![Sample our Language & Literature journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5941/TOC-Subject-Sample-2021-Language-Literature.jpg"})
Abstract
This paper considers the nominal group in Sundanese, a Malayo-Polynesian language of West Java, Indonesia, through Systemic Functional Linguistics. In particular, it builds a meaning-based description of Sundanese nominal groups, born of educational concerns associated with developing literacy programs. To this end, it describes not only the formal syntagms at play, but also their functions; not only the paradigmatic choices that are available, but how they are taken up in text; and not only the grammar by itself, but how it realizes meanings in discourse. The result is a richly co-textualized and metafunctional description of Sundanese nominal groups that is illustrated through texts from a range of genres and registers. It first steps through the nominal group textually, detailing the resources available for tracking participants and for organizing informational prominence. Next it explores the nominal group interpersonally, considering the attitudes and graduation nominal groups realize and the relative social relationships between speakers. Finally, it considers the nominal group ideationally, exploring its resources for organizing experience and the ways it elaborates entities and metaphorically realizes events and occurrences. Together, this builds a detailed view of Sundanese nominal groups that can aid the ongoing development of literacy programs in Indonesia.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
1 The authors wish to thank J.R. Martin and Dongbing Zhang for their extensive comments, as well as David Rose for his detailed interpretations of Sundanese textual meaning in particular.
2 https://systemiclanguagemodelling.wordpress.com/glossing/ Note that the Leipzig morpheme by morpheme glossing is not an SFL analysis – it is simply a gloss used as a lingua-franca for ease of reading, especially for those from outside SFL.
3 Text from: https://www.youtube.com/watch?v=DuKlO75y2yc
4 Here we are calling these embedded clauses as verbal groups can be inserted between the adjective and the linker. However this is only a tentative analysis subject to revision as descriptions of Sundanese develop.
5 In comparison to English, from the perspective of discourse semantics Sundanese does not make a clear-cut distinction between presuming reference, where an entity is marked as being known or recoverable, and presenting reference, where an entity is marked as being not yet known or recoverable (Martin Citation1992). Rather, the main distinction is between entities that are identified in some way – most typically presuming, but also including indefinite or generic marking – and those that have no explicit identification (i.e., they are unmarked). As we will see below, non-identified entities may or may not be recoverable and can occur as a first or subsequent mention.
6 This text, being colloquial spoken Sundanese, involves regular lexical borrowing and some code-switching with Indonesian and Javanese (and at times Dutch and English); this does not affect our discussion as the grammar typically stays Sundanese. When code-switching occurs, the words will not be broken down grammatically unless they have Sundanese affixation or other Sundanese grammatical patterns.
7 As we are not developing a full discourse semantic description of identification in this paper, we will leave aside other types of presuming reference, such as anaphoric bridging (Martin Citation1992) where a participant can be presumed on the basis of their relation with a previous entity or event, e.g., He was stabbed last night. The knife was found in an alleyway. Bridging does of course occur in Sundanese, as suggested by the embedded clause examples (3) and (4) above.
8 The -na/-ana distinction is a morphophonemic variation. -ana occurs when the base it is attached to ends in a nasal, -na occurs elsewhere. This suffix is relatively generalized in its meaning, and so will be glossed as either a definite marker or third person marker depending on the co-text. In addition, it is regularly used for nominalizations and as a linker for possessive nominal groups.
9 There is a question here as to why the suffix -na/ana would be used for such presuming reference rather than, say, a demonstrative Deictic, discussed below. The most likely answer is that the suffix backgrounds the presumption, textually speaking, whereas the use of a demonstrative Deictic likely adds informational prominence. With thanks to David Rose for this interpretation.
10 Urang literally means ‘person’, but is typically glossed as first person plural ‘us’/’we’ (Hardjadibrata Citation2003, 870). Here, however, the speaker is using it to refer to herself in the singular ‘I’. Referential variation in Sundanese pronouns and demonstratives is not uncommon. For example, Müller-Gotama (Citation2001) notes that Eringa (Citation1984, 726) reports the first person singular form kuring ‘I’ is at times used as a plural form ‘us’/’we’. Similarly, our corpus includes a number of instances where demonstratives are used to refer to people (ieu panasaran, ‘I was curious’ lit. ‘this (was) curious’). This highlights significant textual and interpersonal variation that is not easily captured by simple distinctions in person and number.
11 Mwinlaaru (Citation2021) makes a similar distinction for Dagaare, but groups the functions realized by possessives and demonstratives (as well as definite articles) as subtypes of Deictic – definite articles are labeled as Deictic1, possessives are labeled as Deictic2 and demonstratives are labeled as Deictic3. Here we label Mwinlaaru’s Deictic3 as the Deictic (for both those that come before and after the Thing) and Deictic2 as a Possessor. The term Possessor is preferred to Orient, as used by Martin and Shin (Citation2021) for Korean and Zhang (Citation2021) for Khorchin Mongolian, as the Sundanese function is limited to possession.
12 Elliptical nominal groups and embedded clauses are left out of the network here, as they are strictly speaking not a type of nominal group. For similar reasons, the suffixal -na/ana indicator for presuming reference is not included as it is a choice at the word rank.
13 The # before the β indicates that the construction is not iterative – i.e., it is not a serially expanding univariate structure in Halliday’s (Citation1965) terms. Rather, these structures are duplexes, involving only two elements (Rose Citation2001a, Citation2022). The verb that can be used is limited to only a small set, and there is no grammatical or morphological variation that is similar to a clause. For this reason, it is not analyzed as an embedded clause, but rather as a subjacency duplex.
14 Although we will not pursue it here, this co-selection between lemes and kasar forms (interpretable as a speech level prosody), suggests that an interpersonal tier of structure may be appropriate for Sundanese nominal groups (as well as for clauses).
15 As we are not developing a full description of the textual grammar on Sundanese here, we have left out the conjunctions jeung and lamun from our analysis, and have just focused on the participants realized by nominal groups.
16 C.f. Matthiessen (Citation2004, 641) who points out that in English, procedures are a rare type of text where a non-Theme participant is regularly elided: the Goal in imperatives – Cut squid in 2” pieces … ; pat ø very dry on paper towel. Dip ø in flower, then egg, and then in crumbs. As this text shows, Sundanese can also elide the Goal in imperatives but prefers to make it Theme in a receptive declarative.
17 (74) could alternatively be analyzed as the mah marking the entirety of the clause, as a dependent β clause that has been made thematic (see Halliday and Matthiessen Citation2014, 549–554 for similar patterns in English clause complexes). The full clause complex is: batan lésot mah, éta jiret kalah ka beuki meulit ‘rather than loosening mah, the snare only twisted more’. As we are not developing a full description of textual grammar here, we will not pursue this further.
18 The alternative would be to consider these markers as realizing a multivariate function (Halliday Citation1965) within a group; but this would mean the embedded clause would also need to realize a multivariate function within a group. However, the type of group the embedded clause would then occur in and what function it would play is not clear. Indeed, such an analysis would lose the notion that embedded clauses are chosen precisely as an alternative to choosing a group.
19 This is a non-defining relative clause which does not allow for thematic variation. The ‘subject’ of the embedded/relative clause must be the element it is linked to, in this case abdi (me) (see section 4).
20 With thanks to David Rose for this insight.
21 Similarly, we should not expect that any marker such as téh will perform only a single function in text. For example, when it occurs on the final element in a clause, whether this be a nominal group or otherwise, téh often appears to indicate higher-order periodicity – either a Hypertheme or Hypernew (Martin and Rose Citation2007; a similar pattern is noted by Rose [Citation2001b] across languages). This is in fact illustrated in clause (21) from the fable replayed below, where the final element, an embedded clause, is marked by téh. This clause functions as the opening clause of the narrative’s Solution and realizes a Hypertheme that both reintroduces the mouse as the main character and establishes the evaluative prosody of courage that flows through the whole stage:
In other texts, téh marks Hypernews that synthesize the point of a previous section. This illustrates that we should not look for the ‘meaning’ of any single item on their own, but rather consider them relationally as resources for building text (Matthiessen Citation2015).
22 Hanafi (Citation1997) additionally considers it a topic maker because it often occurs early in the clause and is followed by information that has not yet been introduced (i.e., ‘new’ information). We will not be following up this interpretation here.
23 This is a reformulation – the speaker begins by saying kulit urang mah ‘my skin’, with the mah marking the whole nominal group, before reformulating it to be urang mah kulitna téh, ‘I mah, the skin téh’, specifically focusing on the urang ‘I’ as the marked New.
24 Unfortunately publicly available Sundanese corpora are comprised of isolated and constructed individual clauses (Kjartansson et al. Citation2018; Sodimana et al. Citation2018), or instances taken from texts but listed alphabetically, rather than in the order of the text (Leipzig Corpora Collection Citation2017); this makes study of this particle and other textual features difficult.
25 To account for this, a tier of textual structure may be useful when analyzing Sundanese nominal group, which can perhaps complement the prosodic interpersonal tier suggested above. In particular, it seems that the textual metafunction co-opts the logical complexing in order to give prominence to specific elements in the nominal group. This can be compared to Martin’s (Citation1990, 33–36) suggestion that Tagalog draws on logical resources for interpersonal purposes.
26 In addition, various references list a range of alternate forms for a number of the cells in the pronoun paradigm in (e.g., Van Syco Citation1959; Hardjadibrata Citation2003). It is illuminating to note that as far as we can tell, other than Van Syco (Citation1959) no English language reference grammar (e.g., Müller-Gotama Citation2001; Hardjadibrata Citation1985; Hanafi Citation1997, Robins' various papers) has put forward a full pronoun paradigm. Similarly, Sundanese language teaching resources (e.g., Hodson Citation1952; Anderson Citation1996) introduce many of the pronouns in as well as a range of others but in a relatively unordered fashion.
27 Though there are many Epithets that occur as the heads of adjectival groups, realizing Attributes in attributive clauses.
28 Here we are making a sharp distinction between Epithets and Qualifiers based on their use or not of the linker nu/anu as well as their position and other possibilities for realization (see Section 4.2 below). But this sharp distinction misses the similarities between evaluative or descriptive Epithets and Qualifiers, especially when their only realizational distinction is the linker. Although we won’t pursue it here, a possible analysis is to have the Epithet conflated with the Qualifier in examples such as those in , to emphasize this similarity.
29 This is similar to the Measurer in Mandarin Chinese (Wang Citation2020, 66–68). But unlike in Chinese, the standard of measurement, here kilo, is not obligatory. For this reason, we will use the label Numerative.
30 For this reason, a stricter interpretation of these structures might preferably interpret them as a word complex within the Thing, rather than as Thing^Classifiern structures in the nominal group, as suggested by Martin, Doran, and Zhang (Citation2021) for English and illustrated in Martin and Shin (Citation2021) for Korean and Stosic (Citation2021) for Serbian. However, for ease of reading and use in appliable settings, and since no fully developed word grammar is yet available in contemporary SFL, we will leave these as Thing^Classifier structures for now.