Feature 33. Order in appositions

found in question(s): 86a, 86b, 86c

Definition and illustration

Hackstein (2003) proposes a correlation of OV/VO word order with the internal order of nominal appositions, such as King Charles. For the purpose of this database, nominal appositions are understood as the juxtaposition of two independent nominals (words or phrases) that address the same referent but whose relation is not overtly marked. Usually one of the two items is a generic or schematic noun (e.g., king) while the other is a more specific noun or name (e.g. Charles). These will be abbreviated in the following as Nschematic and as Nspecific, respectively. There are two possible word orders:

(1)

a) Nschematic-Nspecific

b) Nspecific-Nschematic

For instance, King Charles is an example for Nschematic-Nspecific order. Note that cases like Standard Mandarin fùmǔ 父母 'parents' should be excluded. In this case, two nouns are juxtaposed, but both are denoting different referents that are are located on the same level of specificity (fù 'father', mǔ 'mother'). We also exclude compounds of the following sort:

(3) Turkish

Türk dil-i

Turkish language-3sg.poss

'Turkish language'

Languages are not necessarily uniform in their appositions, which makes a differentiation into different types necessary. In order to test the correlations, this database includes three different types of appositions. These are titles, such as King George (Question 86a), personal relations, such as my friend Tom (Question 86b), and toponyms, such as Danube River (Question 86c). The titles and toponyms preferrably exclude cases that are borrowed in their entirety from other languages, such as the name Genghis Khan in English or Cengizhan in Turkish. Also excluded should be place names that consist of a noun or name with a modifier like an adjective, such as Rio Grande (literally 'large river') or Mont Blanc (literally 'white mountain').

The three different types are illustrated in the following with Thai and Nepali. In Thai, a Tai (Kra-Dai) language with VO order, there is a mixed pattern with the specific element following in the title (4) and the toponym (6) but preferrably preceding in the relational expression (5).

(4) Thai

a) kasàt/phrácâw cɔ́ɔt

king/hon George

b) *cɔ́ɔt kasàt/phrácâw

George king/hon

'King George'

(5) Thai

a) ?phʉ̂an (khɔ̌ɔŋ) chǎn thom

friend (of) 1sg Tom

b) thom phʉ̂an (khɔ̌ɔŋ) chǎn

Tom friend (of) 1sg

'my friend Tom'

(6) Thai

a) mɛ̂ɛnáam daanúup

river strong

b) *daanúup mɛ̂ɛnáam

Danube river

'River Danube'

In Nepali, an Indo-Aryan (Indo-European) language with OV word order, there is also a mixed pattern. In the title (7), the specific element follows, as it preferrably does in the relational expression (8). However, it necessarily precedes in the toponym (9).

(7) Nepali

a) radza dzərdz

king George

b) *dzərdz radza

George king

'King George'

(8) Nepali

a) mero sathi ram

1sg.gen friend Ram

b) ?ram mero sathi

Ram 1sg.gen friend

'my friend Ram'

(9) Nepali

a) sali nədi

Sali river

b) *nədi sali

river Sali

'Sali River'

In some languages, there are item-specific differences in the available order of the appositions. This sensitive to lexical items can be observed in Turkish. In this language, the title kral 'king' (ultimately from West Germanic) usually precedes the name while the titles sultan 'sultan' (ultimately from Arabic) or han (from Mongolic) preferrably follow. The title padişah 'ruler' (ultimately from Persian) allows both word orders.

Correlations

Based on a small sample of languages, Hackstein (2003: 132) observes “that OV languages characteristically postpose the apposition while VO languages prepose it.” This can be formalized as a testable hypothesis as follows:

(2)

a) if VO then Nschematic-Nspecific

b) if OV then Nspecific-Nschematic

Since English is a language with VO order, the example King Charles is in accordance with the prediction. Note that appositions are a feature that cannot be described in terms of a head-dependent relationship between the two elements. They do not stand in a hierarchical relationship to each other (Payne 2006: 322). Instead, Hackstein (2003: 148) proposes that the correlation in (2b) is based on a semantic relationship:

"A term with a more specific meaning (e.g. apple) is followed by a term with a comparatively less specific or generic meaning (e.g. fruit), in other words, the hyponym is followed by its hyperonym. The hyperonym semantically includes its hyponym. The information flow, i. e. the disambiguation of the sense of the noun phrase, runs from right to left, from the relatively more general to the relatively more specific information. The hyponym is the semantic subordinate of the hyperonym."

References

Author(s)TitleYearPublished in
Hackstein, OlavApposition and word-order typology in Indo-European.2003Brigitte L.M. Bauer & Georges-Jean Pinault (eds.), Language in time and space, 131-152. Berlin: De Gruyter Mouton.
Bauer, Brigitte L. M.Nominal apposition in Indo-European: Its forms and functions, and its evolution in Latin-Romance.2017Berlin: De Gruyter Mouton.
Hackstein, OlavApposition and nominal classification in Indo-European and beyond.2010Wien: Verlag der Österreichischen Akademie der Wissenschaften. [ordered in Stabi Berlin]
Payne, Thomas E.Exploring language structure.2006Cambridge: Cambridge University Press.