Stephen Carlson wrote: ↑July 27th, 2018, 4:22 pm
Jonathan Robie wrote: ↑July 27th, 2018, 8:53 am
First off, there's a third category - open scholarship. It is neither "hobbyist stuff" nor commercial.
***
If we want to have open scholarship in our field, we need to free up these derivative works. That's really at the heart of the problem.
Thanks for your considered and detailed response, too lengthy to quote in full. From my perspective as someone in academia, I fear that the third category basically does not exist, and cannot exist as far as academia is concerned. The mentality in academia is: if it's not peer-reviewed, it's not scholarship. Tenure and promotion committees reward peer-reviewed publications and contributions to open scholarship simply don't count because there is no peer review. In fact, peer review and open scholarship are probably mutually incompatible at a fundamental level. Peer review is ultimately a gate-keeping exercise and the point of "open" is, well, no gate-keeping of any kind.
Thanks for your thoughtful response.
Open data needs gatekeeping too. This third category is precisely what I think needs to be created, and it requires peer-reviewed, high-quality, openly-licensed data. There's an obvious business-model question: if peer-review is essential and the resulting data is freely-licensed, how is it paid for?
There are branches of academia that work this way. Climate science got into trouble because scientists couldn't analyze each other's data and verify results, comparing their analysis in detail to analysis done by other scientists, so they opened it up. Now multiple groups can analyze the same data in various ways, publishing the results in various peer-reviewed journals. The result is more peer-review, not less.
In the Classics, using Perseus certainly does not disqualify you from publishing your research in academic journals. Work done on the OpenText analysis and on Dag Haug's trees have been published in traditional peer-reviewed journals and work done at the ETCBC center in Amsterdam is largely based on open data, but it is also peer-reviewed.
If we really depend on the data, it has to be reviewed. "Scholarship = peer review" is important, and perfectly compatible with freely licensing the results. Consider the open source software world. The reason we can rely on Linux systems for most Web servers is precisely that this software is thoroughly tested, code is reviewed, etc. People have to jump through hoops to become Apache committers, and even then every check-in is tested and reviewed. Many enterprise systems rely deeply on open source software, that's how companies like Red Hat make their money. The open-source enterprise messaging software I worked on was mission-critical for banks and stock-brokers.
I want open with gatekeeping. Peer review is essential, you need to be able to sort the wheat from the chaff. Expertise is essential. Open licensing simply means that you make the results of high-quality work available so others can build on it. Consider the syntax trees produced by Randall Tan or Dag Haug or the discourse analysis of Stephen Levinsohn - these are high-quality work done by serious professionals, they are not toys created by hobbyists.
In our world, if the Nestle-Aland editions were openly licensed, the various treebanks and discourse analyses and morphologies and context-sensitive glosses could all be based on the same text, making it much easier for us to replicate each other's results. And these days, there are freely licensed transcriptions of the many of the important New Testament manuscripts and realistic 2-dimensional images of the original artifacts may or may not be copyrightable, that's not certain. Open GIS data is being created for the ancient world by various groups too.
Stephen Carlson wrote: ↑July 27th, 2018, 4:22 pm
If tenure and promotion committees won't recognize contributions to open scholarship, (academic) scholars might still be incentivized to contribute to open scholarship if they could get grant funding for it. But grant funding review panels are stocked with academics and they still have the "scholarship = peer review" mentality. And the commercial restrictions of the peer-reviewed products would not onerous because the grant money can cover the licensing. (And I'm speaking as someone who has applied for grants to contribute to open scholarships.) Plus, there's hardly any grant money out there in humanities.
We really do need to find funding for something like a high-quality critical edition of the Greek New Testament, and any serious long-term work needs to be funded somehow. We need better ways to fund this work. It's not like people who create our most important data are getting rich under the current system. We might be able to create a funding model based on the value of open data. Or perhaps agencies that get their money from donors could convince donors that open data is valuable.
If we really depend on the data, it has to be reviewed. "Scholarship = peer review" is important, and perfectly compatible with freely licensing the results.
And freely licensing these works is a public good. Shouldn't a granting agency want the results of high-quality research to be as useful as possible? Shouldn't donors to a Bible Society want them to create editions of the Greek New Testament that can be used freely?
So I think one trick is finding business models to fund this kind of work, including grants and donors. Another trick is building structures for peer review and tenure that recognize the value of open science and open data.
Stephen Carlson wrote: ↑July 27th, 2018, 4:22 pm
As for derivative works, the mentality of academics is that everything they do with publications (generally secured by access rights from their library, so they're not paying for it) in their own notes is "fair use" until they publish it, and their peer-reviewed publishers will secure any necessary rights to the publications when they publish. However, since one of the standard criteria for passing peer-review is "originality," most, indeed almost all, of what scholars publish in peer-review outlets are going to be sufficiently transformative so as not to be derivative works.
Originality is actually a requirement for a derivative work too, according to copyright law. If you don't have originality, you don't have a derivative work. A translation or a syntactic analysis or a discourse analysis can be copyrighted precisely because they are original works.
For copyright protection to attach to a later, allegedly derivative work, it must display some originality of its own. It cannot be a rote, uncreative variation on the earlier, underlying work. The latter work must contain sufficient new expression, over and above that embodied in the earlier work for the latter work to satisfy copyright law’s requirement of originality.
But you're right, if you work at a university you have access, you don't feel the pain of getting to a publication. Getting to data is another story. If I want to reanalyze the results of someone's thesis, I need the data and enough information to reconstruct the methods. Jupyter Notebooks are widely used for this purpose in open science.
ἐξίσταντο δὲ πάντες καὶ διηποροῦντο, ἄλλος πρὸς ἄλλον λέγοντες, τί θέλει τοῦτο εἶναι;
http://jonathanrobie.biblicalhumanities.org/