Following Dryer’s (2005j) WALS chapter, this chapter looks at the order of relative clause and noun, as well as at some special, less widely known relative-clause types.
For this chapter, a relative clause is defined as a clause that helps narrow the reference of a noun (the head) and in which the referent of the noun head has a semantic role. (Headless relative clauses are left aside here.) Relative clauses are most often clauses which occur adjacent to the head noun, as seen in examples (1) and (2). (In these and the other examples in this chapter, the head noun is given in boldface, and the relative clause is enclosed in brackets in the gloss line.)
Only seven languages in APiCS have relative clauses of other types, as described in §4-6 below. We distinguish the following five types (a very rare sixth type that Dryer 2005j recognizes does not occur in APiCS).
|Relative clause follows noun||64||8||72|
|Relative clause precedes noun||2||3||5|
|Internally-headed relative clause||0||1||1|
|Correlative relative clause||1||2||3|
|Adjoined relative clause||0||3||3|
Relative clause constructions are further described in Chapters 92-94 from the point of view of the role of the head noun in the relative clause (subject relative clauses, object relative clauses, instrument relative clauses). This aspect is not considered in this chapter.
In the great majority of APiCS languages, the relative clause is adjacent to the head noun and follows it, as in (1) above and in (3).
That this should be overwhelmingly the dominant type in our languages is not surprising, because it is overwhelmingly the dominant type in the indigenous languages of Africa, Europe and Southeast Asia, as well as in Austronesian languages (Dryer 2005j). In APiCS, there are only three languages which do not have postnominal relative clauses (all spoken in Asia: Sri Lankan Malay, Sri Lanka Portuguese, and Chinese Pidgin Russian).
As the lexifiers lack prenominal relative clauses, this construction must be due to the Tamil substrate/adstrate in both languages (note that the WALS map for Sri Lanka only shows Sinhala, which has postnominal relative clauses).
Another Asian language with dominant prenominal order is Singapore Bazaar Malay, as seen in (2) above. In this language, the prenominal order must be due to substrate influence from the Chinese languages of the Bazaar Malay speakers.
Prenominal relative clauses are also found in two mixed languages in the Americas, following the patterns of the indigenous languages:
Internally-headed relative clauses have not been widely known until fairly recently. They are clauses which are not adjacent to the notional head, but contain it inside them. They occur especially in the languages of North America, but also sporadically elsewhere throughout the world. In APiCS, this type occurs only marginally in one language, Ternate Chabacano. Postposed relative clauses are much more common in this language, but (6) shows an internally-headed relative clause.
Three APiCS languages have correlative relative clauses, where the head occurs inside the relative clause together with a relative marker, and which are taken up by a resumptive demonstrative-like element in the main clause. Thus (7) from Pidgin Hindustani is literally ‘Which camp they stayed at, that (was) dirty’.
Correlative relative clauses are particularly well-known from Indo-Aryan languages (Dryer 2005j), so in APiCS we find them in Hindustani-lexified Pidgin Hindustani and in Portuguese-based Korlai (spoken in India, showing strong Indo-Aryan adstrate influence). The presence of the construction in Chinese Pidgin Russian may seem a bit more surprising, but similar constructions are actually found in colloquial Russian.
Three languages have adjoined relative clauses, i.e. relative clauses which do not occur adjacent to the head noun and are not specially marked as relative clauses. That they help narrow the reference of the head noun must be inferred from the context. Adjoined relative clauses are particularly well-known from Australian languages, so we find them in Kriol, but also in Early Sranan and Michif.
We must admit here that the classifications in §4-6 are not particularly certain. Internally-headed relative clauses are often difficult to recognize, and correlative relative clauses could be confused with simple cases of left dislocation of the head noun together with the relative clause, as in (13).
Adjoined relative clauses are the least well-defined. The case of Early Sranan could simply be regarded as a case of relative clause extraposition, for example. The clearest case of an adjoined relative clause is the Kriol example in (10), where the subordinator we is not specific to relative clauses.