(Note: I’m not a copyright expert. If I’ve made any factual errors here, please feel free to point them out. )
Last month, I spent a lot of time putting together an exhibit on the public domain. It was a very interesting process for many reasons.
I’d initially planned to organize it based on date. I wanted to show what had fallen into the public domain at different periods of history, in order to make a strong visual point about the end of the twentieth century, which has been characterized by no new materials entering the public domain. In the end, this was a bit too difficult and precise, so I organized it by medium instead, but as I was doing the research for this exhibit, I found myself thinking about it from a slightly different angle.
Since the 1978 copyright act worked retroactively to extend the length of copyright on everything that wasn’t already in the public domain, nothing has entered the public domain since it passed. Curiously, we have had another copyright extension between then and now, so we’ve pushed it back even further without allowing enough time for works to expire under the previous law. There’s plenty of commentary out there about the problems with extending it so far, but the problem that I think hasn’t been sufficiently discussed is the increased difficulty of identifying public domain material under the current term.
This is a serious problem with basing copyright terms on the life of the author rather than the date of publication. Prior to 1978, works would remain under copyright for a specified amount of time, with one extension allowed. The 1910 law allowed for a 28 year term with a possible extension for another 28. It’s possible to debate about whether that was long enough, too long, or whatever, but at least it was a known fact that after 56 years, a work would be in the public domain.
The 1978 law changed this certainty by basing copyright on the length of the author’s life. It said that copyright would last until the author had been dead for 50 years; in the 1990s, this was extended to 70 years. This means that an additional piece of information is needed to determine the copyright status of a work. In 1955, it was absolutely certain that every work published in 1899 or earlier was in the public domain. A person wanting to create interesting derivative works without a license could easily compile a list of works that had been published by that date and choose among them. After all, publication date is one of a few pieces of basic bibliographic information that is attached to everything that is published and can be easily searched. Works that had been published from 1927 onward might have been in the public domain, but it depended on whether they’d been renewed, so those works would have had to have been evaluated one by one (and finding out whether a copyright was renewed is not as easy as one might hope).
Right now, our situation is fairly similar, except that the dates in question are much more distant. The magic date is 1923; almost anything published before that is public domain (with some weird exceptions for GATT and music recordings, but let’s not get into that now). Works published between then and 1950 may be in the public domain if their copyright was not renewed. However, these works require additional labor to identify and so they get less attention. In general, 1923 is treated as the magic date because it’s a clear line and it’s relatively easy to work with. HathiTrust is trying to do some of this important identification work, but it’s a major undertaking to do anything like this on a major scale.
So, the practical public domain tends to end at the end of what can be easily and clearly identified—in this case, at the magic date of 1923. What does this mean in the future?
In 2019, barring any further copyright extensions, works from 1924 will enter the public domain, giving us a new magic date. (Works that were already under copyright when the 1978 act was passed were granted a 95-year term.) This will carry us through to 2072, when the new magic date becomes 1977—and that will be the new magic date. By then, some works will also have entered the public domain under the new rule—that is, all the work of any author who died in 2002 or earlier. So, a work published in 1979 would still be under copyright if its author had lived beyond that date, but a work published in 2001 by an author who died the following year would be copyright-free. This is odd, but what is more troubling to me is that the latter work would be unlikely to be treated as a public domain work. What’s more likely is that those looking for works in the public domain will focus their efforts on looking through works from the middle of the twentieth century. Works between 1977 and 70 years before the current year will fall into a zone of uncertainty where additional effort is needed to figure out what can and can’t be done. In my example this gap is only thirty years, it’s not a huge problem, but after this point, the magic date will no longer advance, so the gap will only grow.
Now, finding out who the author of a work is and when that person died isn’t a very difficult problem for an individual work. It’s certainly much easier than trying to figure out whether a copyright was renewed. My concern here has more to do with identifying many items at once. Suppose, for instance, that one wanted to engage in large scale digitization projects. With a date-based system, it is very simple to search for a particular date. On the other hand, the date of an author’s death is not immediately considered basic bibliographic information, and it’s quite difficult to get it included, since when a work is published, we do not usually know when the author will die—so the only way to do this is to find out, every time anybody dies, whether he or she had written, painted or recorded anything and then add that information to every record that exists. I’m not sure who is supposed to be responsible for keeping track of that. Even on a smaller scale, it’s not going to be easy to identify items that aren’t already known to be public domain, especially if they are not famous already.
Yes, this is a long term concern, it’s too late to worry about it anyway, and it’s not nearly as pernicious as some other disturbing things that have been happening in the realm of copyright lately, but I do think it’s something that wasn’t adequately considered when these laws were passed.