AILLA:The Archive of the Indigenous Languages of Latin America Heidi Johnson / The University of Texas at Austin
AILLA is a joint project of: Anthropology: Joel Sherzer Linguistics: Anthony C. Woodbury Digital Library Services: Mark McFarland
Pilot site launched March Permanent site launched January, Parallel sites in English and Spanish; someday, we’ll add Portuguese. Audio, video, digital text, scanned manuscript text, photos. Genres ranging from chants to grammars to photo galleries.
Infrastructure Archive housed on library computers, maintained by UT digital librarians. MySQL database; PHP interfaces. Intake & processing done in AILLA’s lab on PCs and Macs by graduate students. Analog media returned to depositor or sent to the Indiana Archive of Traditional Music (
Fun facts Collection is now > 500 G ~ 90 languages from 15 countries 1430 registered archive users 917 resources comprising 5391 files; average resource contains 5-6 files all bundles inc. format variants, eg mp3, pdf ~ 20% of files are password-protected
IMDI-esque metadata Resource = a bundle of files, e.g. recording + annotations, in multiple formats Information about: Depositor: contact info, languages Project: sponsor, contact info. Participants: role, demographic data, languages Resources: orig. media, digi specs, lgs, description Content: genre, content type, description References: related publications
Bundles, aka resources, aka sessions Sets of related files Critical for sharing resources: things that belong together must stay together access restrictions/conditions likely to apply to all files in a bundle metadata describes the set – the whole resource – as well as each component
DELAMAN todo list Define Relations for IMDI & OLAC. NOT mathematical relations; documentary linguistics ones: primary_text (e.g. a recording) translation/transcription/interlinearization illustration commentary
Access management Restrictions are set for individual files Two levels of access: 1: free public access 2: restricted by password To be added: 3: time limit 4: contact info for controller 5: reference/link to license and/or recorded agreement governing the resource
Restricted files at AILLA Who restricts access & why? Students: thesis research in progress Other depositors: pending permission from speakers; recordings include e.g. gossip that needs to be edited out; still uncertain about the whole access issue.
Pros of this simple method It’s EASY: to implement to operate for controllers to share access It’s safe enough for our purposes. Passwords, etc. are part of the metadata – in the database – so it should be easy enough to pass on to another archive.
Cons of the simple method Depositors can be VERY hard to reach. There has to be a time limit and/or phase- out plan, ala the Rosetta protocol. Assume that nearly everything will ultimately be under the archive’s control – what’s the 50-year plan?
Low tech federation We could define a small set of standard “access packages”, implemented via our existing metadata mechanisms: student special: good for 5 years only renew-on-request: expires in 5 years unless you respond to the archive & renew archivists’ discretion: could require e.g. a faxed letter from an indigenous organization; getting that letter is the users’ problem.
Comments gladly received at