
Limit by Language:
How should Orbis software identify records by language?
Go to Summary of Responses and Conclusion at the end of this document.
Situation
Patrons comparing local and Orbis union database search results will sometimes come up with very different results. Limit by Language yields a different set of records in different systems.
When are the results different?
Orbis is set up the same as GFU, L&C, OIT, and UP.
Why are the results different?
Questions
Background
Cataloging records contain codes for the language(s) of the material. The codes are stored in the fixed field "LANG" and also in a variable length field, 041.
OCLC Bibliographic Standards and Formats
field: LANGGuidelines include:
field: 041
subfield a
subfield b: language code of abstract/overprinted title or subtitle
subfield d: sung or spoken text, instead of subf a for sound recordings
subfield e: librettos
subfield f: table of contents
subfield g: accompanying material other than librettos
subfield h: original language and/or intermediate translations of text
Analysis
INCLUDE 041?
When 041 is included in the searching, results will include some materials that the patron presumably doesn't want: materials with very little content in that language, some translations, etc.
This disadvantage - including materials not intended -- is constrained to older cataloging records and records incorrectly coded (where all language codes are in subfield "a" instead of subfields). The problem could be mitigated with a large-scale "fix " project.
Example A. Interested in French language materials about the history of sociology (subject sociology history)? You'll also get books like this:
Example B. Interested in the works of Erica Jong, and limit to English language? You'll also get:
FIXED FIELD ONLY?
When only the fixed field is used, some materials that the patron would want to see are excluded.
You searched for the KEYWORD: dictionary LIMITED TO Latin
16
entries found, entries 1-8 are:
1 Cree's Dictionary of Latin
quotations
2 Dictionarium bibliothecarii practicum : ad usum
internat
3 Dictionarium bibliothecarii practicum : ad usum
internat…
4 A Latin-English dictionary : for the use of junior
students
5 Manipulus vocabulorum: a rhyming dictionary of the
Engl…
6 Mediae Latinitatis lexicon minus. A medieval Latin-Fre…
7
Mediae Latinitatis lexicon minus : lexique latin medieal-
8 Mediae
Latinitatis lexicon minus : lexique latin medieval-
9 Mediae
Latinitatis lexicon minus. A medieval Latin-Fre
10 Mediae Latinitatis lexicon minus : a medieval Latin-Fre
11 Mediae Latinitatis lexicon minus : lexique latin
medieval-francais/ang
12 Orbis pictus Latinus : vocabularius imaginibus illustratus =
13 Selectae e Veteri Testamento, historiae; or, Select passag
14 A smaller Latin-English dictionary,
15 Terminorum musicae index septem linguis redactus = Polyglot
16 Works.
Example D. Interested in materials with keyword "Cajun", in French? You won't get Bogalusa Boogie.
Example E. Interested in bridge design, only works in English? You won't get this one, with text in English and German.
Example F. Looking for Portuguese poetry? Use Keyword "poemas" and limit to "por". You won't get this one:
Summary of responses and Conclusion
Questions
1. How should Orbis limit by language?
(choose one)2. Is it okay for the union catalog (Orbis) and local catalogs to use different rules for limiting by language?
(choose one)Responses
|
ff only |
ff+041 |
uniformity between local & union? |
|
| EOU |
|||
GFU |
ff+041 |
prefer uniformity |
|
| L&C |
ff+041 if not costly |
uniformity is not necessary |
|
| Linfield |
ff+041 |
prefer uniformity |
|
| OIT |
ff+041 |
prefer uniformity |
|
| Reed |
ff+041 |
prefer uniformity |
|
|
SOU |
|||
| UO |
ff+041 | [mixed responses] | |
|
UP |
|||
| UPS |
ff+041 |
prefer uniformity |
|
| WOU |
ff |
uniformity if possible |
|
| Whitman |
ff+041 |
prefer uniformity |
|
| WU |
ff+041 |
local & union should work the same way |
Conclusion
Based on the responses, I believe we should set Orbis to use both the fixed field and 041 for Limit by Language. I will arrange for this profiling change by the end of February. Because there is some interest by local sites for Orbis and local Innopacs to work the same way, there are 4 sites that may want to consider making the same change.
Thank you to all who participated! [apologies if I have not included a reply received after Feb.2]
Return to
Technical and Profiling Information