[Lexicog] polysynthetic languages and dictionaries

jess tauber phonosemantics at EARTHLINK.NET
Tue Jun 1 17:51:33 UTC 2004


Funny you should ask..., but last summer I attended Mark Baker's short course on polysynthesis at the LSA Linguistic Institute in Michigan. Working on a "bipartite" language myself (Yahgan) with the usual instrument/bodypart action/manner prefix and pathway/locational suffix suspects, and always looking for the "big picture" typologically to fit all my weird hypotheses into, I approached Prof. Baker about the observations I had while paying close attention to what he was saying about incorporation, head-marking, etc.

Here are some of the things I noticed: Most Bakerian polysynthetic languages are basically head-marking, with scant case marking of core relations. Conversely, many of the "bipartite" types have substantial case marking. There may be some sort of implicational hierarchical relation regarding the mix of head/dependent marking and how "freshly minted" the "bipartite" construction is- language which have drifted heavily towards lexicalized formations may be the ones which show more head marking, while the newly inducted types (with more open sets of incorporable terms) may prefer dependent marking. This is something I plan to investigate more fully in my next life, if there is time.

As for incorporation, there is also an "inversion" here as well, in that one tends to see demoted nominals in "classical" polysynthesis, indirect objects, circumstantial but not central space/time notions, etc. With the "bipartite" types on the other hand, nominals might be seen as promoted instead- one could think of instrument/bodypart terms as indirect "subjects", and the pathway/location terms are often quite dynamic- the instruments often have causal force, for instance. There is an adverbial flavor to the entire system.

So we have then here the incorporation of lower ranked nominals into the verb, or the incorporation of higher ranked adverbials into the verb (whether or not these adverbials are themselves originally nouns, they have adverbial force by the time they get incorporated, or serialized, compounded as the case may be). A mirror image setup.

Finally, one could see the sendoff, in Bakerian polysynthesis,  of normal nominals into "adjunct-land" as part of this amazing symmetry. It would be interesting then if it were the case that in "bipartism" normal adjuncts start being treated as "nominals".

This might help explain why Bakerian polysynthetic languages have very few ideophones, and little augmentative sound symbolism, but the "bipartite" types have so many, and so much.

In any case the above is an idealization. The real data is far messier, so if the hypothetical symmetry is real then languages  can be in flux between the extreme states, just as is true of with constituent ordering. And by the way, the young "bipartite" languages appear to prefer to be left-headed, the old ones right, while the reverse may hold true for polysynthetic languages. When I discussed these notions with Prof. Baker, he seemed politely bemused, as befits someone who probably realized that the person he was listening to had not likely taken their medication that day.

There is method to the madness, though, even if the true situation is somewhat more complicated. While many languages have left-side instrument and right-side path terms, there are languages which reverse this (such as reconstructed Kartvelian, which you normally wouldn't think of as being this way- a lot of those mysterious "stem extenders" historical linguists seem not to know what to do with appear to be such terms). Secondly, there can also be semantic inversion (where the bodypart terms are patientlike, and the paths may have causal (or possibly more, permissive) force. And these can have either order flanking the regular verb root(s). Muskogean is like this. Such differences may affect preferences for constituent ordering and alignment, but need further investigation.

Jess Tauber
phonosemantics at earthlink.net


------------------------ Yahoo! Groups Sponsor --------------------~-->
Make a clean sweep of pop-up ads. Yahoo! Companion Toolbar.
Now with Pop-Up Blocker. Get it for free!
http://us.click.yahoo.com/L5YrjA/eSIIAA/yQLSAA/HKE4lB/TM
--------------------------------------------------------------------~->


Yahoo! Groups Links

<*> To visit your group on the web, go to:
     http://groups.yahoo.com/group/lexicographylist/

<*> To unsubscribe from this group, send an email to:
     lexicographylist-unsubscribe at yahoogroups.com

<*> Your use of Yahoo! Groups is subject to:
     http://docs.yahoo.com/info/terms/



More information about the Lexicography mailing list