Chapter 8: Patent Protection of Software:
Statutory Subject Matter in the
Supreme Court and the Federal Circuit



A. The Supreme Court Trilogy
on Statutory Subject Matter



Gottschalk v. Benson
United States Supreme Court
409 U.S. 63 (1972)

Mr. Justice Douglas delivered the opinion of the Court.

Respondents filed in the Patent Office an application for an invention which was described as being related "to the processing of data by program and more particularly to the programmed conversion of numerical information" in general-purpose digital computers. They claimed a method for converting binary-coded decimal (BCD) numerals into pure binary numerals. The claims were not limited to any particular art or technology, to any particular apparatus or machinery, or to any particular end use. They purported to cover any use of the claimed method in a general-purpose digital computer of any type. Claims 8 and 13 (set forth in the Appendix to this opinion) were rejected by the Patent Office but sustained by the CCPA.

The question is whether the method described and claimed is a "process" within the meaning of the Patent Act.2

  2.     35 U.S.C. § 100(b) provides:
The term "process" means process, art or method, and includes a new use of a known process, machine, manufacture, composition of matter, or material.

A digital computer, as distinguished from an analog computer, operates on data expressed in digits, solving a problem by doing arithmetic as a person would do it by head and hand. Some of the digits are stored as components of the computer. Others are introduced into the computer in a form which it is designed to recognize. The computer operates then upon both new and previously stored data. The general purpose computer is designed to perform operations under many different programs.

The representation of numbers may be in the form of a time series of electrical impulses, magnetized spots on the surface of tapes, drums, or discs, charged spots on cathode-ray tube screens, the presence or absence of punched holes on paper cards, or other devices. The method or program is a sequence of coded instructions for a digital computer.

The patent sought is on a method of programming a general-purpose digital computer to convert signals from binary-coded decimal form into pure binary form. A procedure for solving a given type of mathematical problem is known as an "algorithm." The procedures set forth in the present claims are of that kind; that is to say, they are a generalized formulation for programs to solve mathematical problems of converting one form of numerical representation to an other. From the generic formulation, programs may be developed as specific applications.

Decimal system

The decimal system uses as digits the 10 symbols 0, 1, 2, 3, 4, 5, 6, 7, 8, and 9. The value represented by any digit depends, as it does in any positional system of notation, both on its individual value and on its relative position in the numeral. Decimal numerals are written by placing digits in the appropriate positions or columns of the numerical sequence, i.e., "unit" (100), "tens" (101), "hundreds" (102), "thousands" (103), etc. Accordingly, the numeral 1492 signifies (1 x103) + (4 x 102) + (9 x 101) + (2 x 100).

Binary system

The pure binary system of positional notation uses two symbols as digits 0 and 1, placed in a numerical sequence with values based on consecutively ascending powers of 2. In pure binary notation, what would be the tens position is the twos position; what would be hundreds position is the fours position; what would be the thousands position is the eights. Any decimal number from 0 to 10 can be represented in the binary system with four digits or positions as indicated in the following table.

Shown as the sum of powers of 2
Decimal   (23 = 8)   (22 = 4)   (21 = 2)   (20 = 1)   Pure Binary
0 = 0 + 0 + 0 + 0 = 0000
1 = 0 + 0 + 0 + 20 = 0001
2 = 0 + 0 + 21 + 0 = 0010
3 = 0 + 0 + 21 + 20 = 0011
4 = 0 + 22 + 0 + 0 = 0100
5 = 0 + 22 + 0 + 20 = 0101
6 = 0 + 22 + 21 + 0 = 0110
7 = 0 + 22 + 21 + 20 = 0111
8 = 23 + 0 + 0 + 0 = 1000
9 = 23 + 0 + 0 + 20 = 1001
10 = 23 + 0 + 21 + 0 = 1010

The BCD system using decimal numerals replaces the character for each component decimal digit in the decimal numeral with the corresponding four-digit binary numeral, shown in the right-hand column of the table. Thus decimal 53 is represented as 0101 0011 in BCD, because decimal 5 is equal to binary 0101 and decimal 3 is equivalent to binary 0011. In pure binary notation, however, decimal 53 equals binary 110101. The conversion of BCD numerals to pure binary numerals can be done mentally through use of the foregoing table. The method sought to be patented varies the ordinary arithmetic steps a human would use by changing the order of the steps, changing the symbolism for writing the multiplier used in some steps, and by taking subtotals after each successive operation. The mathematical procedures can be carried out in existing computers long in use, no new machinery being necessary. And, as noted, they can also be performed without a computer.

The Court stated in Mackay Co. v. Radio Corp., 306 U.S. 86, that "[w]hile a scientific truth, or the mathematical expression of it, is not patentable invention, a novel and useful structure created with the aid of knowledge of scientific truth may be." That statement followed the longstanding rule that "[a]n idea of itself is not patentable." Rubber-Tip Pencil Co. v. Howard, 20 Wall. (87 U.S.) 498. "A principle, in the abstract, is a fundamental truth; an original cause; a motive; these can not be patented, as no one can claim in either of them an exclusive right." Le Roy v. Tatham, 14 How. (55 U.S.) 156. Phenomena of nature, though just discovered, mental processes, and abstract intellectual concepts are not patentable, as they are the basic tools of scientific and technological work. As we stated in Funk Bros. Seed Co. v. Kalo Co., 333 U.S. 127, "He who discovers a hitherto unknown phenomenon of nature has no claim to a monopoly of it which the law recognizes. If there is to be invention from such a discovery, it must come from the application of the law of nature to a new and useful end." We dealt there with a "product" claim, while the present case deals with a "process" claim. But we think the same principle applies.

Here the "process" claim is so abstract and sweeping as to cover both known and unknown uses of the BCD to pure binary conversion. The end use may (1) vary from the operation of a train to verification of drivers' licenses to researching the law books for precedents and (2) be performed through any existing machinery or future-devised machinery or without any apparatus.

In O'Reilly v. Morse, 15 How. (56 U.S.) 62, Morse was allowed a patent for a process of using electromagnetism to produce distinguishable signs for telegraphy. But the Court denied the eighth claim in which Morse claimed the use of electromagnetism, "however developed, for marking or printing intelligible characters, signs, or letters, at any distances." The Court in disallowing that claim said:

If this claim can be maintained, it matters not by what process or machinery the result is accomplished. For aught that we now know, some future inventor, in the onward march of science, may discover a mode of writing or printing at a distance by means of the electric or galvanic current, without using any part of the process or combination set forth in the plaintiff's specification. His invention may be less complicated, less liable to get out of order, less expensive in construction, and in its operation. But yet, if it is covered by this patent, the inventor could not use it, nor the public have the benefit of it, without the permission of this patentee.

In The Telephone Cases, 126 U.S. 1, the Court explained the Morse case as follows:

The effect of that decision was, therefore, that the use of magnetism as a motive power, without regard to the particular process with which it was connected in the patent, could not be claimed, but that its use in that connection could. Bell's invention was the use of electric current to transmit vocal or other sounds. The claim was not "for the use of a current of electricity in its natural state as it comes from the battery, but for putting a continuous current, in a closed circuit, into a certain specified condition, suited to the transmission of vocal and other sounds, and using it in that condition for that purpose." The claim, in other words, was not "one for the use of electricity distinct from the particular process with which it is connected in his patent." The patent was for that use of electricity "both for the magneto and variable resistance methods." Bell's claim, in other words, was not one for all telephonic use of electricity.

In Corning v. Burden, 15 How. (56 U.S.) 252, the Court said, "One may discover a new and useful improvement in the process of tanning, dyeing, etc., irrespective of any particular form of machinery or mechanical device." The examples given were the "arts of tanning, dyeing, making waterproof cloth, vulcanizing India rubber, smelting ores." Those are instances, however, where the use of chemical substances or physical acts, such as temperature control, changes articles or materials. The chemical process or the physical acts which transform the raw material are, however, sufficiently definite to confine the patent monopoly within rather definite bounds.

Cochrane v. Deener, 94 U.S. 780, involved a process for manufacturing flour so as to improve its quality. The process first separated the superfine flour and then removed impurities from the middlings by blasts of air, reground the middlings, and then combined the product with the superfine. The claim was not limited to any special arrangement of machinery. The Court said:

That a process may be patentable, irrespective of the particular form of the instrumentalities used, cannot be disputed. If one of the steps of a process be that a certain substance is to be reduced to a powder, it may not be at all material what instrument or machinery is used to effect that object, whether a hammer, a pestle and mortar, or a mill. Either may be pointed out; but if the patent is not confined to that particular tool or machine, the use of the others would be an infringement, the general process being the same. A process is a mode of treatment of certain materials to produce a given result. It is an act, or a series of acts, performed upon the subject matter to be transformed and reduced to a different state or thing.

Transformation and reduction of an article "to a different state or thing" is the clue to the patentability of a process claim that does not include particular machines. So it is that a patent in the process of "manufacturing fat acids and glycerine from fatty bodies by the action of water at a high temperature and pressure" was sustained in Tilghman v. Proctor, 102 U.S. 707. The Court said:

The chemical principle or scientific fact upon which it is founded is, that the elements of neutral fat require to be severally united with an atomic equivalent of water in order to separate from each other and become free. This chemical fact was not discovered by Tilghman. He only claims to have invented a particular mode of bringing about the desired chemical union between the fatty elements and water.

Expanded Metal Co. v. Bradford, 214 U.S. 366, sustained a patent on a "process" for expanding metal. A process "involving mechanical operations, and producing a new and useful result," was held to be a patentable process, process patents not being limited to chemical action.

Smith v. Snow, 294 U.S. 1, and Waxham v. Smith, 294 U.S. 20, involved a process for setting eggs in staged incubation and applying mechanically circulated currents of air to the eggs. The court, in sustaining the function performed (the hatching of eggs) and the means or process by which that is done, said:

By the use of materials in a particular manner, he secured the performance of the function by a means which had never occurred in nature and had not been anticipated by the prior art; this is a patentable method or process.... A method, which may be patented irrespective of the particular form of the mechanism which may be availed of for carrying it into operation, is not to be rejected as "functional" merely because the specifications show a machine capable of using it.

It is argued that a process patent must either be tied to a particular machine or apparatus or must operate to change articles or materials to a "different state or thing." We do not hold that no process patent could ever qualify if it did not meet the requirements of our prior precedents. It is said that the decision precludes a patent for any program servicing a computer. We do not so hold. It is said that we have before us a program for a digital computer but extend our holding to programs for analog computers. We have, however, made clear from the start that we deal with a program only for digital computers. It is said we freeze process patents to old technologies, leaving no room for the revelations of the new, onrushing technology. Such is not our purpose. What we come down to in a nutshell is the following.

It is conceded that one may not patent an idea. But in practical effect that would be the result if the formula for converting BCD numerals to pure binary numerals were patented in this case. The mathematical formula involved here has no substantial practical application except in connection with a digital computer, which means that if the judgment below is affirmed, the patent would wholly preempt the mathematical formula and in practical effect would be a patent on the algorithm itself.

It may be that the patent laws should be extended to cover these programs, a policy matter to which we are not competent to speak. The President's Commission on the Patent System rejected the proposal that these programs be patentable:

Uncertainty now exists as to whether the statute permits a valid patent to be granted on programs. Direct attempts to patent programs have been rejected on the ground of nonstatutory subject matter. Indirect attempts to obtain patents and avoid the rejection, by drafting claims as a process, or a machine or components thereof programmed in a given manner, rather than as a program itself, have confused the issue further and should not be permitted.

The Patent Office now cannot examine applications for programs because of a lack of a classification technique and the requisite search files. Even if these were available, reliable searches would not be feasible or economic because of the tremendous volume of prior art being generated. Without this search, the patenting of programs would be tantamount to mere registration and the presumption of validity would be all but nonexistent.

It is noted that the creation of programs has undergone substantial and satisfactory growth in the absence of patent protection and that copyright protection for programs is presently available.

If these programs are to be patentable, considerable problems are raised which only committees of Congress can manage, for broad powers of investigation are needed, including hearings which canvass the wide variety of views which those operating in this field entertain. The technological problems tendered in the many briefs before us indicate to us that considered action by the Congress is needed.

Reversed.

Appendix

Claim 8 reads:

The method of converting signals from binary coded decimal form into binary which comprises the steps of

(1) storing the binary coded decimal signals in a reentrant shift register,

(2) shifting the signals to the right by at least three places, until there is a binary "1" in the second position of said register,

(3) masking out said binary "1" in said second position of said register,

(4) adding a binary "1" to the first position of said register,

(5) shifting the signals to the left by two positions,

(6) adding a "1" to said first position, and

(7) shifting the signals to the right by at least three positions in preparation for a succeeding binary "1" in the second position of said register.

Claim 13 reads:

A data processing method for converting binary coded decimal number representations into binary number representations comprising the steps of

(1) testing each binary digit position "1," beginning with the least significant binary digit position, of the most significant decimal digit representation for a binary "0" or a binary "1";

(2) if a binary "0" is detected, repeating step (1) for the next least significant binary digit position of said most significant decimal digit representation;

(3) if a binary "1" is detected, adding a binary "1" at the (i + 1)th and (i + 3)th least significant binary digit positions of the next lesser significant decimal digit representation, and repeating step (1) for the next least significant binary digit position of said most significant decimal digit representation;

(4) upon exhausting the binary digit positions of said most significant decimal digit representation, repeating steps (1) through (3) for the next lesser significant decimal digit representation as modified by the previous execution of steps (1) through (3); and

(5) repeating steps (1) through (4) until the second least significant decimal digit representation has been so processed.

Notes

1. For those interested in the background of this decision, the prosecution history and briefs in the CCPA are available at this link (providing further links to them): Benson at the CCPA.

Can you puzzle out the "masking" and "shifting" in a shift mentioned in the claims? If you shift a binary number x in a shift register to the left by one position, you get as a result two times the original number x. If you now shift x two more places to the left you get eight times the original number x. If you add the first and second results, you get ten times the original number (2x + 8x = 10x). Can you now figure out the great Benson algorithm? According to the Government, in a nutshell (and carried out below only to the hundreds place) it is:

N = U + 10 (10 H + T) = 100 H + 10 T + 1 U,

where N is the answer; U is the units-place digit; T, the tens-place digit; and H, the hundreds-place digit.

In other words, take the sum of 1 times the units-place digit, 10 times the tens-place digit, 100 times the hundreds-place digit, and so on.

The Government's briefs deprecated the inventiveness that it took to define a decimal (BCD) number, for purposes of converting it to another base, as U + 10T +100H + ... In a six-page appendix to its reply brief, the Government undertook to demonstrate that the portion of the respondents' brief describing their "mathematical discovery" was nothing but an obfuscated recitation of how a number such as 253 in decimal notation is (3 x 1) + (5 x 10) + (2 x 100), and that you can convert it into binary by carrying out the foregoing summation in binary arithmetic. (You replace all BCD numerals by their binary equivalents and replace all tens and its powers by 1010 [binary ten] and its powers.) Counsel for the assignee of the patent application, Bell Labs, tried to paint the algorithm as a great mathematical discovery, while Government counsel went to equal pain to portray the algorithm as so trivial that perhaps even a Justice could have devised it if asked to solve the problem. The Court's opinion did not explicitly mention the point. (But see the opinion's table of powers of 10.)

Was the triviality or great intellectual advance of the algorithm material to the ratio decidendi of the case? To its outcome? Why the emphasis? Does it matter whether the claimed invention is based on E=mc2 or is based on a+(b+c) = (a+b)+c? Does it tilt the equities? Should it?

2. Now, here's a less jaundiced view of the Benson algorithm and how its works. It actually was an improvement over brute-force methods for converting BCD to binary. The Bell Labs patent lawyer handling the case in the CCPA was more successful in explaining this to that court than Washington DC counsel in the Supreme Court was. This is how the Benson algorithm works: By way of background, a normal brute-force method of converting BCD to binary works this way: Consider the decimal number 1492, which is 1x103 + 4x102 + 9x101 + 2x100. In BCD notation, that is 0001 0100 1001 0010. We can convert 1492 to binary as follows: We have a table of powers of 2: 2, 4, 8, 16, 32, 64, 128, 256, 512, 1024, 2048 . . . We’ll use the table to look up values as we proceed.

The largest power of 2 smaller than 1492 is 1024. We subtract 1492-1024, which = 468. That remainder is smaller than 512, so the first two binary digits are 10. (There is no 512 place.)

We move to the 256 place to get the next binary digit. We subtract 468-256, which = 212, the remainder after subtracting the largest power of 2 available. Now, we know the first three binary digits are 101. The remainder is 212, which we use next.

Proceeding along, 212-128=84, which gives us another 1, so that we have 1011 and a remainder of 84.

Next, 84-64=20, so we have a 1 for the 64 place. The remainder is only 20, and we therefore bypass 32 for another 0 in the running figure for the binary answer, so we now have binary 101110.

Next 20-16=4, so we have binary 1011101. We bypass 8, since 8>4, so now we have binary 10111010. The remainder of 4 gives us a 1 as next binary digit in the 4 place, so now we have binary 101110101. There is no remainder, so the last two places (the 2 and 1 places) are 0s, for an answer of 10111010100.

That is decimal 1492 in binary, and it is a conversion of BCD 0001 0100 1001 0010 into binary. This and another brute-force method of decimal-binary conversion are described at link. (We converted BCD to binary via decimal, then converted decimal to binary.)

Now, after considering that ugly approach, consider this: Start with BCD 1492, i.e., 0001 0100 1001 0010. Remember, this is the sum of 1x103 + 4x102 + 9x101 + 2x100. Recall, also, that binary eight is 0100, binary 2 is 0010, and 8+2=10. Also, we can rewrite 1x103 + 4x102 + 9x101 + 2x100 as 10{10[10(1)+4]+9}+2. We are going to expand that from the inside—i.e., [10(1)+4] out. We will multiply by ten by first shifting once to the left to multiply by two, then shift two more places to the left to multiply by eight, and then sum those two to get 8+2=10 times the starting number. This is the basis of the process or algorithm that Benson claimed to have invented. Here goes:

Now, take the first (leftmost) digit, decimal 1, BCD 0001. Shift it once to the left, getting BCD 0010; put that in a register to hold it. Shift it another two times to the left, getting 1000. Add that to the register getting a sum of 1010. That is ten times the first (leftmost) digit.

Add the next digit, decimal 4, BCD 0100, to the register. The sum is now 1110. (Decimal 14.)

Shift the sum once to the left, store result; shift another two times, taking sum. That is 10001100. (Decimal 140.) Add the next digit, which is decimal 9, BCD 1001. Result is 10010101. (Decimal 149.)

Shift the sum once to the left, store result; shift another two times, taking sum. That is 10111010010. (Decimal 1490.) We are almost there—two more to go. Add the last digit, decimal 2, binary 0010, to get decimal 1492 in binary: 10111010100. Same result as in note 1, but accomplished with just a series of shifts and adds. Re-read note 1 and try to decide whether the picture of this algorithm painted there and in the opinion of the Court is fair. Does it matter to the rationale of the Court's opinion and to the policies at stake whether the Benson algorithm is a great advance in thinking, or even whether Benson originated the idea?

The Patent Office apparently did not know this, but a generally similar technique was common at least as early as 1964. See J. Nuttall, A Computer Process for the Conversion of Binary-coded Decimal Numbers to Pure Binary Form, The Radio and Electronic Engineer, November 1965, pp. 317. Table 1 illustrates the method as performed with pencil and paper. The author describes the process in these terms:

A common form of [BCD to binary] conversion is to multiply the left-most character by ten (in binary) and to add the next character to it. The result is then multiplied by ten and the next character added, and so on until the number is in binary form.

The author also discusses the concept of multiplying by the base of a number system by shifting to the left, although his interest is in binary-coded octal. The author describes what he considers a superior method of doing BCD to binary conversion, and asserts that it is “twice as fast as the more conventional approach” (that previously described) because it requires fewer addition steps. The Benson patent appplication was filed October 9, 1963. The Nuttall paper was submitted for publication six months later. How much earlier the “conventional approach” was in use is unclear.

As explained in note 1, above, the Government argued that Benson’s putative invention was that a number in the decimal system is 1 times the units place plus 10 times the tens place plus 100 times the hundreds place, and so on. In light of note 2, however, would it be more accurate to say that the invention was that 101 = 21 + 23?

3. One of the recurrent metaphors in algorithm cases makes its appearance here. That is the powerful metaphor of the artisan's tool. Thus, the Court says, "Phenomena of nature, though just discovered, mental processes, and abstract intellectual concepts are not patentable, as they are the basic tools of scientific and technological work." The Court fails to explain what conclusion is supposed to follow, perhaps on the theory that it was obvious. Is it?

The Government had said:

The underlying rationale of the [law's] limitation of patentability to the pragmatic application of ideas is that ideas as such -- scientific principles, laws of nature, mental processes, and abstract intellectual concepts -- are the basic currency or mental tools of scientific and technological development, whose free exchange must not be hindered by indiscriminate granting of patent monopolies, lest the result be to obstruct, rather than to promote, "the Progress of Science and useful Arts." For that reason, the Pythagorean Theorem, the binomial theorem, Gibbs' vectors, the Laplace Transform, the general theory of relativity, and Russell's theory of types, for example, even though the products of great intellectual effort, or a flash of genius, are not patentable under our law. Mathematical theorems, abstractions, ideas, and laws of nature are the property of everyone and the exclusive right of no one.

The implication of this passage, and the corresponding passage in the Court's opinion, is that granting patents (gasp, monopolies!) on these basic tools would bring technological progress to a halt, or at least significantly retard it. After all, if you take his screwdriver and calipers away from a mechanic, or take his plumb bob and trowel away from a mason...then what? What if you take his voltmeter and oscilloscope from an electrical engineer? His spreadsheet from an MBA? His C compiler from a programmer? His CAT-scan software from a radiologist? (Note that the word take is now omitted. Why? What does that mean? Did he have them 100 years ago?)

Imagine that I newly invent the screwdriver or plumb bob. (If that is too hard to swallow, how about the Philips-head screwdriver? Philips-head screws, and presumably screwdrivers for them, appeared on the scene in the 20th century.) Do I get patents? What social consequence?

Is the tool metaphor faulty? Is E = mc2 properly analogized to a screwdriver or calipers? What about the proposed legal rule for which the metaphor of the mechanic's tool was invoked?

4. The Court relied on the precedent in O'Reilly v. Morse. As you will recall, Morse claimed that his invention, on which he was entitled to a patent, was "the use of the motive power of the electric or galvanic current, which I call electromagnetism, however developed, for marking or printing intelligible characters...at any distances." The Court allowed that Morse was entitled to the earlier apparatus-limited claims such as claim 1, but not claim 8, which had no apparatus limitations. How is that relevant to the issues in the Benson case?

5. What rule of law the Court announced in Benson was not clear. The famous "nutshell" near the end of the opinion was too opaque to be a guide for decisions:

It is conceded that one may not patent an idea. But in practical effect that would be the result if the formula for converting BCD numerals to pure binary numerals were patented in this case. The mathematical formula involved here has no substantial practical application except in connection with a digital computer, which means that if the judgment below is affirmed, the patent would wholly preempt the mathematical formula and in practical effect would be a patent on the algorithm itself.

If the claimed subject matter has no substantial practical application except in connection with a digital computer, does that make it unpatentable? To fail under Benson must the proposed patent, if allowed, "wholly preempt" the use of the algorithm? What if the applicant claims (“preempts”) only some fields of use of the algorithm and abandons (or gerrymanders out) the other fields? The uncertainty led to further litigation.

6. What rule of law the Government sought in Benson was also unclear. At times—for example, very early on—it seemed that the Government would have been satisfied if the respondents had been willing to limit the claim to use of the procedure in the specific environment of a PBX (telephone switching) system. One problem that the Patent Office had was that it is infeasible to search for prior art except in terms of the subject matter categories into which the Office has classified prior art—the so-called shoes. (It would be like trying to look something up in the West Digest System when there is no key number for the subject on which you want to find cases, or trying to look up a word in the dictionary when you do not know how to spell it.) A field-of-use limitation and/or an apparatus limitation would thus have gone a long way toward overcoming the problem of impossibility to search for something not in the classification categories. (Whether the search would have been productive and would have furthered the purposes of the patent examination system is another question.)

In this vein, the Government relied on the telegraph (Morse) and telephone (Bell) cases to support the following proposed rule of law (Benson Main Govt. Br.):

These cases show that claims for methods exploiting natural principles or laws (including the laws of physics or mathematics), if they are patentable at all, are patentable only when they contain proper limitations both (1) confining the monopoly over the application of the method to a specified field of technology and (2) confining the grant to a particular type of apparatus or hardware to be used to carry out the method.

(Observe that both a field-of-use [“shoe”] limitation and an apparatus limitation are demanded.) Suppose claim 8 of the patent application had read this way: The method of converting number-dial signals from a telephone from binary coded decimal form into binary which comprises the steps of (1) storing the binary coded decimal signals in a reentrant shift register, said shift register located in a telephone network, . . . (where the bodface and underlined words are added, and the claim continues after that as it originally did). How would that claim have fared under the foregoing proposed rule?

The respondents argued that they were just seeking a patent on a PBX (Private Branch eXchange) telephone system, so that they really satisfied the test demanded. The Government countered, however, that although the respondents mentioned “PBX” two dozen times in their brief they didn’t mention it once in the claims.

The respondents then pointed to process-patent cases in each of which the process patent was not limited to particular apparatus, such as the cases involving the patents on vulcanization of rubber, conversion of fat to soap, and grinding of flour. Thus, in Cochrane v. Deener, 94 U.S. 780, the Court had said of the patented process that it made no difference whether the flour was pulverized with a hammer, pestle and mortar, or flour mill. But that was not the point of the invention. The point was to sieve out the finer part of the flour, however pulverized, subject the remainder to an air blast, and then regrind the part of the remainder left after the air blast carried away the “pulverulent impurities.”

The Government replied that the inventions in all of the cases where the apparatus was immaterial gained the concreteness and tangibility necessary to qualify for a patent from the fact that the claimed process operated on a physical object and transformed it into a different physical object, in contrast with the intangible entities (electrical signals) addressed in the telephone and telegraph cases. In those case, apparatus limitations (for example, Morse's repeater apparatus or Bell's magneto or variable resistance apparatus) were essential to lend tangibility and concreteness to the claimed subject matter. Since the method in the Benson case operated on intangible information, not rubber, fat, flour, eggs, sheet metal, or some other tangible object as in the other pure-process cases, the Government argued that method needed apparatus and end-use limitations to be eligible for patenthood. Thus, in its Reply Brief, the Government said that “we submit that the cases follow such a rule––implicitly or explicitly––and that they cannot be rationalized otherwise.” Id. at 9. (This was the first explicit articulation of what eventually became known as the machine-or-transformation test.)

The Court refused to resolve the question. It said:

It is argued that a process patent must either be tied to a particular machine or apparatus or must operate to change articles or materials to a “different state or thing.” We do not hold that no process patent could ever qualify if it did not meet the requirements of our prior precedents.

The Court also mentions several other things that “we do not hold.” What did the Court hold, besides Bell Labs loses and the Patent Office wins this round?

7. Several other themes are found in this case that will recur in the subsequent episodes of the Algorithm War. One concerns “the old piano-roll blues.” [Second link.] The respondents argued that machine patents should be allowed for algorithms when claimed as programmed digital computers, because putting a new computer program into a computer turns the latter into a "new" machine. Inspired by what copyright precedent did the Government respond that this is like saying that if you put a new piano roll into an old player piano you get a new player piano?

The very end of the Benson opinion elaborates on a major recurrent theme: “leave it to Congress.” What are the comparative institutional advantages of judicial and legislative legislation, respectively, in this field? Has Congress already acted, by enacting § 101 in 1952, or was it in the Act of Apr. 10, 1790, ch. 7, § 1, 1 Stat. 109?

Another recurrent theme appearing here is the relationship between §§ 101 and 103 in algorithm cases. In its main brief in Benson, the Government had said:

[T]here is no claim here that the mathematical principle respondents claim synergizes with the computer or cooperates with it in some new and unconventional way that makes for invention. Without novel and unconventional cooperation between the claimed mathematical principle or programs embodying it and the computer, the claimed method amounts to no more than another conventional use for a known machine, comparable to the insertion of a new piano roll into an old player piano. The development of such a use is not a patentable “discovery.”

The respondents then challenged the Government for trying to inject § 103 obviousness considerations into a § 101 statutory subject matter case. The Government replied (Benson Reply Br.):

Respondents misunderstand the government's references to synergism and obviousness. We do not contend here that the patent should not be granted because respondents' mathematical theorem is obvious. We contend that the patent should not be granted because (1) what is claimed amounts to a direct translation of their theorem into another language (that of computer programming) and (2) there is no additional claim of an unobvious implementing device, nor any synergism between a device and the algorithm, that raised the claimed algorithm above the unpatentable status of a theorem.

What does that mean? Why isn't it an injection of § 103 matters into a § 101 inquiry?

8. Perhaps the most important recurring theme in these opinions is preemptiveness—the concern that patents on abstract, sweeping formulations of claimed inventions will preempt some entire field of technology and enclose it against the opportunity for others to claim patent rights in the field. Thus, Morse’s claim to the exclusive right to the use of the electromotive force would preempt the invention of TV and the fax machine, and thereby discourage innovative activity in that technology. See Brenner v. Manson, 383 U. S. 519 (1966), for a more detailed discussion of this policy issue. In The Telephone Cases, 126 U.S. 1 (1888), the Court upheld Bell’s process claim because the use of the process was limited to specific magneto and variable resistance apparatus:

We see nothing in Morse’s case to defeat Bell’s claim; on the contrary, it is in all respects sustained by that authority. It may be that electricity cannot be used at all for the transmission of speech except in the way Bell has discovered, and that therefore, practically, his patent gives him its exclusive use for that purpose; but that does not make his claim one for the use of electricity distinct from the particular process with which it is connected in his patent. It will, if true, show more clearly the great importance of his discovery, but it will not invalidate his patent. [126 U.S. at 535.]

Thus, even if subsequent events proved that Bell de facto preempted all practicable use of the underlying principle, his claim did not in principle do so, as Morse’s claim 8 had. We thus see the Court making a distinction bertween de facto and de jure preemption based on claim language.

9. In the next software case that went to the Supreme Court, Dann v. Johnston, see ch. 9, the Government faced the CCPA's distinguishing Benson away on the ground that its precedent applied only to method claims, not apparatus claims. In Johnston, the applicant claimed a system for facilitating the sorting of checks into separate categories (rent, fuel, medical expenses, etc.). The claims were written in the format of "means for processing so-and-so," rather than a method of doing so-and-so by such-and-such steps. Thus, instead of a series of steps in the form—(1) doing A, (2) doing B, (3) doing C—we have this series: (1) means for doing A, (2) means for doing B, (3) means for doing C. The CCPA and respondent contended that the claims were directed to a machine, because a "means" is an apparatus is a machine, and thus immune to Benson. Judge Rich dissented because he considered this a misreading of Benson and said that the CCPA should not disregard the Supreme Court’s rulings on patent eligibility by resorting to the legal fiction that a means-for claim described a machine. Application of Johnston, 502 F.2d 765 (CCPA 1974).

The Government said that this was a machine devised not by an artisan but by a patent lawyer. (In an excess of partisan zeal, the Government also argued so vigorously that the claimed subject matter was obvious that the Supreme Court never addressed the Government's main issue—patentability of the subject matter. Instead, the Court simply held Johnston's claimed invention obvious.)

The piano-roll blues thus became a major theme on appeal, since the lower court (enthused by the piano-roll blues) had held the subject matter statutory on the ground that Benson did not apply to an algorithm-related procedure claimed in "means for," machine format (3-2, over an interesting dissenting opinion by Judge Rich). The claimed "machine," of course, was a conventional general-purpose digital computer programmed to carry out Johnston's method.

The Supreme Court's opinion in Johnson is not highly informative as to statutory subject matter, since it merely holds that "the gap between the prior art and respondent's system is simply not so great as to render the system nonobvious to one reasonably skilled in the art." The government's brief on patentability in that case is condensed immediately below, however, because it further develops and possibly marks the maximal extension of the § 101 / § 103 arguments first raised in Benson, then partly adopted in Flook (the second Supreme Court algorithm decision), and then retreated from (5-4) in Diehr, the third and last Supreme Court algorithm decision. After Flook and Diehr, the CCPA lost interest in singing the piano-roll blues. As will appear, however, the piano-roll blues much later made a big comeback, in the CAFC in 1994, after several decades of eclipse.

10. Before moving on, however, let’s take a step backwards—about thirty years—to a case that sets the stage for the Flook case and the revival of its doctrine in the 21st century. What happens when the implementation of a natural principle, idea, or other inherently unpatentable (“patent-ineligible”) thing is implemented in a machine or other physical product, but one so conventional, trivial, or obvious that there is no conceptual space separating the implementation from the patent-ineligible idea? Suppose that implementation effectively covers any possible implementation of the idea? Does the patent effectively become a patent on the idea?



Funk Bros. Seed Co. v. Kalo Inoculant Co. Unitd States Supreme Court
333 U.S. 127 (1948)

Mr. Justice Douglas

Through some mysterious process leguminous plants are able to take nitrogen from the air and fix it in the plant for conversion to organic nitrogenous compounds. The ability of these plants to fix nitrogen from the air depends on the presence of bacteria of the genus Rhizobium, which infect the roots of the plant and form nodules on them. These root-nodule bacteria of the genus Rhizobium fall into at least six species. No one species will infect the roots of all species of leguminous plants. But each will infect well-defined groups of those plants. Each species of root-nodule bacteria is made up of distinct strains which vary in efficiency. Methods of selecting the strong strains and of producing a bacterial culture from them have long been known. The bacteria produced by the laboratory methods of culture are packaged for sale to and use by agriculturists in the inoculation of the seeds of leguminous plants. This also has long been well known.

It was the general practice, prior to the Bond patent, to manufacture and sell inoculants containing only one species of root-nodule bacteria. The inoculant could therefore be used successfully only in plants of the particular inoculation group corresponding to this species. Thus if a farmer had crops of clover, alfalfa, and soy beans he would have to use three separate inoculants. There had been a few mixed cultures for field legumes. But they had proved generally unsatisfactory because the different species of the Rhizobia bacteria produced an inhibitory effect on each other when mixed in a common base, with the result that their efficiency was reduced. Hence it had been assumed that the different species were mutually inhibitive. Bond discovered that there are some strains of each species of root-nodule bacteria which do not exert a mutually inhibitive effect on each other. He also ascertained that those mutually noninhibitive strains can, by certain methods of selection and testing, be isolated and used in mixed cultures. Thus he provided a mixed culture of Rhizobia capable of inoculating the seeds of plants belonging to several cross-inoculation groups.

Representative claim 1 reads: An inoculant for leguminous plants comprising a plurality of selected cultures of different species of bacteria of the genus Rhizobium, one of said cultures being Rhizobium trifolii alpha, said cultures being substantially unaffected by each other in respect to their ability to fix nitrogen in the leguminous plant for which they are specific.*

      * Claim 5 more generally covers: An inoculant for leguminous plants comprising a plurality of strains of different species of bacteria of the genus Rhizobium, at least one of which strains is unaffected by the others in respect to its ability to fix nitrogen in the leguminous plant for which it is specific.


Bond did not create the state of inhibition or noninhibition in the bacteria. Their qualities are the work of nature. Those qualities are of course not patentable. For patents cannot issue for the discovery of the phenomena of nature. See Le Roy v. Tatham, 14 How. 156, 175. The qualities of these bacteria, like the heat of the sun, electricity, or the qualities of metals, are part of the storehouse of knowledge of all men. They are manifestations of laws of nature, free to all men and reserved exclusively to none. He who discovers a hitherto unknown phenomenon of nature has no claim to a monopoly of it which the law recognizes. If there is to be invention from such a discovery, it must come from the application of the law of nature to a new and useful end.

Bond's discovery of the fact that certain strains of each species of these bacteria can be mixed without harmful effect to the properties of either is a discovery of their qualities of noninhibition. It is no more than the discovery of some of the handiwork of nature and hence is not patentable. The aggregation of select strains of the several species into one product is an application of that newly discovered natural principle. But however ingenious the discovery of that natural principle may have been, the application of it is hardly more than an advance in the packaging of the inoculants. Each of the species of root-nodule bacteria contained in the package infects the same group of leguminous plants which it always infected. No species acquires a different use. The combination of species produces no new bacteria, no change in the six species of bacteria, and no enlargement of the range of their utility. Each species has the same effect it always had. The bacteria perform in their natural way. Their use in combination does not improve in any way their natural functioning. They serve the ends nature originally provided and act quite independently of any effort of the patentee.

There is, of course, an advantage in the combination. The farmer need not buy six different packages for six different crops. He can buy one package and use it for any or all of his crops of leguminous plants. And, as respondent says, the packages of mixed inoculants also hold advantages for the dealers and manufacturers by reducing inventory problems and the like. But a product must be more than new and useful to be patented; it must also satisfy the requirements of invention or discovery. The application of this newly-discovered natural principle to the problem of packaging of inoculants may well have been an important commercial advance. But once nature's secret of the noninhibitive quality of certain strains of the species of Rhizobium was discovered, the state of the art made the production of a mixed inoculant a simple step. Even though it may have been the product of skill, it certainly was not the product of invention. There is no way in which we could call it such unless we borrowed invention from the discovery of the natural principle itself. That is to say, there is no invention here unless the discovery that certain strains of the several species of these bacteria are noninhibitive and may thus be safely mixed is invention. But we cannot so hold without allowing a patent to issue on one of the ancient secrets of nature now disclosed. All that remains, therefore, are advantages of the mixed inoculants themselves. They are not enough.

Reversed.



Note

Suppose I discover that a chemical (trypsin) has a desirable therapeutic effect (it is an anti-inflammatory) when absorbed from the small intestine (the ileum), but stomach acid (HCl) destroys it when it is ingested. What should I do to make a useful oral medication? I know, and perhaps you do too, that putting an enteric coating around a pill will protect it from stomach acid and let it reach the small intestine where it can dissolve and be absorbed. So, I claim an oral dosage unit comprising a therapeutically effective amount of trypsin enterically coated, as well as a method of alleviating inflammation comprising administering to a patient suffering from inflammation an oral dosage unit of enterically coated trypsin. In my specification, I point out that I have discovered that when a concentrated dosage of trypsin is placed in the ileum, absorbability of such an extent occurs that an anti-inflammatory effect is achieved by some still mysterious transfer to the point of injury. I not only discovered the absorbability of trypsin by the ileum but also the anti-inflammatory effect of that absorption. See Armour Pharmaceutical Co. v. Richardson-Merrell, Inc., 396 F.2d 70 (3d Cir. 1968). The Third Circuit said:

Our reading of the Supreme Court's opinion in Funk leads us to conclude that the test of patentability of a natural phenomenon is as follows: Would an artisan, knowing the newly discovered natural phenomenon require more than ordinary skill to discover the process by which to apply that phenomenon as the patentee had done? Once nature’s secret that the ileum would absorb trypsin was uncovered, any artisan would have known the process of enterically coating the trypsin to enable it to pass through the acidic environment of the stomach and continue into the ileum. … [The patentee’s] application of that newly discovered principle would itself have to be inventive in order to sustain [the] patent.



Government Brief — Dann v. Johnston
Excerpt from Brief for Petitioner in Dann, Commissioner of Patents & Trademarks v. Johnston,
425 U.S. 219 (1976)

[An introductory portion of this brief is found in ch. 7 as a tutorial.]

II. As a new idea for using existing general-purpose digital computers, respondent's claimed invention does not constitute patentable subject matter

The gist of respondent's claimed invention is the idea of having banks modify their usage of existing computers so as to include and process an additional item of information (the category code) associated with each transaction in its customers' checking accounts. Thus, as is often the case with new developments in the field of computer applications, the assertedly imaginative element consists essentially of the idea of combining and relating specific types of data in a particular manner.

* * * * *

However unobvious they may be, ideas have always constituted "nonstatutory" and thus unpatentable subject matter. What may be patentable is the specific and tangible application of novel ideas in some concrete form to a specific use, in which the idea is concretely embodied in a novel and inventive machine, article of manufacture, chemical or other physical composition, or process for making or using such machinery, manufactures, or compositions. The rationale of the rule against patenting ideas per se is that hindering the free exchange and spread in use of mathematical, scientific and other ideas by an indiscriminate grant of patent monopolies would obstruct, rather than promote, "the Progress of Science and useful Arts." U.S. Const., Art. I, § 8, Cl. 8.

Thus, this Court stated in Benson that "[p]henomena of nature, though just discovered, mental processes, and abstract intellectual concepts are not patentable, as they are the basic tools of scientific and technological work." The same principlehas been applied to business ideas, because patents on methods of transacting business would destroy legitimate competition, just as patents on scientific ideas would impede scientific progress.

For example, the idea of the self-service cafeteria, like Morse's idea of transmitting information at a distance by means of electromagnetic force cannot be patented. No matter how novel, such ideas can be the basis for a patent, if at all, only in terms of, and to the extent of, their embodiment in a novel machine (such as Morse's telegraph apparatus or the coin-operated food dispensing machine used in certain cafeterias or in a novel process.

Even when an idea is unobvious its implementation must be unobvious, as well, to justify the grant of a patent monopoly, whether the implementation involves a process or a product. The patent laws require that not only must there be the application of the idea to a new and useful end (embodied in a specific, tangible product or process), but the conception and devising of the end product or process must display the inventive faculty. The means or mode of implementing the idea must extend beyond the routine, conventional, or obvious applications of the idea that follow once the idea is known. See, e.g., Benson; Funk Bros. Seed Co. v. Kalo Inoculant Co., 333 U.S. 127, 132;* Armour Pharmaceutical Co. v. Richardson Merrell, Inc., 396 F.2d 70 (C.A. 3);** National Lead Co. v. Western Lead Products Co., 324 F.2d 539 (C.A. 9); Davison Chemical Corp. v. Joliet Chemicals, Inc., 179 F.2d 793 (C.A. 7), cert. denied, 340 U.S. 816; Loew's Drive-In Theatres, Inc. v. Park-In Theatres, Inc., supra, 174 F.547 (C.A. 1), cert. denied, 338 U.S. 822.

  *     In Funk the patent covered a particular mixture of strains of bacteria useful in fixing nitrogen in soil, because the strains were not mutually inhibitory. The discovery that such strains existed was not inventive, because it was a natural phenomenon or principle. As for making the patented product, this was obvious "once nature's secret of the non-inhibitive quality of certain strains of the species of Rhizobium was discovered," and the patent was therefore invalid.

  **     In Armour the court stated: "[T]he test of patentability of a natural phenomenon is as follows: Would an artisan, knowing the newly discovered natural phenomenon, require more than ordinary skill to discover the process by which to apply that phenomenon as the patentee had done? ...Once nature's secret that the ileum [intestine] would absorb trypsin [the enzyme] was uncovered, any artisan would have known the process of enterically coating the trypsin to enable it to pass through the acidic environment of the stomach and continue into the ileum. ...Assuming arguendo that ... [the patentee had discovered that trypsin absorbed in the ileum had an anti-inflammatory effect, his] application of that newly discovered principle would itself have to be inventive in order to sustain a patent.

In other words, the process or apparatus implementing a new idea must itself satisfy the conditions of patentability, in which case a patent monopoly may be granted, not on the idea, but on the inventive process or product.

III. The implementation of respondent's new idea for computer usage is not patentable

Respondent seeks a patent by reason of having devised a new use for an existing machine, the conventional general-purpose digital computer. While one who invents a new use for an existing machine may under some circumstances be entitled to a patent monopoly, respondent's new use does not qualify for such a grant.

Since the enactment of the first patent law in 1790, Congress and the courts have adhered to the principle that a new use for an old machine can never be patented as a machine.

Like much of American patent law, as embodied both in statutes and in decisions of the courts, this principle was first expressed by Thomas Jefferson. In his famous letter to Isaac McPherson in 1813 (cited in Graham v. John Deere Co., 383 U.S. 1, 8 n.2, 10 n. 3). Jefferson said:

If one person invents a knife convenient for pointing our pens, another cannot have a patent right for the same knife to point our pencils. A compass was invented for navigating the sea; another could not have a patent right for using it to survey land. A machine for threshing wheat has been invented in Scotland; a second person cannot get a patent right for the same machine to thresh oats, a third rye, a fourth peas, a fifth clover, etc. A string of buckets is invented and used for raising water, ore, etc., can a second have a patent right to the same machine for raising wheat, third oats, a fourth rye, a fifth peas, etc.?

[VI Writings of Thomas Jefferson 176-177 (Washington ed. 1854).]

He added that one of the general rules established by the patent board, of which he had been a member, was that a machine of which we were possessed, might be applied by every man to any use of which it is susceptible and that this right ought not to be taken from him and given to a monopolist, because the first perhaps had occasion so to apply it.

The insightful analysis Jefferson addressed to the simpler machines of his day applies with equal force to machines as complex and versatile as the modern general-purpose digital computer. When the principal benefit of an existing machine lies in the versatility of its applications, no conventional use of the machine for a new purpose entitles anyone to an exclusive patent monopoly covering the machine.

In revising the Patent Code in 1952, Congress explicitly recognized limited patent protection for one who invents a new use for an existing machine. Thus, in § 100(b) Congress defined "process" as including "a new use of a known...machine," while § 101 provides that one who invents or discovers "any new and useful process, ... or any new and useful improvement thereof, may obtain a patent therefor," subject to the other conditions and requirements of the Patent Code (e.g., novelty and non-obviousness). Enactment of § 100(b) did not affect the established principle that a new use of an old machine is not patentable as a machine; rather, it simply meant that a new use could be patentable, if at all, only as a process.

Respondent, however, has abandoned his "process" claims for the new use of computers involved here. As a result, he has eliminated what would have been the only possible basis for patentability of a new use of existing general-purpose digital computer. Abandonment of the process claims was well advised and their retention would not have availed respondent here -- because in Benson this Court held that a program that had no substantial practical application except in connection with a general-purpose digital computer was not a patentable "process" within the meaning of §§ 100(b) and 101 of the Patent Code.

Thus, in Benson this Court reversed a judgment of the court below, which had allowed a patent monopoly over a method for using a conventional, general-purpose digital computer to convert one form of numerical information (binary-coded decimal) to another (pure binary). In response to the suggestion that its decision "precludes a patent for any program servicing a computer," the Court stated that it did not so hold, explaining that it dealt with a program for a digital computer and did not extend its holding against patentability to analog computers. The Court concluded that: "[i]f these programs are to be [made] patentable," then Congress must deal with the matter.

While the outer boundaries of the Benson ruling may not be precisely defined, we believe that it covers the instant case. Respondent claims an idea for programming, in the ordinary way, a conventional, general-purpose digital computer. At a minimum, Benson precludes patent monopolies on such programs as processes -- the only relevant category -- regardless of the possible patentability of other types of computer programs.

The rationale of Benson is that such programs are not sufficiently removed, inventively, from the ideas they embody. The translation of an idea into a program for a conventional, general-purpose digital computer no more imports patentability to the idea than does the translation of the idea from English to French or from English to Fortran (or any other computer language). Programs for use of conventional, general-purpose digital computers are unpatentable under Benson, unless perhaps, at the very least, they "synergize" with the computer, so that the two co-act in an improved, new, and unobvious manner combining "ordinary elements in an extraordinary way" (Deepsouth Packing Co. v. Laitram Corp., 406 U.S. 518) in which "the whole in some way exceeds the sum of its parts." Great Atl. & Pac. Tea Co. v. Supermarket Equipment Corp., 340 U.S. 147; Anderson's Black Rock, Inc. v. Pavement Salvage Co., 396 U.S. 57. Respondent's programming idea does not meet this test.

Respondent has attempted, by inventive semantics and a facile drafting device, to circumvent the foregoing barriers to obtaining a patent on his idea for programming existing digital computers. Thus, rather than seeking a machine patent by claiming discovery of a new use for an existing machine (barred by long-standing authority), he claims that an existing digital computer programmed to implement his idea is a "new machine," and rather than seeking a process patent on his idea for use of a computer (barred by Benson), he presses his claims in terms of a "machine system" consisting of the programmed computer. These efforts are without merit.

In no meaningful sense has respondent invented a "new machine" or "machine system." The only actual machine involved is the familiar general-purpose digital computer, which is inherently capable of performing a wide variety of data processing tasks when so instructed through appropriate programs. If one could call such a computer, each time its program is modified, a "new machine," it would be so only in a purely linguistic or rhetorical sense, not in any pragmatic and realistic sense; it would be a "new machine" devised not by a mechanic or artisan but by one skilled in "the highly developed art of drafting patent claims."

The court below elevated form over substance in declaring Benson not "applicable" merely because respondent's patent application characterized his idea as a "machine system." That holding would provide a simple means for circumventing the important principle applied in Benson. As in that case, respondent's idea involves, at best, merely a new use for an existing machine, and that idea should no more be patentable when described here as a new "machine system" than when described in Benson as a new "process." As Judge Rich observed, below, any competent patent draftsman can readily convert claims from one format to the other.

Respondent contends that he has "devised a [new] single special-purpose computer." He argues that his "claims per se do not raise any issue of computer programs; that they do not call for the use of a computer program in the construction of respondent's special-purpose computer," because a "wired computer-circuitry version" of respondent's program could instead be used to carry out the idea of performing the bookkeeping service, and that, in any event, "a program inserted in a general-purpose computer creates a new machine." The latter view appears to be based upon Application of Bernhart, 417 F. 2d 1395, 1400 (C.C.P.A.). As applied to respondent's patent application, at least, such an analysis cannot withstand scrutiny.

First, respondent's own description of his alleged invention, in the court below, conceded that this invention "is being sold as a computer program to banks and to other data-processing companies so that they can perform these data-processing services for depositors." In addition, respondent's application, and his disclosure required under 35 U.S.C. 112, are expressly founded upon the use of existing automatic data-processing equipment employing general purpose digital computers for bank record-keeping; thus, the court below observed that the general control for the bookkeeping system, "as allegedly reduced to practice, is in the form of a software program" and that the application illustrates its use by means of a complete program and flow chart "for use with a known commercial general purpose computer, the IBM 1400 series, to provide the special-purpose computer which forms the apparatus of appellant's invention."

More importantly, respondent has not claimed or disclosed a new "hardware" special purpose machine in his application. As Judge Rich recognized in dissent, respondent has proposed only a software program for use in the existing hardware of general-purpose computers. If respondent's invention is embodied in the program, it simply cannot encompass in any relevant way the various existing machines with which it may be used.

Second, respondent's suggested use of wired computer circuitry, instead of programming, is without substantial practical application. Respondent sells his idea to banks as a program, to use in conjunction with the banks' existing program routines for processing checks and deposits. Banks would have no interest in destroying the versatility of their general-purpose computers (often leased) by building "wired computer-circuitry versions" of the program into the machine. This is as unrealistic as the argument, rejected in Benson, that the conversion method in issue there could be of practical use in anything but a general-purpose digital computer.

Finally, respondent's premise (based on Bernhart) that a program in a general purpose computer creates a new special purpose machine, is unsound. As Judge Rich pointed out, "the legal doctrine that a new program makes an old general-purpose digital computer into a new and different machine ... partakes of the nature of a legal fiction when it comes to drafting claims." The Board of Appeals properly observed that, even assuming the correctness of Bernhart, supra, there is no structural or operational change in the physical identity of the computer simply because respondent, the bank, or the bank's customer has decided to associate a new meaning with a particular physical state of memory cells within the computer, i.e., that a particular combination of signals in a particular context is to be deemed "rent" or "dividends." Nor does a general-purpose digital computer become a "different" machine simply because respondent, the bank, or the bank's customer chooses to have such an association made.

There is no basis in the Patent Code, its legislative history, or the prior decisions of this Court for this strained and unreasonable conception of identity and difference.* A new program in an old computer no more turns the computer into a "new" machine (let alone a new and unobvious one) than: putting a new piano roll into an old player piano makes it a "new" piano; putting a new recording on an old phonograph turns it into a "new" phonograph; inserting into a "mag card" typewriter a new magnetic card coded with the text of a new letter turns the machine into a "new" typewriter; or shaking an old kaleidoscope turns it into a "new" one. In each case it is precisely the ability to perform in such varied ways that is the essence of the old machine.

  *     Respondent's arbitrary assignment of meaning to "same" and "different" is reminiscent of Heraclitus' concept that one can never step into the "same" river twice because "different" waters will be flowing past.

The general-purpose digital computer used to implement respondent's idea (e.g., the IBM 1400) is not respondent's creation. When the patents covering such machines expire, their subject matter will fall into the public domain. Under the decision of the court below, however, respondent would receive, in effect, a second and further partial monopoly on "making, using or selling" such machines, which pro tanto would impermissibly withdraw from the public its right of "free access to materials already available." We do not believe that Congress intended the present patent laws to make available such a monopoly.

We acknowledge that the dynamic interrelationship between general-purpose digital computers and the software programs through which they are applied to numerous and varying tasks may call for new measures or legal concepts to further reward and stimulate innovation in programming. This Court recognized as much in Benson when it called upon Congress to determine "[i]f these programs are to be patentable" or otherwise to be protected. But, as the law now stands, however high "a reach of imaginative ingenuity" a program for a general-purpose digital computer may have demanded from its creator, it is nonetheless not patentable "because it will not be within the terms of the statute." Old Town Ribbon & Carbon Co. v. Columbia Ribbon & Carbon Mfg. Co., supra. See also Deepsouth Packing Co. v. Laitram Corp., supra.

Notes

1. Critics of the theory advanced in this and similar Government briefs have called it a two-invention theory. First, the inventor has to make invention #1. Then, to get a patent he has to make invention #2. Is that a valid point?

2. Consider Rubber-Tip Pencil Co. v. Howard, 87 U.S. (20 Wall.) 498 (1874). Rubber erasers were known. But how do you find the eraser when you need it? Blair invented, "as a new article of manufacture, an elastic, erasive pencil head" consisting of a rubber cap with a longitudinal shaft, slightly narrower in inside diameter than the outside diameter of a pencil to be inserted into it, and adapted to be placed over the nonwriting end of the pencil and held thereon by the inherent elasticity of the rubber. The Court said that "it is very evident that the essential element of the invention as understood by the patentee was the facility provided for attaching the [rubber eraser] head to the pencil." The Court held that there was no invention in making a piece of rubber with a hole in it that is smaller than what you are going to stick into the hole. Everybody knows that rubber is elastic and that, if you stick into a hole in rubber a thing slightly bigger than the hole, the rubber will stick to the thing. "What, therefore, is left for this patentee," the Court concluded —

but the idea that if a pencil is in serted into a cavity in a piece of rubber smaller than the rubber itself the rubber will attach itself to the pencil, and when so attached become convenient for use as an eraser? An idea of itself is not patentable, but a new device by which it may be made practically useful is. The idea of this patentee was a good one, but his device to give it effect, though useful, was not new. Consequently, he took nothing by his patent.

Do you consider that a fair disposition of Blair's invention?

What does the Court mean by "new" (next to last sentence of quoted passage)?

This is what counsel for the patentee argued to the Court:

Lead-pencils have very long — longer than any living man remembers — been used to make marks. India-rubber has very long — longer than any living man remembers — been used to rub them out. But never until lately was india-rubber used for this purpose except in a form disconnected from the pencil. But on a summer's morning of 1867, one Blair, a poor artist of Philadelphia, seeing that it will be more convenient to use it on his pencil than off, puts in a certain way, a piece of a certain shape, on the pencil, and finding a great advantage in thus using such a piece, shows what he has done. Behold! thousands, hundreds of thousands and millions of rubber-tipped pencils at once appear. Very rich companies, like the Rubber-Tip Pencil Company, are incorporated. Great capital is invested in the matter, and rubber-tipped pencils become a manufacture of the nation. How can it be said that there is no invention here? So far as the patent laws are concerned, utility, as ascertained by the consequences of what is done, is the test of invention, and when utility is proved to exist in any great degree, a sufficiency of invention to support the patent must be presumed. In such a case it is vain to talk about the small amount of ingenuity shown or to say that the arrangement and application are so simple and obvious that anybody could see them.

3. Johnston felt that he was being denied equal protection of the law by the Government's attitude in Dann v. Johnston. He put into his brief a section captioned, "Equal protection requires that respondent's machine invention be given the same patent protection afforded any new and unobvious machine." Was he right?

What do you think the Government said? (Among other things, it said that the same short shrift would have been given to Faraday, Willard Gibbs, Einstein, Bertrand Russell, and Boole. See Reply Brief for the Petitioner in Johnston, at 10 n.7.)



Parker (Acting Commissioner of Patents) v. Flook
United States Supreme Court
437 U.S. 584 (1978)

Mr. Justice Stevens.

Respondent applied for a patent on a "Method for Updating Alarm Limits." The only novel feature of the method is a mathematical formula. In Gottschalk v. Benson, 409 U.S. 63, we held that the discovery of a novel and useful mathematical formula may not be patented. The question in this case is whether the identification of a limited category of useful, though conventional, post-solution applications of such a formula makes respondent's method eligible for patent protection.

I

An "alarm limit" is a number. During catalytic conversion processes, operating conditions such as temperature are constantly monitored. When any of these "process variables" exceeds a predetermined "alarm limit," an alarm may signal the presence of an abnormal condition indicating either inefficiency or perhaps danger. Fixed alarm limits may be appropriate for a steady operation, but during transient operating situations, such as start-up, it may be necessary to "update" the alarm limits periodically.

Respondent's patent application describes a method of updating alarm limits. In essence, the method consists of three steps: an initial step which merely measures the present value of the process variable (e.g., the temperature); an intermediate step which uses an algorithm* to calculate an updated alarm-limit value; and a final step in which the actual alarm limit is adjusted to the updated value. The only difference between the conventional methods of changing alarm limits and that described in respondent's application rests in the second step the mathematical algorithm or formula. Using the formula. an operator can calculate an updated alarm limit once he knows the original alarm base, the appropriate margin of safety, the time interval that should elapse between each updating the current temperature (or other process variable), and the appropriate weighting factor to be used to average the original alarm base and the current temperature.

  *     We use the word "algorithm" in this case, as we did in Gottschalk v. Benson, 409 U.S. 63, 65, to mean "[a] procedure for solving a given type of mathematical problem..."

The patent application does not purport to explain how to select the appropriate margin of safety, the weighting factor, or any of the other variables. Nor does it purport to contain any disclosure relating to the chemical processes at work, the monitoring of process variables, or the means of setting off an alarm or adjusting an alarm system. All that it provides is a formula for computing an updated alarm limit. Although the computations can be made by pencil and paper calculations, an abstract of disclosure makes it clear that the formula is primarily useful for computerized calculations producing automatic adjustments in alarm settings.

The patent claims cover any use of respondent's formula for updating the value of an alarm limit on any process variable involved in a process comprising the catalytic chemical conversion of hydrocarbons. Since there are numerous processes of that kind in the petrochemical and oil-refining industries, the claims cover a broad range of methods. They do not, however, cover every conceivable application of the formula.

II

[The patent examiner and Board of Appeals of the PTO rejected the patent application.]

The CCPA reversed. It read Benson as applying only to claims that entirely preempt a mathematical formula or algorithm, and noted that respondent was only claiming on the use of his method to update alarm limits in a process comprising the catalytic chemical conversion of hydrocarbons. The court reasoned that since the mere solution of the algorithm would not constitute infringement of the claims, a patent on the method would not preempt the formula.

The Acting Commissioner of Patents and Trademarks filed a petition for a writ of certiorari, urging that the decision of the CCPA will have a debilitating effect on the rapidly expanding computer "software" industry, and will require him to process thousands of additional patent applications. Because of the importance of the question, we granted certiorari.

III

This case turns entirely on the proper construction of § 101 of the Patent Code, which describes the subject matter that is eligible for patent protection. It does not involve the familiar issues of novelty and obviousness that routinely arise under §§ 102 and 103 when the validity of a patent is challenged. For the purpose of our analysis, we assume that respondent's formula is novel and useful and that he discovered it. We also assume, since respondent does not challenge the examiner's finding, that the formula is the only novel feature of respondent's method. The question is whether the discovery of this feature makes an otherwise conventional method eligible for patent protection.

The plain language of § 101 does not answer the question. It is true, as respondent argues, that his method is a "process" in the ordinary sense of the word."* But that was also
  *     The statutory definition of "process" is broad. An argument can be made, however, that this Court has only recognized a process as within the statutory definition when it either was tied to a particular apparatus or operated to change materials to a "different state or thing." See Cochrane v. Deener, 94 U.S. 780. As in Benson, we assume that a valid process patent may issue even if it does not meet one of these qualifications of our earlier precedents.
true of the algorithm, which described a method for converting binary-coded decimal numerals into pure binary numerals, that was involved in Gottschalk v. Benson. The holding that the discovery of that method could not be patented as a "process" forecloses a purely literal reading of § 101. Reasoning that an algorithm, or mathematical formula, is like a law of nature, Benson applied the established rule that a law of nature cannot be the subject of a patent. Quoting from earlier cases, we said:

"A principle, in the abstract, is a fundamental truth; an original cause; a motive; these cannot be patented, as no one can claim in either of them an exclusive right." Le Roy v. Tatham, 14 How. 156, 175. Phenomena of nature, though just discovered, mental processes, and abstract intellectual concepts are not patentable, as they are the basic tools of scientific and technological work.

The line between a patentable "process" and an unpatentable "principle" is not always clear. Both are "conception[s] of the mind, seen only by [their] effects when being executed or performed." In Benson we concluded that the process application in fact sought to patent an idea, noting that --

[t]he mathematical formula involved here has no substantial practical application except in connection with a digital computer, which means that if the judgment below is affirmed, the patent would wholly preempt the mathematical formula and in practical effect would be a patent on the algorithm itself.

Respondent correctly points out that this language does not apply to his claims. He does not seek to "wholly preempt the mathematical formula," since there are uses of his formula outside the petrochemical and oil-refining industries that remain in the public domain. And he argues that the presence of specific "post-solution" activity the adjustment of the alarm limit to the figure computed according to the formula distinguishes this case from Benson and makes his process patentable. We cannot agree.

The notion that post-solution activity, no matter how conventional or obvious in itself, can transform an unpatentable principle into a patentable process exalts form over substance. A competent draftsman could attach some form of post-solution activity to almost any mathematical formula; the Pythagorean theorem would not have been patentable, or partially patentable, be cause a patent application contained a final step indicating that the formula, when solved, could be usefully applied to existing surveying techniques. The concept of patentable subject matter under § 101 is not "like a nose of wax which may be turned and twisted in any direction." White v. Dunbar, 119 U.S. 47.

Yet it is equally clear that a process is not unpatentable simply because it contains a law of nature or a mathematical algorithm. For instance, in Mackay Radio & Telegraph Co. v. Radio Corp. of America, 306 U.S. 86, the applicant sought a patent on a directional antenna system in which the wire arrangement was determined by the logical application of a mathematical formula. Putting the question of patentability to one side as a preface to his analysis of the infringement issue, Mr. Justice Stone, writing for the Court, explained: "While a scientific truth, or the mathematical expression of it, is not patentable invention, a novel and useful structure created with the aid of knowledge of scientific truth may be." Funk Bros. Seed Co. v. Kalo Co., 333 U.S. 127, expresses a similar approach: "He who discovers a hitherto unknown phenomenon of nature has no claim to a monopoly of it which the law recognizes. If there is to be invention from such a discovery, it must come from the application of the law of nature to a new and useful end." Mackay Radio and Funk Bros. point to the proper analysis for this case: The process itself, not merely the mathematical algorithm, must be new and useful. Indeed, the novelty of the mathematical algorithm is not a determining factor at all. Whether the algorithm was in fact known or unknown at the time of the claimed invention, as one of the "basic tools of scientific and technological work," see Gottschalk v. Benson, it is treated as though it were a familiar part of the prior art.

This is also the teaching of our landmark decision in O'Reilly v. Morse, 15 How. 62. In that case the Court rejected Samuel Morse's broad claim covering any use of electro-magnetism for printing intelligible signs, characters, or letters at a distance. In reviewing earlier cases applying the rule that a scientific principle cannot be patented, the Court placed particular emphasis on the English case of Neilson v. Harford, Web. Pat. Cases 295,371 (1844) which involved the circulation of heated air in a furnace system to increase its efficiency. The English court rejected the argument that the patent merely covered the principle that furnace temperature could be increased by injecting hot air, instead of cold into the furnace. That court's explanation of its decision was relied on by this Court in Morse:

It is very difficult to distinguish it [the Neilson patent] from the specification of a patent for a principle, and this at first created in the minds of the Court much difficulty; but after full consideration, we think that the plaintiff does not merely claim a principle, but a machine, embodying a principle, and a very valuable one. We think the case must be considered as if the principle being well known, the plaintiff had first invented a mode of applying it. We think this case must also be considered as if the principle or mathematical formula were well known.

Repondent argues that this approach improperly imports into § 101 the considerations of "inventiveness" which are the proper concerns of §§ 102 and 103. This argument is based on two fundamental misconceptions.

First, respondent incorrectly assumes that if a process application implements a principle in some specific fashion, it automatically falls within the patentable subject matter of § 101 and the substantive patentbility of the particular process can then be determined by the conditions of §§ 102 and 103. This assumption is based on respondent's narrow reading of Benson, and is as untenable in the context of § 101 as it is in the context of that case. It would make the determination of patentable subject matter depend simply on the draftsman's art and would ill serve the principles underlying the prohibition against patents for "ideas" or phenomena of nature. The rule that the discovery of a law of nature cannot be patented rests, not on the notion that natural phenomena are not processes but rather on the more fundamental understanding that they are not the kind of "discoveries" that the statute was enacted to protect. (The underlying notion is that a scientific principle, such as that expressed in respondent's algorithm, reveals a relationship that has always existed, such as the law of gravitational attraction, F = m1·m2/d2, which existed before Newton announced it.) The obligation to determine what type of discovery is sought to be patented must precede the determination of whether that discovery is, in fact, new or obvious.

Second, respondent assumes that the fatal objection to his application is the fact that one of its components the mathematical formula consists of unpatentable subject matter. In countering this supposed objection, respondent relies on opinions by the CCPA which reject the notion "that a claim may be dissected, the claim components searched in the prior art, and, if the only component found novel is outside the statutory classes of invention, the claim may be rejected under 35 U.S.C. § 101." Our approach to respondent's application is, however, not at all inconsistent with the view that a patent claim must be considered as a whole. Respondent's process is unpatentable under § 101, not because it contains a mathematical algorithm as one component, but because once that algorithm is assumed to be within the prior art, the application, considered as a whole, contains no patentable invention. Even though a phenomenon of nature or mathematical formula may be well known, an inventive application of the principle may be patented. Conversely, the discovery of such a phenomenon cannot support a patent unless there is some other inventive concept in its application.

Here it is absolutely clear that respondent's application contains no claim of patentable invention. The chemical processes involved in catalytic conversion of hydrocarbons are well known, as are the practice of monitoring the chemical process variables, the use of alarm limits to trigger alarms, the notion that alarm limit values must be recomputed and readjusted, and the use of computers for "automatic monitoring-alarming." Respondent's application simply provides a new and presumably better method for calculating alarm limit values. If we assume that that method was also known, as we must under the reasoning in Morse, then respondent's claim is, in effect, comparable to a claim that the formula 2πr can be usefully applied in determining the circumference of a wheel. As the CCPA has explained, "if a claim is directed essentially to a method of calculating, using a mathematical formula, even if the solution is for a specific purpose, the claimed method is nonstatutory."

To a large extent our conclusion is based on reasoning derived from opinions written before the modern business of developing programs for computers was conceived. The youth of the industry may explain the complete absence of precedent supporting patentability. Neither the dearth of precedent, nor this decision, should therefore be interpreted as reflecting a judgment that patent protection of certain novel and useful computer programs will not promote the progress of science and the useful arts, or that such protection is undesirable as a matter of policy. Difficult questions of policy concerning the kinds of programs that may be appropriate for patent protection and the form and duration of such protection can be answered by Congress on the basis of current empirical data not equally available to this tribunal.
It is our duty to construe the patent statutes as they now read, in light of our prior precedents, and we must proceed cautiously when we are asked to extend patent rights into areas wholly unforeseen by Congress. As Mr. Justice White explained in writing for the Court in Deepsouth Packing Co. v. Laitram Corp., 406 U.S. 518:

[W]e should not expand patent rights by overruling or modifying our prior cases construing the patent statutes, unless the argument for expansion of privilege is based on more than mere inference from ambiguous statutory language. We would require a clear and certain signal from Congress before approving the position of a litigant who, as respondent here, argues that the beachhead of privilege is wider, and the area of public use narrower, than courts had previously thought. No such signal legitimizes respondent's position in this litigation.

The judgment of the CCPA is reversed.

[Link to audio (mp3) of announcement of opinion by Justice Stevens and of dissent by Justice Stewart]


Appendix to Opinion of the Court

Claim 1 of the patent describes the method as follows:

1. A method for updating the value of at least one alarm limit on at least one process variable involved in a process comprising the catalytic chemical conversion of hydrocarbons wherein said alarm limit has a current value of Bo + K, wherein Bo is the current alarm base and K is a predetermined alarm offset, which comprises:

(1) Determining the present value of said process variable, said present value being defined as PVL;

(2) Determining a new alarm base B1, using the following equation:

B1 = Bo(l.0-F) + PVL(F)

where F is a predetermined number greater than zero and less than 1.0;

(3) Determining an updated alarm limit which is defined as B1 + K; and thereafter

(4) Adjusting said alarm limit to said updated alarm limit value.

In order to use respondent's method for computing a new limit, the operator must make four decisions. Based on his knowledge of normal operating conditions, he first selects the original "alarm base" (Bo); if a temperature of 400 degrees is normal, that may be the alarm base. He next decides on an appropriate margin of safety, perhaps 50 degrees; that is his "alarm offset" (K). The sum of the alarm base and the alarm offset equals the alarm limit. Then he decides on the time interval that will elapse between each updating; that interval has no effect on the computation although it may, of course, be of great practical importance. Finally, he selects a weighting factor (F), which may be any number between 99% and 1%, and which is used in the updating calculation.

If the operator has decided in advance to use an original alarm base (Bo) of 400 degrees, a constant alarm offset (K) of 50 degrees, and a weighting factor (F) of 80%, the only additional information he needs in order to compute an updated alarm limit (UAV), is the present value of the process variable (PVL). The computation of the updated alarm limit according to respondent's method involves these three steps:

First, at the predetermined interval, the process variable is measured; if we assume the temperature is then 425 degrees, PVL will then equal 425.

Second, the solution of respondent's novel formula will produce a new alarm base (B1) that will be a weighted average of the preceding alarm base (Bo) of 400 degrees and the current temperature (PVL) of 425. It will be closer to one or the other depending on the value of the weighing factor (F) selected by the operator. If F is 80%, that percentage of 425 (340) plus 20% (1-F) of 400 (80) will produce a new alarm base of 420 degrees.

Third, the alarm offset (K) of 50 degrees is then added to the new alarm base (B1) of 420 to produce the updated alarm limit (UAV) of 470.

The process is repeated at the selected time intervals. In each updating computation, the most recently calculated alarm base and the current measurement of the process variable will be substituted for the corresponding numbers in the original calculation, but the alarm offset and the weighing factor will remain constant.


Notes

1. Justice Stevens uses the terms "formula" and "algorithm" interchangeably. Is that sound? For every formula, is there an algorithm? Consider the formula for the Pythagorean Theorem: x2 + y2 = h2. A corresponding algorithm might be:

  • To find the hypotenuse of a right triangle, first take the shortest side and square its length.
  • Then take the next shortest side and square its length.
  • Sum the two squares.
  • Take the square root of the foregoing sum.
  • That is the hypotenuse.

Is there a formula for every algorithm? Does the formula-algorithm distinction, if any, affect the validity of the Stevens legal analysis?

2. Justice Stevens says that the Mackay and Funk decisions point to the proper analysis for algorithm-related patents: the process, itself, not merely the algorithm, must be new and useful. We can probably disregard the word "useful" because there would be little or no point in the applicant's going to the bother and expense involved if the process and/or algorithm were useless. Hence, can we properly restate this rule (which we may designate for mnemonic purposes as the McFunky rule) as follows: the process utilizing the algorithm must be new; whether the algorithm, itself, is new does not matter.

Is that all? Farther along, Justice Stevens says that McFunky requires "an inventive application of the principle" and that "the discovery of a phenomenon cannot support a patent unless there is some other inventive concept in its application." Does that mean that novelty of the whole subject matter, in the patent-law sense (which is failure to disclose the entire subject matter identically in a single reference) is not enough? How would you therefore restate the prior statement of the McFunky rule?

3. Comment on how Justice Stevens turns around "considered as a whole" in countering the argument that he is importing §§ 102 and 103 into § 101, and improperly dissecting the seamless web of the patent claim to root out its non-statutory point of novelty. He says that he is considering it as a whole — A + B + C + D — and, when A is recognized as an unpatentable algorithm and is therefore disregarded (erased, per Nielson v. Harford, as explained in Morse), the whole (B + C + D, which remains) is not novel and is thus unpatentable. Does that reply to and overcome the gestaltist criticism?

Here is another way of putting it:

“The chemical processes involved in catalytic conversion of hydrocarbons are well known, as are the practice of monitoring the chemical process variables, the use of alarm limits to trigger alarms, the notion that alarm limit values must be recomputed and readjusted . . .” (Flook opinion)

Let A = "The chemical processes involved in catalytic conversion of hydrocarbons"
Let B = "the practice of monitoring the chemical process variables"
Let C = "the use of alarm limits to trigger alarms"
Let D = "the notion that alarm limit values must be recomputed and readjusted"
Let E = smoothing algorithm - the Flook algorithm

  • Flook claims A + B + C + D + E
  • Flook does not claim any novel cooperation among A+B+C+D+E or any subcombination thereof (see Flook opinion)
  • It is conceded that A+B+C+D is old (Flook opinion)
  • Per Morse and Neilson, E must be regarded as if old
  • Accordingly, A+B+C+D+E is conceded to be old
  • Therefore, Flook does not even purport to claim a NEW and useful invention as § 101 requires
  • Therefore, Flook's patent application fails per § 101
4. This is a § 101 case, not a § 103 case. In § 103, the statute states that the inquiry as to obviousness must be directed to the invention (claimed subject matter) considered as a whole. Where in § 101 does the statute command consideration of something as a whole? Ergo?

5. Relate these equations to the McFunky test as stated in Flook:

[SSM + Implementation = Sum]
0 + 0 = 0
0 + 1 = 1

6. The following is an excerpt from the Government's main brief in Flook (footnotes and most citations omitted):

2. Novel and Inventive Use

Even if the end-use limitations in respondent's claims were sufficient to distinguish his application from the one involved in Benson, he would still not be entitled to a patent under § 101, To merit monopoly protection, it is not sufficient that a claim embody a concrete, tangible application of a mathematical or scientific principle; it must also extend beyond the routine, conventional, or uninventive applications that follow once the applicant possesses the idea.

Thus in Funk Bros. Seed Co. v. Kalo Inoculant Co., the patentee had discovered the law of nature that particular strains of bacteria used to inoculate certain plants did not have an inhibitory effect on one another. He then applied this discovery by preparing a new mixture of bacteria cultures. Despite the fact that this new mixture was a concrete product, newly devised to be employed for a specific end use in a specific art or technology, the Court held that it nevertheless "fell short of invention within the meaning of the patent statutes." The reason, as the Court explained, was that there was nothing inventive in preparing the mixture of bacteria to carry out the discovery, once the scientific discovery (unpatentable in itself) was grasped — "however ingenious the discovery may have been, the application of it is hardly more than an advance in the packaging of the inoculants."

Thus, regardless of the novelty of the basic scientific or mathematical principle, its application must also be inventive for the method or product to qualify for a patent grant. Unless that test of inventiveness is continued, the public will be deprived of its right freely to utilize the ordinary applications that follow from abstract ideas and laws of nature.

Respondent's application of the algorithm for determining alarm limits does not meet this test of inventiveness. The only inventive contribution in respondent's method is in the mathematical formula by which alarm limits are calculated. Once respondent worked out the equations or algorithm for computing the values, each aspect of the implementation of this algorithm was entirely straightforward and conventional just as the aggregation of certain bacteria in Funk and [reference to another case] were straightforward and conventional implementations of the scientific principles involved in those cases.

Since respondent's method involves neither a newly devised apparatus nor any other inventive manner of implementing an idea, there exists nothing to raise the innovative mathematical algorithm from the unpatentable status of an idea to the patentable status of an "invent[ion] ... [of a] new and useful process under 35 U.S.C. § 101.

7. The following is the Government's reply brief in Flook, most footnotes and citations omitted:

1. Respondent errs in asserting that our argument confuses the standard of non-obviousness prescribed in 35 U.S.C. § 103 and the requirement of statutory subject matter under 35 U.S.C. § 101. We do not contend that respondent's particular algorithm for computing updated alarm limits is not novel or is obvious within the meaning of 35 U.S.C. §§ 102 or 103. We simply contend that the subject matter he seeks to patent is unpatentable under 35 U.S.C. § 101, because it is not an "invent[ion] or discover[y]" within the meaning of that section.*

The plain language of § 101 requires that the application of a mathematical algorithm involve invention or discovery for it to be patentable. It states that patents may issue only to one who "invents or discovers any ... process, mahine, manufacture, or composition of matter." This language dates from the original Patent Act of 1790. In none of the subsequent amendments to the patent statute has Congress altered this basic requirement.

Yet respondent would have the courts ignore this explicit language and adopt a new rule that would allow patents to issue to anyone who "[applies for a patent on] any...process, machine, manufacture, or composition of matter...." Congress could have changed the language of § 101 [in 1952] to broaden the statutory standards of patentability, but it did not ....

By denying the instant patent application, the Patent and Trademark Office is carrying out the statutory command that patents be granted only for the "invent[ion] or discover[y]" of patentable subject matter a command that derives from the constitutional authorization, and its limitation on Congress' power, that patents be granted only to "Inventors" for "Discoveries," and "[t]o promote the Progress of Science and useful Arts."

2. Respondent further contends that acceptance of petitioner's test for patentable subject matter would preclude the patenting of any invention whose conception preceded its reduction to practice.** But again, this argument misconceives the scope of petitioner's position. A test of inventiveness of implementation is appropriate for claims that are tantamount to claims to mathematical or scientific principles, in the light of the policy that patents must not be awarded on claims that preempt basic scientific concepts.

  *     We thus put aside, as unresponsive and immaterial, the contention that the test of inventiveness that we urge fails to consider the obviousness of the claimed subject matter "as a whole." We may assume that this is the test appropriate in determining obviousness under 35 U.S.C. § 103. But it is not apposite here, where the question is whether the claimed subject matter constitutes the "invent[ion] or discover[y]" of a process under 35 U.S.C. § 101.

  **     This argument is apparently based on the erroneous theory that the application of any idea, once conceived, is necessarily routine and noninventive. This theory ignores, for example, the inventiveness of Morse's and Bell's respective forms of apparatus that they devised to carry out their ideas to use laws of nature for a novel and useful purpose.

Funk and Benson indicate that this important policy dictates that an "invention" that does no more than apply such a basic principle in a straightforward, conventional manner does not merit a grant of a patent monopoly.

In the present case respondent claims a mathematical formula, the solution of which is conventionally applied to conventional technology. The formula itself is, concededly, unpatentable, but respondent seeks to transform it into a patentable "invention" by the conventional application of its results to monitor existing processes. To require the award of patent monopolies for such claims would create an unwarranted burden for all who use and develop computer software based on mathematical algorithms.

8. The Flook opinion holds that a field-of-use limitation (for example, I claim the Pythagorean theorem as applied to land surveying) is ineffective, at least when the claim covers (preempts) that field. Recall that at the very beginning of the Benson case the government would have been satisfied with a field limitation that would have let it go to a shoe to search for prior art. That moderation evaporated on further reflection. Consider this: Inventor A claims the use of the Flook algorithm in petrochemical reactors (as Flook's assignee did). Inventor B claims the use of the Flook algorithm in weather prediction. Inventor C claims the use of the Flook algorithm in investing in stocks and financial analysis. Inventor D claims the use of the Flook algorithm in maintaining grocery store inventory on a "just in time" basis. Or to make a long story short, A claims the use of the Flook algorithm in in Field F, while B claims the use of the Flook algorithm in the field of everything except F.

9. For more on the mathematics of Flook's alleged invention, see this NIST tutorial on exponential smoothing as a forecasting technique, or look that subject up with a search engine. Robert G. Brown is credited with the "invention" of this technique, which is described. among other places, in Robert G. Brown, Smoothing, Forecasting, and Prediction of Discrete Time Series (1963). Another source states that exponential smoothing was first suggested by C.C. Holt in 1957. Flook's application, Ser. No. 194,032, was filed October 29, 1971. In the figure shown at right, exponential smoothing of data (red) is shown with coefficients of 0.5 (blue) and 0.1 (green). This illustrates the result of giving greater (blue) or less (green) weight to recent data. The less heavily one weights recent data, the smoother and less responsive to change the curve is. (The Patent Office did not know about this when the patent application was pending. I did not know about it until years after I briefed the case.)




Link to next part of section A of chapter 8

Return to Table of Contents