Sample Metadata Record

oai:TDProject.sr.language-archives.org:agreement

XML format

<olac:olac>
   <dc:title>Agreement and alignment in the languages of the world</dc:title>
   <dc:contributor xsi:type="olac:role" olac:code="editor">Siewierska, Anna</dc:contributor>
   <dc:contributor xsi:type="olac:role" olac:code="author">Siewierska, Anna</dc:contributor>
   <dc:contributor xsi:type="olac:role" olac:code="author">Bakker, Dik</dc:contributor>
   <dc:contributor>Bakker, Dik</dc:contributor>
   <dc:creator>Siewierska, Anna</dc:creator>
   <dc:creator>Bakker, Dik</dc:creator>
   <dc:date>2001-05-01</dc:date>
   <dc:description>The database on agreement in the languages of the world has been developed
by Anna Siewierska over the last 8 years. It is constantly being corrected
and updated, both in terms of new languages and new variables. A number of
articles and conference papers have been based on it, and are in progress.
Currently, the database contains close to 100 variables on agreement and
relevant syntactic and morphological aspects of around 400 languages. A
major subset of these languages conforms to the sampling criteria of
Rijkhoff and Bakker (1998). This subset comprises 45 of the 50 languages
defined by the Typological Database Project as the standard sample.
Furthermore, a subset of 200 languages conform to the sample for the World
Atlas of Language Structure project of the Max Planck Institute in Leipzig.
 The variables are coded in the comprehensive system of the Computer
Programs for Language Typology (CPLT), introduced in Bakker (1994), which
have been used in the cause of the EuroTyp project. The usual way of
representing typological data, the rectangular data matrix, has an atomic
value in each cell, i.e. the combination of a language and a variable. See
example (1) below.
(1)	a. Dutch/Basic Word Order = SVO
b. Dutch/Alternative Word Order = OVS
In contrast to this, the CPLT system allows for multiple values per cell,
as well as the addition of scales and syntagmatic conditions. Cf. example (2).
(2)	Dutch/Main Clause Order = SVO - (basicness = basic) \
                                                         OVS - ((basicness
= marked),(O = contrastive focus))

This method of coding has been used to a large extent in the database under
discussion. Examples may be found under (3) and (4) below, for Maung and
Ngiti respectively.
(3)  SO(1+2)\OS(3+3,1+2)\both
(4) pref\(inal_1sg+plinc+3)\suf(inal)\indep(al)
(3), for variable V52_6 in the database (i.e. order of S and O prefixes on
verb) should be interpreted as: the order is SO when the S is first person
and the O is second person but OS when either both are third person or the
O is first person and the S second person; when S is a first or second
person and the O third or vice versa, either order may be used
(4), for variable V60_1 (i.e. form of pronominal possessor) should be
interpreted as: pronominal possessors if marking inalienable possession are
marked either by a prefix or a suffix; the prefix is used for the 1st
person singular and plural inclusive  and for the third person, other wise
a suffix is used; alienable possession is expressed by independent pronouns.
As for the current variables in the agreement database, these are
distributed over the following sections:
Section A. General information on the language
Section B. Agreement for grammatical relations
Section C. Agreement marking of possessors in adnominal possessive
constructions
Section D. Agreement marking for adpositions 
Section E.  Free personal and possessive pronouns
</dc:description>
   <dc:format>CPLT format</dc:format>
   <dc:format>Microsoft Access/Excel</dc:format>
   <dc:subject>Typological data on several types of agreement in a genetically and areally stratified sample of the languages of the world.</dc:subject>
   <dc:subject>Agreement</dc:subject>
   <dc:subject>Subject agreement</dc:subject>
   <dc:subject>Object agreement</dc:subject>
   <dc:subject>Possessor-possessed agreement</dc:subject>
   <dc:subject>Adpositional agreement</dc:subject>
   <dc:type xsi:type="dcterms:DCMIType">Dataset</dc:type>
</olac:olac>

Display format

Metadata quality analysis

OLAC metadata records are scored for metadata quality on a 10-point scale explained in OLAC Metadata Metrics. The score for the above record (along with comments on changes that could improve the score) is as follows: