Introduction to the Special Issue on “Possession in the languages of Wallacea”

1 Linguistic Wallacea

Located in what is now eastern Indonesia and East Timor, Wallacea has long been recognized by naturalists and biologists as a transitional area between the Asian and Oceanic biological areas. Zoogeographic Wallacea falls between the ancient continents of Sunda and Sahul, bounded to the west by the Wallace Line, which separates the islands of Sulawesi and Lombok from the Philippines, Borneo, and Bali, and to the east by Lydekker’s Line, running between New Guinea and Australia on the one hand and insular Southeast Asia on the other. The Weber Line, denoting the boundary between Asian and Australasian types for most plants and vertebrate animals, runs through the center of the zone. Linguistic Wallacea likewise constitutes a transitional zone between the Southeast Asian linguistic type and the Melanesian linguistic type. Linguistic Wallacea, described and argued for in detail by Schapper (2015, 2017, is shifted slightly to the east of Biological Wallacea, excluding Lombok and Sulawesi but adding in much of northwest New Guinea, as far as the eastern coast of Cenderawasih Bay. Linguistic Wallacea includes both Austronesian and Papuan languages, and comprises much of the overlap of two larger proposed linguistic areas: Linguistic Melanesia, whose western boundary largely coincides with that of Wallacea but which extends southeast as far as New Caledonia and Fiji (Schapper 2020), and the Mekong-Mamberamo macro-area stretching from the Mekong River in mainland Southeast Asia to the Mamberamo River in central-western New Guinea (Gil 2015). Map 1 shows Linguistic Wallacea relative to the Wallace, Weber, and Lydekker Lines.

Map 1:

Linguistic Wallacea in context (adapted from Schapper 2017).

Linguistic Wallacea includes Austronesian languages from the traditional Central Malayo-Polynesian and the South Halmahera-West New Guinea subgroups. These are spoken alongside Papuan languages from the Greater West Bomberai, North Halmahera, East, West, and South Bird’s Head, Hatam–Mansim, Yawa-Saweru, and Mairasic families, as well as a number of poorly described smaller groups and isolates. (See Schapper and Gasser, this volume, for a brief discussion of subgrouping.) While the languages of the area are both genetically and typologically very diverse, longstanding language contact has led to extensive convergence in many areas of morphosyntax and lexical semantics. The most widespread and well-documented of these are semantic alignment (Donohue 2004) and neuter gender (Schapper 2010), which are shared between unrelated languages across the region; synchronic metathesis (Schapper 2015) and colexification of do and give (Gil 2017; Reesink 2013) have more limited distributions but likewise cross family boundaries. Like any other language area, the boundaries of Linguistic Wallacea are fuzzy: Features typical of Linguistic Wallacea can spill into surrounding languages, while not all languages in Linguistic Wallacea necessarily have a given feature.

This volume focuses on grammatical systems of possession in Linguistic Wallacea. The Austronesian languages of the region are known for having possessive systems that differ from those typical of Western Austronesian in a number of ways, most notably in their Possessor–Possessum word order and alienability distinctions (Donohue and Schapper 2008; Himmelmann 2005; Schapper and Gasser, this volume). The idea that Papuan languages are responsible for many of the innovative features in eastern Austronesian languages is nothing new (see, e.g., Klamer et al. 2008; Reesink 1998 for overviews of features of (parts of) the area). The Wallacean development of Possessor–Possessum word order in place of the conservative Austronesian Possessum–Possessor order has been attributed to contact with Papuan languages at least since Brandes (1884). Papuan influence on Austronesian systems is not limited to possession in Linguistic Wallacea: Other features that are typical of Papuan languages and also found in eastern Austronesian languages include prefixal agreement on verbs, neuter gender, and the absence of (morphological) voice. That such ‘Papuan’ features arise in Austronesian languages across the same region as that in which the innovative possessive features emerge offers additional evidence that the Austronesian languages in the region were subject to substratal Papuan influence. That said, while the direction of grammatical influence is primarily Papuan-to-Austronesian, it is not exclusively so. For example, Arnold (this volume) proposes that split inalienable possessive systems were spread via Biak, an Austronesian language, though their ultimate origin is less clear. Overall, the picture is one of broad inter- and intra-family structural convergence, both on a macro (areal) and micro (local) scale.

2 This issue

This issue is the result of a workshop held at the KITLV, Leiden in December 2019. A group of scholars working on languages of Linguistic Wallacea came together to discuss issues around grammatical possession in the region. Papers in this issue represent a combination of presentations developed out of the workshop and additional contributions solicited subsequently.

The first article in the volume, by Antoinette Schapper and Emily Gasser, is a survey of adnominal possessive systems in 85 Austronesian and Papuan languages of Linguistic Wallacea, providing a broad areal context for the studies that follow. By focusing solely on possession rather than a broader grammatical typology and expanding the number of languages considered, this survey is able to develop a more granular picture of possessive constructions in Linguistic Wallacea languages than hitherto. The main features discussed are patterns in adnominal possessive word orders, locus of possessive marking, possessive classification systems, and multifunctionality of adnominal possessive markers. These languages show several globally rare patterns, including a widespread use of contrastive direct/indirect possessive constructions to encode alienability, found broadly across the area, as well as zero-marked possessive constructions and possessive-like attribute constructions with a more limited distribution. The authors draw attention to the many subregional patterns in possessive constructions that are of typological interest and seek to understand them as the result of diachronic change, looking especially at the role of contact. Across related languages, they look for changes in historically related form-function pairs in the possessive domain and use these in order to posit progressive historical changes in the morphosyntax of possessive constructions of Wallacean languages.

Laura Arnold’s article describes a type of differential possessive marking not previously recognized in the theoretical or typological literature. Languages with Split Inalienable Coding (SIC) have two or more possessive coding strategies that are closely or exclusively associated with expressing inalienable possession. Arnold considers a sample of 189 languages spoken in and around Linguistic Wallacea, comprising 19 families and seven isolates, and finds that 13 of them have semantically-based Split Inalienable Coding distinguishing body part nouns from kin terms. She argues that Austronesian languages are predisposed to develop SIC, due to a structurally distinct nominal subclass of kin terms inherited from an earlier stage of Austronesian, and that contact has played an important role in its spread, especially in Northwest New Guinea. Arnold suggests that because the paradigmatic differences involved in SIC are so subtle, it is easy to overlook and is likely to be more widespread in the world’s languages than is currently recognized.

The following article by Antoinette Schapper adds to the growing body of typological research on attributive constructions in which adjectival attributes are observed to be marked as if they were possessed nouns in possessive constructions. Schapper looks at the historical relationship between attributive and possessive constructions in the Papuan languages of the Timor-Alor-Pantar family and shows that in one language, Teiwa, the possessive marker has extended from marking attributes to restrictive relative clauses. This finding is contrary to some of the existing literature on the diachronic sources of relative clause markers that suggest that despite widespread possessive-relative identity in the world’s languages, relative clause markers do not originate in possessive markers. In establishing a new diachronic pathway between possessive and relative clause markers by way of attributive markers, Schapper’s article illustrates the continuing importance of data from genealogically and structurally diverse languages for understanding grammaticalization processes.

Moving eastwards, David Gil analyses the possessive paradigm used in Roon for non-paired body parts, which exhibits considerable inter- and intra-speaker variation. This paradigm is unusual, as it is the only place in the grammar where contrastive tone is found. He argues that the tonal contrast and spatial deictic semantics developed diachronically from the grammaticalization of a High Extended Intonation contour common to many languages of the region. This study highlights the striking paradigmatic variation that can appear in even a small language community when spoken under conditions of contact and bilingualism, and demonstrates how data from prosody and syntax can be usefully combined in synchronic and diachronic analyses.

Finally, Linn Iren Sjånes Rødvand gives a description of the possessive system of Patani, a heretofore largely undescribed Austronesian language of South Halmahera. This article constitutes the first in-depth description of the possessive system of a South Halmahera language with the typologically unusual edibility distinction. It thus fills an important gap in our understanding of the South Halmahera-West New Guinea subgroup. Patani differs from its closest well-documented neighbour, Taba, in maintaining a morphologically distinguished alienability contrast. To this it adds a further distinction whereby the alienable class of nouns is divided into two categories based on edibility, each marked by a different possessive classifier. Alimentary possession is well-documented in the Oceanic languages and attested in Wallacea, but is highly unusual elsewhere. Finally, Rødvand describes predicative possession in Patani, a category of constructions that are sorely under-described in Linguistic Wallacea.