Daniel Strypey Bruce
[This is a piece from a member the Creative Commons Aotearoa NZ community., and is licensed CC-BY-SA. If you want to write for us, please email firstname.lastname@example.org.]
When Land Information New Zealand officials, prompted by the support in NZGOAL for free culture CC licenses, started liberating the publicly-funded map data they hold under CC-BY, they may have anticipated that their data could end up in OpenStreetMap.org (OSM). OSM is the libre licensed, crowdsourced, world mapping project (formerly CC-BY-SA, now Open DataBase License) created as a free-as-in-speech alternative to Google Maps, which licenses copyright map and image data from other companies. What they probably never imagined was their data appearing on community sites like the new NSW government funded TimeBanking platform and in maps powered by Leafet, the free code, mobile-friendly, web map engine which uses OSM as the default source for map data. Such are the unintended benefits that flow from sharing in the commons.
Just as OSM contributed to the development of websites using Leaflet, the cultural commons also has a contribution to make to the development of free code speech recognition software. Douglas Bagnall gave a great talk about free code speech recognition engines at Kiwi PyCon 2012, and one of the main non-coding challenges he identified is the lack of a suitably large corpus of spoken speech with which to train’ the software (the interesting bits for non-coders starts at about 3:00), particularly for less common accents like New Zealand English.
Douglas mentioned a site called VoxForge.org and encouraged us all to record some speech for it. VoxForge collects samples of speech to “compile them into acoustic models for use with Open Source speech recognition engines”, with both the acoustic models and the source data are made available under the GNU General Public License, the free code software license used by the Linux kernel.
Voxforge is a great resource, but it would be a lot of work to drive enough ‘Nu Zylnd’ accented speakers there. and it occurred to me recently that there is another source of spoken speech for which there is strong argument for libre licensing; public interest radio. Releasing audio programs made for National Radio, student radio, and community access radio, under a libre license (eg CC-BY), has always seemed like a no-brainer to me. Even before I set up the CC-NZ email list, I was contacting community radio people and telling them about the potential benefits of podcasting and CC licensing for increasing program exposure, as part of my work with Aotearoa.Indymedia.org.
National Radio are publicly-owned and funded, and currently podcast all their programs in both MP3 and Ogg Vorbis formats. Student radio stations have also been experimenting with recording their shows and turning them into podcasts, and if they libre licensed the sometimes substantial spoken word parts of recordings like the Radio1 shows shared on MixCloud, they’d be unlikely to miss out on any commercial return.
Community access radio stations, as their name suggests, exist to give special interest groups in our communities access to the airwaves, so their unique voices have a chance to be heard in broadcast media. Access stations are heavily subsidised by public broadcasting funding, and their programs aren’t expected to have commercial potential.
Whenever these programs are rebroadcast, both the program producers, and the public who subsidize their production, get more bang for their buck. Over the last decade or so, access radio stations have been building up the infrastructure to digitally record programs, so they can be made available as podcasts, on sites like accessradio.org.
What I notice in all three cases though, is that despite NZGOAL-inspired wave of CC license adoptions across public service organisations from LINZ, to schools and tertiary institutions, to archives and museums, I see no sign of CC licenses on locally made radio so far. I’m guessing one issue is legal confusion and risk aversion.
For example, if I want to CC my radio program, what about the copyright songs I played during the program? Who owns the copyright on ads, and is it acceptable to remove them? Radio stations and rebroadcast sites can pay a fixed fee to collection organisations like NZRIA and APRA for permission to play the songs, but they can’t pass on that permission to anyone who rebroadcasts the program.
What can we do to show public radio stations that libre licensing – CC or otherwise – is a good fit for their organizational mission, and help them negotiate the legal minefields around copyright? Between them, the RadioNZ, community access, and students stations have vast numbers of hours of recordings, covering a wide range of kiwi accents spoken regularly on ethnic community radio programs. Having these available under a libre license would have benefits beyond just sharing the information and perspective included in public interest radio programs.
It could also be tremendously helpful to people working on getting free code/ open source speech recognition software to recognise New Zilnd accents, which in turn would be tremendously helpful for any kiwi wanting a speech-operated computer that respects their software freedom. The same would be true for speakers of Te Reo Māori, and other Polynesian languages, who, like NZ English speakers, are also less likely to record enough speech at sites like VoxForge for them to generate a useful acoustic model.
In ‘The Coming Civil War over General Purpose Computing‘, Cory Doctorow asks us to consider the implications if the proprietary software running on many of our personal computers was instead running on our bionic legs;
“If I can lo-jack your legs, I can lease them to you with the confidence of my power to repo them if you default on payments. If I can’t, I may not lease you legs unless you’ve got a lot of money to begin with. But if your legs can decide to walk to the repo-depot without your consent, you will be totally screwed the day that muggers, rapists, griefers or the secret police figure out how to hijack that facility.”
Like bionic legs, accessibility tools like speech recognition and text-to-speech are augmentations which restore equalit, helping people with conditions from blindness to quadraplegia to have the full use of computers and the internet that most of us take for granted. Such essential tools should not be subject to the price pumping and hidden agendas that go with proprietary sofware.
[box size=”large” style=”rounded” border=”full”]
Daniel is a community organiser and long-time Creative Commons Aotearoa NZ supporter, based in Ōtepoti (Dunedin). Daniel set up InterActive, a media activist resource centre and was one of the founders of pioneering citizen journalism website Aotearoa Indymedia. He also set up the Aotearoa Permaculture Network. Daniel also helped to kick off the Aotearoa/ New Zealand localisation of the CreativeCommons licenses with the cc-nz email list. You can read Daniel’s full bio at Disintermedia.