In this paper I will discuss some of the major copyright issues that have arisen as a result of experiences of the eLib.
This paper assumes readers are familiar with the basic ground rules about who owns copyright, what it is, the restricted acts that are given to the copyright owner, the exceptions such as fair dealing and the library copying provisions, the lifetime of copyright, and so on. I would just emphasise a couple of points of law. Firstly, many people mistakenly call the library permissions "fair dealing ". "Fair dealing" is a defence against a copyright infringement action. An individual can claim this defence so long as the copying was "fair" - in other words does not damage the legitimate commercial interests of the copyright owner, and so long as it was for one of several purposes, such as for research or private study. The onus is on the person who copied to prove that it was for one of the specified purposes and that what was done does not damage the copyright owner. If he or she cannot prove this, then what he or she has done is infringement. Each case is taken on its merits, and there are NO safe limits. Copying 10% of one text may be fair, but of another may not be.
Fair dealing, then, is a risky defence. In contrast, academic and some other libraries have their library privileges which guarantees them against an infringement action. The copyright owner cannot even start the action, even if it wanted to. That's the up side; the down side is the library has to insist on a personally signed form, money must change hands, and the library must not supply more than one copy of the document to the patron at any time for essentially the same purpose - you know the rules. They are important. If you do not follow the rules, you are taking your chances that you may be sued. So please do not call library privileges "fair dealing" - the two are very different beasts.
The second point of law I would remind you of is that in UK law, no difference exists between print and machine readable data. Therefore, the fair dealing and library privileges apply just as much to electronic materials as they do to print. Don't be alarmed or put off by statements from some publishers that there is no such thing as fair dealing in an electronic environment. Fair dealing DOES apply to electronic data. This was confirmed by the recent conference in Geneva to update the Berne Convention, but already applied to UK law. So the real question is, what is "fair" in an electronic environment? I do not think anyone can doubt that what might be "fair" when photocopying will not be fair if you copy the same material and put it on a server. So "fair dealing" in an electronic environment will be tighter than in a print environment. Equally, if you are a librarian, you are permitted to make a copy of, say, a single item from a CD ROM or an online database and supply it to a patron who has signed the form, paid the money and so on. In all probability your contract with the CD ROM supplier or online host allows you to do this anyway, and indeed allows the librarian to ignore the requirement for a form or for payment, but just in case the librarian's contract is ambiguous or leaves this matter unsaid, librarians have that right.
In this paper, I highlight some of the issues that have arisen in eLib, and how they are being tackled, or (as in some cases) how they are proving to be serious obstacles. Some of the text is based on reports sent to me by eLib projects.
Some of the electronic journal  projects have involved capturing and displaying publishers' data. The projects receive data from many publishers. Typically, journal issues consist of article headers marked up in SGML and the full articles in PDF, or the entire article in SGML. From the SGML headers data items are extracted and stored in the object database that controls the application.
There are inconsistencies between what different publishers include within the copyright tags. Some publishers do not include a year with the copyright, others do not include a copyright string at all; some publishers have the same copyright string for all their titles, whereas others have a different copyright string for each journal. In some cases, the publisher has the year at the end of the copyright string, whereas others put the year at the start.
There are also problems with including the copyright, such as: how many times do you put the copyright notice up to protect the interests of the publishers, but not annoy the users as they navigate; for example, one publisher requested that a copyright notice be shown on every page of an article. This is presumably so that if a user prints off just one page of an article s/he will see the copyright. However the printed copy of the journal does not have the copyright on each page and someone could photocopy a page of that. Is there any justification for such an approach?
Some electronic journals include multimedia content with some articles, for example sound clips from the BBC Nine O'clock news, with accompanying photos of the speakers. This raises the question of whether a copyright notice for multimedia items should be displayed and where it should be displayed - by hypertext link? The copyright owner for multimedia clips is probably not the publisher or even the article author. In the case of the BBC clips, we know that they have given permission for their inclusion, but no-one knows who owns the copyright. There are also serious Moral Rights issues that arise in the case of multimedia - see my earlier article in Ariadne  on this topic.
Another major issue for electronic journals, and for other projects, is the inclusion of hypertext links on the Web to other materials. There are two issues here - is it an infringement of the copyright or Moral Rights of the original author of the article to amend his or her article by embedding such links? And is it an infringement of the copyright or the Moral Rights of the item to which you are linking to offer such links?
The questions are not easily answered. As far as the original author is concerned, the best way forward is to get their approval to make such amendments as are necessary to embed such links. It is much less easy to get the agreement of the owners of materials to whom you are linking. The current Shetland News versus Shetland Times  case in Scotland should act as a warning for everyone that you cannot willy nilly put links between your WWW pages and someone else's without careful thought [ed - you sure about this?]. I would urge everyone to exercise great care in the links they set up. Some caution also needs to be applied in the use of hypertext links for students. They can be used to direct students to services such as the Electronic Telegraph ; thereafter its up to the student to register and abide by the terms and conditions attached. Links can be used to guide the students to various URLs. However, unless express permission exists you should not apply links to third party material (articles, etc. that belong to someone else). However, URLs can be given to students along with advice such as that they may find this article (or whatever) interesting etc. This way, the student is being given a fact (a URL; no copyright in a fact) and they can decide for themselves if to copy it.
Copyright issues arise when a user requests a document to be delivered to them. The various EDD  projects offer such documents by means of scanning in the original, transmitting the scanned material to a remote site, and then printing it off. Typically, the document would be a serial article (or some pages from a monograph perhaps) and not copyright paid. In other words, we are dealing with library privileges here.
The user should be able to locate, order and receive documents in a simple, unmediated process from a place of work where he or she has network access. There is no question that the relevant technology is available. However, copyright has forced the various projects to incorporate inelegant solutions on users.
When the user creates an order for such a document, he or she is required to complete a copyright declaration form, sign it, and send it to the library. Nothing is done until receipt of the copyright declaration.
Documents are then sent to the server and stored pending use. Users are notified of arrivals when they next log on or optionally by email. Documents are printed by the administrator and the file is then deleted. The copy is collected by the user or goes by internal post. This means that the document is printed before it is seen by the user, and that the machine-readable copy is destroyed. This is a waste of paper if the user, on giving one look, says "this is not what I want"; it is also a waste of time if user wants legitimately to use machine readable extracts; remember, fair dealing applies equally to electronic formats as to print.
Frankly, you have no choice in this matter. The library privileges are designed to protect the interests of librarians, and all this formal bureaucracy is your protection. At present, the law insists that you receive a form, personally signed, and that money must change hands in some way before a document delivery request can be fulfilled. This may seem a nonsense in the electronic age, but that's the way it is. Likewise, you have no choice but to destroy to scanned material when the request is fulfilled. It is not the library's business to store copies of documents in this way, and if it did so, it would be outside the library privileges. It is no secret that publishers are deeply concerned by the growth of the document delivery services, and would vigorously resist any attempts to change the law - allowing, say, electronic signatures or e mail requests, or permitting the library to store the scanned material for future requests. They may be more amenable regarding libraries using their privileges to deliver a single copy of an article in machine readable rather than printed form - one of the JISC/PA Working Parties is working on this; for the moment though, although the law supports this approach, I would be wary of antagonising publishers by offering such a service.
These are two closely related areas  . Both relate to heavily used materials in academic institutions. In the case of electronic reserves this is analogous to short loan materials, where the student is told what is on short loan, and can borrow the material for a short time. An electronic reserve offers the material in machine readable form for the student to download, view or print. Typically, such materials will be needed over a very short time and then never needed again. In slight contrast, on demand publishing is an electronic equivalent to course packs. These are collections of photocopies of articles and book chapters that students can borrow. At present, they are created under a CLA licence. Material is scanned in, and then printed out on demand, and at a certain cost to the student. A printout is always the result. Course packs may be longer lasting than electronic reserves in that they represent basic background data that may change slowly over the years.
eLib has been funding experiments in both areas. Projects which are very short term - say, one year - pointed out to publishers that scanned material would only be held for one year and that we would write again if we wished to extend the life of the system. Such approaches result in rapid positive responses. On the other hand, longer term projects have tended to result in outright refusal, or demands for impossibly high royalties.
All this may be mere coincidence, but I conclude that permission is more easily obtained for short time spans. It is clearly important for administrators of electronic reserves to identify the period an item is required on the system and to be very careful not to ask for a longer time than is strictly required for the course/module.
These projects also threw up other copyright problems. Many publishers did not know if they held the rights to offer material to be scanned. All too often, their contracts with their authors are ambiguous on this. With the increasing aggressiveness of authors regarding the retention of rights, publishers are nervous about giving permissions, or else offer permissions but will not accept responsibility if it turns out the permission was not in their gift. This is quite unacceptable. Publishers must agree always to a warranty that they have the right to offer the permissions they give, and if it turns out they do not, they must accept liability.
A particular problem arose in project SCOPE  as follows: SCOPE wished to scan in some pages from a textbook plus some photos. The publisher gave permission for the words, but said it could not give permission regarding the photo, as it did not own the copyright to that. OK, let's not scan the photo thought SCOPE. But then there is a danger of infringing the Moral Rights of the author who might regard the photo as a key component of his or her document, and could sue for derogatory treatment if the photo was omitted. SCOPE had to drop all idea of including that material, caught as it was in this dilemma. This, I think will be a common problem in all cases of scanning copyright materials. I do urge you not to under-estimate the dangers of Moral Rights issues.
A final issue that arose in these projects was escrow. Most publishers initially demanded that once the eLib project was completed, the project team must destroy the machine readable data. This is frankly a silly request. Much time, trouble and expense have been expended in creating the data. Instead, the publishers should agree that at the end of the Project the data be placed in an escrow deposit, controlled by the publisher but with copyright in the machine readable data belonging to the Project (in other words, the publisher cannot make copies of the data without the project's agreement, but the project cannot release the data without the publisher's agreement.) If you are involved in any scanning licence deal, make sure you have such a clause.
One of the action lines under FIGIT has been educating and training library staff in aspects of the digital environment and its implications. This programme, too, has been hit by copyright problems!
One project involves workshops for HE librarians. They have nearly 100 HE librarians involved in taster and full programmes. The workshop leaders dish out handouts before, and during the workshops. These include articles from journals and pages from texts. Initially the materials were produced on site under the local CLA licence. However, as the project move to deliver the programme in the regions, they have been advised by some of the institutions involved that we cannot assume their CLA licences cover the Project team to have copying made for use in the workshops in various training venues around the country. This is because the trainees are all staff from HE institutions but they are not registered as a student on a course in any of the institutions where the workshops take place. This is just the tip of a broader iceberg regarding CLA licenses when materials are dished out to third parties off site, but one eLib project has been caught up in it.
One very clear message from these projects is the need for a strategy for clearing any third party material, and that it always takes more time and money than you would expect. Securing your clearances can take many months, and you might not receive a positive (or any) response at all.
You should decide very early on which third party material you wish to use, (plus substitutes), how you want to use it, and method of delivery. Third party materials need to be cleared for all planned uses, media and territories.
You need to have a procedure that includes identifying rights required for your project; time-scale for clearances that would include a cut off date; chase up date; reporting refusals, inserting substitutes; cancellations; paying for clearances and record keeping for future access and use of all assets cleared. If you send a request that requires them to construct a reply letter then they may just put it aside, whereas a letter that they can just sign and return gets it off their desk quicker and back to you quicker.
It is important to acknowledge all third party material, particularly in electronic media where ownership can be easily lost. As soon as you move outside text, the problems multiply. These are different industries, each with their own clearance culture and traditions.
As if that were not enough, another problem is growing. IT has moved the balance of what is possible towards exploiters and away from the creators. Many wish to exploit the technological possibilities to move information around freely without rewarding the creators. If the balance is moved too far towards access, then those who create will have less incentive to so do. You should be aware that authors' groups world-wide are flexing their muscles on these issues, and will not hesitate to sue both publishers or librarians if they feel their rights have been trampled on. So things will get more complex in the future for LIS people in the future, as they will have to deal with both publishers and authors!
What we need, of course, is a central clearance house representing HE in clearing materials dealing with a central clearing house of rights to copyright materials. Buying rights should be made no more difficult than going shopping in a supermarket. Alas, this ideal scenario of a simple, fast, reliable method of rights clearance and permissions is some way off.
There is a need to promote a greater understanding of intellectual property issues within the HEI community and - perhaps surprisingly - amongst publishers. Many institutions pass responsibility for copyright issues on to a member of library staff, who is expected to take care of copyright amongst a raft of other responsibilities. There is also a need for professional training relevant to the needs of HE staff and publishers. Arguably, there is a need for a central advisory unit to advise Universities on copyright problems.
There is an urgent need for more efficient and cost effective clearance mechanisms.
Things have improved vastly, but there is a need for an on-going dialogue between the HEIs, authors and copyright owners. Rights owners sometimes do not understand the needs of HE for quick and fairly costed clearance. Equally, academics tend to view publishers as money grabbing profiteers and do not understand many of the issues and problems facing the industry.
Finally, I must stress that things are not all doom and gloom. Initially publishers were wary and sometimes outright hostile to eLib projects, but are now much more relaxed. I am not saying all scepticism has vanished, but there is no question there has been a sea change in attitude in the last three years. Although problems have been encountered, the number of successful collaborations between HEIs and copyright owners is impressive and is growing all the time. eLib has led to a number of meetings at a high level between the academic community and the copyright owners. The JISC/PA Working Parties are progressing well, and should have all completed their work by April. The result will be a set of ground rules that all can agree on.
All told, comparing the situation now to three years ago, I would say things have gone better than I expected, vastly better than I feared, but copyright remains a serious problem, and has thrown up some unexpected difficulties in places. There are still many issues to be resolved, but the outlook is bright.
I would like to thank all those involved in eLib projects who responded to to my plea for help and provided me with the examples of copyright related problems that I have used as the basis of this article.
 Listing of Electronic Journals projects in the eLib programme
 Moral Rights and the Electronic Library - an article from a previous Issue of Ariadne
 Copyright battles: The Shetlands - Charles Oppenheims description of the copyright and legal dispute between two on-line newspapers in the Shetland Islands
 The Electronic Telegraph Web Site: a parallel version (with links to sites related to news items) of the Broadsheet Newspaper
 Listing of Electronic Document Delivery projects in the eLib programme
 Listing of Electronic Reserve projects in the eLib programme
 Listing of On Demand Publishing projects in the eLib programme
 Web pages for the Scope: Scottish On Demand Publishing Enterprise eLib project