back to the future

In the mid-80s, I worked at the U.S. Congress Office of Technology Assessment, or "OTA". OTA's job was to advise the Congress on technology policy issues ranging from nuclear energy to educational technology to biotech. A study I worked on had the ostentatious title Intellectual Property Rights in an Age of Information and Electronics, and it concerned the impact of digital media on copyright.

OTA no longer exists, but the leavings of the agency abound on the internet. Princeton has archived all or most of the studies published, and Googling around produces about 196,000 hits.

I was playing around with MSN's new search tool, performing the definitive test of any search engine (searching one's own name), and came upon an old speech I gave at the Library of Congress in 1987.

Now, please forgive the exercise in egoism (which, after all, is what a blog often is), but this lazy writer has decided to use that speech for today's post.

Its title is prosaic enough, but for a piece that was written pre-internet it is remarkably sturdy. I think I was successful enough framing the problem. When I have success framing the solution, I will be a wealthy man.

INFORMATION AS PROPERTY: Capitalism in the Information Age

Your participation in this conference, and your use of this medium, are very much at the center of an emerging upheaval in the concept of information as property; an upheaval that occurs, ironically, just as we enter an Information Age. The source of wealth for this information age rests, in part, on an obscure area of the law known as copyright.

Copyright is suffering from a growing number of anomalies - deviations from a general rule or policy -- which, like the epicycles invoked to preserve the Ptolemaic view of the solar system, can be understood as indications of some very fundamental difficulties with the concept of property rights in information. The problem can, I believe, be summed up in the following way: the copyright system grew out of the printing press, which placed the printer/publisher at the center of the known technological universe, but computers and telecommunications are creating a user-centered universe that copyright has yet to come to terms with.

Copyright and information technology, in contrast to their present antagonism, were originally wedded together, and coexisted in bliss for nearly three centuries. This marriage has, however, been on the rocks since the beginning of this century, and the technologies of the next century will require a new reconciliation between the interests of creators and those of the public.

COPYRIGHT

We can begin to lay the groundwork for this discussion by defining some of the terms we will use. And, where better to begin than with copyright itself. By copyright, I understand a set of positive, statutory rights (as opposed to natural rights) which attach to the particular expression or manifestation of information.

Information is a more nettlesome concept, to which I'll return in a moment. It is often said that the function of copyright is as an incentive to create and distribute information. This is true, but what we really mean by this is that the function of copyright is to make information behave in the marketplace as if it were tangible property -- hard goods like shoes, refrigerators, and automobiles.

The way in which copyright accomplishes this is by use of a tautology; the res, the thing that is owned, is an original work of authorship, and an original work of authorship is one that is not a copy. Now this definintion presents little difficulty for cases of "knock off" duplication, where a copy is a copy is a copy.

But it becomes considerably more difficult in cases, such as the recent computer related litigation in Whelan v. Jaslow, where what is a copy must be determined ad hoc by identifying the so-called "expression" peculiar to a given original. Expression defines what is owned.

INFORMATION AS PROPERTY

If, for the time being, we ignore these complexities, we can say very simply that copyright turns information into property. But information is a very reluctant form of property.

Information, which I will define as the meaningful concatenation of symbols, images, or sounds, is neither naturally scarce nor naturally exclusive. It is not scarce because, unlike shoes, refrigerators, and automobiles, information can be given or taken with no diminution in the number of "pieces" of it available once it is produced (and there's the rub). Unlike tangible property, where 1-1=0, information has a strange arithmetic: 1-1=2.

A corollary to this is that information is not exclusive; it is not the case that either *you* or *I* must possess it at any given moment, as is the case with tangible goods. Instead, both you and I can possess it, and my possession is not in derogation of yours. Note that I'm not saying that copying a computer program doesn't displace a sale, nor that harm to a producer's market hasn't occured; only that, in the case of information, the stolen goods need never leave the warehouse.

Scarcity and exclusivity, which do not come naturally to information, are nevertheless fundamental to the notion of property, and at the very heart of a market economy. It is, after all, how we tell buyers from sellers -- sellers have, and buyers want!

And this is where copyright comes in. Because the Framers of the Constitution, and before them, the House of Lords, decided that having lots of books and maps and charts around was a good idea, and because the free market was so adept at providing other types of goods demanded by society, some mechanism had to be found to remedy the market failure of information.

Copyright was a truly ingenious solution; it made information plentiful by making it scarce and available to all by making it exclusive. This was possible because information was always and everywhere to be found embedded in tangible goods -- copies -- which went proxy for information in the marketplace. What was bought and sold in the marketplace was wood pulp with ink stains -- never mind the fact that what you were really after was the complete works of Thomas Paine or Charles Dickens. What we had going was a product compromise: information could be sold like goods as long as it acted like them.

THE AGE OF STATIC MEDIA

This compromise, which is today coming unglued, really began in the mid-1400s, when Johannes Gutenberg concocted a clever, high volume, low overhead scheme for selling church indulgences; the moveable, interchangeable, type printing press. The invention of the printing press necessitated the invention of copyright. Because books could be mass produced in days rather than years, because printing allowed for the standardized, canonical version of works to appear, rather than being scattered in fragments throughout the monasteries of Europe, and because authors could be identified with their works, writings became a commercially hot item, and piracy followed in due course.

But if the printing press made copyright necessary, it also made it possible. Printing was a capital intensive, highly visible, 16th century high tech business. It formed a perfect bottleneck, or chokepoint, by which the King, through licensing, could control infringements -- or more importantly at the time, sedition. Of equal was the fact that, once words were printed on a page, they *stayed* on that page -- fixed, static, immutable, petrified on paper -- never again to be word processed. And, at the time, clear and commonsense distinctions existed between hardware and software, between inventions and writings; no one could ever mistake the book explaining the construction of a coke oven for the cast iron real McCoy.

During this period, which I call the age of Static Media, copyright was essentially a form of commercial regulation, since the ability to "print, publish, and vend" a writing -- the original copyright rights -- lay exclusively with commercial enterprises, rather than private individuals. Infringement, where it occurred, was a business proposition, and not a matter of casual button pushing by private individuals.

THE AGE OF DYNAMIC MEDIA

The age of Static Media ended abruptly near the beginning of the 20th Century with the invention of new ways of moving information in intangible, electromagnetic signals -- the telegraph, the radio, and the television -- and with new ways of liberating information from the package in which it was sold - the paper copier, the tape recorder, and the camera. During this period, which in contrast to its predecessor might be called the age of Dynamic Media, copyright lost control over the bottleneck. It was no longer possible or adequate for copyright to control the sale and distribution of *copies*.

Instead, the Copyright Act of 1976, which was really a response to the dynamic media of 50 years prior, sought to control the *use of the work* itself. The distinction between the work and the copy in which it resides is a point belabored in sections 102 and 202 of the Copyright Act, and is a recognition of the fact that information is devolving back into its elemental, non-property, form.

THE AGE OF DIGITAL MEDIA

Today, with the emergence of what I call the age of Digital Media in the late 20th Century, we may have come full circle, returning in a strange way to a pre-Gutenberg era, with fragments of full text, searched and summarized by AI editors, floating in bit streams across national borders.

The full impact on copyright of optical mass storage, computers and computer networks, analogue to digital conversion devices, satellite communications, and broadband fiber optic highways into the home will probably not be felt for some time. Nevertheless, I believe the enduring effects of modern technology on copyright can be sorted into basically two types: effects on practical matters such as enforcement and permissions, and effects on theoretical matters such as: "what is it that is owned, anyway?"

One of the charms of copyright was that it was essentially self-enforcing. Rights holders could spot infringing copies, and bring the infringers to justice with administrative mechanisms no more complex that the federal courts. Technology itself imposed limits on the ability of private individuals to avoid the marketplace. I suspect, for example, that the paper copier was never a real challenge to a finely bound book. But with information in digital media, the copy of the work does not degrade from generation to generation.

Moreover, electronic communications allows copies to be transported anywhere on the planet at little less than the speed of light, without the hassles associated with cars and airplanes. For a particularly compelling illustration of the scope of this power, imagine that a computer program was "shared" on NWIIS by one person with two of his friends. These friends, in turn, "shared" with two of their friends through Gateways to other networks, and so on, once every 15 minutes. In just 32 iterations taking just over 8 hours, the entire population of the planet could, in principle, be blanketed with 4.29 billion copies of the program: The chain letter has returned with a vengeance!

---

Of course, this scenario supposes that the public is largely dishonest, which, according to a public survey conducted for OTA's intellectual property report, is not really true. Instead, the survey revealed that the public is by and large unaware of some of the more basic principles of copyright, and is apt to believe that they can do as they wish with their own possessions.

But, even on the supposition that we have a law abiding, informed public, making full use of the available technology poses extremely cumbersome problems for the individual seeking permission from copyright holders. The CD-ROM, which holds over 550 megabytes of data, or over 200,000 pages of text, or copious amounts of music or still and motion picture video, is a case in point.

Imagine that I am a law abiding CD-ROM based, value-added, information purveyor who wants to store a multitude of different texts, musical compositions, and images on my CD-ROM disk. To store anything on disk -- possibly even for personal use – involves reproducing it under the terms of the copyright act. I have to get permission, and very likely pay royalties. Provided I know who they are, this may be as simple as making a phone call to hundreds or thousands of authors, copyright holders, or their assignees to negotiate their "OK." "Simple," of course, provided that I have a WATS line or plenty of spare change available.

If the attorneys of the copyright owners I are able to reach some accord on the reproduction rights involved in my CD-ROM venture, we still have distribution, performance, and display to think about. Most CD-ROMs work in conjunction with computers, and it is a simple matter to have computers work in conjunction with communications facilities. Any communication of a work stored on CD-ROM (or any other medium) is probably either a performance or a display, whether sent to 10 people in Japan via satellite, or to 1000 people in your business via Local Area Network. Now, if calling the distribution of a work over phone lines a performance or display sounds like an overly legalistic stretch, consider this: electronic distribution is not distribution at all -- at least in the legal sense -- since one can distribute only copies of a work, and copies are material objects under the law.

In any event, I'm back on the phone negotiating with an attorney, who probably want to know how I intend to control the use of the material on the CD-ROM once the disk is sold. I probably can't answer...truthfully at least.

TRANSACTIONS COSTS

The enforcement and permissions problems are two sides of one coin known as transactions costs. The question in the both cases is whether it will cost me more to enforce my rights in a work or to gain permission to use it than the revenues that that work generates.

One way around the problems of transactions costs is to create a compulsory license and have the government pick up the tab for the costs associated with pooling and distributing income, but this will not work well for markets where there are an excessive number of hard to identify buyers or sellers, as is the case with the audio and video cassette markets, and probably the microcomputer software market as well. In this case, a tax can be imposed on blank media and revenues doled out to copyright holders based on an estimate of their share of the market which copying supplants.

The problem with this is that, while single purpose media such as video or audio tape and possibly even floppy disks, may submit to market substitution analyses, versatile media, such as CD-ROM and eventually, erasable-programmable compact disk, do not -- there is simply no way of estimating fairly how much these disks are used for recording Michael Jackson and how much they are used for DBase III.

THE CASE OF COMPUTER SOFTWARE

As difficult as the practical problems for copyright may be, the truly thorny problems are theoretical. In case of software, for example, I believe we have a choice between too little protection and too much; the protect able expression can either be the literal line by line code, in which case the protection is trivial; or the literal code can be interpreted in terms of the processes which it executes in a computer, in which case we have endowed the program with patent like protection without a showing of novelty or advance over prior art (even supposing that a record of prior art exist, which it doesn't).

To see how this is so, go back to the watershed case of Baker v. Selden, which held that, though the petitioner's design for an accounting book could be copyrighted, the system of accounting that the book implemented could not. "There is a clear distinction between the Book, as such, and the art it is intended to illustrate," said the Supreme Court, and the latter is protected, if at all, by letters patent, and not copyright.

Now imagine that Selden's account ledger was written in computer code, and try to separate 'the book, as such, from the art it was intended to illustrate.' I cannot, and I suggest to you that the clear distinction has collapsed.

It is not clear yet what effect this confusion may have on the software industry. I understand, however, that Lotus Development Corporation, which is rightly jealous of its rights in its excellent 1-2-3 software, is presently being sued by members of its intellectual and marketplace ancestor.

But I suspect this is the tip of the iceberg. The day is not long in coming when, within the limits of our ability to formalize the syntax and semantics of natural language, computers will execute programs based on commands in spoken English, and we will be face to face with the question of whether the logical structure of algorithms is copyrightable.

In a way, such questions are here today: can the factory foreman who runs a robot arm through a series of steps, which are simultaneously recorded in code in computer memory, claim copyright in the procedure for welding a chassis to a frame?

WORKS OF FACT

But software is only part of the problem, and an argument can be made that software and other *works of function*, such as the nucleotide sequence that controls the manufacture of insulin in a microbe, are more coherently treated as patentable inventions, rather than copyrightable writings.

However, "works of fact", such as stock market information, news stories, telephone directories, and the like are most emphatically not patentable, and their protection, absent trade secret, falls to copyright. Copyright in works of fact ostensibly protects only the organization, arrangement, design, and selection in works of fact, and not the underlying information.

But computers are arrangers and designers par excellence; this after all, is the great power of text editing, spreadsheet, word processing, and list processing programs. It would seem that copyright is a slender reed upon which to hang the protection of computer process able works of fact.

The recent case of West Publishing v. Mead Data, which held that West's system of pagination in its online database was copyrightable, has muddied the waters somewhat and I think we can anticipate more litigation on this subject as time goes on.

CONCLUSION

I'd like to close by hazarding a guess about where all this is headed.

For computer programs and for machine process able works of fact, I suspect that courts will continue in the direction that they are already going. That is, the focus will be on infringing conduct, rather than infringing works. Under this approach, which hearkens back to the common law doctrine of misappropriation, similarity between works becomes an indication of malfeasance on the part of a defendant, rather than the sine qua non of infringement, and it is the defendant's acts which constitute the important object of proof. This is a subtle, yet very significant shift, for it turns copyright on its head. It may in fact be the best way of avoiding the difficulties of treating information as property, while at the same time providing the software and database developer with the protection they need to conduct business.

The question is whether case law can be fully developed within the confines of the current copyright law. Although I once believed that software and associated developments in technology were the Gordian knot for copyright, I suspect the issues of enforcement, permissions, and transactions costs will loom larger as the first large scale attempts at an Integrated Services Digital Network begin in the early 1990s, and as computer networks proliferate and become common, and as digital audio and video tape, optical storage media, expert systems, and a host of other technologies converge in an interconnected information utility.

Technology itself may provide some of the stopgap measures, with embedded copy or transmit disabling signals, public key encryption, and so forth. Compulsory licensing and collecting societies may also help to preserve some semblance of copyright, by providing for network access tariffs and sampling of usage. Contracts between the electronic publisher and its clients may also help to keep legal reigns on the problem.

But, the real question is whether we want to continue to find patchwork solutions for the sake of preserving copyright, or whether there isn't some better way of taking full advantage of all the technology can offer, while at the same time observing the old adage: to every cow her calf.

----------------
Author's note:
Robert Kost is somewhat gainfully employed as a Legal Analyst for the U.S. Congress' Office of Technology Assessment (OTA), whose mission is to assist Congress in anticipating and planning for the social consequences of technological change. The views represented in this article do not necessarily represent those of the OTA. If proven wrong, even Mr. Kost may disown them. He welcomes all comments on this article, hostile or sympathetic.