Notes on current work
(1) I have been adding new material from a number of Outlier languages. The most notable has been from Albert Davletshin (Moscow), who has recently returned from fieldwork on Nukeria (Nuguria), an atoll north of Bougainville in Papua New Guinea where one of the most poorly documented Polynesian languages is spoken. The existing material in POLLEX was from Ray (1916) based on late 19th and early 20th century sources. Albert has generously allowed me to add material from his unpublished vocabulary to POLLEX, and I have naturally found connections with Takuu (Richard Moyle’s recent dictionary) and Sikaiana (Bill Donner’s dictionary, now online at http://www.sikaianaarchives.com/), as well as revisiting Anne Salmond’s (1975) vocabulary of Luangiua and finding many additional reflexes. Of this closely related group of Outliers, only Nukumanu now remains almost totally undocumented.
For various reasons I have also recently “revisited” West Futuna (Capell 1974) and West Uvea (Hollyman 1987), correcting small errors and adding much that had earlier been overlooked. (In the process -- and with apologies to non-French-speaking users -- I have re-converted many of Hollyman’s glosses from English to French. I felt that the glosses ought to be consistent, reflecting the published source. Translating them all will be a project for some future date.)
(2) I have also been re-structuring the “Notes” panel of the Reconstruction page. This provides information about both internal and external relations of reconstructed Polynesian words. The information is structured in several different fields labelled with *N, where N is a digit from 0 to 8.
The fields in the revised system are as follows. Fields 0 and 1 give information about relations among different reconstructed forms within Polynesian:
0: This is meant to show relations of filiation (descent), where one reconstructed form has developed into another, whether through an irregular sound change or through morphological processes (prefixation, suffixation, compounding). I have tried to show the connections both ways. For example, the note under PCE goio.1a “Common Noddy” shows:
0 << PN gogo.1
which means that PCE goio.1a has developed (by an irregular sound change) from PPN gogo.1. The corresponding notation under PPN *gogo.1 is:
0 >> CE goio.1a
The double chevrons indicate the direction of development. The lack of glosses in this example means that the meaning has not changed.
1: Notes in this field begin with “Cf.”, and give cross-references to other reconstructed forms which may be related in some way – either through some form/meaning resemblance or (in some cases) just synonymy. I use this field when the relation between the forms is less than certain; and also when their morphological relation is pre-Polynesian. (So although PPN fana and ma-fana, both meaning “hot”, are clearly related, both the prefixed and the simple form have existed at least since Proto-Oceanic, so they are simply cross-referenced here.) In fields 0 and 1, I have eliminated the P (standing for “Proto-“) before starred forms. This is in part just to simplify the display of information, but also because some of the “level” codes now in use do not refer to postulated subgroups, but simply to distribution of reflexes. Thus “XO” means that reflexes occur in several Outlier languages (and not elsewhere), but I do not propose the existence of an “XO” subgroup or a corresponding “PXO” proto-language. Likewise, since 0 and *1 are purely Polynesian-internal, I use no level codes higher than PN in citing reconstructions. So a reconstruction which shows “OC” or “MP” in the general listing will simply be “PN” in these fields.
Fields 2 - 7 give information about reconstructed forms which are ancestral to the Polynesian word. Authors of these external reconstructions are usually identified in parentheses. ). Starred forms (reconstructions) are identified as belonging to proto-languages such as PCP (Proto Central Pacific), POC (Proto Oceanic), PAN (Proto Austronesian).
The reconstructions proceed by successively higher levels (larger language groups
*2: Fijic and Central Pacific (PFJ, PCP) -- Polynesian plus Fijian and/or Rotuman
*3: Eastern Oceanic (PEO) – including languages of Vanuatu, New Caledonia and Micronesia as well as Central Pacific; and sometimes Remote Oceanic (PROC) which adds the languages of the Southeast Solomons.
*4: Oceanic (POC). Many PO reconstructions cited here are from The Lexicon of Proto-Oceanic, a series of volumes edited by Malcolm Ross, Andrew Pawley and Meredith Osmond, published by Pacific Linguistics: Volume 1, Material Culture (1998), Volume 2, Physical Environment (2003), Volume 3, Plants (2008), Volume 4, Animals (2011) – at least two more volumes are to follow. Page references to these books are given as LPO followed by volume and page number.
*5: Malayo-Polynesian (PMP)
*6: Austronesian (PAN)
*7: Collateral groups not directly ancestral to Polynesian, e.g. Proto-Micronesian, Proto-Western Malayo-Polynesian.
Finally, *8 includes miscellaneous comments
Those not familiar with the Austronesian family and its subgroups might want to refer to Ethnologue’s enormous list: http://www.ethnologue.com/subgroups/austronesian or Wikipedia articles such as http://en.wikipedia.org/wiki/Oceanic_languages http://en.wikipedia.org/wiki/Austronesian_languages and the print sources referenced therein.
Ross Clark 03-02-2014
In my previous notes, I neglected to mention a very important source of information about the earlier (pre-Polynesian) history of these words: Robert Blust's monumental Austronesian Comparative Dictionary. Although still very much a work in progress, this has been made into a highly useful web site by Steve Trussel (http://www.trussel2.com/acd/). Many reconstructions (mainly PAN and PMP) in the Notes panel are now marked (ACD) to refer to this site -- much more convenient than having to reference Blust's numerous scattered publications of the last 40 years. For those with access to the journal Oceanic Linguistics, Blust and Trussel have published a paper in the latest issue explaining in detail the origins of the ACD and how the site is organized.
I am currently adding cognates from Richard Moyle's newly published Takuu Grammar and Dictionary (Pacific Linguistics 634, 2011 -- see http://pacling.anu.edu.au/ for publication and ordering information). The dictionary is rich in ethnographic detail, and the electronic version on the accompanying CD includes numerous photos and video clips to illustrate various entries.
On the subject of electronic marvels, I should take the opportunity to mention Bill Donner's new Sikiana Archives site (http://www.sikaianaarchives.com/). Bill is an anthropologist who worked on Sikaiana in the early 1980s. He produced a fine small Sikaiana-English dictionary, which unfortunately was published in very limited numbers in Honiara. My photocopy of it disappeared a while ago, so it was just in time that the new site arrived, with the entire dictionary accessible to all online. There is also a wealth of cultural and historical information, photos, songs, etc.
Putting in cognates from these two sources often involves removing (overwriting) entries from sources which at one time were our mainstays for these languages: for Takuu, the unpublished vocabulary by Irwin Howard from the 1970s; for Sikaiana, field notes by Peter Sharples for his M.A.thesis (1968).
I feel some regret at effacing the contributions of these pioneers; but to avoid excessive duplication here, I follow a policy of entering a given word only once, and that from the most accessible (not necessarily the earliest) source. Two or more sources are cited only where they provide significantly different evidence on the form or meaning.
Finally, a note on the glosses given here. Ideally, it would be good to give the full entry for each word cited from the source. For practical reasons, however, I often edit them. Ellipsis (...) in the gloss indicates that potentially significant information has been omitted. This has not been consistently applied throughout, however, and those interested in the semantic nuances and extensions in a given language should always seek out the source for fuller information.
Dear Friends of POLLEX,
Even though most of you would not have known him personally, I wanted to pass along the news that Peter Ranby died last Wednesday in Howick (Auckland).
As a student at the University of Auckland in the 1960s, Peter was part of the team which produced the first published version of POLLEX (Walsh & Biggs 1966). He did field work on the Nanumea dialect of the Tuvalu language, and a description of its syntax was his M.A.thesis (1973). Another product of this research was his Nanumea Lexicon (Canberra, 1980), in which he provided an etymology for each word - in the process proposing many new reconstructions.
After a time in Wellington, he returned to Auckland in the 1980s, where he taught Linguistics 306 (Comparative Polynesian Linguistics) for some years. He would lead his students on cognate-hunting expeditions through the lesser-known reaches of the Outlier world, which resulted in still more new reconstructions. An exchange of papers with Bruce Biggs in the journal Te Reo (1979-80 and 1982) highlighted the complex phonological history of the Anuta language. This was a period when numerous notes, memoranda, proposals, critiques and lists of cognates (often hand-written) passed back and forth among a group including Peter, Bruce Biggs, Andy Pawley and myself, all leading eventually to much improvement and enlargement of our beloved database.
Although Peter had been in retirement for some years, 61 reconstructions on the web site labelled Rby still attest to his work.
Hello and welcome to the POLLEX-Online test server. We're still ironing out bugs and glitches.