adjacent vs. verbal 'only'

[Martin Haspelmath]

Peter,

I think the contrast between (adjacent) tol'ko/tik and (verbal) only/te- 
is an important one, and the grammatical typology of such focus 
quantifiers is a fascinating topic that is just barely beginning to be 
explored. So I look forward to seeing your work on Lithuanian.

But I'm not so sure that the contrast between "affixal" and 
"non-affixal" only-expressions is going to be very interesting, because 
we don't have good criteria for telling affixes apart from words. 
Couldn't Lithuanian tik be regarded as an affix as well, or te as a free 
word? Apparently the main reason why tik is not considered an affix is 
that it is non-selective (or "promiscuous") with respect to its host, 
i.e. it can occur next to words of diverse categories, whereas te only 
occurs next to verbs. But that is best described as an adjacency 
difference (focus quantifier is adjacent to its focus or occurs in a 
fixed position close to the verb), not an indepenent difference in 
"affixhood". Note that many of the "suffixes" in the Gast & van der 
Auwera database would have to be considered words, not suffixes, by the 
non-selectivity criterion (it seems that the main criterion for 
classifying them as suffixes is the spelling).

Incidentally, another well-known language that has a (partly) verbal 
'only'-marker is French:

Ivan ne lit que des livres.
'Ivan only reads books (and nothing else).'

In French, the 'only' marker is bipartite, with one verbal part (ne) 
(which is often argued to be affixal), and one adjacent part (que) 
(which is rarely argued to be affixal, due to its non-selective character).

Martin

peterarkadiev schrieb:
> Dear colleagues,
>
> I would like to inquire whether you could give me examples of languages exhibiting an affixal expression of meanings similar to English 'only'. Let me explain what I mean.
>
> 1. In Russian, 'only' is expressed by a particle _tol'ko_, which must precede the constituent it has scope over, thus
>
> (1a) Tol'ko Ivan chitaet knigi 'Only Ivan (and nobody else) reads books'
> (1b) Ivan chitaet tol'ko knigi 'Ivan reads only books (and nothing else)'
> (1c) Ivan tol'ko chitaet knigi 'Ivan only reads books (and does nothing else)', but not *'Ivan reads only books'
>
> Are there languages where 'only' in sentences like (1a) and (1b) would be expressed as an affix on the noun?
>
> 2. In English, as is well known, the particle _only_ can appear in the preverbal position while having narrower scope (depending on intonation, as far as I may judge from the literature):
>
> (1d) John only reads books 
> - 'John reads books and does nothing else'
> - 'John reads books and reads nothing else'
>
> Are there languages where 'only' in sentences like (1d) would be expressed as a genuine affix on the verb? More importantly, I am especially interested in situations when such affix would allow readings similar to the second interpretation of (1d), i.e. with the narrow scope.
>
> I know of one such language, namely Lithuanian (not an 'exotic' language, as it were). Here there are two expressions of 'only': one is a particle _tik_ in all relevant respects similar to the Russian _tol'ko_, and the other is a prefix _te-_, which, notably, shows scopal properties more similar to those of the English _only_ (i.e. it allows almost any type of constituent to be in its scope, including the subject, the direct and indirect objects, obliques, and even can scope into the embedded clause with some matrix predicates). Now I am looking for typological parallels to this situation.
>
> Many thanks in advance!
>
> Best wishes,
>
> Peter Arkadiev
> Institute of Slavic Studies
> Russian Academy of Sciences
>
>
>