| 
  • If you are citizen of an European Union member nation, you may not use this service unless you are at least 16 years old.

  • You already know Dokkio is an AI-powered assistant to organize & manage your digital files & messages. Very soon, Dokkio will support Outlook as well as One Drive. Check it out today!

View
 

Estonian Leipzig

Page history last edited by Anne Tamm 14 years, 4 months ago

Morphology based glossing of Estonian and the Leipzig glossing rules: problems and solutions for Uralic languages

 

 

Ülle Viks and Anne Tamm 

 

1. Introduction

 

One of the characteristics of the Uralic languages is its rich morphology. There are a number of clear functional categories that are expressed in morphology, such as possession, which are expressed by a separate set of morphemes - possessive suffixes.

 

2. Finno-Ugric/Uralic tradition

 

The traditional scholars of the Finno-Ugric or Uralic languages presume that the readership of their articles is well educated in Finno-Ugric and Uralic morphology. Providing linguistic examples with translations has been an adequate way of presenting linguistic material in a scientific article.  Outside Uralistics, not many “outsiders” seemed to be interested in the rich tradition of Finno-Ugric or Uralistic studies.

 

3. English-based traditions

 

In theoretical linguistics, the linguistic examples have been predominantly from English. Perhaps more than is the tradition in other generative schools, LFG has concentrated on a wide variedy of languages other than English, and this trend has spread to other streams of generative linguistics. Those researchers typically do not have previous schooling in those languages, therefore, if they discuss a linguistic example, there is always a line of glossing sandwiched between the example and its translation.

 

4. A surge of interest in smaller languages brings along a problem

 

In this century, theoretical linguists, typologists, neurolinguists and others have started to become interested in data from smaller languages, including the Uralic languages. However, previous valuable work is practically useless for those who are interested but fall out of the Uralic tradition. The problem with earlier writings is the lack of exact correspondence between the linguistic example and its translation. The more general researchers are interested in the exact, word-by-word, morpheme-by-morpheme, category-by-category correspondences.

 

5. Standardization does not provide a solution

 

Standardization is needed, and in this century, the so-called Leipzig glossing rules have been established for standard interlinear glossing to solve several problems of mutual intelligibility. Sporadically, they are applied in writings on the Uralic languages. However, several categories that are necessary in the discussions of the Uralic language are missing in the Leipzig convention. There are articles that have solved the glossing by a general reference to the webpage on the Leipzig rules and another, individual author’s webpage with additional, Uralic glosses.

 

6. An inventory of problems, both general and specific

 

This presentation discusses the general and more specific problems of uniting the glossing conventions form the

  • syntactic
  • constructional
  • morph(no)ological
  • phonological
  • functional
  • semantic
  • diachronic

 perspective.

 

 

Problems appear in all modules of language description – semantics, pragmatics, syntax, morphology, phonology. Some of the problems appear exactly because they involve linguistic information that spans across modules and temporal evolvement of the languages in question.  Here are some examples:

  • cross-linguistic categories (is abessive always caritive in the Uralic languages, is caritive the same as privative in the Australian languages),
  • category change (is the partitive marked present participle epistemic modality or infinitive)
  • case syncretism (is there an accusative in Estonian, or is it just genitive and nominative)
  • morphological versus non-morphological glossing (what is the relationship between form and function, when is it better to gloss something as VAT versus EPIST_MOD or PART_EVID)
  • the problem of encoding lenghth (‘välde’) (shall we find a uniform way of distinguishing the length in Inari Sami, Estonian and Nganasan)

 

These and other items need to be discussed.

 

7. The proposal of an exact procedure

 

We will propose an exact procedure to deal with these problems that are recurrent across our languages. We propose the following procedure, which we detail for Estonian:

 

  • start with morphological glossing based on morphological descriptions of the languages in question
  • revise standard grammars
  • revise automatic text processing labels used for the language, morphological codes etc
  • revise leading articles in leading journals or other publications on the specific languages and using interlinear glossing
  • find another model that has been applied to a morphologically rich language, comparing the results to other languages
  • extend the morphology based glossing to instances that are either more marginal or more restricted (glossing for historical linguistics with many examples of category change)

 

Appendix1. The codes used in automatic rule-based morphology of Estonian. The explanations are in Estonian as well.

 

 

VERB:

 

 

Inf

infinitive

infinitiiv e da-infinitiiv e da-tegevusnimi

Ger

gerund

gerundium e des-vorm

Sup

supine

supiin e ma-infinitiiv e ma-tegevusnimi

Pts

participle

partitsiip e kesksõna

Ps

personal voice

personaal e isikuline tegumood

Ips

impersonal voice

impersonaal e umbisikuline tegumood

Pr

present

preesens e olevik

Pt

past

preteeritum e (üld)minevik

Ipf

imperfect

imperfekt e lihtminevik

Pf

perfect

perfekt e täisminevik

Ppf

pluperfect

pluskvamperfekt e enneminevik

Ind

indicative

indikatiiv e kindel kõneviis

Kvt

quotative

kvotatiiv e kaudne kõneviis

Knd

conditional

konditsionaal e tingiv kõneviis

Imp

imperative

imperatiiv e käskiv kõneviis

Sg 1

1. person singular

singulari e ainsuse 1. pööre

Sg 2

2. person singular

singulari e ainsuse 2. pööre

Sg 3

3. person singular

singulari e ainsuse 3. pööre

Pl 1

1. person plural

pluurali e mitmuse 1. pööre

Pl 2

2. person plural

pluurali e mitmuse 2. pööre

Pl 3

3. person plural

pluurali e mitmuse 3. pööre

Af

affirmative

afirmatiiv e jaatav kõne

Neg

negative

negatiiv e eitav kõne

 

 

 

 

NOUN:

 

 

Sg

singular

singular e ainsus

Pl

plural

pluural e mitmus

Nom

nominative

nominatiiv e nimetav

Gen

genitive

genitiiv e omastav

Part

partitive

partitiiv e osastav

Adt

aditive

aditiiv e suunduv (e lühike sisseütlev)

Ill

illative

illatiiv e sisseütlev

In

insessive

insessiiv e seesütlev

El

elative

elatiiv e seestütlev

All

allative

allatiiv e alaleütlev

Ad

adessive

adessiiv e alalütlev

Abl

ablative

ablatiiv e alaltütlev

Tr

tranlative

tranlatiiv e saav

Ter

terminative

terminatiiv e rajav

Es

essive

essiiv e olev

Ab

abessive

abessiiv e ilmaütlev

Kom

comitative

komitatiiv e kaasaütlev

 

Ü. Viks. Eesti keele avatud morfoloogiamudel. -- Arvutuslingvistikalt inimesele (toim T. Hennoste). Tartu Ülikooli üldkeeleteaduse õppetooli toimetised 1. Tartu 2000, lk 9--36. (http://www.eki.ee/teemad/avatud_mrf.html)

Ü. Viks. http://www.eki.ee/tarkvara/morf_lisa.html

Comments (0)

You don't have permission to comment on this page.