Cover Image

Not for Sale



View/Hide Left Panel

14

Final Plenary Discussion

MR. REICHMAN: I want to make a couple of observations about the discussion session that looked at the scenario in which Congress decides to enact a strong property rights model protecting databases. I was a little bit uncertain and a little disappointed about the transformative-use discussion, given the findings that I thought came out of the discussion this morning that were new to me and rich with significance, including the movement toward machine-automated searching of other databases, the need to mine data, and the incredible serendipities that were coming out even in the private sector, especially in biotechnology, from doing these kinds of data operations.

Given this new information, the possibilities for a strong database right to interfere with the scientific community's ability to recombine data in complex new databases would wreak even more havoc than we had previously predicted. Everyone who has looked into this problem has said, look out for the danger that a so-called redistribution right can just disrupt the ability that scientists have now to take databases that they have paid to access and then take a piece of that and pieces of other things and make something new. I think that the consequences of getting in the way of that customary practice would be very grave. I am a little skeptical of the statement I heard that the party who owns the data will surely license them if there was a convenience.

I would like to call attention to the Feist case. The crux of the Feist case was that there was a very innovative new database being put together from multiple telephone directories, and a sole-source provider —one of the telephone companies—refused to deal, refused to grant one of the multiple licenses needed to produce the value-adding directory. I doubt that case would have gotten past the first-level court if there had not been a basic hard-nosed refusal to deal to begin with. As long as you are talking about widespread sole-source owners of data, that is a risk that has to be factored into the equation.

DR. STEFIK: I just want to build on that last comment by Jerry Reichman. I think I am saying the same thing in a different way. A lot of what we heard is that a tremendous value lay in the way that databases were combined. This was something that was really different in terms of what is happening inside the computer versus what is happening inside your head when you read lots of journal articles or multiple books. But that is part of the machinery of access.

There is a sense in which, if one database contributes a tiny piece to a larger thing that is used in some kind of combination, now you get a dilution of its contribution so that there is some threshold below which you no longer have the complexity of too much licensing or too much of any other kind of restriction on how it is being used. Some principle about that seems to be important for preserving the value that is so important to databases for scientific use. I think that needs to be a basic observation from the scientific community that we would like to see reflected in whatever protective means we are moving forward on. Certainly there are some provisions about that, but this becomes part of the rationale about what is special about the observations from this community.

DR. BERRY: I want to push a little bit further on this issue from the perspective of the user/generator, specifically the user/generator who is being supported by government funds.

We have to go back to the economic basis for funding from government. This research is



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 307
PROCEEDINGS OF THE WORKSHOP ON PROMOTING ACCESS TO SCIENTIFIC AND TECHNICAL DATA FOR THE PUBLIC INTEREST: AN ASSESSMENT OF POLICY OPTIONS 14 Final Plenary Discussion MR. REICHMAN: I want to make a couple of observations about the discussion session that looked at the scenario in which Congress decides to enact a strong property rights model protecting databases. I was a little bit uncertain and a little disappointed about the transformative-use discussion, given the findings that I thought came out of the discussion this morning that were new to me and rich with significance, including the movement toward machine-automated searching of other databases, the need to mine data, and the incredible serendipities that were coming out even in the private sector, especially in biotechnology, from doing these kinds of data operations. Given this new information, the possibilities for a strong database right to interfere with the scientific community's ability to recombine data in complex new databases would wreak even more havoc than we had previously predicted. Everyone who has looked into this problem has said, look out for the danger that a so-called redistribution right can just disrupt the ability that scientists have now to take databases that they have paid to access and then take a piece of that and pieces of other things and make something new. I think that the consequences of getting in the way of that customary practice would be very grave. I am a little skeptical of the statement I heard that the party who owns the data will surely license them if there was a convenience. I would like to call attention to the Feist case. The crux of the Feist case was that there was a very innovative new database being put together from multiple telephone directories, and a sole-source provider —one of the telephone companies—refused to deal, refused to grant one of the multiple licenses needed to produce the value-adding directory. I doubt that case would have gotten past the first-level court if there had not been a basic hard-nosed refusal to deal to begin with. As long as you are talking about widespread sole-source owners of data, that is a risk that has to be factored into the equation. DR. STEFIK: I just want to build on that last comment by Jerry Reichman. I think I am saying the same thing in a different way. A lot of what we heard is that a tremendous value lay in the way that databases were combined. This was something that was really different in terms of what is happening inside the computer versus what is happening inside your head when you read lots of journal articles or multiple books. But that is part of the machinery of access. There is a sense in which, if one database contributes a tiny piece to a larger thing that is used in some kind of combination, now you get a dilution of its contribution so that there is some threshold below which you no longer have the complexity of too much licensing or too much of any other kind of restriction on how it is being used. Some principle about that seems to be important for preserving the value that is so important to databases for scientific use. I think that needs to be a basic observation from the scientific community that we would like to see reflected in whatever protective means we are moving forward on. Certainly there are some provisions about that, but this becomes part of the rationale about what is special about the observations from this community. DR. BERRY: I want to push a little bit further on this issue from the perspective of the user/generator, specifically the user/generator who is being supported by government funds. We have to go back to the economic basis for funding from government. This research is

OCR for page 307
PROCEEDINGS OF THE WORKSHOP ON PROMOTING ACCESS TO SCIENTIFIC AND TECHNICAL DATA FOR THE PUBLIC INTEREST: AN ASSESSMENT OF POLICY OPTIONS producing a public good. A public good normally is one that does not lose value by use. In this particular situation, we are dealing with a public good that gains value by use. The intent, presumably, of the government supporting the research is to enhance this public good. Any activity that inhibits the distribution of information coming from this is inhibiting the generation of the public good. Anything consistent with that enhancement, any private use, is fine. Any private use that inhibits the distribution, inhibits the growth of the public good, is acting against the government's intent in providing the original support. This provides us with a criterion for defining what is an acceptable contract, what is an acceptable relationship between the private exploiter of the data and the community. This only applies to government-supported research, but that is certainly a significant part of what we have been talking about for the last 2 days. MR. LEAVITT: May I respond to that? There are some interesting aspects to it, in that the government provides the data and then you can use them but don't own them. Let's say the data on the temperatures outside are provided to me by the federal government and it has a source of information that becomes very productive and leads to a lot of work. The government's philosophy, which I subscribe to, is that it provided the temperature information and that, therefore, everybody has the right to see the source of the work. That is fine, and I accept that. What if I would like to keep the source of the work? I could go out and stick a thermometer right next to the government's thermometer and I will provide myself with the exact same information, which now would be mine. DR. BERRY: If you duplicate, that is fine. Certainly I am not going to go out and do a high-energy physics experiment to duplicate one done at a government laboratory. DR. SERAFIN: What about a book that is based upon government data, which is freely and openly available? It becomes a best seller and it falls under all the copyright laws, etc. Is that in any way in conflict with your thoughts on public good? DR. BERRY: Not at all. As long as the authors contributing to that book are free to distribute the information as they wish in their own channels, then that is consistent with continuing to sell the book as a best seller. MR. REICHMAN: There was one other point that is related to this issue. In the first breakout session there was some discussion of the government data problem. There was some concern about what I have been calling “no-capture,” some no-capture provision. There is a problem with government data when they go into the private sector, where they may be lost often because the government itself doesn't retain further copies of them and they are only available in the private firm's value-added format. One of the things we were toying with during the Senate Judiciary Committee negotiations last summer was a provision that would require maintenance on a cost-recovery basis of the original data set for scientific and educational purposes in case the data were not available anywhere. All I am saying is that there is a concern that once data have become privatized and value added, that the original data sets not be lost for scientific and educational purposes. DR. SAXON: Yes, it was in response to that, Jerry, that we recognized that there is tremendous risk associated with the no-capture problem. When you are transferring a data activity to a private entity, and they cease to exist or stop their data activity, the integrity of the data you are providing to be maintained cannot be guaranteed. DR. SERAFIN: I would like to address that in a slightly different realm. I think there are many circumstances where, in fact, the government will choose to toss away its database because it is provided by some other provider at a cost. I have experienced this in our own laboratory,

OCR for page 307
PROCEEDINGS OF THE WORKSHOP ON PROMOTING ACCESS TO SCIENTIFIC AND TECHNICAL DATA FOR THE PUBLIC INTEREST: AN ASSESSMENT OF POLICY OPTIONS where we gave up a certain type of service in order to provide a new one. Any time that is done, those people who use that old service are very concerned about it. Yet, the limitation on resources requires that those kinds of judgments be made from time to time. MR. JASZI: I think that one of the great advantages of today's session in helping clarify this debate is the opposition that was developed in the formulation of the panels in the two models of database protection legislation; one session focused on the strong property rights model, and the other on the unfair competition model. As I have listened to the summaries of those sessions and some of the discussions we have had since then, it occurs to me that it might be possible on that basis to summarize the apparent—apparent to me, at least—advantages of the second of those two approaches, that is, the focus on unfair competition. The paramount advantage of this approach, it seems to me, is that it would seek directly to address and identify problems of piratical or parasitical competition and the consequent disincentives to invest or create new knowledge, which that sort of unfair competition might bring about. At the same time, when compared to the strong property rights model, this unfair competition model would offer relatively greater clarity in terms of its terminology and concepts, and a relatively greater ease of interpretation without the likelihood that there would, in any new legislation, be ambiguity as to the meaning of terms and concepts. The unfair competition model could be done with greater certainty and clarity. It would require relatively few exceptions, exemptions, or carve outs, because the scope of the initial prohibition itself would be relatively narrow. Therefore, in particular, it would leave scientific, educational, and cultural uses, as well as noncompetitive, transformative commercial uses essentially untouched, since they, too, would fall outside the scope of the prohibitions. Finally, along the same lines, such a model would substantially eliminate the very real problems of term of protection, registration, and deposit that we have been worrying about in each of our different panels. I would say that one bit of progress that I see us having made here today is having developed what strikes me as a relatively powerful case for considering the advantages of the unfair competition model. DR. SERAFIN: Would anyone, a proponent of the first of those two models, like to speak? MR. BAUMGARTEN: I guess I don't think this is progress. I think it is regression. I think it is unfortunate that regression is the accepted notion of the unfair competition model as defined by the strictest definition as an economic issue between two parties. It leaves open wide gaps in protection. Misuse by someone who just wants to disseminate the information over the Internet because they happen to believe personally that that should be a free environment is open, as is misbehavior by customers. It requires buyers to invest in government contracts that not only require universities to impose additional contracts, but also requires universities to police their students. I really hesitate to say what I am going to say, because I think it is in Professor Perlman's area, but frankly, I am not sure that is the definition of competition that has been read out of the unfair competition law. Certainly it was read out of the trademark law. So, I would hope that the concept of competition is more economic than an activity that impedes the ability to invest and compete in the marketplace, rather than the sort of old-fashioned prevention of competition. It may give the advantage of a more finite field of operation, and therefore give you more specificity, or at least minimize the risks, but the area it leaves totally uncovered is a real concern.

OCR for page 307
PROCEEDINGS OF THE WORKSHOP ON PROMOTING ACCESS TO SCIENTIFIC AND TECHNICAL DATA FOR THE PUBLIC INTEREST: AN ASSESSMENT OF POLICY OPTIONS DR. SERAFIN: Since Professor Perlman is mentioned, do you want to comment? MR. PERLMAN: Well, it is all in the definition. I think the current bills that talk about unfair competition as actual or potential markets are, in fact, eliminated. It is not talking about improving investments; it is talking about exploiting all the economic gain from the investments. It may be much more—hopefully much more—than the initial investment itself. It seems to me that any activity that directly prevents the database owner or creator from exploiting the direct market would be actionable. I don't view it as requiring any kind of commercial benefit on the part of the defendant. I didn't hear anything in what he summed up that was contrary. There are nuances of definition and I think those certainly need to be worked on. To say that it is open to any kind of exploitation, I think, is much broader than even you are trying to suggest that you want. MR. BAUMGARTEN: We didn't set any restrictions on the definition of potential marketplace. MR. PERLMAN: I understand. If one were going to pursue a limited view, then the devil is in the details of how you define the scope. MR. SERAFIN: The devil is always in the details. MR. KLIPPER: (Remark off microphone.) MR. JASZI: I would certainly agree that the LaMacchia situation is of interest and concern. I am not at all sure that it could not be dealt with, as Harvey suggests, within an unfair competition law approach. MR. MAURER: We are interested in having enough incentives available for databases. You can always add more. I think people need to think about what the existing incentives are, whether they need to be more extensive, not that we give everybody every possible thing. DR. SERAFIN: This particular National Research Council committee and this particular workshop are focusing on the scientific and technical community's access to data. The principal incentive there is the advancement of knowledge. MR. LEAVITT: It is my understanding that when copyrighted data or information gets incorporated into a publication and the copyright remains as a restriction on the use of that package, even as the original document is enhanced by further research or whatever, that as long as that data still appear in the work, there is some ownership. This is the pattern as long as the information is still identified. Is that more or less correct in copyright law? MR. BAUMGARTEN: I think so. I just wouldn't use the word “information.” As long as there is copyrighted expression in a derivative work, the derivative work is limited by the rights under license. MR. LEAVITT: Somewhere in my dim recollection I recall that in patent law—and you can correct me on this—if you take and revise a patent and then, using the same device, make a significant improvement on it that was not obvious from the original device, then you can patent that second device without violating the patent right of that original holder. MR. BAUMGARTEN: You might be able to patent it, just like you can copyright derivative work. The question is whether you can exploit it beyond the terms of license to use the underlying work or the underlying facts. I am not sure that copyright differs. MR. MILES: I think I would disagree with that. The proposition is that data—the facts—are not protected by copyright. The expression is protected. So, someone else can use the factual information in your work, and put it in another work. The part that is pure factual information is not yours to control. Presumably, the publication has a new finding that is then protected by the copyright for that publication.

OCR for page 307
PROCEEDINGS OF THE WORKSHOP ON PROMOTING ACCESS TO SCIENTIFIC AND TECHNICAL DATA FOR THE PUBLIC INTEREST: AN ASSESSMENT OF POLICY OPTIONS DR. STEFIK: There was a question that came up earlier about copyright. If we have a new regime using the very clear example of a thermometer, where someone was duplicating the weather measurements from the government, it was said that you could go ahead and make your own database based on that. My understanding of copyright is that you are really talking about the expression of something and if you copy a form of that expression from someplace else, that would be an infringement. For databases, it strikes me as being clouded in this case if the information, no matter where you got it, happens to be the same information. If you get the source from a previous database versus measuring it yourself, are the rights in what you can do with it different? I am interested in any comments regarding the legal model here, the sui generis model versus the other regimes, whether they differ on this issue of the source of the information. MR. MAURER: I think the fundamental question here involves a little bit of a misunderstanding about copyright. Yesterday, Jim Lohr from Chemical Abstracts was worried about their copyright. The Chinese are copying the pages, the layout, the fonts, the expression in the words. All those things are copyright protected. If I go in there and I take the facts out, the same way I take out facts about World War II to write my own new book, those facts are not protected. Copyright gives substantial protection against that Chinese-style attack. That is valuable; it is part of the ongoing strategy. But the facts themselves are not protected, and that is the distinction. DR. STEFIK: I knew that was true of copyright law. The question is relative to how you treat the facts, whether any of the new regimes differ in their definition of what amounts to being the facts. Is it the source matter, if you measure it yourself? MR. MAURER: Independent invention is usually available as a defense in these things. DR. SERAFIN: I want to make another statement, then, about the distinction between patents and copyrights. A patent does, in my opinion, protect new knowledge that goes into the literature, and that literature is out there. That same new knowledge is not protected under a copyright, if that happens to be the only means through which it was published. Is that so? MR. MAURER: That is right, although you can't patent facts. DR. SERAFIN: I understand that, but once that new idea is put into print, it becomes fact, and the patent protects it. If it were first put into a normal publication, it would not be protected by the copyright. So, there is a distinction here, as I understand it. MR. JASZI: I would just take a crack at Mark Stefik's question. I think there is a difference and we owe it again to the contrast between the two groups of models under discussion. In the strong property rights model—which in my opinion incorporates both the European sui generis approach and the approach of H.R. 2652—the functional focus of protection is the fact or facts as such. That is, the prohibited act is the unauthorized extraction and use of factual data. It may be that the goal of protection is to enhance the status of the database as a whole, but the means of protection is a prohibition against the use of factual data. That is very different from copyright. I think it is also quite different from the way in which the focus on the unfair competition model functions. DR. STEFIK: Could you characterize the difference between unfair competition and sui generis relative to the thermometer example? Is there a difference, or do they both come out the same? MR. JASZI: I think they both treat that the same. MR. BAUMGARTEN: I think Peter would agree to amend his statement by the fact that

OCR for page 307
PROCEEDINGS OF THE WORKSHOP ON PROMOTING ACCESS TO SCIENTIFIC AND TECHNICAL DATA FOR THE PUBLIC INTEREST: AN ASSESSMENT OF POLICY OPTIONS you have to take enough of the facts to represent the taking of a substantial proportion of the collection, not just individual facts. At least, I hope you will. MR. JASZI: What the H.R. 2652 proposal says is that the taking has to be qualitatively or quantitatively substantial. The proposal also makes it clear that the taking of one fact cannot be a wrong. It does not deal explicitly with the question of whether, for example, the taking of two facts could be a wrong. MR. ONSRUD: I want to raise an issue in the context of higher education. It deals with contracting away rights, particularly relevant to government-funded projects in universities. In essence, some university chancellors had a discussion about university policies and decided that it shall be the policy of these universities that professors, within the scope of their employment, shall retain nonexclusive rights in their works of authorship. One could add, and in data sets they produce. Has there been any discussion of that? This doesn't seem to destroy any incentives. We can still publish, and deal with private publishers. Yet, there is also the safety net that if the publisher chooses to assert this new right, there are still nonexclusive rights in the data set or the intellectual work with the author at the university. Does that significantly harm incentives for publishers, is one question. The other issue is helping to ensure that government-funded work is always available and open to inquiry. DR. BERRY: I think your question of whether that makes it less attractive to publishers must be a second-order question. The first-order question, particularly coming back to government-supported research, is in what way does it affect the growth and spread of information within that community? I think sometimes we seem to take for granted, in this discussion, that every opportunity or potential opportunity should attract a commercial publisher. I think this is an incorrect position. There will be lots of databases, potential databases, that might be published commercially that won't be profit making, or will be marginal, that publishers should be ready to turn down. That is part of their responsibility when they assess it in the first place. MR. ONSRUD: What I was trying to do was look at another potential balance between assuring openness among academic researchers but also allowing this kind of marketplace interface. DR. BERRY: Let's just take that logical step further. Just because no commercial publisher chooses to publish it doesn't mean that it should not be distributed. In other words, if they don't, then the responsibility for distribution should fall to the funding agency or the researcher in the first place. MR. LAMBERT: My concern is that while patents and copyrights provide exclusive rights in exchange for public disclosure of information (e.g., how to make an invention or the expression of ideas), the proposed legislation will allow information to be removed from the public domain and sequestered in a database. DR. SERAFIN: Thank you. I found these last 2 days to be very interesting and well worth my time. I hope that the committee is going to be able to come to some rational and worthwhile conclusions that we will publish under the auspices of the National Research Council, and that this will help the nation deal with the issues raised here.

OCR for page 307
PROCEEDINGS OF THE WORKSHOP ON PROMOTING ACCESS TO SCIENTIFIC AND TECHNICAL DATA FOR THE PUBLIC INTEREST: AN ASSESSMENT OF POLICY OPTIONS Appendixes