relative strandability of adpositions

Steven Schaufele fcosw5 at MAIL.SCU.EDU.TW
Mon Oct 26 22:55:53 UTC 1998

Bingfu Lu wrote:

> Tsunoda et al. (1995: 759) observe that adposition stranding is found
> in a fair number of prepositional languages, but rarely found in post-
> positional languages.  In 130 languages they investigated, stranding
> is explicitly attested in 8-10 % of the prepositional languages, but
> dubiously only 2% of the postpositional languages.  The stranding phe-
> nomenon suggests that postpositions are more bound than prepositions in
> general.  This may be a by-phenomenon of the tendency that 'postposed
> grammatical material is more likely to affix than preposed grammatical
> materials' (Bybee 1990:3).  In other words, a postpositional bound
> morpheme appears more bound than its semantically counterpart
> prepositional morpheme.

10 years ago i presented a paper at the South Asian Languages Analysis
Roundtable (and i may even be able to find a copy of it somewhere around
here somewhere!) in which i examined the syntactic repercussions of the
traditional (i.e., mostly Delbruck, who took his cue from discussions of
similar issues in Classical Greek grammar) distinction between
`authentic' and `inauthentic' ("echte" vs. "unechte") adpositions in
Sanskrit, particularly with regard to Vedic, the earliest recorded stage
in the language's history. In my paper, i reported that (1)
`inauthentic' adpositions (i.e., adpositional/adverbial lexemes
transparently derived from nominal stems) were less exclusively
postpositional that `authentic' ones (i.e., adpositions apparently
inherited as such from PIE, without any evident derivation from any
other lexemes), (2) even when postpositional, `inauthentic' adpositions
were more likely to be stranded then `authentic' ones, (3) nevertheless,
Vedic Sanskrit had certain common syntactic phenomena (e.g.,
topicalization) that could easily result in the stranding of any
adpositional lexeme at all.

Be it noted that none of these facts survive into the Classical
language. Indeed, the whole notion of the distinction between
`authentic' and `inauthentic' adpositions appears to be completely
meaningless synchronically in any stage of Sanskrit's history after the
end of the Vedic period (ca. 1000 B.C.E.).  On the other hand, as some
of my colleagues in Sanskrit studies have commented, there is some
serious question as to the extent to which post-Vedic Sanskrit should be
thought of as having been a `natural'/`living' language.


Steven Schaufele, Ph.D., Asst. Prof. of Linguistics, English Department

Soochow University, Waishuanghsi Campus, Taipei 11102, Taiwan, ROC

(886)(02)2881-9471 ext. 6504     fcosw5 at

        ***O syntagmata linguarum liberemini humanarum!***

        ***Nihil vestris privari nisi obicibus potestis!***

More information about the Lingtyp mailing list