International Business Machines Corporation v Commissioner of Patents

[1991] FCA 811

Re: INTERNATIONAL BUSINESS MACHINES CORPORATION
And: PATRICK ANSELM SMITH, COMMISSIONER OF PATENTS
No. G40 of 1990
FED No. 811
Patents
(1992) AIPC 90-853
(1991) 105 ALR 388, (1991) 22 IPR 417
(1991) 33 FCR 218

COURT

IN THE FEDERAL COURT OF AUSTRALIA

NEW SOUTH WALES DISTRICT REGISTRY
GENERAL DIVISION
Burchett J.(1)

CATCHWORDS

Patents - appeal under s. 52 of the Patents Act 1990 from decision of a Delegate of the Commissioner of Patents refusing to accept an application and specification - principles applicable to such a refusal - discussion of United States authorities dealing with the Freeman test in relation to a claim reciting an "algorithm" - relationship of invention to principles of science and mathematics - interpretation of claim in context of specification.

Patents Act 1990, ss. 52 and 234(2)

Commissioner of Patents v Microcell Limited (1959) 102 CLR 232

National Research Development Corporation v Commissioner of Patents (1959) 102 CLR 252

In re Freeman (1978) 197 USPQ 464

In re Walter (1980) 205 USPQ 397

Diamond, Commissioner of Patents and Trade Marks v Diehr and Lutton (1981) 209 USPQ 1

Burroughs Corporation (Perkins') Application (1974) RPC 147

HEARING

SYDNEY

#DATE 13:12:1991

Counsel for the Applicant: Mr D.K. Catterns

Solicitors for the Applicant: Messrs Williams Niblett

Counsel for the Respondent: Mr D. Yates with Mrs J. Stuckey-Clarke

Solicitors for the Respondent: Australian Government Solicitor

ORDER

The decision of the Delegate be set aside and that the application and complete specification be remitted to be dealt with according to law.

Note: Settlement and entry of orders is dealt with in Order 36 of the Federal Court Rules.

JUDGE1

This is an appeal, so called, under s. 52 of the Patents Act 1990. The proceeding is in reality original: Kaiser Aluminium and Chemical Corporation v The Reynolds Metal Company (1969) 120 CLR 136 at 142. The decision that is challenged is the decision of a delegate of the Commissioner of Patents to refuse to accept an application and specification lodged on 1 June 1987, in relation to an United States basic application filed 13 June 1986. The applicable legislation is the Patents Act 1990, by virtue of s. 234(2).

1. The principles upon which an application and specification may be refused acceptance were stated in Commissioner of Patents v Microcell Limited (1959) 102 CLR 232 at 244-245:
   

"(I)t is well settled that the Commissioner ought not to refuse acceptance of an application and specification unless it appears practically certain that letters patent granted on the specification would be held invalid. As Menzies J. has pointed out, he will not normally have before him the material necessary for the formation of a concluded opinion. Moreover, whereas refusal of acceptance is final, acceptance is not: ... and, if a patent is granted, its validity is open to attack in proceedings for infringement or for revocation. So, in McDonald v Commissioner of Patents (1913) 15 CLR 713 (a case in which the validity of a patent granted on the specification may be thought to have been very doubtful) this Court allowed an appeal from the Commissioner, Griffith C.J. saying: `I think that it is only in a clear case, where it is obvious that a patent cannot be granted, that the Commissioner should reject an application altogether'. On the other hand, it is not to be overlooked that the Commissioner has a duty to the public as well as to the applicant for a patent, and, if it appears manifest that a valid patent could not be granted, the Commissioner not merely has power, but is under a duty, to reject the application ... ."

As Evershed J. said in Rantzen's Application for a Patent (1946) 64 RPC 63 at 66, it is the duty of the court "always in such a matter to give to the applicant the benefit of the doubt".

1. In the present case, the complete specification related to what was described as a "method and apparatus for producing a curve image", regarding which it stated: "The present invention relates to computer graphics and more specifically to a method and apparatus for generating curves on computer graphics displays." The recital of the prior art commenced by pointing out that the "evolution of computer technology has resulted in the creation of a sophisticated technical area devoted to the representation of graphical information generated by computers", an area termed computer graphics. The method generally used for the generation of smooth curves in computer graphics is then said to have the disadvantage that a curve "will in general have perceptible discontinuities in slopes of curvature at joints between successive curve segments unless specific curve types or forms are used". Difficulties "in generating the coordinate values" are also mentioned, and compounded inaccuracies resulting from accepted techniques. The nature of the solution described in the specification has been summarised by the Delegate, as follows:
   

"The invention involves the use of forward differencing or incrementing techniques to produce sets of integral numbers for the control points in each dimension. The numbers are produced by incremental scaling of initial parameters using integer arithmetic, which increases the speed and accuracy of the computations. Integer arithmetic multiplication operations are performed by binary shifting operations in paths connecting registers in computer curve generator circuits, and integer arithmetic adding operations are performed by adders placed serially with the registers such that the computations may be performed in a pipelined fashion. A final computational stage includes a division operation only requiring binary shifting and adding rather than a floating point operation."

1. The difficulties of the case stem from the terms of claim one, which reads as follows:
   "1. A method for producing a visual representation of a
   

curve image from a set of control points which define the curve and which are input for each dimension and a number of intervals of the curve to be computed, said method comprising the steps of:

(a) computing a set of scaled vector coefficient

integers for each dimension from the set of input control points for that dimension and from a scaling parameter;

(b) computing forward difference interval

coefficient integers for each dimension for each interval from the scaled vector coefficient integers for that dimension and the interval integer number;

(c) computing the curve coordinate values for each

interval for each dimension from the forward difference interval coefficient integers for that dimension for each interval and the scaling parameter; said computing steps being carried out without the use of floating point arithmetic; and

(d) displaying the curve by displaying curve

coordinate points in accordance with the computed curve coordinate values for each dimension and a plurality of straight lines which successively connect said computed curve coordinate points."

1. It was objected that this claim did not define a "manner of manufacture". Rather, claim one was said to recite a mathematical algorithm, which it then wholly pre-empted. An algorithm, named for the Arab mathematician whose work on algebra popularised Arabic numerals in Europe, may be relevantly defined as a procedure for solving a given type of mathematical problem. The notion that a claim which recites and wholly pre-empts such an algorithm will be bad derives from the United States decision in In re Freeman (1978) 197 USPQ 464. The rationale of the decision itself stems from the judgment of the Supreme Court of the United States in Gottschalk, Acting Commissioner of Patents v Benson (1972) 409 US 63, and is elaborated by the Supreme Court in Parker, Acting Commissioner of Patents and Trade Marks v Flook (1978) 198 USPQ 193. In the former case, Douglas J. (at 67) referred to "the long standing rule that `(a)n idea of itself is not patentable'". He added:
   

"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 US 127, 130, `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.'"

In the latter case, a corollary was asserted (at 199):

"Respondent's process is unpatentable ... 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."

1. In the present case, the Delegate referred to later refinements of the Freeman test, and particularly to In re Walter (1980) 205 USPQ 397 and In re Abele and Marshall (1982) 214 USPQ 682. As so refined, the test involves two steps: (1) "(I)t must be determined whether the claim directly or indirectly recites an `algorithm' (in the sense I have stated)" (Freeman, at 471); (2) The second step was stated in Walter (at 407), as follows:
   

"Once a mathematical algorithm has been found, the claim as a whole must be further analyzed. If it appears that the mathematical algorithm is implemented in a specific manner to define structural relationships between the physical elements of the claim (in apparatus claims) or to refine or limit claim steps (in process claims), the claim being otherwise statutory, the claim passes muster under 101. If, however, the mathematical algorithm is merely presented and solved by the claimed invention, as was the case in Benson and Flook, and is not applied in any manner to physical elements or process steps, no amount of post-solution activity will render the claim statutory; nor is it saved by a preamble merely reciting the field of use of the mathematical algorithm.

Various indicia are helpful in determining whether a claim as a whole calls merely for the solution of a mathematical algorithm. For instance, if the end-product of a claimed invention is a pure number, as in Benson and Flook, the invention is nonstatutory regardless of any post-solution activity which makes it available for use by a person or machine for other purposes. If, however, the claimed invention produces a physical thing, such as the noiseless seismic trace in In re Johnson (what this case involved will be indicated infra) the fact that it is represented in numerical form does not render the claim nonstatutory."

1. The Delegate then considered the arguments addressed to him on behalf of the applicant, who argued that the invention involved more than an algorithm; a curve without discontinuity is a vendible product that is not the same as a set of numbers, and when displayed, is a physical thing having many uses. The delegate concluded that claim one was not "directed to an industrial process". He said:
   

"The features of claim 1 clearly relate to mathematical `computing' operations on `integers', `parameters' and `numbers'; the claim is to a set of mathematical steps and therefore to an algorithm in the Benson, Flook and Walter sense."

He referred to Walter as indicating "that claims including mathematical algorithms are too broad and not patentable if they embrace `any and all arts' and are not limited to a particular industrial environment." He thought it would be easier to limit an invention to a particular industrial "environment" where that environment had nothing to do with mathematics, but more difficult where, as in this case, "the environment itself is computer technology", which involves mathematics. Although he accepted that computer graphics "is a vendible product", he added:

"However, considering claim 1, I am drawn to the conclusion that there is no environmental limitation on the claim; the `curve image' is not defined as being shown on any particular apparatus for example, the `computations' are not specified as being done in any particular environment, a hand-held calculator for example could be used, and there is no definition of how the `control points' are `input'."

He concluded:

"Thus, in applying the Freeman test to claim 1, I find that:

1. The claim recites a mathematical algorithm because it is directed to a set of mathematical steps.

2. The claim in its entirety wholly pre-empts the algorithm because it is too broad and not limited to a field of use."

He added: "If it can be shown that the curve image is defined in a proper environment, then I think it would come closer to the seismic traces." (This was a reference to the United States decision in In re Johnson, Parrack, and Lunsford (1978) 200 USPQ 199, where the invention involved methods for removing undesired "noise" from seismic data by "processes which include the performance of mathematical calculations as but one of a sequence of substantive steps" (at 210). In Johnson, this was held patentable.) The delegate did not pursue that point because he had already decided that "claim 1 fails the Freeman test". The other claims which are in dispute failed consequentially. But certain of the claims were held good. The delegate said:

"The feature of `multiplication by only binary shifting and addition', in claim 4 for example, is related to the operation of a computer and this feature limits these claims to a computer environment, in a way similar to the process limitation of claim 6 of Abele. A method of operating a computer is a patentable invention ... ."

1. A large gap would appear in this reasoning if claim one were construed as limited to the use of computers, or, to use the language of the Delegate, "to a computer environment". In Decor Corporation Pty Ltd v Dart Industries Inc. (1988) 13 IPR 385 at 400 Sheppard J. said:
   

"It is not legitimate to confine the scope of the claims by reference to limitations which may be found in the body of the specification but are not expressly or by proper inference reproduced in the claims themselves. ... Nevertheless, in approaching the task of construction, one must read the specification as a whole." (Emphasis added.)

If the present specification is read as a whole, it seems to me that there is a necessary inference confining claim one to the operation of computers. One should not pick one's way through a claim wearing blinkers that exclude from sight the context in which the claim is set. In the modern cases, it is frequently said that a specification should be given a purposive construction rather than a purely literal one (see for example Tye-Sil Corporation Ltd v Diversified Products Corporation (1991) 20 IPR 575 at 585). In the older cases, the same result was reached, although without reference to that formula: Martin v Scribal Proprietary Limited (1954) 92 CLR 17 at 102-103, where Taylor J. said:

"(W)hile it is clear that `the function of the claims is to define clearly and with precision the monopoly claimed, so that others may know the exact boundaries of the area within which they will be trespassers' (per Lord Russell of Killowen in Electrical and Musical Industries Ltd. v Lissen Ltd. (1939) 56 RPC 23, at p 39), it is equally clear that it is the duty of the court to endeavour to ascertain from the language used the true meaning of any claim or claims. It is not the function of the court lightly to discard the claims in a specification on the ground that the language used is vague and uncertain."

When the present specification is read, the whole of the context rises up to insist that claim one is talking about the operation of computers. Almost the first words of the specification inform the reader that the invention "relates to computer graphics and more specifically to a method and apparatus for generating curves on computer graphics displays". The recital of the prior art reinforces the clear impression thus conveyed from the beginning of the document. Claim one itself uses the terminology of computers - "input", "intervals of the curve to be computed", "computing ... from the set of input control points", "displaying" and "computed". The figures quite plainly involve diagrammatic representations of the functioning of computer programs.

1. At any rate, I think it would be impossible, consistently with the passage I have cited earlier in these reasons from Microcell, to refuse to entertain, for the purposes of this application, the reading relied on by the applicant. The applicant should have, in the words of Evershed J., "the benefit of the doubt".

2. Before considering further the precise point, grounded in the United States cases, in reliance on which the delegate refused the application, it is desirable to re-examine the broad principle which is in question. The leading authority is National Research Development Corporation v Commissioner of Patents (1959) 102 CLR 252, where the High Court considered the patentability of a new process for ridding crop areas of certain kinds of weeds, without damaging the crops, by the use of compounds the biochemical properties of which were inimical to the weeds but harmless to the crops. The inventiveness lay in the perception that the compounds would have this selective effect by reason of differences in the enzyme systems of the crops and the weeds. The Deputy Commissioner of Patents took the view that several claims should be deleted on the basis that they were claims "to the mere use of known substances - which use ... does not result in any vendible product". Dixon C.J., Kitto and Windeyer JJ., in their joint judgment (at 261), referred to Microcell and commented:
   

"(S)ince a requirement that claims be deleted from a specification is equivalent to a refusal of the patent for the invention claimed thereby the discretion ought not to be exercised by making such a requirement except in circumstances which would justify the refusal of a patent on a specification containing those claims and no others."

They pointed out (at 262) that if a new use of a known substance "consists in taking advantage of a hitherto unknown or unsuspected property of the material", the situation is that

"there may be invention in the suggestion that the substance may be used to serve the new purpose; and then, provided that a practical method of so using it is disclosed and that the process comes within the concept of patent law ultimately traceable to the use in the Statute of Monopolies of the words `manner of manufacture,' all the elements of a patentable invention are present: see the Microcell Case. It is not necessary that in addition the proposed method should itself be novel or involve any inventive step ... ."

Their Honours made plain (at 263-264) their view that "it is not decisive - it is not even helpful - to point out in such a case that beyond discovery of a scientific fact nothing has been added except the suggestion that nature, in its newly ascertained aspect, be allowed to work in its own way." They said: "The truth is that the distinction between discovery and invention is not precise enough to be other than misleading in this area of discussion." These observations seem to me to be as applicable to the laws of mathematics as they are to those of biochemistry.

1. The judgment (at 269) emphasizes the "breadth of the concept which the law has developed by its consideration of the text and purpose of the Statute of Monopolies", so that the test today whether a claimed invention falls within that concept is "Is this a proper subject of letters patent according to the principles which have been developed for the application of s. 6 of the Statute of Monopolies?"

2. After some discussion of the cases, the court (at 271) said:
   

"The truth is that any attempt to state the ambit of s. 6 of the Statute of Monopolies by precisely defining `manufacture' is bound to fail. The purpose of s. 6, it must be remembered, was to allow the use of the prerogative to encourage national development in a field which already, in 1623, was seen to be excitingly unpredictable. To attempt to place upon the idea the fetters of an exact verbal formula could never have been sound. It would be unsound to the point of folly to attempt to do so now, when science has made such advances that the concrete applications of the notion which were familiar in 1623 can be seen to provide only the more obvious, not to say the more primitive, illustrations of the broad sweep of the concept."

The court added (at 272): "In the varying applications of which the word `manufacture' is capable analogy has always played a considerable part."

1. The court's conclusion (at 275) was that
   

"a process, to fall within the limits of patent-ability ..., must be one that offers some advantage which is material, in the sense that the process belongs to a useful art as distinct from a fine art ... - that its value to the country is in the field of economic endeavour."

1. It is by this, by the production of some useful effect, that patent law has distinguished, so far as it has distinguished, between the discovery of a principle of science and the making of an invention. Buller J., as far back as 1795, expressed the point in Boulton and Watt v Bull (1795) 2 H Bl 463 at 485; 126 ER 651 at 662 when he said:
   

"The very statement of what a principle is, proves it not to be a ground for a patent. It is the first ground and rule for arts and sciences, or in other words the elements and rudiments of them. A patent must be for some new production from those elements, and not for the elements themselves."

The words "some new production" seem almost presciently broad: though, doubtless, Buller J. had a very different kind of production in mind, his language is perfectly apt to include the production of a curve image on the display screen of a computer. Cf. Rantzen (supra, at 66).

1. The decision of the High Court in National Research and Development Corporation was followed by the Patents Appeal Tribunal (Graham and Whitford JJ.) in Burroughs Corporation (Perkins') Application (1974) RPC 147. Graham J. (delivering the decision of the Tribunal) referred (at 154) to the proposition that "claims for a method of operating a computer could not be a manner of new manufacture within the Act since the end product of the method was `merely intellectual information' which does not fall within the meaning of `product' as defined ... ." This seems to be just the other side of the coin examined in the United States cases dealing with the use of an algorithm. Graham J. (at 158) said:
   

"(I)t is not enough to take a narrow and confined look at the `product' produced by a method. Of course, if a method is regarded purely as the conception of an idea, it can always be said that the product of such a method is merely intellectual information. If, however, in practice the method results in a new machine or process or an old machine giving a new and improved result, that fact should in our view be regarded as the `product' or the result of using the method, and cannot be disregarded in considering whether the method is patentable or not."

The Tribunal concluded (at 160-161) as follows:

"(W)e consider the correct conclusion in the present case is as follows: If a claim, whatever words are used, namely, whether the claim is for example for `a method of transmitting data ...', `a method of controlling a system of computers' or `a method of operating or programming a computer ...', is clearly directed to a method involving the use of apparatus modified or programmed to operate in a new way, as the present claims are, it should be accepted. ... Claims in this form in truth meet Mr Aldous' argument that there must be an artificial end product or effect before a method can be patentable. Whether the argument is right or not, there is in fact such an effect in this case. It is plain that the question of patentability must in every case be considered on the facts before a conclusion can be reached as to whether the claim covers a mere idea or method, or is more than that and results in fact in some improved or modified apparatus, or an old apparatus operating in a novel way, with consequent economic importance or advantages in the field of the useful as opposed to the fine arts."

Their Honours held, accordingly, that a method of transmitting data over a communication link between a central computer and slave computers connected to it, embodied in the apparatus and program described in the specification, was patentable. They added: "It also follows from what we have said, though this is not strictly necessary for the purposes of our decision, that in our view computer programs which have the effect of controlling computers to operate in a particular way, where such programs are embodied in physical form, are proper subject matter for letters patent."

1. In the present case, it seems to me that the use of the algorithm is not different conceptually from the use of the compounds involved in National Research and Development Council. Just as those compounds were previously known, so here, it is not suggested there is anything new about the mathematics of the invention. What is new is the application of the selected mathematical methods to computers, and in particular, to the production of the desired curve by computer. This is said to involve steps which are foreign to the normal use of computers and, for that reason, to be inventive. The production of an improved curve image is a commercially useful effect in computer graphics.

2. Nor, in my opinion, do the United States authorities preclude this view. The Supreme Court of the United States returned to the subject in Diamond, Commissioner of Patents and Trade Marks v Diehr and Lutton (1981) 209 USPQ 1. The decision of the court was delivered by Rehnquist J. (as he then was). What was involved was the use of a mathematical equation in a programmed digital computer to control a certain precision moulding process in relation to synthetic rubber products. Rehnquist J. (at 8) said:
   

"It is now commonplace that an application of a law of nature or mathematical formula to a known structure or process may well be deserving of patent protection."

He held (at 9) the mathematical equation was "not patentable in isolation, but when a process for curing rubber is devised which incorporates in it a more efficient solution of the equation, that process is at the very least not barred at the threshold ... ." He added (at 10):

"We recognize, of course, that when a claim recites a mathematical formula (or scientific principle or phenomenon of nature), an inquiry must be made into whether the claim is seeking patent protection for that formula in the abstract. A mathematical formula as such is not accorded the protection of our patent laws. ... Because we do not view respondents' claims as an attempt to patent a mathematical formula, but rather to be drawn to an industrial process for the molding of rubber products, we affirm the judgment of the Court of Customs and Patent Appeals."

Similarly here, the formula is applied to achieve an end, the production of the improved curve image. A method of producing that by computer, which is novel and inventive, is entitled to the protection of the patent laws.

1. For these reasons, the appeal succeeds. It is ordered that the decision of the Delegate be set aside and that the application and complete specification be remitted to be dealt with according to law.