I’ve done a bit more reading on this, and I think I may have figured out an answer to this myself. I’ll start by listing the properties of topics, as given by Li and Thompson:bradrn wrote: ↑Wed Jun 17, 2020 7:32 am I’ve been reading Li and Thompson’s 1976 article about subject-prominent and topic-prominent languages. In it, they distinguish four types of language:
In the article, they discuss types (i) and (ii) quite extensively. However, they don’t seem to comment on how languages of type (iii) or (iv) work. Does anyone have any more information on this?Li & Thompson wrote: According to our study, there are four basic types of languages: (i) languages that are subject-prominent … (ii) languages that are topic-prominent; (iii) languages that are both subject-prominent and topic-prominent; (iv) languages that are neither subject-prominent nor topic-prominent.
- The topic always has the same role in the sentence, namely ‘specifying the domain within which the predication holds’
- The topic does not necessarily need to be an argument of the predicate
- The topic does not play a role in grammatical processes
- Topics are always definite
- Topics are always in sentence-initial position
- The subject must be an argument of the predicate
- The subject plays a role in grammatical processes such as reflexivisation, passivisation, Equi-NP deletion, verb serialisation and imperativisation
- Subjects may be indefinite
- Topic-comment sentences are usually more basic (less marked)
- They have ‘double subject’ constructions (where a topic is followed by a subject)
- They have extensive zero-anaphora (Li & Thompson don’t explicitly say this, but Huang (2000) does), typically controlled by the topic rather than the subject
- They lack a passive (or have only a marginal passive)
- They lack dummy subjects
- The unmarked word order is typically verb-final
- They lack ‘double subject’ constructions
- They have a passive
- They may have dummy subjects
- The topic (when specified) may be greatly constrained to a specific set of choices
As for type (iv) languages (neither Tp nor Sp): using the definitions above, and reading a bit about them, I conclude that these are languages which do not consistently mark either topic or subject. The examples Li and Thompson give are all Philippine languages, which makes sense: their crazy voice systems mean that there isn’t any consistent way to define a ‘subject’ (see Schachter for details), and (like English) they don’t have any surface coding for topics either. (There is something which usually gets called the ‘topic’, but it doesn’t satisfy the criterion listed above of ‘specifying the domain within which the predication holds’.)