Licensing Product Times

Vol. 3, No. 4, pp. 1-8

Computer Software and the Evolution of Intellectual Property Rights

©1991 by Charles M. Nelson

For years we have been told by economic prognosticators that we are entering a postindustrial information age. By this they mean that new industrial processes, technological progress and economic prosperity depend on the application of complex information systems. The two principle models which forecast the information age are genetics and cybernetics, projected at the industrial level as computer and biotechnology. In biotechnology, deciphering the DNA genetic code (information system) and understanding its functions is revolutionizing medicine and agriculture. In computer technology, large-scale integration of hardware, firmware and software has led to intense competition in robotics, artificial intelligence and real-time simulation.

Technological processes and information structures in computer and biotechnology are very different from those envisaged in traditional systems of intellectual property (IP) rights. As a result, the traditional system is being transformed on a global scale. In the US, this revision takes the form of administrative and court decisions fashioned Case-by-case by individuals who are expert in the law and the patent system, but untrained in technological information systems, and inexperienced in biochemistry, genetics, computer architecture or advanced programming languages. Internationally, revision takes the form of negotiations conducted by teams who seek to position their countries strategically in the global economic system. Negotiations focus primarily on the conditions under which business is conducted rather than technological and information systems which give rise to the business, itself.

The outcome of this process is uncertain, but will structure much of the world's economic activity in the next century, determine the politics of conflict in international competition and influence the level of our nation's economic prosperity. This article focuses on the process as it is revealed by the rapidly evolving IP framework for computer software in the US. This is a part of the IP system which combines trademark, patent, copyright and trade secrets protection in a fashion which disrupts the computer industry, inhibits its growth and limits its economic productivity.

The Trouble with Software

A software package consists of reference and learning manuals, a set of disks which contain computer programs and a licensing agreement. The manuals are copyrighted just like other publications, but the disks pose a real problem. When operating the computer with the programs contained on the disk, the user sees little or nothing of what the disks actually contain. Instead, the user sees and interacts with a series of "screens" displayed on a video monitor. Frequently, the user modifies the contents and appearance of the screen while using the software. For some screens, such as those where program activities are selected, the modification is negligible and not intended to be printed, saved or otherwise used. Other screens are designed to be used as a framework within which the user creates a document, such as a publication or work of art. Some programs even allow the user to design functional screens, such as menus including command names which govern selection of program operations. In addition, programs routinely allow the user to write information back into special files on the program disk to control printing and other aspects of program operation.

What are the IP rights of the creator of a software package? Who owns the disks and their contents? How freely may the disks and the information which they contain be used by the individual who purchases the software? Who owns the products which are produced using the software?

To preserve their rights, creators of software license each copy which is distributed. Before you open the package which contains the program, you are confronted with a statement which begins something like this: "BY OPENING THIS PACKAGE, YOU AGREE TO BECOME BOUND BY THE TERMS OF THIS LICENSE. IF YOU DO NOT AGREE TO THE TERMS OF THIS LICENSE, RETURN THE SOFTWARE FOR A FULL REFUND." It goes on to explain that you do not own the software, but merely have purchased the right to use it in certain ways. You are prohibited from reading the program, itself. However, the products you create using the software belong to you subject only to the usual copyright or trademark restrictions.

A computer program can be regarded as a piece of technology in itself. For example, it is possible to take a program and "burn" it into the structure of a silicon chip which is used on the circuit board of a computer. Word processors are available in this form. Once installed, the chip provides instant access to the word processor without reading the program from a disk. This is called "firmware" rather than "software" because the program resides on a chip in the computer. It is not called "hardware" because the microprocessor of the computer still accesses the chip in much the same way as it does a disk in a disk drive and uses the program in the same way that it would the software version. Is this a technological device which should be patented or merely a document written on a silicon chip rather than a piece of paper and therefore subject to copyright law instead?

Our IP system was designed 200 years ago and refined in an era when ideas had to be translated into products using technological processes that involved mechanical manipulation of materials. Creative ideas were expressed visually and in sound so they could be perceived directly with little effort. But what happens when creative ideas, information systems and specific pieces of information all become tangible, operating components of the technology itself? This is exactly what has happened. As a result, our system of IP rights, with its carefully balanced privileges and obligations, its notion of public good and its sense of fair play, is traversing a chaotic period of fundamental readjustment. Corporations are attempting to control the outcome of this readjustment through litigation. The government is attempting to control its outcome by promulgating new regulations and by giving advice to the courts in the hope of influencing the course of litigation. Meanwhile, those who create programs believe these approaches will restrict the productivity of the software business and retard the development of technology generally. They would prefer that a rational system more sensitive to the public good and long-term national economic goals be imposed legislatively, a system which would recognize the hybrid nature of information dominated technologies and create a set of IP categories and rules which define such technologies and allocate IP rights in a reasonable fashion.

The Way Things Were

Twenty years ago programmers and software developers gave little thought to IP rights. They were content if competitors did not copy the coded text of their programs or misappropriate their name or trademarks. Incorporating large sections of program concepts taken from others was fair game if you wrote your own unique program code and created a product which was not identical to a competitors; close, even very close, was all right. Computer languages, utility programs to make computer operation more convenient and simple user applications were placed in the public domain. Nearly everyone thought that the pace of development would make programs obsolete so rapidly that more IP protection was unnecessary.

Programmers developed remarkably sophisticated packages by building on one another's software. For example, the BASIC computer language was developed out of Fortran to make programming less abstract and easier for more people to comprehend. The Visicalc® spreadsheet and its formula logic were developed partly from the command structure and syntax of BASIC. Lotus 1-2-3® incorporated and was compatible with most of Visicalc. This was exactly what hardware manufacturers wanted. Their computers were far more sophisticated than the available programs and it was essential to create more friendly software products to sell their computers. But today, Lotus dominates and jealously guards the spreadsheet market; the software development business runs by very different rules. Why?

Maturation of the Software Industry

Most creators of software believe that their industry is still young and expansive, with unlimited potential for future development. Programmers are incredulous when told the storm of litigation which appears to be consuming the software industry is simply a phase of market consolidation which heralds the onset of a mature industry. But external observers find current developments at once mundane and economically inevitable. Some compare it to the broadcasting industry. Early in its maturation cycle, the networks and electronic giants joined to form RCA to foster a monopoly which could be extended to all aspects of the industry. Though initially successful, this soon led to a broad range of antitrust actions by the government, including the divestment of RCA and many other holdings, the eventual protection of individual artists through ASCAP and the creation of public broadcasting.

Today, emerging powers in microcomputing hardware and software are forming strategic alliances and using their economic power to secure market position through litigation. The theory is relatively simple at the strategic level:

    Form working relationships to create and enforce industry standards which coincide with proprietary hardware and software, making it difficult for competitors to introduce new technology.

    Aggressively enforce and defend hardware and software patents, copyrights and trade secrets.

    Use the ambiguous structure and content of hardware and software, which were never contemplated during the creation of our IP system, to make favorable changes in the law through litigation.

    Create a universe of favorable case law by engaging in unequal courtroom battles with smaller, strategically selected companies unlikely to have the resources to mount effective defenses or to appeal unfavorable decisions. Rely on these precedents to win later appeals at the Appellate and Supreme Court levels.

If the history of the broadcasting industry is a reasonable gauge, industry giants may at first succeed in their effort to enforce as wide a monopoly as possible. If so, expect the revolution of engineers, programmers and users to force political and legal actions to break the monopolies now coming into existence. Meanwhile, exploration of numerous avenues of hardware and software development will be temporarily foreclosed, slowing economic growth and reducing productivity within the computer industry.

The Way Things Seem Today

During the past decade, court decisions have added enormously to the range and depth of patent and copyright protection afforded software, but it is unclear if the new standards will persist or fall as courts at all levels view information technologies with increasing sophistication.

In the copyright arena, a number of legal theories contend for acceptance. Two are the "entertainment" approach with its "look and feel" theory and the "accounting" approach with its "concept vs. expression" theory. The concept vs. expression theory considers a piece of software to be like an accounting manual from which forms may be duplicated and filled out in order to implement an accounting system. Much copyright law was built around this analogy between 1880 and 1920, when the courts consistently ruled that ledger forms and accounting terms embodied the concepts of accounting and could not be considered copyrightable expression. Accounting systems were conceptual structures and could not be copyrighted. The role of expression was narrowly confined. Although it included explanatory commentary and art work used to embellish forms, it excluded accounting terms and direct graphic representations of the conceptual relationships between such terms, such as ledger sheets. Extension of this theory to computer programs would severely restrict which aspects of a program could be copyrighted. This approach has not been very successful, but still has many ardent supporters.

The look and feel theory derives in part from an analogy to interactive games. What counts is what the user sees and manipulates on the screen. If it looks like spaceman and acts like spaceman, then the game is spaceman regardless of how many details vary. This makes sense when you are thinking about video games, but is less plausible when you think of a spreadsheet or a word processor. Nevertheless, since some of the first successfully litigated cases dealt with games and relied on this argument, it became a model for reasoning in other software rulings. Using this approach, Lotus was able to obtain an infringement judgment against Paperback Book Software, creators of VP-Planner, a Lotus 1-2-3 look-alike. The ruling concluded that menu structures and contents of Lotus 1-2-3 defined an overarching structure for the software user interface which could be copyrighted as a whole. Paperback had infringed the Lotus copyright even though its program differed substantially in the details of its menu structure and contents. Lotus used this victory to bring suit against Borland International and argues that the court should extend the look and feel approach to the command language and programming structure of its user interface. If successful, this will extend the look and feel doctrine to the overarching, non-literal components of the computer program which control screen displays and printing. Most programmers and a number of court decisions would view this as copyright protection for the algorithms underlying the program structure, an area which is currently subject to patent rather than copyright law since algorithms are ideas rather than expressions.

In the patent arena, the Supreme Court ruled a decade ago that algorithms were patentable. An algorithm is a logical command structure which can be expressed in symbols. This structure, like any idea, can be translated into any language which contains the necessary vocabulary and grammar. Hence, if you hold a patent for the logic required to move a cursor in a screen display, any commercial program which uses that logical approach in any programming language will infringe your patent. Potentially, this could extend to programming languages as well as programs written using those languages.

Many cybernetics experts and the patent office, itself, I believe this ruling extends patent rights much too far into the realm of fundamental logic and natural order. In their view, algorithms are more discoveries then they are inventions and so should not be patentable. They point out that the Supreme Court decision which established the patentability of algorithms was on a five to four vote and that it contained murky logic. With so many new justices, many believe the court will reverse itself when given the opportunity.

Consequences for Software Development

Experienced programmers and software developers agree that consolidation of the status quo will make it difficult to develop software under royalty and contract structures which prevail in the industry. They point out that any new software package with the sophistication of a spreadsheet or word processor is likely to infringe about 100 patents and copyrights. The programmer usually gets a royalty of between 5% and 15%. After satisfying other business expenses, this generally leaves the developer between 10% and 20% of the sales price from which to take profits and to pay any secondary royalties. This is between $40 and $80 on a $400 product. If the developer takes $20, only a 5% profit on a $400 item, this will leave between $20 and $60 with which to satisfy 100 royalty claims, an average royalty payment of .05% or less and far, far below what is considered acceptable in the industry today.

In the short run, such a situation will favor large developers who can afford to defend their own copyrights and patents while making it extremely expensive for small developers or individual programmers to prosecute them for appropriating their IPs. Small players will be forced to fight back by licensing their properties to contending giants interested in using litigation to muscle aside their competitors. In the long run, it will foster the development of monopolies by making market-oriented software development the domain of a few large corporations.

Radical Solutions: Are They Possible?

Some programmers, such as those in the League for Programming Freedom, believe that the industry should be accorded only the most rudimentary forms of patent and copyright protection. They argue that a minimal level of protection is sufficient to maintain market position in an industry where product life is relatively short and the potential for developing new product lines is ever present. Others are busily patenting as many algorithms as they can conjure in hopes of cashing in when other programmers stumble onto the same ideas when constructing real-world programs.

The majority who fall between these extreme views believe that the piecemeal restructuring of the industry through litigation is wasteful of resources, slows the development of new software by creating unacceptably high levels of uncertainty and is creating an irrational structure which will poorly serve the industry and the nation as long as it is in place. And what do they favor?

    Programming languages of all types and at all levels should not be protected by patent or copyright.

    Place basic utilities, such as computer virus identification and eradication programs, in the public domain.

    Patent protection should be afforded to complex algorithms and clear rules should be established on scientific merit which distinguish "natural" algorithms, treated as discoveries, from more complex "artificial" algorithms, treated as inventions.

    Copyright protection should be afforded to program code of all kinds.

    Extend copyright protection to aspects of the look and feel of program output, but not those which make the program operate efficiently or to components, such as command words, icons and display locations, which become de facto industry standards.

    Trade secrets should be afforded substantial protection within a framework which logically derives from the nature of software and hardware development and application.

    Trademark protection should be made as extensive and economical as possible.

Such a system cannot be created in a framework which assumes that mental ideas are always clearly different from their technological expression, that symbols have no direct technological utility and that information can have no tangible, technological presence. Since information technologies violate all of these assumptions, no amount of tinkering through litigation will create a system which achieves the goals of the patent system. These are to make as much technology available to as many people as possible as fast as possible by creating the optimum balance between the interests of individual inventors, corporate developers of technology and the public welfare. The results of piecemeal litigation have already thrown the industry seriously out of balance. How can this balance be restored?

Two Roads to the Future

Ultimately, any solution to the software problem will require federal legislation, but Congress will act only if forced by a crisis or to reflect a strongly enunciated industry consensus. Neither of these conditions exists today. It will be at least another five to ten years before litigation produces conditions severe enough to require legislative intervention. Meanwhile, there is no industry-wide consensus at the corporate level.

The Patent and Trademark Review Commission has solicited recommendations in 13 problem areas, including patent protection for computer technology. More than half of all responses deal with this point alone, showing that most individual programmers and engineers understand the nature of the problem and are extremely concerned. Yet the National Association of Manufacturers, through its Council on High Technology, could only weakly support the status quo because of divided opinion at the corporate level. There is currently no effective forum in which programmers, engineers, software developers and corporate end users can fashion a meaningful consensus.

The Software Iceberg

In Norton Juster's Phantom Tollbooth, there is an island named Conclusions located in the Sea of Knowledge off the Kingdom of Wisdom. Individuals who leap to unwarranted conclusions find themselves suddenly upon the desolate beaches of that island. The only way back to Wisdom is an icy dip in the Sea of Knowledge. Many are stranded on Conclusions for months because they fear to swim, though few are known to have perished in the attempt. Today, our Island of Conclusions is a slippery iceberg floating in the Sea of Tropical Litigation. It is crowded with programmers and small software development corporations. Like it or not, they are being pitched into the sea as the berg melts. They're going to be a good deal more knowledgeable and very angry by the time they reach shore. In time, they will be followed by numerous lawyers and corporate executives.

This is hardly a climate conducive to industry growth and prosperity, much less one favoring discourse and consensus. But concerted action is needed now if the US is to maintain its standing in the world computer industry or its leadership role in the development of international IP law. This means establishing a computer industry intellectual property forum charged with developing a consensus which can be effectively promoted in Congress.

The author wishes to acknowledge the thoughtful and patient discourse of numerous professionals whose ideas have contributed to this simplistic synthesis of an extraordinarily complex problem, but most especially Dan Bricklin, Judge Glenn Groenewold, Joseph Iandiorio, Barry Shein, Richard Stallman, and Eric Ulman for spending so much of their time and sharing so many of their insights. Readers interested in a radical view of the situation should contact The League for Programming Freedom (617) 243-4091. Those wanting an overview which reflects what lawyers believe you should be doing to protect your software IP rights can contact the office of Joseph Iandiorio at (617) 890-5678 and ask for Pamphlet XII, Software Protection. Useful industry discussion can be found in the Communications of the Association for Computing Machinery, Computer Design and Boston Computer Society Update. Scholarly legal analysis is available in the High Technology Law Journal, Technology Law Bulletin and The Computer Lawyer.