ke need not appear. Thus, the action of eating armadillo suggests that the agent does it gladly and without being forced. In (41b), on the other hand, the subject is an affected participant, since in the Ka’apor culture, to eat owl always involves being affected. Thus, sentence (41b) is a clear context where the subject of an action transitive verb, although being an agent, is also affected by the event. As a consequence, it must be marked with ke, in order to encode that the transitive subject is an affected agent, that is, it is not a prototypical agent subject.

Given these data, we can arrive at the conclusion that the affected agents share a common semantic system: they are all the recipients of some causing event and constitute the goal toward which the action is directed. More precisely, these agents have some of the properties that one typically expects of patients and goals, as they are not only agents but also patients of the event represented by verbs such as ‘shout’ and ‘eat’, among others.

Based on the data examined thus far, we can conclude the following: since affectedness is the unmarked property of objects, the subjects of action verbs will always be case-marked with the particle ke when these subjects are not directly associated with agenthood. Notice that a markedness reversal system emerges in such contexts due to the fact that affectedness must be interpreted as an unmarked property for definite objects. Markedness here means that the subject receives a differential morphological marking to encode that, although the subject occupies an external syntactic position of a transitive action verb, it is not a prototypical agent but an affected agent. In line with this, I will propose another scale/dimension in relation to the DSM, since the prototypical agent subject is not case-marked, but the affected/patient agents are. This favors a view of DSM in Ka’apor that is not related to DOM. The reason has to do with the fact that DSM does not make direct reference to multiple dimensions, to both animacy and definiteness, but only to the affectedness dimension. In this sense, the complex case system found in objects that are [+definite]/[+animate] cannot be extended to DSM in Ka’apor. Thus, my hypothesis is that DSM is required in contexts in which the subject is associated with lower agenthood. In other words, whenever subjects of unergative and transitive verbs exhibit the affectedness properties of objects, they will be obligatorily case-marked with ke. Pursuing this line of reasoning, one can assume that DSM in Ka’apor emerges as an example of a markedness reversal, since affectedness is not a typical property of subjects, but of objects. Thus, DSM in Ka’apor is determined by the fact that it is only those subjects which resemble typical patient objects that get overtly case-marked by ke. Thus, we can state that the affected agent subject of unergative and transitive verbs may be overtly case-marked by means of the case marker ke.

WHAT HAPPENS WHEN BOTH DSM AND DOM COMPETE?

In contexts where the subject is an affected agent rather than a prototypical agent, there are at least three grammatical patterns observed.

The first occurs when the subject is higher than the object in the animacy scale. In such contexts, the affected agent subject is obligatorily case-marked by the particle ke. This grammatical possibility is exemplified in the example (42), where the animate subject, being affected by the event of mashing manioc, receives the case-marker ke. Note that the object remains unmarked, since its referent is indefinite and inanimate. As a consequence, one may predict that the DSM emerges when the subject corresponds to an affected agent and is higher than the object in the animacy scale.

NP human / affected / ke   >   NP indefinite / affected /

Additionally, DSM and DOM can simultaneously co-occur in the same sentence in contexts in which the object is definite and the affected agent subject is pronominal. Thus, in the sentences below, although the objects are lower than the subjects on the animacy scale, they are both case-marked by the particle ke. Thus, this semantic property shows that the particle ke can case-mark both the subject and the object simultaneously, thereby producing both DSM and DOM, as is shown in the following sentences:

NP human / affected / human / ke   >   NP definite / affected / non - human / ke

The third pattern involves contexts in which the subject and the object are both high in the animacy scale; for example, both of them could be overtly realized either as first or second person pronouns or as animate human NPs. In such environments, a reversed system emerges in such a way that only the subject is case-marked. Thus, the generalization observed is that subjects and objects cannot both be case-marked when they are prominent in the animacy hierarchy. In such context, the subject will have preference over the object, as follows:

NP human   person   pronoun / affected / ke   >   NP definite / human   person   pronoun /

Note that the semantic interpretation of the subject in the examples above is ambiguous, since it corresponds to an affected agent. Therefore, in (47), the subject is affected because the action of making someone get up involves physical effort.

In sum, one can conclude that it is the lack of the association with prototypical agenthood that leads to the occurrence of DSM system in Ka’apor. In this sense, affectedness is not the unmarked property for transitive subjects; from this fact we derive an explanation for why affected agent subjects of verbs of activity are systematically case-marked with the particle ke. This thus explains why affected agent subjects are formally marked with the same particle that is used to encode objects. The reason for the emergence of this differential subject and object marking has to do with the fact that affectedness is the unmarked property for objects, whereas definiteness and agenthood is the unmarked pattern for subjects.

ON THE GRAMMATICAL STATUS OF ke

Based on distribution of ke presented thus far, one may conclude that this item belongs to the class of Case particles due to the following reasons: (i) it is syntactically dependent on the noun phrase over which it has grammatical scope and (ii) it can only occur in a post-nominal position, usually after adjectives and plural markers. Evidence in favor of this proposal relates to the fact that it has a fixed position within the D/NP, a property that clearly differs ke from affixes, adverbs and quantifiers. This is reinforced by the fact that it cannot be moved to other syntactic position nor be floated away from the noun phrase that it modifies. In this sense, I will claim that ke is a functional particle with clear semantic denotation and a fixed position in the sentence. Thus, it cannot be categorized as an inflectional affix, since grammatical inflection in Ka’apor usually occurs at the left position in the morphological template of noun roots. This is reinforced by the fact that the morphemes related to person agreement in nouns systematically occur as prefixes. This is the situation below, in which the prefix {r} encodes that the possessor is immediately adjacent to the noun root, whereas the prefix {h-} indicates that the third person possessor has been omitted from the noun phrase, as follows:

Another piece of evidence regards the fact that adjectives and plural particles can interpolate between ke and the noun, a situation that clearly signals that the canonical word order is {Noun+Adjective+Plural Marker+ke} within the nominal domain. This in turn indicates that this particle really forms a phonological word with the NP to which it phonologically adjoins. Compare the examples below in which the head of the NP co-occurs with the adjective panɛm ‘unlucky’ and the plural marker ta:

Note that, when the particle ke appears in the phrases above, it must be obligatorily following the adjective and the quantifier, thereby signaling that it must follow the noun, the adjective and the plural marker, as is exemplified below:

Additionally, when the universal quantifier upa ‘all’ combines with the nominal phrase above, it must be positioned after ke, thereby indicating that it occupies a higher position in the extended functional structure of the D/NP, as follows.

In sum, one may conclude that, as ke occupies an intermediate position in the phrases above, it clearly indicates that it cannot be analyzed as a suffix, but as a Case particle, since affixes cannot occur far away from the root it must adjoin to. More to the point, adjectives and plural marker particles cannot interpolate between affixes and the stem.

Based on these empirical assumptions, I will propose that ke is a morphological realization of the abstract Case that transitive subjects, direct objects and goals receive in the course of the syntactic derivation. Following Bittner and Hale’s (1996)BITTNER, Maria; HALE, Ken. The structural determination of case and agreement. Linguistic Inquiry, Canada, v. 27, n. 1, p. 1-68, 1996. and Baker’s (2012)BAKER, Mark. On the relationship of object agreement and accusative case: evidence from Amharic. Linguistic Inquiry, Canada, v. 43, n. 2, p. 255-274, 2012. DOI: http://dx.doi.org/10.1162/LING_a_00085.
https://doi.org/10.1162/LING_a_00085...
theory, according to which KP is the Case functional extension of D/NP projection, I will assume that the particle ke heads such a Case projection. In line with this theory, I will propose that NPs marked with the enclitic particle ke exhibit the syntactic structure depicted in the diagram below:

CONCLUDING REMARKS

In conclusion, this paper shows that the DOM mechanism in Ka’apor is regulated both by the animacy scale and by the definiteness scale. This phenomenon is better explained if we adopt a two-dimensional approach, according to which animacy and definiteness play an important role in determining whether or not objects will be marked within clauses. In this regard, Ka’apor grammar requires extensive case-marking for personal pronouns and for human non-pronominal objects. Thus, one of the roles of DOM in Ka’apor is to disambiguate subject from object in transitive action predicates, especially when they compete for the syntactic slots within the vP argument structure. As such, only definite objects are case-marked, regardless of whether they are human, animate or inanimate. Thus, it seems plausible to propose that case-marking of definite NPs is obligatory in Ka’apor rather than optional. This in turn explains why the local persons (first and second), definite human/animate NPs and definite inanimate NPs are case-marked by ke, whereas indefinites cannot be case-marked, regardless of whether they correspond to either animate or inanimate objects.

Additionally, it is shown that the particle ke can mark the subject of agentive verbs whenever it does not correspond to a prototypical agent, but instead to an argument whose theta role is hybrid in nature. More to the point, one can conclude that, since affectedness is the unmarked property of objects, subjects will be always case-marked with the particle ke, when they are not directly associated with agenthood. A further conclusion is that DSM occurs as an example of a markedness reversal, since affectedness is typical of objects but not of subjects. In other words, the affected agent subject of unergative and transitive verbs will always be overtly case-marked by means of the case marker ke. Furthermore, this particle can be interpreted as a morphological spell-out of an abstract Case that is used to mark animate and definite objects and affected transitive subjects with reduced control.

In line with this theory, I assume that, due to its multifunctional status, the particle ke corresponds to a dative Case in the sense of Woolford (1997WOOLFORD, Ellen. Four-way case systems: ergative, nominative, objective and accusative. Natural Language and Linguistic Theory, [S.l.], v. 15, n. 1, p. 181-227, Feb. 1997., 2000WOOLFORD, Ellen. Object agreement in Palauan: specificity, humanness, economy, and optimality. In: PAUL, Ileana; PHILLIPS, Vivianne; TRAVIS, Lisa. Formal Issues in Austronesian Linguistics. Dordrecht: Kluwer, 2000. p. 215-245., 2006)WOOLFORD, Ellen. Lexical case, inherent case, and argument structure. Linguistic Inquiry, Canada, v. 37, n. 1, p. 111-130, 2006. DOI: https://doi.org/10.1162/002438906775321175.
https://doi.org/10.1162/0024389067753211...
in that it is a semantically oriented Case, assigned to core arguments in the v-VP domain. In sum, it is a type of inherent Case that fits well in the paradigm of inherent Case theory as proposed originally by Woolford (2006)WOOLFORD, Ellen. Lexical case, inherent case, and argument structure. Linguistic Inquiry, Canada, v. 37, n. 1, p. 111-130, 2006. DOI: https://doi.org/10.1162/002438906775321175.
https://doi.org/10.1162/0024389067753211...
, since this Case is directly related to the following semantic properties:

  1. affectedness of subject of unergative and transitive verbs;

  2. definiteness/animacy of the referent of the argument in object position.

ABREVIATIONS

  • 1
    In this paper, abstract Case will be spelled out with capital letter, while morphological case will remain with lowercase letter. See details on the difference between abstract Case and morphological case in Ura (2001)URA, Hiroyuki. Case. In: BALTIN, Mark; COLLINS, Chris (ed.). The handbook of contemporary syntactic theory. Oxford: Blackwell Publishing, 2001. p. 334-373..
  • 2
    Aissen (2003, p. 436-439)AISSEN, Judith. Differential object marking: iconicity vs. economy. Natural Language and Linguistic Theory, [S.l.], v. 21, n. 3, p. 435-483, 2003. posits that higher ranked direct objects tend to be more marked than the lower ranked ones. Also, “If any inanimate objects are case-marked in a language with DOM, then at least some animate objects will be case-marked; […] then at least some definite objects will be marked, etc. […] The intuition is that high prominence which motivates DOM for objects is exactly the prominence which is unmarked for subjects. Thus, it is those direct objects which most resemble typical subjects that get overtly case-marked. Because of the association of subjects with agenthood, on the one hand, and topicality on the other hand, animacy and definiteness are unmarked properties for subjects (Keenan, 1976KEENAN, Edward. Towards a universal definition of subject. In: LI, Charles (ed.). Subject and Topic. New York: Academic Press, 1976. p. 303-333.). But they are not unmarked properties for objects. In fact, they seem to be marked properties for objects, in part perhaps because of the pressure to maximally differentiate subject and object. Thus, exactly what is marked for objects is unmarked for subjects, and vice versa - an instance of what has been termed markedness reversal”.
  • 3
    I refer the reader to Lopes (2009)LOPES, Mário Alexandre Garcia. Aspectos gramaticais da língua ka’apor. Tese (Doutorado em Linguística) – Universidade Federal de Minas Gerais, Belo Horizonte, 2009. and Cabana (2015)CABANA, Nasle Maria. Marcação diferencial dos argumentos nucleares na língua Ka’apor. 2015. Tese (Doutorado em Linguística) – Universidade Federal de Minas Gerais, Belo Horizonte, 2015. for a more comprehensive description on the Ka’apor Language.
  • 4
    Saksena (1980SAKSENA, Anuradha. The affected agent. Language, [S.l.], v. 56, n. 4, p. 812-826, Dec. 1980., p. 821 apud Duarte, 2014DUARTE, Fábio Bonfim. On the semantics of affectedness and its implications for argument structure in the Ka’apor language. Revista Linguística, Rio de Janeiro, v. 10, n. 1, p. 99-122, jun. 2014., p. 109) assumes that affected agents “[…] undergo a change of state physically (as in the activity expressed by running) or psychologically (as in the activity of studying). In other words, these agents have some of the properties that one typically expects of patients. These agents are not only doers (performers of their activities) but also doees (recipients of these same activities)”.
  • 5
    Butt (2006, p. 84)BUTT, Miriam. The dative-ergative connection. In: HOFHERR, Patricia Cabredo; BONAMI, Olivier (ed.). Empirical issues in syntax and semantics 6. The Hague: Thesus, 2006. p. 69-92. proposes that the dative may be interpreted both “[…] as a goal (place) and, in contrast to another case marker, as an agent with reduced control over the action. […]. In Urdu, the dative contrasts with the ergative. In Bengali, the genitive contrasts with the nominative. Given that Bengali has no ergative case, the nominative acts as the default marker for agents; and the genitive in contrast with the nominative indicates reduced control over the action”.
  • 6
    See Butt (2006)BUTT, Miriam. The dative-ergative connection. In: HOFHERR, Patricia Cabredo; BONAMI, Olivier (ed.). Empirical issues in syntax and semantics 6. The Hague: Thesus, 2006. p. 69-92. for a detailed analysis on Urdu Case system.
  • 7
    Notice that under Butt’s (2006, p. 84-85)BUTT, Miriam. The dative-ergative connection. In: HOFHERR, Patricia Cabredo; BONAMI, Olivier (ed.). Empirical issues in syntax and semantics 6. The Hague: Thesus, 2006. p. 69-92. proposal, “This is expected as part of language change when new case markers enter the language or engage in competition in a system of semantic contrasts […]”. According to Butt’s theory, “[…] if a Case marker can express both low control (affectedness) and the dimension of place and path, then this Case marker can take over the semantic space of the accusative as well as the dative, thus resulting in homophony of the accusative and dative […]”.
  • 8
    I refer the reader to Cabana (2015)CABANA, Nasle Maria. Marcação diferencial dos argumentos nucleares na língua Ka’apor. 2015. Tese (Doutorado em Linguística) – Universidade Federal de Minas Gerais, Belo Horizonte, 2015. for a similar analysis regarding the animacy scale.

ACKNOWLEDGEMENTS

I would like to thank two anonymous reviewers of Boletim do Museu Paraense Emílio Goeldi. Ciências Humanas, who generously offered their constructive critiques which contributed greatly to improving this article. An earlier version of this paper was presented at Differential Object Marking Conference, which held at the University of Tromsø, in May 23-24, 2013 and at the Conference on Semantics of Under-Represented Languages in the Americas (SULA) 7, which was held at Cornell University, May 4-6, 2012. The research represented here has been funded by CAPES-Brazil (grant # 88881.120815/2016-01); by FAPEMIG (grant # 19901); by CNPq (grant # 302674/2009-8) and by the Pró-Reitoria de Pesquisa da Universidade Federal de Minas Gerais (PRPq/UFMG). Most part of it was developed further during my stay as a visiting scholar in the Department of Linguistics at University of Toronto, during the years 2017-2018. Home pages: www.letras.ufmg.br/fbonfim; www.letras.ufmg.br/laliafro. I take full responsibility for all possible errors in the content of this paper.

REFERENCES

  • AISSEN, Judith. Differential object marking: iconicity vs economy. Natural Language and Linguistic Theory, [S.l], v. 21, n. 3, p. 435-483, 2003.
  • BAKER, Mark. On the relationship of object agreement and accusative case: evidence from Amharic. Linguistic Inquiry, Canada, v. 43, n. 2, p. 255-274, 2012. DOI: http://dx.doi.org/10.1162/LING_a_00085.
    » https://doi.org/10.1162/LING_a_00085
  • BITTNER, Maria; HALE, Ken. The structural determination of case and agreement. Linguistic Inquiry, Canada, v. 27, n. 1, p. 1-68, 1996.
  • BUTT, Miriam. The dative-ergative connection. In: HOFHERR, Patricia Cabredo; BONAMI, Olivier (ed.). Empirical issues in syntax and semantics 6 The Hague: Thesus, 2006. p. 69-92.
  • BUTT, Miriam; KING, Tracy Holloway. The status of case. In: DAYAL, Veneeta; MAHAJAN, Anoop (ed.). Clause structure in south Asian languages Dordrecht: Springer, 2004. v. 61, p. 153-198.
  • BUTT, Miriam. Case systems: beyond structural distinctions. In: BRANDNER, Ellen; ZINSMEISTER, Heike (ed.). New perspectives on case theory Stanford: CSLI Publications, 2003. p. 53-87.
  • BUTT, Miriam; KING, Tracy Holloway. Semantic case in Urdu. In: DOBRIN, L.; NICHOLS, L.; RODRIGUEZ, R. M. (ed.). Papers from the 27th Regional Meeting of the Chicago Linguistic Society Chicago: CLS, 1991. p. 31-45.
  • CABANA, Nasle Maria. Marcação diferencial dos argumentos nucleares na língua Ka’apor 2015. Tese (Doutorado em Linguística) – Universidade Federal de Minas Gerais, Belo Horizonte, 2015.
  • CALDAS, Raimunda Benedita. Uma proposta de dicionário para a língua Ka’apor 2009. Tese (Doutorado em Linguística) – Universidade de Brasília, Brasília, 2009.
  • CALDAS, Raimunda Benedita. Aspecto, modo de ação e modalidade na língua Ka’apor 2001. Dissertação (Mestrado em Letras: Linguística e Teoria Literária) - Universidade Federal do Pará, Belém, 2001.
  • COMRIE, Bernard. Language universals and linguistic typology 2nd ed. Oxford: Basil Blackwell, 1989.
  • CROFT, William. Typology and universals 4th ed. Cambridge: Cambridge University Press, 1990.
  • CROFT, William. Agreement vs case marking and direct objects. In: BARLOW, Michael; FERGUSON, Charles. (ed.). Agreement in natural language: approaches, theories, descriptions. Chicago: University of Chicago Press, 1988. p. 159-179.
  • DUARTE, Fábio Bonfim. On the semantics of affectedness and its implications for argument structure in the Ka’apor language. Revista Linguística, Rio de Janeiro, v. 10, n. 1, p. 99-122, jun. 2014.
  • ENÇ, Mürvet. The semantics of specificity. Linguistic Inquiry, Canada, v. 22, n. 1, p. 1-25, 1991.
  • GIVÓN, Talmy. Topic, pronoun and grammatical agreement. In: LI, Charles (ed.). Subject and Topic New York: Academic Press, 1976. p. 149-188.
  • KAKUMASU, James. Urubu-Ka’apor. In: DERBYSHIRE, Desmond; PULLUM, Geoffrey (ed.). Handbook of Amazonian languages New York: Mouton de Gruyter, 1986. v. 1, p. 326-403.
  • KEENAN, Edward. Towards a universal definition of subject. In: LI, Charles (ed.). Subject and Topic New York: Academic Press, 1976. p. 303-333.
  • KLAIMAN, M. H. Bengali dative subjects. Lingua, [S.l], v. 51, n. 4, p. 275-295, Aug. 1980.
  • LOPES, Mário Alexandre Garcia. Aspectos gramaticais da língua ka’apor Tese (Doutorado em Linguística) – Universidade Federal de Minas Gerais, Belo Horizonte, 2009.
  • NGUNGA, Armindo Saúl Atelela; DUARTE, Fábio Bonfim; CAMARGOS, Quesler Fagundes. Differential object marking in Mozambican languages. In: PAYNE, Doris L.; PACCHIAROTTI, Sara; BOSIRE, Mokaya (ed.). Diversity in African languages Berlin: Language Science Press, 2016. p. 333-354. DOI: http://dx.doi.org/10.17169/langsci.b121.489.
    » https://doi.org/10.17169/langsci.b121.489
  • SAKSENA, Anuradha. The affected agent. Language, [S.l], v. 56, n. 4, p. 812-826, Dec. 1980.
  • SILVA, Tabita Fernandes. Classes verbais e algumas questões pragmáticas em Ka’apor Dissertação (Mestrado em Linguística) – Universidade Federal do Pará, Belém, 2001.
  • URA, Hiroyuki. Case. In: BALTIN, Mark; COLLINS, Chris (ed.). The handbook of contemporary syntactic theory Oxford: Blackwell Publishing, 2001. p. 334-373.
  • WOOLFORD, Ellen. Lexical case, inherent case, and argument structure. Linguistic Inquiry, Canada, v. 37, n. 1, p. 111-130, 2006. DOI: https://doi.org/10.1162/002438906775321175.
    » https://doi.org/10.1162/002438906775321175
  • WOOLFORD, Ellen. Object agreement in Palauan: specificity, humanness, economy, and optimality. In: PAUL, Ileana; PHILLIPS, Vivianne; TRAVIS, Lisa. Formal Issues in Austronesian Linguistics Dordrecht: Kluwer, 2000. p. 215-245.
  • WOOLFORD, Ellen. Four-way case systems: ergative, nominative, objective and accusative. Natural Language and Linguistic Theory, [S.l], v. 15, n. 1, p. 181-227, Feb. 1997.

Publication Dates

  • Publication in this collection
    02 Dec 2019
  • Date of issue
    Sep-Dec 2019

History

  • Received
    12 Apr 2019
  • Accepted
    21 Oct 2019
MCTI/Museu Paraense Emílio Goeldi Coordenação de Pesquisa e Pós-Graduação, Av. Perimetral. 1901 - Terra Firme, 66077-830 - Belém - PA, Tel.: (55 91) 3075-6186 - Belém - PA - Brazil
E-mail: boletim.humanas@museu-goeldi.br