TOPIC, FOCUS and relatives. (Longish)

[Helge Dyvik]

Two points, one about existentials and relatives, the other about 
TOPIC vs. FOCUS:


1. Existentials and relatives

I may not have fully understood the point about existentials. Why 
would the assumption that the relativized element is a TOPIC lead us 
to expect that the demoted subject cannot be relativized? Demoted 
subjects certainly can be topicalized; cf. Norwegian (where 
existentials occur with a wider range of verbs than in English):

("I am looking for some newspapers. Do you have any?")
Aviser     ligger det        på bordet
newspapers lie    there.SUBJ on the-table (NB! Not interpretable as 
"Newspapers lie *there*, on the table")
'As for newspapers, there are some lying on the table'

As for Joan's ill-formed examples:

  #Someone who there is being interviewed...
  ??It's JOHN who there is being interviewed.

it seems to me that their ill-formedness is accounted for by the 
referential/scope-related properties of the correlate (someone/John), 
which are not compatible with the restrictions on the demoted subject 
in existentials. True, one can say "There is someone being 
interviewed", but crucially 'someone' is in the scope of 'there' in 
that sentence. Non-referential bare nouns, and narrow-scope-inducing 
'any', on the other hand, seem (admittedly to a non-native speaker) 
to work well:

"John is hopeless. He said he couldn't find ice cream, which there is 
in the freezer, or beer, which there is in the fridge."
...any mistakes which there may have been... (Dick Hudson)


2. TOPIC vs. FOCUS

Further examples which suggest that relatives pattern with topic 
constructions come from topic prominent languages. In a topic 
prominent language like Vietnamese - and actually also in more topic 
prominent stages of a Germanic language like Old Norse - we find a 
topic-comment construction where the relationship between topic and 
comment is such that the topic does not necessarily fill a syntactic 
function within the comment. In Vietnamese, for example, we may have:

Cây  này  lá   lón quá
tree that leaf big exceed, i.e. "That tree the leaves are too big"
'As for that tree, the leaves are too big.'

In Old Norse (Old Icelandic + Old Norwegian) we sometimes find the 
same phenomenon:

Thæssa luti sem nu hæfir thu spurda tha ma ec thec æigi sannfrodan 
gøra til fullz
these things that now have you asked then may I you not knowledgable make fully
'As for the things you have asked about now, I am not able to make 
you fully informed'

In both languages relative clauses display the same phenomenon: we 
find relative clauses in which there is no 'gap' corresponding to the 
head of the NP:

Môt cái cây  mà  lá  lón  quá...
one CL  tree REL leaf big exceeding...
'a tree whose leaves are too big'

Minnzt thærs at marghr lifir sa litla rid er længi lifir athæfi hans 
æptir hann
remember that that many lives he little while REL long lives 
reputation his after him, i.e. "Remember that many a man lives a 
little while that his reputation lives long after him"

The fact that such relatives seem to cooccur with such topic-comment 
constructions suggests that the basic relation is the same: both the 
topic and the NP head "limit the applicability" of the predication of 
the comment and the relative clause, respectively, "to a certain 
restricted domain" (to paraphrase Wallace Chafe's old 
characterization of the topic relation).


Helge Dyvik
--------------------------------------------------------------------
Helge J. Jakhelln Dyvik
Department of Linguistics and Comparative Literature
Section for Linguistic Studies
University of Bergen    	   Phone: 	  +47 55582261
Sydnesplass 7   	           Fax:   	  +47 55589354
N-5007 Bergen, Norway              E-mail: helge.dyvik at lili.uib.no
--------------------------------------------------------------------