Boston — Had the sorcerer's apprentice cataloged and filed all the wizardry cascading from his out-of-control creativity, he might have wound up with just what Uncle Sam has today -- a United States Patent and Trademark Office of incredibly mushrooming magnitude.
This extraordinary agency of the US Department of Commerce is the largest depository of patent and technical information in the world. Patent applications that were only a trickel when the nation began have become a tidal wave.
There are now 4.5 million US patents on file, 8 million cross-references (the same patent may relate to such diverse items as garden hoses and vacuum cleaners), 10 million foreign patents, and 1.5 million articles from the world's technical literature. It adds up to 24 million documents, all in one place. That's not all. Every week some 2,160 new US patent applications arrive, or about 112,300 a year. Approximately 65 percent of this total, or 73,000 patents , are issued annually.
Even the founding fathers, who firmly planted the American patent system in the Constitution, would probably have been staggered by such an avalanche of inventions.
Since 1790, when the Patent Office flung open its doors, it's been dutifully squirreling away every patent document springing from the country's creative genius. While the average patent runs eight pages, "jumbo patents" are thousands of pages long -- thicker than metropolitan telephone directories. And every one of these files is on paper!
Yet the Patent Office -- this reservoir of innovation -- still hasn't found a computer system to fit its needs. So it files and retrieves almost all its records by hand. Less than 1 percent of its patent information storage systems are computerized.
This causes a serious weakness in an otherwise admirable patent system: an inability to maintain the integrity of its files.
This makes it extremely difficult even for the Patent Office's corps of highly skilled patent examiners, let alone the bewildered public, to search the files and ensure with absolute certainty that an invention has not already been patented -- especially since the workload permits Patent Office examiners only 15 hours to conduct such a hunt.
With the soundness of a patent hinging on the effectiveness of searches, this problem is causing concern. Some patent attorneys bluntly warn there is now "a lack of confidence in the patent system."
What inventor wants to run the risk of gearing up to manufacture and market a clever gimmick if it turns out the patent will not protect it after all? Some giant corporation might invest hundreds of hours in searching the Patent Office files and come up with some early, long-forgotten patent that would render the little guy's claim invalid. The inventor could lose his shirt.
The inventor could fight it out in court, but lawsuits could drag on for years and cost thousands of dollars.
Only a small fraction of all US patents is ever challenged in court. That in itself is a measure of the respect they command. However, today half of those that are litigated are invalidated. Some patent attorneys say they figure that a 50- 50 chance of success is not a bad deal. But this makes a lot of people anxious.
The fact that last December Congress passed an omnibus patent reform bill indicates that the nation's lawmakers, too, want to make it easier for new ideas to break out into the open marketplace.
Here's how a patent search works:
The idea for a new kind of rocking chairs hits you. You want to hurry up and protect it with a patent. The question you have to answer is: Is your widget really a new idea or are you reinventing the wheel?
So off to your public library. A number of them in the US are repositories for Patent Office files. There you will find every US patent ever issued -- all neatly arranged in chronological order by patent number. No. 1 may relate to safety pins. No. 2 to chemistry. No. 3 to electronics, etc. The rocking chairs are in there somewhere rocking, but can you get them out without flipping through 4.5 million patents? You can try.
You turn to the "class lists" issued by the Patent Office, which list patent numbers by subject. Under class 297 (chairs and seats) are 800 rocking chair patents broken down into 22 subclasses: collapsible, convertible, platform, etc. If you have the endurance, you plow through the 4.5 million patents and pull out all 800 to see if any of them predate your inspiration. That's one way to search -- the hard way. As a search tool, professional searchers regard the local library as "totally useless." It takes so long to use that it is too expensive.
The other way (which also is not cheap) is to go to Washington and cross the Potomac to Arlington, Va., where a cluster of glass office towers called Crystal City sprang up during the 1960s next to the Washington National Airport.
The Patent Office's cavernous public search room is so large it fills the entire first floor of two buildings. In the center section that connects the two buildings are tables where the public can study patents they have found in the files. The rest of the area is crowded with row after row of library-like stacks full of shelves jammed with US patents and their millions of cross-references.
Upstairs, Patent Office examiners have their own separate, even more mammoth file scattered through 19 floors of the two buildings. This collection includes not only US but also foreign patents and technical literature.
These buildings are the only place where US patents are filed according to the sophisticated classification system, which has no less than 107,000 classes and subclasses. A standing joke is that tea mixers and cement mixers share the same classification. But in general, the classification manual is helpful, guiding you to the very spot where all 800 patents on rocking chairs are filed together. If these are on the shelves in their "shoes" -- little trays that hold a bunch of them -- you pull them all out, take them to a table, and zip through them to determine whether you are the inventor you think you are.
But if rocking chairs are a hot item -- like solar energy devices are today -- you will not be the only party interested in that file. There may be as many as 500 other seekers in the search room that day, pulling patents out, refiling them -- a librarian's nightmare. The very file you want may be in use. Worse still, it may be missing.
Though it's forbidden to remove patents from the search room, it is unrealistic to assume that with all those hands roaming at will in this paper jungle, there will not be some patents missing -- misfiled, surreptitiously borrowed, or filched. It is estimated that at any given moment between 5 and 30 percent of the documents are missing. The more active the file, the greater the pilferage. Theft has become so serious that the Patent Office now has an alarm system. To date it has proved only partly successful.
If your idea proves in fact to be a dazzling flash of light, you would then apply for a patent. Only after a Patent Office examiner makes his independent search will a patent be issued. That process, which the office thinks should take no more than 18 months, now takes 22 -- nearly two years.
In his hunt, the examiner includes not only foreign rocking chair patents but also the world's technical literature. Uncle Sam is stricter than most nations in insisting that patents be granted only on truly new contributions to society. Isaac Fleischmann, director of information services for the Patent Office, puts it this way: "If the invention was described in the Dead Sea Scrolls, and that information is available to our examiners, you can't get a US patent on it."
Is the purpose of a patent system to protect inventors? Most people would probably say yes. but framers of the Constitution provided that protection only as the means to a much broader end. National progress, they maintained, would be based on a sharing of knowledge, not the hoarding of it in trade secrets. So they gave Congress "power to promote the progress of science and useful arts by securing for limited times to authors and inventors the exclusive right to their respective writings and discoveries."
A patent is in effect a bargain that promises: Share with us what you have invented and for 17 years the exclusive right to exploit it will be yours. So the system is not designed for the sake of the inventor but to promote science and art through disclosure. Until the day a patent is issued, an invention is private property. The moment it is issued, the idea is public knowledge for the benefit of all.
A major advantage is the prevention of duplication. Everybody is alerted to what the latest ideas are. Knowing what someone else has designed should stir an inventor off in other directions. Or it may suggest to inventors how the work of competitors can be improved or how the same thing can be achieved in a different way. Thus one idea builds on another. This keeps the inventor of the original device always looking for ways to improve it. Such churning and competition spur invention on.
Mr. Fleischmann's 35 years with the Patent Office have taught him that "whenever there is a national emphasis in a specific area of need, there is greater inventive activity in those areas."
"What is an invention?" he asks. "It is meeting of a need for the solution to a problem. America is a nation of improvisers and people who want to make living conditions better.
"Our early period was agrarian. Almost from the very beginning there was growth in population with a need for food. . . . So inventions right up through the middle of the last century were mainly in the agricultural field. They just grew and grew and grew because of need."
Not long ago the emphasis was on environmental inventions, still an active and important field. Now it's on energy -- windmills, solar cells, etc.
"I dare say that in 10 or 15 years from now what we think are the latest developments in solar and other fields will be old hat," Mr. Fleischmann says. "Areas we are unaware of today will be explored. . . . I believe that a generation from now perhaps oil will be obsolete. Faced with a problem, we have got to solve the problem. This has been our history. I hope it will always maintain this impetus for development."
Between 1961 and 1970, the number of patent applications received each year rose steadily from 85,000 to 106,000. But during the 1970s this growth flattened out, reaching only 107,400 10 years later.
Last year recorded a welcome leap to 112,315 applications. And thus far into 1981, this higher rate is continuing. The Patent Office views the jump as significant.
Patent attorneys all over the country long for the Patent Office search files to be computerized and put "on line." Instead of journeying to Washington, they could conduct their searches in their own offices via computer. This would create an always complete, hands-off, reliable filing system in Crystal City.
Nevertheless, patent attorneys say the office's examiners are doing "a heroid job under the circumstances." When attorneys get stuck and can't find what they need, they sometimes appeal to examiners, who point to a drawer and say, "Look there."
"Examiners are just phenomenal!" exclaims George A. Herbster, a patent attorney with the Boston firm of Ceasari & McKenna, who for years has worked to stir interest in computerizing the agency's files.
Bar members see the Patent Office as underfunded and short-handed. In 1974 there were 1,200 examiners. Now, because of budget cuts, there are fewer than 850. Mr. Herbster, who nevertheless commends the office's classification system , points out that it "automatically excludes a vast amount of information from a searcher." If he is hunting a patent that was not deemed important enough to be classified, he realizes that "I'm never going to find it unless some night I have a terrific dream that says, 'Go look there.'"
So why hasn't the Patent Office a long time ago called up some data-processing firm and told it to come on over and computerize the place?
It isn't that easy. For one thing, cost estimates -- just wild guesses at this point -- range from $45 million to $400 million.
Beyond that, a workable computer system is based on a well-defined terminology: You have to define something the same way every time you refer to it. It's fairly easy to define a chemical concept, but there might be many ways to define a mechanical or electrical concept or device (most patents fall into one of these three braod categories). A computer will only cough up a patent if the searcher describes it exactly as it is in the computer's files.
And if a computer spits out 500 documents in unclassified form, the manual searcher may not be any better off than he is now with two able hands and one good head.
As far back as the mid-1950s, the office began tinkering with automated retrieval systems on an experimental basis. At one time, the Patent Office has as many as 30 systems on trial simultaneously. None panned out as well as was hoped. Only two have survived.
During all the years this experimentation has been going on, computer technology has been making giant strides, brightening hopes for some ideal solution. At present the office is looking into the use of microfilm.
Even if the green light flashed, the question is, who should install the system? The Patent Office or private industry?
Naturally, many companies in the private sector are eager for the job. Several companies already offer computerized patent search systems.
Attorney Kenneth E. Madsen, chairman of the ABA's information retrieval committee, points out that under legislation signed Dec. 12 the Patent Office has two years to study the problem and report back with a plan for computerization if its evaluation warrants it. One issue to be decided is whether the full text of a patent or simply a brief abstract should be placed in the computer's memory to best aid the searchers.
Meanwhile, let's see, where did we refile those 800 patents on rocking chairs?