WEBVTT 00:00:00.000 --> 00:00:03.255 Now I would like to introduce our great speakers. 00:00:03.255 --> 00:00:08.745 Philip Young is the repository manager at VTechWorks, 00:00:08.745 --> 00:00:10.320 which provides global access 00:00:10.320 --> 00:00:11.955 to Virginia Tech scholarship. 00:00:11.955 --> 00:00:14.430 He also provides outreach to the university on 00:00:14.430 --> 00:00:17.565 open access, ORCID and Perma.cc. 00:00:17.565 --> 00:00:19.920 Jimmy Ghaphery is Associate Dean for 00:00:19.920 --> 00:00:22.110 Scholarly Communications and publishing at 00:00:22.110 --> 00:00:24.510 Virginia Commonwealth University Libraries. 00:00:24.510 --> 00:00:26.175 He and his team coordinate 00:00:26.175 --> 00:00:27.990 library technology operations and 00:00:27.990 --> 00:00:29.745 advocacy for all things open, 00:00:29.745 --> 00:00:32.685 including open access, open repositories, 00:00:32.685 --> 00:00:34.620 open publishing, open educational 00:00:34.620 --> 00:00:36.640 resources, and open data. 00:00:36.640 --> 00:00:39.260 Without further ado, this is their presentation, 00:00:39.260 --> 00:00:41.330 public domain and paywall copyright 00:00:41.330 --> 00:00:42.800 statements on journal articles 00:00:42.800 --> 00:00:45.095 by US government employees. 00:00:45.095 --> 00:00:48.820 Well, thank you for the introduction, Ellyn. 00:00:48.820 --> 00:00:51.920 I'm Jimmy Ghaphery and Philip and I are 00:00:51.920 --> 00:00:55.010 thrilled to be here at the Kraemer Copyright Conference. 00:00:55.010 --> 00:00:58.190 The sessions have been fantastic so far, 00:00:58.190 --> 00:01:00.290 so much of what I've been able to 00:01:00.290 --> 00:01:04.320 attend is really builds on what 00:01:04.320 --> 00:01:06.330 we want to talk about from 00:01:06.330 --> 00:01:10.010 Sandra and Kyle's session that I was able to 00:01:10.010 --> 00:01:12.740 go to talking about government works, to 00:01:12.740 --> 00:01:16.040 Pia's excellent keynote about author's rights. 00:01:16.040 --> 00:01:20.990 I just sat in on Kevin Smith and Kenneth Crews' session, 00:01:20.990 --> 00:01:22.550 which really goes to the heart of 00:01:22.550 --> 00:01:24.110 a lot of what we're talking about. 00:01:24.110 --> 00:01:28.395 Anyway, happy to be here and next slide. 00:01:28.395 --> 00:01:31.280 What we hope to do today is to give 00:01:31.280 --> 00:01:33.020 some brief background on 00:01:33.020 --> 00:01:35.990 the copyright status of US government authored works. 00:01:35.990 --> 00:01:38.750 I'll share some preliminary research results 00:01:38.750 --> 00:01:40.190 of a study we've been doing 00:01:40.190 --> 00:01:41.450 about how these works are 00:01:41.450 --> 00:01:45.140 labeled by commercial journal publishers. 00:01:45.140 --> 00:01:47.120 Talk about the implications of 00:01:47.120 --> 00:01:49.475 this research for libraries and access. 00:01:49.475 --> 00:01:53.240 Most importantly, we're thrilled to be talking to a group 00:01:53.240 --> 00:01:55.760 of librarians and copyright experts 00:01:55.760 --> 00:01:57.965 and we look forward to getting your feedback. 00:01:57.965 --> 00:02:02.230 For background and this is as much to let you all 00:02:02.230 --> 00:02:06.070 know that Philip and I are not copyright lawyers, 00:02:06.070 --> 00:02:09.550 but just to share some of our understanding of 00:02:09.550 --> 00:02:11.560 copyright law is that 00:02:11.560 --> 00:02:13.420 copyright protection is not 00:02:13.420 --> 00:02:15.535 available for any work of the US government. 00:02:15.535 --> 00:02:19.565 This is for 17 USC 105. 00:02:19.565 --> 00:02:26.470 The US government work is defined in USC 101 00:02:26.470 --> 00:02:29.860 as a work that's prepared by an officer or employee 00:02:29.860 --> 00:02:31.375 of the United States government 00:02:31.375 --> 00:02:33.505 as part of their official duties. 00:02:33.505 --> 00:02:36.925 It's interesting looking at the legislative history, 00:02:36.925 --> 00:02:40.270 that this doesn't prevent 00:02:40.270 --> 00:02:41.780 a government employee from securing 00:02:41.780 --> 00:02:44.580 copyright even in their field of expertise, 00:02:44.580 --> 00:02:48.845 if this isn't part of their official duties. 00:02:48.845 --> 00:02:51.050 Backup one more slide. 00:02:51.050 --> 00:02:58.380 The last piece that we want to mention is 17 US 103, 00:02:58.690 --> 00:03:03.790 where there is no requirement for any copyright labeling. 00:03:03.790 --> 00:03:05.540 It's interesting to note that 00:03:05.540 --> 00:03:09.850 this only dates back to 1989. 00:03:09.850 --> 00:03:12.330 Before then it was a requirement to have 00:03:12.330 --> 00:03:13.735 the copyright label and as 00:03:13.735 --> 00:03:16.505 Kenneth Crews said, it was very draconian. 00:03:16.505 --> 00:03:19.920 He just said that in the past presentation. 00:03:19.920 --> 00:03:25.160 This included the requirement to state whether part 00:03:25.160 --> 00:03:30.570 of the work was part of the US government employees, 00:03:30.570 --> 00:03:32.210 and if that was mislabeled, 00:03:32.210 --> 00:03:35.795 that would throw the copyright completely out the window. 00:03:35.795 --> 00:03:39.485 The other interesting thing about US government works, 00:03:39.485 --> 00:03:40.910 and you'll often see this on 00:03:40.910 --> 00:03:45.390 the copyright labeling, is foreign protection. 00:03:45.390 --> 00:03:48.860 Even though these works are in the public domain, 00:03:48.860 --> 00:03:50.569 in the United States, 00:03:50.569 --> 00:03:53.299 the United States government is reserving 00:03:53.299 --> 00:03:56.945 the right to assert copyright overseas. 00:03:56.945 --> 00:04:00.290 This is seen when you look at 00:04:00.290 --> 00:04:05.105 the legislative history from the US House report, 00:04:05.105 --> 00:04:07.940 that there are no valid reasons for 00:04:07.940 --> 00:04:11.615 denying such protection to the United States government. 00:04:11.615 --> 00:04:14.240 It's interesting to think of this as 00:04:14.240 --> 00:04:17.360 merely theoretical that the US government 00:04:17.360 --> 00:04:21.290 would assert the copyright 00:04:21.290 --> 00:04:23.440 of these works overseas or not. 00:04:23.440 --> 00:04:26.060 But we had a great email conversation 00:04:26.060 --> 00:04:27.290 with Peter Hirtle with 00:04:27.290 --> 00:04:31.160 this question and he helped us in our thinking on it. 00:04:31.160 --> 00:04:33.650 He noted as an exercise, 00:04:33.650 --> 00:04:37.445 he looked at the copyright system in Canada, 00:04:37.445 --> 00:04:40.520 in which he found that the US government has actually 00:04:40.520 --> 00:04:42.290 registered a few copyrights for 00:04:42.290 --> 00:04:44.615 CDC publications in Canada. 00:04:44.615 --> 00:04:48.845 One of the best resources we found about 00:04:48.845 --> 00:04:53.410 US government works and copyright are the CENDI FAQs. 00:04:53.410 --> 00:04:57.065 CENDI is an interagency organization. 00:04:57.065 --> 00:04:59.750 It's just a really nice overview. 00:04:59.750 --> 00:05:01.850 It explains the various exceptions, 00:05:01.850 --> 00:05:05.210 including specific agencies that are exempt from this, 00:05:05.210 --> 00:05:09.845 as well as how this all fits in with contractors. 00:05:09.845 --> 00:05:11.825 The one thing that was 00:05:11.825 --> 00:05:13.670 the most reassuring to Philip and me, 00:05:13.670 --> 00:05:16.595 we had lots of questions about co-authorship. 00:05:16.595 --> 00:05:18.910 What happens if you're in 00:05:18.910 --> 00:05:20.900 a co-authorship situation where 00:05:20.900 --> 00:05:23.330 one author is from the US government and one 00:05:23.330 --> 00:05:26.305 is from private industry or another university? 00:05:26.305 --> 00:05:29.510 This remains confusing to us to 00:05:29.510 --> 00:05:33.245 some degree as does I think co-authorship in general. 00:05:33.245 --> 00:05:34.520 It was reassuring that 00:05:34.520 --> 00:05:38.510 the CENDI site had notes that much of this is 00:05:38.510 --> 00:05:42.740 unsettled in courts and 00:05:42.740 --> 00:05:45.515 in such situations you should talk to a real lawyer. 00:05:45.515 --> 00:05:49.460 The last thing I'll note on the CENDI FAQ that's 00:05:49.460 --> 00:05:51.710 posted is their 00:05:51.710 --> 00:05:53.720 actual copyright notice that they include, 00:05:53.720 --> 00:05:55.850 which is similar to some of the notices we've 00:05:55.850 --> 00:05:59.135 seen where they say this is a work of the US government. 00:05:59.135 --> 00:06:01.220 It's not subject to copyright protection 00:06:01.220 --> 00:06:02.615 in the United States, 00:06:02.615 --> 00:06:04.610 foreign copyrights may apply. 00:06:04.610 --> 00:06:07.550 It was interesting to us that even on this FAQ, 00:06:07.550 --> 00:06:09.440 they asserted that possibility 00:06:09.440 --> 00:06:12.650 for foreign copyright. Philip. 00:06:12.650 --> 00:06:15.155 Thanks Jimmy. 00:06:15.155 --> 00:06:19.160 I want to talk a little bit about our research prompts. 00:06:19.160 --> 00:06:22.100 As a repository manager 00:06:22.100 --> 00:06:23.240 on part of my job is to getting 00:06:23.240 --> 00:06:25.190 content into the repository. 00:06:25.190 --> 00:06:27.260 When I see articles that are 00:06:27.260 --> 00:06:28.910 marked public domain in the United States, 00:06:28.910 --> 00:06:30.560 I feel relatively comfortable that I can go 00:06:30.560 --> 00:06:32.615 into the institutional repository, 00:06:32.615 --> 00:06:34.760 and have the articles of 00:06:34.760 --> 00:06:37.280 Virginia Tech authors in there and that's great. 00:06:37.280 --> 00:06:40.490 Then I noticed that there's lots of other articles by 00:06:40.490 --> 00:06:42.110 government authors that aren't 00:06:42.110 --> 00:06:44.585 marked public domain in the United States. 00:06:44.585 --> 00:06:47.445 Is it still okay to put them in the IR? 00:06:47.445 --> 00:06:49.400 Or are we depending on 00:06:49.400 --> 00:06:52.085 publishers to correctly mark these? 00:06:52.085 --> 00:06:54.530 What's the risk of putting these articles 00:06:54.530 --> 00:06:56.950 in the institutional repository? 00:06:56.950 --> 00:06:58.920 Am I going to hear from publishers? 00:06:58.920 --> 00:07:01.545 Am I going to get a takedown request? 00:07:01.545 --> 00:07:03.840 In 2019 I posted to 00:07:03.840 --> 00:07:07.455 the IR Managers Google group about this issue. 00:07:07.455 --> 00:07:11.200 The general response was confusion. 00:07:11.200 --> 00:07:13.745 I remember two of their responses. 00:07:13.745 --> 00:07:16.820 One of them was from another manager who said, 00:07:16.820 --> 00:07:18.320 "Yes, we are doing this and we've 00:07:18.320 --> 00:07:21.089 been doing it for quite a while." 00:07:21.860 --> 00:07:26.270 The other response was from a manager of 00:07:26.270 --> 00:07:29.690 a repository at a government agency, he said, "Yes, 00:07:29.690 --> 00:07:31.190 I'm very interested in knowing 00:07:31.190 --> 00:07:33.080 the answer to this because I have 00:07:33.080 --> 00:07:34.550 lots of authors writing 00:07:34.550 --> 00:07:37.620 articles and I'd like to put them into my repository." 00:07:38.510 --> 00:07:41.915 We decided to collect some data. 00:07:41.915 --> 00:07:46.565 How are publishers currently labeling these articles? 00:07:46.565 --> 00:07:48.980 When they label these articles 00:07:48.980 --> 00:07:50.450 or if they label these articles, 00:07:50.450 --> 00:07:53.030 is access provided when they're 00:07:53.030 --> 00:07:57.280 public domain in the United States? 00:07:58.010 --> 00:08:01.200 How do publishers differ on labeling, 00:08:01.200 --> 00:08:02.950 is one publisher always 00:08:02.950 --> 00:08:05.420 labeling and another publisher never labeling? 00:08:05.420 --> 00:08:08.550 But what if all of the authors are government employees? 00:08:08.550 --> 00:08:13.290 That would throw out the joint works confusion there. 00:08:13.290 --> 00:08:15.850 There also, we wanted to look at trends 00:08:15.850 --> 00:08:18.505 in a type of copyright assertion, 00:08:18.505 --> 00:08:20.560 as well as disciplinary differences, 00:08:20.560 --> 00:08:22.600 which we'll see in just a moment. 00:08:22.600 --> 00:08:24.970 We're not the only ones who have looked at this. 00:08:24.970 --> 00:08:26.740 Others have looked at the issue of 00:08:26.740 --> 00:08:29.065 copyright marking practices before. 00:08:29.065 --> 00:08:32.140 For example, Paul Royster, in commenting on NISO's 00:08:32.140 --> 00:08:35.845 proposed open access metadata standards in 2014, 00:08:35.845 --> 00:08:37.780 provided eight examples of 00:08:37.780 --> 00:08:40.725 incorrect copyright assignment in scholarly journals, 00:08:40.725 --> 00:08:44.045 all of which dealt with US government works. 00:08:44.045 --> 00:08:46.835 April Hathcock in 2016, 00:08:46.835 --> 00:08:49.055 noted conflicting copyright statements 00:08:49.055 --> 00:08:50.690 on the same work as 00:08:50.690 --> 00:08:51.920 part of her critique of 00:08:51.920 --> 00:08:55.529 the ethical dimensions of copyright and reuse. 00:08:56.000 --> 00:08:58.970 We collected two datasets, 00:08:58.970 --> 00:09:01.595 one from the National Cancer Institute 00:09:01.595 --> 00:09:04.790 and the other from the US Department of Agriculture. 00:09:04.790 --> 00:09:08.870 Both datasets from the calendar year 2019. 00:09:08.870 --> 00:09:12.549 We derive this from the Web of Science. 00:09:12.549 --> 00:09:15.765 These articles do include coauthors. 00:09:15.765 --> 00:09:19.310 This search retrieves any article that 00:09:19.310 --> 00:09:23.620 has a government author on it from those two agencies. 00:09:23.620 --> 00:09:27.589 You can see our search strategy at the bottom. 00:09:27.589 --> 00:09:32.735 We use the affiliation data for the two agencies. 00:09:32.735 --> 00:09:35.810 Then we filtered by 00:09:35.810 --> 00:09:38.090 the article type and then 00:09:38.090 --> 00:09:41.180 refined by the publication year 2019. 00:09:41.180 --> 00:09:43.415 For the National Cancer Institute, 00:09:43.415 --> 00:09:46.790 this resulted in 2,378 articles issued in 00:09:46.790 --> 00:09:51.600 2019 and for USDA, over 6,000 articles. 00:09:52.980 --> 00:09:56.800 These are large datasets and we wanted to do 00:09:56.800 --> 00:10:00.580 a random sample to look at these manually. 00:10:00.580 --> 00:10:03.310 To get a 95 percent confidence level 00:10:03.310 --> 00:10:05.440 and five percent margin of error, 00:10:05.440 --> 00:10:12.040 we got the resulting random samples of 318 articles for 00:10:12.040 --> 00:10:16.930 the NCI dataset and 362 articles for 00:10:16.930 --> 00:10:22.000 the USDA dataset and we each looked at these manually, 00:10:22.000 --> 00:10:26.890 Jimmy taking the NCI dataset and I took the USDA dataset. 00:10:26.890 --> 00:10:31.150 We decided to look for the copyright statement on the PDF 00:10:31.150 --> 00:10:33.865 and so we were always looking at the PDF 00:10:33.865 --> 00:10:36.610 for whatever copyright statement was there. 00:10:36.610 --> 00:10:39.130 That's frequently referred to as the version of 00:10:39.130 --> 00:10:42.010 record and of course it's portable and 00:10:42.010 --> 00:10:44.740 gets passed around from person to person and so it's 00:10:44.740 --> 00:10:46.510 important to have a good copyright statement 00:10:46.510 --> 00:10:48.505 on that document. 00:10:48.505 --> 00:10:53.230 Of course, we are both fortunate to be at 00:10:53.230 --> 00:10:55.960 a major research institutions where we have 00:10:55.960 --> 00:11:00.130 access to Web of Science as well as the PDFs. 00:11:00.130 --> 00:11:01.915 One thing I should mention is 00:11:01.915 --> 00:11:03.610 that we would have preferred to use 00:11:03.610 --> 00:11:05.590 a non-proprietary source for 00:11:05.590 --> 00:11:08.515 the bibliographic information such as Lens. 00:11:08.515 --> 00:11:10.720 I did do this search for 00:11:10.720 --> 00:11:13.360 USDA on the Lens site and only retrieved 00:11:13.360 --> 00:11:15.385 about a third of the articles 00:11:15.385 --> 00:11:18.460 and so we decided to go with Web of Science instead. 00:11:18.460 --> 00:11:20.470 Both Jimmy and I always 00:11:20.470 --> 00:11:22.780 found that each of the articles did 00:11:22.780 --> 00:11:24.580 have a government author on there so we feel 00:11:24.580 --> 00:11:27.770 like the accuracy was quite good. 00:11:28.470 --> 00:11:31.030 When we manually looked at these articles, 00:11:31.030 --> 00:11:33.820 we put them into one of five designations. 00:11:33.820 --> 00:11:36.280 The first was copyright statement missing, 00:11:36.280 --> 00:11:37.600 we could not find a copyright 00:11:37.600 --> 00:11:39.475 statement anywhere on the PDF. 00:11:39.475 --> 00:11:41.875 The second was a generic statement, 00:11:41.875 --> 00:11:44.199 something along the lines of copyright, 00:11:44.199 --> 00:11:47.245 Wiley 2019 something like that. 00:11:47.245 --> 00:11:49.870 Then we collapsed all the Creative Commons 00:11:49.870 --> 00:11:52.945 licenses into one Creative Commons category. 00:11:52.945 --> 00:11:57.280 Then of course, we want to look at the ones that 00:11:57.280 --> 00:12:01.705 did recognize Public Domain in the US on the article. 00:12:01.705 --> 00:12:04.060 Finally, we had a category 00:12:04.060 --> 00:12:08.350 for CC0, public domain dedication. 00:12:08.350 --> 00:12:10.780 In a few cases there were 00:12:10.780 --> 00:12:13.945 more than one copyright statement on the article. 00:12:13.945 --> 00:12:16.840 For example, there were a few articles that said 00:12:16.840 --> 00:12:20.460 public domain in the United States and CC-BY. 00:12:20.460 --> 00:12:23.340 When that happened, we always put the article in 00:12:23.340 --> 00:12:28.670 the public domain category. Jimmy. 00:12:29.160 --> 00:12:35.800 This is what we found and looking across these samples, 00:12:35.800 --> 00:12:39.985 roughly combined with the NCI and the USDA, 00:12:39.985 --> 00:12:43.450 13 percent actually recognize 00:12:43.450 --> 00:12:45.980 the US government authorship. 00:12:46.410 --> 00:12:50.785 More interestingly, across both sets, 00:12:50.785 --> 00:12:52.180 I don't know if it's more interesting, 00:12:52.180 --> 00:12:57.235 but 28 percent use the Creative Commons. 00:12:57.235 --> 00:13:00.715 Almost half of the articles in the sample 00:13:00.715 --> 00:13:04.375 had a generic copyright statement and that would be 00:13:04.375 --> 00:13:07.300 no recognition at all that there was 00:13:07.300 --> 00:13:12.130 a US government author and very 00:13:12.130 --> 00:13:17.785 few with a CC0 Public Domain dedication. 00:13:17.785 --> 00:13:20.920 We're going to talk later 00:13:20.920 --> 00:13:25.120 about the ones we found that had no copyright statement. 00:13:25.120 --> 00:13:27.820 The other thing that I'll just mention 00:13:27.820 --> 00:13:29.650 is that it was very 00:13:29.650 --> 00:13:34.719 interesting to us that with disciplinary differences, 00:13:34.719 --> 00:13:39.010 that these sets were pretty similar in terms of 00:13:39.010 --> 00:13:41.530 copyright labeling practices for government 00:13:41.530 --> 00:13:45.280 authored works and we found that somewhat surprising. 00:13:45.280 --> 00:13:49.270 Some examples of copyright assertions 00:13:49.270 --> 00:13:53.050 where there were clearly governmental authors. 00:13:53.050 --> 00:13:57.040 Here's the first one where the lead author 00:13:57.040 --> 00:13:59.050 was from the National Cancer Institute and 00:13:59.050 --> 00:14:01.780 four of the six authors were from the NCI. 00:14:01.780 --> 00:14:03.805 That simply says copyright, 00:14:03.805 --> 00:14:05.650 American Society for Clinical 00:14:05.650 --> 00:14:08.420 Oncology, all rights reserved. 00:14:08.600 --> 00:14:12.795 Likewise, the second example, 00:14:12.795 --> 00:14:15.975 only two co-authors from the NCI, 00:14:15.975 --> 00:14:20.190 no recognition of government authorship and pretty 00:14:20.190 --> 00:14:23.215 severe copyright statement that 00:14:23.215 --> 00:14:26.035 the society reserves all rights. 00:14:26.035 --> 00:14:30.220 The third one, an example where NCI was 00:14:30.220 --> 00:14:33.820 the lead author and four of seven authors 00:14:33.820 --> 00:14:37.765 were from NCI and Springer Nature simply says yes, 00:14:37.765 --> 00:14:39.700 we got the copyright here. 00:14:39.700 --> 00:14:43.270 Likewise, we saw the same type of thing for 00:14:43.270 --> 00:14:47.689 USDA and not much disciplinary difference 00:14:49.080 --> 00:14:54.940 whether it's all USDA co-authors or clearly 00:14:54.940 --> 00:14:59.380 some significant contribution from the USDA and this was 00:14:59.380 --> 00:15:04.480 across all of societies and publishers. 00:15:04.480 --> 00:15:06.760 The examples of government authorship 00:15:06.760 --> 00:15:09.130 were also really interesting. 00:15:09.130 --> 00:15:11.740 There are some similarities here, 00:15:11.740 --> 00:15:14.650 but bolded it's interesting what 00:15:14.650 --> 00:15:19.550 the different publishers are considering is copyrighted. 00:15:19.680 --> 00:15:22.435 Is it the work and its text? 00:15:22.435 --> 00:15:24.565 Is it just the article? 00:15:24.565 --> 00:15:29.815 There's also the copyright symbol on some of these. 00:15:29.815 --> 00:15:32.830 It was an even variety 00:15:32.830 --> 00:15:34.060 as you can see in the middle of 00:15:34.060 --> 00:15:36.700 this chart, within Wiley. 00:15:36.700 --> 00:15:41.725 A very interesting example from Taylor and Francis, 00:15:41.725 --> 00:15:47.110 where they have the same boilerplate information at 00:15:47.110 --> 00:15:49.510 the top of the copyright statement that the work was 00:15:49.510 --> 00:15:53.065 offered as part of the contributors official duties, etc. 00:15:53.065 --> 00:15:59.560 But then below, different open access addendums, 00:15:59.560 --> 00:16:02.515 I would call them probably incorrectly. 00:16:02.515 --> 00:16:05.080 But one that has 00:16:05.080 --> 00:16:07.870 OA Creative Commons public domain mark 00:16:07.870 --> 00:16:11.305 and the other that lists Creative Commons, 00:16:11.305 --> 00:16:13.900 non-commercial no derivatives license. 00:16:13.900 --> 00:16:15.580 This is interesting in a couple of 00:16:15.580 --> 00:16:20.080 ways since the article is in the public domain in the US, 00:16:20.080 --> 00:16:22.360 how that could conflict with 00:16:22.360 --> 00:16:25.270 a non-commercial no derivatives licensed and 00:16:25.270 --> 00:16:29.110 also since these Creative Commons licenses 00:16:29.110 --> 00:16:30.729 are international licenses, 00:16:30.729 --> 00:16:32.500 how that conflicts with 00:16:32.500 --> 00:16:34.180 the US government reserving 00:16:34.180 --> 00:16:36.160 the right to assert copyright. 00:16:36.160 --> 00:16:39.820 Were the NCI authors in a position to even 00:16:39.820 --> 00:16:45.320 agree to International Creative Commons license? 00:16:45.900 --> 00:16:48.745 We also wanted to look at 00:16:48.745 --> 00:16:50.905 the Elsevier articles which 00:16:50.905 --> 00:16:53.185 was the largest by any publisher. 00:16:53.185 --> 00:16:56.785 As you can see on the graph on the far left, 00:16:56.785 --> 00:17:00.939 almost a quarter of the sample was published by Elsevier. 00:17:00.939 --> 00:17:02.710 But there are a couple of other things 00:17:02.710 --> 00:17:03.730 that caught our eye that 00:17:03.730 --> 00:17:06.160 I've circled on this graph 00:17:06.160 --> 00:17:08.920 and the first one is 00:17:08.920 --> 00:17:11.050 that of the articles 00:17:11.050 --> 00:17:13.450 that were missing a copyright statement, 00:17:13.450 --> 00:17:18.054 a large number of them were from Elsevier, 00:17:18.054 --> 00:17:20.455 in fact 75 percent. 00:17:20.455 --> 00:17:22.975 We found it unusual that 00:17:22.975 --> 00:17:25.915 the largest publisher of academic journals in the world 00:17:25.915 --> 00:17:27.160 was simply not putting 00:17:27.160 --> 00:17:31.675 any copyright statement at all on their articles. 00:17:31.675 --> 00:17:36.670 The other thing was that almost none, 00:17:36.670 --> 00:17:39.580 in fact just one of the Elsevier articles 00:17:39.580 --> 00:17:40.870 and I'll just say that the articles in 00:17:40.870 --> 00:17:42.985 the public domain in the United States. 00:17:42.985 --> 00:17:45.070 Between these two things, 00:17:45.070 --> 00:17:47.260 we thought it was quite egregious for 00:17:47.260 --> 00:17:49.750 the world's largest publisher 00:17:49.750 --> 00:17:55.310 to do such a poor job of labeling these articles. 00:17:56.520 --> 00:17:59.350 Of these missing articles, 00:17:59.350 --> 00:18:00.610 a large number had something 00:18:00.610 --> 00:18:02.275 like published by Elsevier at the bottom, 00:18:02.275 --> 00:18:05.260 but nothing in the way of the copyright statement. 00:18:05.260 --> 00:18:08.170 And we realize that there's no legal requirement to do so, 00:18:08.170 --> 00:18:11.830 but we are suggesting that it's a good practice in 00:18:11.830 --> 00:18:13.090 scholarly communication to have 00:18:13.090 --> 00:18:16.255 a copyright statement on the PDF. 00:18:16.255 --> 00:18:20.170 This doesn't encourage scholarly sharing 00:18:20.170 --> 00:18:22.705 or use in the least. 00:18:22.705 --> 00:18:24.850 Thirteen of these articles 00:18:24.850 --> 00:18:27.370 have sole affiliations from the USDA, 00:18:27.370 --> 00:18:30.205 and so again, nothing 00:18:30.205 --> 00:18:34.460 acknowledging the public domain whatsoever. 00:18:34.970 --> 00:18:38.415 We also wanted to look at the open-access publishers. 00:18:38.415 --> 00:18:41.130 One of the things that caught my eye in the USDA 00:18:41.130 --> 00:18:44.765 set were the PLoS journals. 00:18:44.765 --> 00:18:48.430 Almost all of them except for 00:18:48.430 --> 00:18:53.290 one had a CC0 public domain dedication. 00:18:53.290 --> 00:18:56.320 I emailed PLoS about this and one of 00:18:56.320 --> 00:18:58.630 their editorial managers emailed me back and said, 00:18:58.630 --> 00:19:02.455 ''Yes indeed, when there's a US government author on one of our articles, 00:19:02.455 --> 00:19:07.495 we assign the CC0 public domain dedication." 00:19:07.495 --> 00:19:09.880 The numbers in parens here are the numbers in 00:19:09.880 --> 00:19:12.910 the combined sample datasets. 00:19:12.910 --> 00:19:17.035 We had 18 articles from PLoS between NCI and USDA. 00:19:17.035 --> 00:19:19.060 I also noticed the CC0 from 00:19:19.060 --> 00:19:22.255 PeerJ and emailed them as well, 00:19:22.255 --> 00:19:25.614 and they said it was always author choice. 00:19:25.614 --> 00:19:27.970 We only had two articles by them, 00:19:27.970 --> 00:19:32.695 but one set of authors chose CC0 and the other one CC BY. 00:19:32.695 --> 00:19:36.820 The other large open access publishers, MDPI, Frontiers, 00:19:36.820 --> 00:19:41.410 and BioMed Central use their standard CC BY license. 00:19:41.410 --> 00:19:43.090 But it's interesting that there's 00:19:43.090 --> 00:19:45.250 no acknowledgment from any of 00:19:45.250 --> 00:19:47.530 these publishers really that these 00:19:47.530 --> 00:19:48.700 are works that are in the public 00:19:48.700 --> 00:19:50.095 domain in the United States. 00:19:50.095 --> 00:19:53.710 At least PLoS is recognizing it by assigning 00:19:53.710 --> 00:19:58.580 an international license. Jimmy. 00:19:59.880 --> 00:20:02.500 It's also interesting as we 00:20:02.500 --> 00:20:05.170 think about disciplinary differences. 00:20:05.170 --> 00:20:09.415 When we back up and look at the entire data set, 00:20:09.415 --> 00:20:11.170 not just the sample, 00:20:11.170 --> 00:20:19.900 the full 8,523 articles that were published in 2019, 00:20:19.900 --> 00:20:22.795 where we found very little differences 00:20:22.795 --> 00:20:25.150 of copyright labeling practices. 00:20:25.150 --> 00:20:26.890 These articles did have 00:20:26.890 --> 00:20:33.535 a very different open access portrait. 00:20:33.535 --> 00:20:37.660 Whereas 21 percent of 00:20:37.660 --> 00:20:41.170 the articles from the NCI were paywalled, 00:20:41.170 --> 00:20:46.360 not open access. Twice as many, 00:20:46.360 --> 00:20:51.115 43 percent from the USDA were paywalled. 00:20:51.115 --> 00:20:53.950 We think this is largely due to 00:20:53.950 --> 00:20:58.300 the NIH mandate for deposit and 00:20:58.300 --> 00:21:00.220 our definition of being 00:21:00.220 --> 00:21:03.100 paywalled versus open access 00:21:03.100 --> 00:21:05.515 comes from the Web of Science, 00:21:05.515 --> 00:21:12.190 Our Research classification that is there. 00:21:12.190 --> 00:21:17.320 Combined in 2019 of these US offered works, 00:21:17.320 --> 00:21:22.870 more than a third were not available on open access. 00:21:22.870 --> 00:21:26.260 This gets even more interesting as we 00:21:26.260 --> 00:21:29.755 look historically at what 00:21:29.755 --> 00:21:34.780 a large chunk of literature is bottled 00:21:34.780 --> 00:21:39.805 up behind paywalls that should be in the public domain. 00:21:39.805 --> 00:21:41.860 We're talking about two pretty 00:21:41.860 --> 00:21:43.720 big US government agencies, 00:21:43.720 --> 00:21:46.660 but we're not talking about all of them. 00:21:46.660 --> 00:21:52.615 Between the NCI and the USDA over 50 years, 00:21:52.615 --> 00:21:57.415 almost 200,000 articles are behind paywalls. 00:21:57.415 --> 00:21:59.890 Not surprisingly, we see 00:21:59.890 --> 00:22:06.850 a larger percentage back in the '70s and '80s, 00:22:06.850 --> 00:22:11.380 before a lot of open access advocacy. 00:22:11.380 --> 00:22:14.665 We also wanted to look at the OA status 00:22:14.665 --> 00:22:18.445 of the articles or the public domain statement on them. 00:22:18.445 --> 00:22:20.260 From the USDA set, 00:22:20.260 --> 00:22:23.800 there was 51 articles or 14 percent of the sample set, 00:22:23.800 --> 00:22:27.040 and for NCI it was 35 of 00:22:27.040 --> 00:22:30.550 the articles in that set or 11 percent. 00:22:30.550 --> 00:22:32.950 We use the Unpaywall simple 00:22:32.950 --> 00:22:34.990 query tool and we're very grateful to 00:22:34.990 --> 00:22:37.810 Unpaywall where you can get 00:22:37.810 --> 00:22:41.590 a set of DOIs and enter them into that web tool, 00:22:41.590 --> 00:22:44.425 and they email the spreadsheet 00:22:44.425 --> 00:22:46.360 of the OA status for those works. 00:22:46.360 --> 00:22:50.470 You can see in the table below that over half, 00:22:50.470 --> 00:22:52.435 now I'm on the far right, 00:22:52.435 --> 00:22:55.310 the published version was open. 00:22:56.100 --> 00:22:58.870 But a few had 00:22:58.870 --> 00:23:03.070 the accepted version was the only one that was open. 00:23:03.070 --> 00:23:05.335 Then for submitted version or pre-prints, 00:23:05.335 --> 00:23:07.720 the only two came from the NCI dataset. 00:23:07.720 --> 00:23:10.030 But the category that really interested me the 00:23:10.030 --> 00:23:12.040 most as a repository manager is 00:23:12.040 --> 00:23:13.450 the column that says 00:23:13.450 --> 00:23:17.290 no OA version and so they couldn't find any OA version, 00:23:17.290 --> 00:23:20.590 18 articles in the USDA sample set and 00:23:20.590 --> 00:23:24.175 five in the NCI sample set. 00:23:24.175 --> 00:23:27.340 From the perspective of a repository manager, 00:23:27.340 --> 00:23:29.920 if you can provide the only OA copy, 00:23:29.920 --> 00:23:31.510 then you're probably going to get a lot 00:23:31.510 --> 00:23:34.340 of views and downloads. 00:23:34.770 --> 00:23:37.750 We also wanted to look at cases 00:23:37.750 --> 00:23:41.064 where all of the authors are US government employees. 00:23:41.064 --> 00:23:43.960 We did this for the USDA's sample, 00:23:43.960 --> 00:23:49.735 which again was 362 articles out of the larger dataset. 00:23:49.735 --> 00:23:53.395 We found that 35 of them or 10 percent 00:23:53.395 --> 00:23:56.845 were authored only by USDA authors. 00:23:56.845 --> 00:24:00.850 10 of those were marked as the public domain in the US, 00:24:00.850 --> 00:24:05.260 23 were paywalled, or 00:24:05.260 --> 00:24:07.510 about two-thirds and that 00:24:07.510 --> 00:24:12.140 included six that were marked public domain in the US. 00:24:12.360 --> 00:24:15.160 We can see a divide here between 00:24:15.160 --> 00:24:17.470 the licensing and the access, 00:24:17.470 --> 00:24:19.990 where if you see a work that's in the public domain 00:24:19.990 --> 00:24:21.370 in the United States, and you can't get 00:24:21.370 --> 00:24:24.025 access to it, it's rather frustrating. 00:24:24.025 --> 00:24:26.875 That's what we see on the screenshot. 00:24:26.875 --> 00:24:29.425 This is from a Wiley journal. 00:24:29.425 --> 00:24:31.720 Although we were looking at PDFs, 00:24:31.720 --> 00:24:35.275 this one is clearly marked on the website as well. 00:24:35.275 --> 00:24:37.270 This article has been contributed to by 00:24:37.270 --> 00:24:38.860 US government employees and their work 00:24:38.860 --> 00:24:41.290 is in the public domain in the USA. 00:24:41.290 --> 00:24:44.529 If you go to click the little PDF icon, 00:24:44.529 --> 00:24:47.455 then you get this. You don't have access. 00:24:47.455 --> 00:24:49.480 You need to log in to an institution 00:24:49.480 --> 00:24:53.030 or pay a fee to get access to that article. 00:24:53.670 --> 00:24:56.230 One of the other interesting things that we 00:24:56.230 --> 00:24:58.195 found in this research was that 00:24:58.195 --> 00:25:00.130 the US Forest Service has 00:25:00.130 --> 00:25:01.840 a repository where they are 00:25:01.840 --> 00:25:04.255 actively taking these articles, 00:25:04.255 --> 00:25:06.520 whether they have a statement on them or not, 00:25:06.520 --> 00:25:07.945 whether they're open or not, 00:25:07.945 --> 00:25:10.510 and putting them in their repository, 00:25:10.510 --> 00:25:13.135 it's called Treesearch. 00:25:13.135 --> 00:25:15.010 They say that they are taking 00:25:15.010 --> 00:25:16.870 articles published by the agency, 00:25:16.870 --> 00:25:18.490 as well as those published by others, 00:25:18.490 --> 00:25:20.530 including papers appearing journals, 00:25:20.530 --> 00:25:22.405 conference proceedings, or books. 00:25:22.405 --> 00:25:25.000 And they take a much stronger stand than CENDI does 00:25:25.000 --> 00:25:28.045 on the public domain status of these works. 00:25:28.045 --> 00:25:31.000 "US copyright regulations require that 00:25:31.000 --> 00:25:33.790 publications by the federal government or authored 00:25:33.790 --> 00:25:35.800 by federal employees must remain in 00:25:35.800 --> 00:25:40.190 the public domain." Jimmy. 00:25:42.210 --> 00:25:47.275 An interesting piece on the NCI data set, 00:25:47.275 --> 00:25:52.345 because of the disciplinary nature 00:25:52.345 --> 00:25:55.360 of biomedical research, 00:25:55.360 --> 00:26:00.715 is it really shows the co-author confusion. 00:26:00.715 --> 00:26:04.480 A number of many of the articles in the data set had 00:26:04.480 --> 00:26:08.424 many different co-authors and this is just one example, 00:26:08.424 --> 00:26:12.235 but it was not the example with the most co-authors, 00:26:12.235 --> 00:26:16.465 but here's one article that has 35 different co-authors, 00:26:16.465 --> 00:26:20.500 two of which work at NCI. 00:26:20.500 --> 00:26:22.870 In those 35 co-authors were 00:26:22.870 --> 00:26:25.285 talking about nine different countries, 00:26:25.285 --> 00:26:27.355 a mix of universities, 00:26:27.355 --> 00:26:31.855 governmental or non-profit, employers. 00:26:31.855 --> 00:26:34.510 The copyright statement on 00:26:34.510 --> 00:26:37.060 this article is, copyright the authors, 00:26:37.060 --> 00:26:38.530 under exclusive license to 00:26:38.530 --> 00:26:43.945 Springer Nature America, Incorporated, 2019. 00:26:43.945 --> 00:26:48.310 The attribution statement of 00:26:48.310 --> 00:26:53.320 credit in terms of author contributions, 00:26:53.320 --> 00:26:56.515 notes that, EV and RS, 00:26:56.515 --> 00:26:58.465 that's highlighted on the slide, 00:26:58.465 --> 00:27:01.670 those were the NCI authors. 00:27:02.010 --> 00:27:07.645 Well, they contributed to writing the manuscript, 00:27:07.645 --> 00:27:10.450 and the contribution statement also notes 00:27:10.450 --> 00:27:13.765 that all authors discussed and approved the manuscript. 00:27:13.765 --> 00:27:17.830 It's hard to know how much work they actually did and 00:27:17.830 --> 00:27:19.810 whether two out of 35 should put 00:27:19.810 --> 00:27:22.090 this in the public domain, 00:27:22.090 --> 00:27:25.645 but it's also hard to note, 00:27:25.645 --> 00:27:28.900 did Springer Nature actually get 35 signed 00:27:28.900 --> 00:27:31.390 copyright clearance statements from 00:27:31.390 --> 00:27:33.025 all the authors or not. 00:27:33.025 --> 00:27:37.045 So some of our preliminary conclusions 00:27:37.045 --> 00:27:40.540 based on this earlier research, is that, 00:27:40.540 --> 00:27:43.180 there's a significant number of incorrect assertions of 00:27:43.180 --> 00:27:47.080 copyright upwards to 46 percent, 00:27:47.080 --> 00:27:53.425 and that this inconsistent labeling practices 00:27:53.425 --> 00:27:56.795 for US works especially. 00:27:56.795 --> 00:27:59.175 We found this to be consistent 00:27:59.175 --> 00:28:01.605 across two very different disciplines and 00:28:01.605 --> 00:28:03.120 agencies that had 00:28:03.120 --> 00:28:05.520 very different open access profiles 00:28:05.520 --> 00:28:08.315 for their publications. 00:28:08.315 --> 00:28:14.875 Beyond the rational preliminary conclusions, 00:28:14.875 --> 00:28:20.650 you'll give me a brief moment for a wider thinking. 00:28:20.650 --> 00:28:23.425 It definitely raises the question whether copyright, 00:28:23.425 --> 00:28:26.709 to me, is a good fit for scholarship. 00:28:26.709 --> 00:28:30.235 I'll quote Kevin Smith, 00:28:30.235 --> 00:28:34.735 where in the previous session, where he said, 00:28:34.735 --> 00:28:36.370 copyright is a tool that gives 00:28:36.370 --> 00:28:39.160 monopoly power that's leveraged 00:28:39.160 --> 00:28:41.140 against the academy and to me, 00:28:41.140 --> 00:28:44.695 I think we see this partially in this study. 00:28:44.695 --> 00:28:48.474 I think about the Budapest Open Access Initiative goals, 00:28:48.474 --> 00:28:51.520 that literature that should be freely accessible 00:28:51.520 --> 00:28:53.890 online is that which scholars 00:28:53.890 --> 00:28:56.785 give to the world without expectation of payment. 00:28:56.785 --> 00:29:00.160 I also was really struck 00:29:00.160 --> 00:29:03.565 by a recent book chapter by Charlotte Roh, 00:29:03.565 --> 00:29:07.810 Harrison Inefuku and Emily Drabinsky in the MIT book 00:29:07.810 --> 00:29:11.320 Reassembling Scholarly Communications that also 00:29:11.320 --> 00:29:15.145 goes to the social justice aspect of this, 00:29:15.145 --> 00:29:17.470 and we don't want to suggest at 00:29:17.470 --> 00:29:19.480 all that open access is going to 00:29:19.480 --> 00:29:21.985 cure everything, and they say, 00:29:21.985 --> 00:29:24.430 while open access publishing advances 00:29:24.430 --> 00:29:27.715 equitable access to reading scholarly work, 00:29:27.715 --> 00:29:29.650 it does not automatically reverse 00:29:29.650 --> 00:29:32.290 the biases and norms of scholarship 00:29:32.290 --> 00:29:35.470 itself and that was really resonant with 00:29:35.470 --> 00:29:39.160 us and if you haven't checked out this recent book, 00:29:39.160 --> 00:29:40.750 I would also recommend that, 00:29:40.750 --> 00:29:44.870 just a lot of great chapters. Philip. 00:29:46.380 --> 00:29:51.490 Some of the implications for libraries include, 00:29:51.490 --> 00:29:53.920 that the inconsistent labeling practices 00:29:53.920 --> 00:29:56.304 contribute to confusion on reuse. 00:29:56.304 --> 00:29:59.230 Particularly when statements are missing, 00:29:59.230 --> 00:30:01.870 you have obviously a government author 00:30:01.870 --> 00:30:05.360 on the article and there's no recognition of that. 00:30:05.390 --> 00:30:07.590 We also realized that there is 00:30:07.590 --> 00:30:10.440 a complicated risk analysis. 00:30:10.440 --> 00:30:13.940 Neither of us are lawyers, 00:30:13.940 --> 00:30:16.900 but as we saw earlier, 00:30:16.900 --> 00:30:19.090 international law is complicated, 00:30:19.090 --> 00:30:21.250 this is a jurisdictional public domain, 00:30:21.250 --> 00:30:22.570 and so what about 00:30:22.570 --> 00:30:25.270 the articles in the international arena, and of course, 00:30:25.270 --> 00:30:26.920 if we take one of the articles and put 00:30:26.920 --> 00:30:29.530 them in one of our repositories, 00:30:29.530 --> 00:30:34.630 that's available worldwide, and as we saw 00:30:34.630 --> 00:30:40.765 the joint work issue with co-authors is also a problem. 00:30:40.765 --> 00:30:42.550 If you have 100 authors and 00:30:42.550 --> 00:30:46.070 only one government author, what does that mean? 00:30:46.070 --> 00:30:50.040 There's also the issue of texts versus the work. 00:30:50.040 --> 00:30:52.035 For example, 00:30:52.035 --> 00:30:54.690 some publishers might claim copyright in the formatting, 00:30:54.690 --> 00:30:58.930 particularly in other countries internationally. 00:30:59.180 --> 00:31:02.025 As Jimmy mentioned earlier, 00:31:02.025 --> 00:31:04.770 license collisions for lack of a better word, 00:31:04.770 --> 00:31:07.890 we saw a few with two statements, 00:31:07.890 --> 00:31:09.710 public domain and CC, 00:31:09.710 --> 00:31:13.750 or public domain and copyrighted, that sort of thing, 00:31:13.750 --> 00:31:17.410 particularly looking at public domain 00:31:17.410 --> 00:31:18.685 versus some of the more 00:31:18.685 --> 00:31:20.770 limiting Creative Commons licenses, 00:31:20.770 --> 00:31:25.100 like the NC and ND licenses, how does that work? 00:31:26.130 --> 00:31:28.570 We have lots of questions. 00:31:28.570 --> 00:31:30.400 The first and perhaps most important 00:31:30.400 --> 00:31:32.440 coming out of this presentation is that, 00:31:32.440 --> 00:31:35.380 if we are comfortable with public domain in 00:31:35.380 --> 00:31:37.390 the US status of 00:31:37.390 --> 00:31:40.510 these articles with any government author on them, 00:31:40.510 --> 00:31:42.580 should there be a concerted effort to 00:31:42.580 --> 00:31:44.215 put them in open repositories, 00:31:44.215 --> 00:31:47.080 because as we've seen, many of them are not open? 00:31:47.080 --> 00:31:50.530 Could this effort apply to states where 00:31:50.530 --> 00:31:54.370 government authored work is in the public domain? 00:31:54.370 --> 00:31:56.020 So both Jimmy and I are in Virginia, 00:31:56.020 --> 00:31:58.090 which I believe is one of the states 00:31:58.090 --> 00:32:00.490 where government works are in the public domain, 00:32:00.490 --> 00:32:02.230 and so could an employee of 00:32:02.230 --> 00:32:07.030 our Fish and Wildlife Service be put in 00:32:07.030 --> 00:32:09.200 an open repository? 00:32:09.630 --> 00:32:11.770 And as you saw in 00:32:11.770 --> 00:32:14.380 Jimmy's slide about going back through the decades, 00:32:14.380 --> 00:32:16.420 could public domain in the US be a source 00:32:16.420 --> 00:32:18.745 of open access for older articles? 00:32:18.745 --> 00:32:21.160 This has been one of the confounding things 00:32:21.160 --> 00:32:22.465 about open access 00:32:22.465 --> 00:32:24.460 as it advances every 00:32:24.460 --> 00:32:27.040 year and there's more and more work that's open access. 00:32:27.040 --> 00:32:28.840 We're still looking back at 00:32:28.840 --> 00:32:30.520 a huge back catalog, so to 00:32:30.520 --> 00:32:33.655 speak, of works that are behind paywalls. 00:32:33.655 --> 00:32:37.000 And could there be a technological solution for 00:32:37.000 --> 00:32:39.730 this divide between the licenses and access, 00:32:39.730 --> 00:32:44.500 as we saw on the screenshot of the Wiley journal article 00:32:44.500 --> 00:32:46.840 which says that the articles in 00:32:46.840 --> 00:32:48.280 the public domain in the US, 00:32:48.280 --> 00:32:50.185 but no, you don't have access to it. 00:32:50.185 --> 00:32:53.740 Could publishers potentially recognize 00:32:53.740 --> 00:32:57.070 IP addresses coming from the United States, 00:32:57.070 --> 00:33:03.160 for example, and offer access to those IP addresses. 00:33:03.160 --> 00:33:07.375 There are also other opportunities for libraries. 00:33:07.375 --> 00:33:10.240 Perhaps some of these issues could be putting 00:33:10.240 --> 00:33:14.604 license addendums to specify labeling best practice, 00:33:14.604 --> 00:33:17.845 to have rights to republish public domain material, 00:33:17.845 --> 00:33:19.945 and perhaps most importantly, 00:33:19.945 --> 00:33:23.890 to remove paywalls for the public domain material. 00:33:23.890 --> 00:33:26.949 What might a solution at scale look like? 00:33:26.949 --> 00:33:29.815 As Jimmy pointed out on one of his slides, 00:33:29.815 --> 00:33:31.030 we're maybe looking at 00:33:31.030 --> 00:33:34.160 hundreds of thousands of articles here. 00:33:34.470 --> 00:33:40.240 One thing we could potentially do is use Unpaywall to 00:33:40.240 --> 00:33:43.390 identify the "no OA copy" articles 00:33:43.390 --> 00:33:46.525 and prioritize those for IR upload. 00:33:46.525 --> 00:33:49.225 Because there is no open access version, 00:33:49.225 --> 00:33:50.560 and so that could be 00:33:50.560 --> 00:33:52.570 the low-hanging fruit that we focus on first, 00:33:52.570 --> 00:33:54.175 to get into IRs. 00:33:54.175 --> 00:33:56.125 We may also be able to help 00:33:56.125 --> 00:33:58.180 Unpaywall mark these as public domain US. 00:33:58.180 --> 00:33:59.740 When we got our spreadsheets back, 00:33:59.740 --> 00:34:03.205 we noticed that very few were marked PD, 00:34:03.205 --> 00:34:07.975 and so we might be able to help improve the data set. 00:34:07.975 --> 00:34:11.140 Also, Unpaywall has relationships with 00:34:11.140 --> 00:34:13.870 Web of Science and Scopus, and so part 00:34:13.870 --> 00:34:15.895 of that relationship might be 00:34:15.895 --> 00:34:18.520 getting the affiliation data from 00:34:18.520 --> 00:34:20.830 those two services and using 00:34:20.830 --> 00:34:25.970 that to inform the license metadata. 00:34:26.910 --> 00:34:29.410 We want to offer our thanks. 00:34:29.410 --> 00:34:31.450 A couple of months ago we 00:34:31.450 --> 00:34:34.300 attended the ASERL Copyright Office Hours, 00:34:34.300 --> 00:34:36.925 the Association of Southeast Research Libraries 00:34:36.925 --> 00:34:38.905 with a lot of these questions, 00:34:38.905 --> 00:34:42.100 and we are grateful for feedback from Brandon Butler and 00:34:42.100 --> 00:34:46.030 Laura Burtle at those office hours. As Jimmy mentioned, 00:34:46.030 --> 00:34:47.740 Peter Hirtle was very gracious in 00:34:47.740 --> 00:34:50.500 replying by email to some of our questions. 00:34:50.500 --> 00:34:52.840 Our scholarly communications colleagues for 00:34:52.840 --> 00:34:55.630 engaging the question on LISTSERV, 00:34:55.630 --> 00:34:57.670 and of course, the Web of Science for 00:34:57.670 --> 00:35:00.985 the affiliation data and Unpaywall, 00:35:00.985 --> 00:35:03.970 Our Research for the open access data. 00:35:03.970 --> 00:35:06.910 But we do want to make it clear that any errors and 00:35:06.910 --> 00:35:10.910 misunderstandings are totally ours. 00:35:11.820 --> 00:35:15.174 We have some selected references 00:35:15.174 --> 00:35:16.840 to the documents that were referred to, 00:35:16.840 --> 00:35:19.915 for example, the CENDI documents, the FAQs, 00:35:19.915 --> 00:35:21.700 and several others with 00:35:21.700 --> 00:35:24.805 links to those works that we'll be providing, 00:35:24.805 --> 00:35:27.175 and so we want to thank you very much for listening, 00:35:27.175 --> 00:35:31.160 and we're anxious to take your questions.