This page presents Heinrich Puschmann's Nhewenh spelling proposal for Mapudungun (or Mapusdugun), the language of the Mapuche people in Chile. Throughout this and other pages by the same author, `Mapuche' is defined to mean `vernacular speaker of Mapudungun'.
`Nhewenh' means net. You may think of it referring to either a social web of urban Mapuche or to the internet, as you wish. Since there are many different spelling conventions for Mapudungun, I believe that each proposal should be given a short identifier.
Various aspects on the Mapuche people and their language (including a Mapuche Interest Page) may be found on the Mapuche Page, edited by Jennifer Arnold. An incomplete chronological bibliography on the language is available online.
Almost all literate Mapuche (some 40% of the active speakers) are living in a state of diglossia, using the Spanish language for almost all written purposes. Until very recently (about 1980), `learning how to write' meant to Mapuche people `learning written Spanish', and was therefore considered quite a formidable task. Due to this situation, Mapudungun has no writing standard that is used by a majority of its speakers. Several de-facto or formal competing spelling proposals have been around, notably Augusta, `UniLing', and to a lesser extend Rangileo, Aflayay. (Identifiers are not standard)
The proposed Nhewenh spelling system has been evolving since 1975, the time when I started compiling field notes for personal use. Of course, I owe many adopted ideas to other authors, but then I have quite specific reasons for preferring this very combination of graphemes. My utmost thanks go to Silvia Marín, Luciano Ñancucheo, Sofía Painiqueo, Timothy Sandvig, Jorge Olivos, and many others for their invaluable help.
| Distr. | GROUP | front unrounded | midback unrounded | back rounded |
|---|---|---|---|---|
| V | Low | e | a | o |
| V | Syllabic high | i | v (*) | u |
| F | Final high | y (*) | w (*) | |
| I | Initial high | y | w |
| Distr. | GROUP | labial | dental | middle | palatal | velar |
|---|---|---|---|---|---|---|
| I | Occlusives | p | td (*) | t | k | |
| I | Affricates | tr (*) | ch (*) | |||
| IF | Fricatives | f | sd (*) | r (*) | sh (*) | q (*) |
| IF | Nasals | m | nd (*) | n | nh (*) | g (*) |
| IF | Laterals | ld (*) | l | lh (*) |
SYLLABIC STRUCTURE of Mapudungun must be of the form IVF, IV, VF, V, where (as in the table) a (nonsyllabig) initial sound I may be any consonant or semivowel, a syllabig vowel sound is specified as V, and a (nonsyllabig) final sound F may be any fricative, nasal, lateral, or the high vowels i and u. However, some of these combinations may not occur (a syllable iq does not exist).
When acting as final sounds F, the high vowels i, u, are proposed to be spelled y, w, respectively, in order to allow only one vowel for each syllable. This agrees with the UniLing proposal.
ARTICULATION SHIFT is often semantically relevant by itself, meaning that specific phoneme shifts within a word allow for a neutral, diminutive, empatic, or negative perception or sentiment expressed by the speaker (compare narki, nashki, and nasdki, all three meaning `cat').
Some articulation shifts may account for phonemes that do not exist in a neutral context, but only within a diminutive or emphatic connotation. It is not clear whether `sh' exists in a neutral context, or whether the corresponding neutral phoneme should be `sd' or even `s', a central fricative. Catrileo also includes a neutrally nonexistent palatal occlusive (which we might spell out `tt') in her phonemic system.
Syllabic stress seems to be relevant, but a word may frequently vary its stress, depending on its context within the sentence. The phenomenon is not yet well understood, and no current spelling proposal uses stress markers.
Voicing is not phonemically relevant. Occlusives and affricates are unvoiced, fricatives vary, and nasals and laterals are voiced.
October 1995
Last modified: 1997 Apr 15
© This proposal is copyrighted to Heinrich Puschmann.
It may only be disseminated if mentioning its source.