Peeter P

1
????????????? ?????????????? ?????? ? ?????????
?????????? ?????

• Peeter Päll
• 9th Baltic Division meeting
• Jurmala, October 2005

2
Overview

• Existing collections in Estonia
• Two types of collections (systematic databases
  archived materials)
• Digitization options
• ETOP, the digital archive
• Data structure, encoding problems
• Named features identification
• Perspectives

3
History of names collections

• Various initiatives in the 19th century (1888 M.
  J. Eisen, 1895 J. Jung, 1901 F. Kuhlbars)
• Systematic collection since 1920s by the Mother
  Tongue Society
• Collections by scholarship, field expeditions by
  researchers and students (1930s1990s)

4
Current collections

• Place Names Archive at the Institute of Estonian
  Language over 500,000 cards (integrates also
  collections of the Mother Tongue Society)
• Endel Vareps archive at EKI (ca. 95,000 A5
  pages)
• M. J. Eisens collections at the National Museum
  (dispersed copy in Helsinki in 3 vol.)
• Collections at various museums and institutes
• Main Estonian-Swedish collections at Uppsala

5
Place Names Archive at EKI

• Collections since 1920s, arranged by parishes
  general alphabetic index
• Data on cards
• name in standard spelling and pronunciation
• locative case forms (external or internal)
• references to other connected names
• short description of the named feature (maps
  usually not included)
• background information, examples of name usage
• village, parish, informant, collector

6
(No Transcript)
7
Place name card (1930s)
8
Alphabetic index card
9
Place name card (1960s)
10
Place name maps
11
Problems

• Parish collections are uneven by the amount of
  cards and their quality
• Usually no map references are given
• Using the collection for research requires a lot
  of time

12
(No Transcript)
13
(No Transcript)
14
Digital resources

• Place Names Database at EKI (KNAB) 90,000
  records (244,000 names)
• Toponymic Database for historic Võrumaa by the
  Institute of Võru ca. 10,000 records
• National Place Names Register test version only
• Databases of private publishers (e.g. Regio road
  map index contains ca. 11,000 names)

15
KNAB (www.eki.ee/knab/knab.htm)
16
(No Transcript)
17
(No Transcript)
18
(No Transcript)
19
Võrumaa (www.ekk.ee/avka/)
20
Two types of collections

• 1. Archives
• unsystematic
• uneven
• excessive data
• 2. Systematic databases
• processed data, based on various sources
• standard quality criteria
• structured data, even coverage

21
Digitization of names archives

• Aims
• to protect against destruction of data
• to give better access to data
• to enable further processing
• Models
• Finland (partial type-in of card information)
• Sweden (scanning digital headwords)
• Norway (type-in and scanning)

22
ETOP, the digital archive of EKI

• Data format SGML-text
• Character format Unicode-enabled
• Uniformity with KNAB, the systematic place names
  database
• Recognition of international standards on digital
  gazetteers (Alexandria Digital Library, ADL
  Gazetteer Protocol)
• Availability through the Internet (both for
  entering and accessing the data)

23
(No Transcript)
24
Data fields

• ltnimigtname label commentlt/nimigt
• ltvargtname variants lt/vargt (includes
  pronunciation in standardized transcription the
  original notation)
• ltkohgtuse of locative formslt/kohgt
• ltvrdgtreferences to other nameslt/vrdgt
• ltselgtdescription of the featurelt/selgt
• ltlkoodgtfeature typelt/lkoodgt (new information)
• lttxtgtbackground, examples, textlt/txtgt
• ltallgtname source (informant)lt/allgt
• ltkhkgtparishlt/khkgt
• ltlähikgtvillage, collection pointlt/lähikgt
• ltautgtcollecting person, owner of collection,
  typistlt/autgt

25
Principles of digitization

• inclusion of all the content of the original
• maximum preservation of original structure
• all additions and corrections are clearly marked
• no interpretations, but comments are allowed

26
Phases of digitization

• All new cards will be processed first, these are
  not copied for alphabetic index file any more
• Other materials (not in a card format) that can
  be integrated, obtained from other sources
• New contributions may be provided through web
  digitally
• Entering the data of main collections starts from
  sample parishes
• After the main collection is entered, comparison
  with the alphabetic index (and subsequently this
  may be discarded)

27
Perspectives

• Estimated time-length of digitization 1020
  years
• Possible additions to data geocoding
• Possible further processing linking records of
  identical features
• Research continues..