Contents
I. New holding as to creativity and the scopeof protection in computer programs
On June20, 1989 the Tokyo High Court partially sustained an appeal from the denial of a petition for a preliminary injunction against a copy-right infringement (Case No. (ra) 327, 1989, Tokyo District Court Case No. (yo) 2531 % 2551) but Rejected appellant's claim that one of appellees' programs was an adaptation of one of appellant's programs.In its statement of reasons, the court explained as follows concerning creativity and the scope of protection in programs:
"In finding that a program infringes the copyright of a program work, it is necessary that the combinations of instructions in the program work have portions that can be found creative, and that the combinations of instructions in the later created program be similar to those in the program work in the portions that can be found creative."
"Because the symbols expressing programs are extremely limited and the relevant system (grammar) is rigorous, any attempt to make a computer functional so as to achieve a certain result in a more effective way will necessarily lead to many similar combinations of instructions. Therefore, we must be careful in finding infringement of a program copyright."
"The'processing flow' in a program, which, being itself an algorithm, i.e., a 'solution' as provided for in Art. 10 (3) (iii)1)of the Copyright Law, is a portion that is not protected as a work and therefore has no relevance to the creativity of the program."
II. Creativity, the idea/expression dichotomy
It is common knowledge that a computer program is protected as a copyrightable work under the Copyright Law. The 1985 Amendment to the Law makes this explicit while at the same time providing that no programming language, rule or algorithm is protected (Art.10 (3))2).
A computer program is protected as a copyrightable work for "its expression of thought or sentiment", just as with any other work (Art. 2(1) (i)) 3).
In determining a computer program's protectability and scope of protection, there are two key points: (1) Does it have creativity or originality, and (2) Is it expression or idea? The above-quoted provision that no programming language, rule or algorithm is protected is presumably a mere confirmative or informative provision based on the assumption that those aspects fall under the category of ideas.
III. Approaches to infringement in the United States and West German cases
Based on the above key points (1) and (2), approaches to copyrightability and the scope of protection of a specific
computer program seem to differ between common law countries and civil
law countries. In common
law countries, such as the United States, no high level of originality is required: a work can be original
if it is not an imitation of another work. It naturally follows that the
dispute in an infringement case is not so much over the copyrightability of
the plaintiff's program as over the scope of protection, with the
expression/idea dichotomy at its core. In the Whelan decision
of 1986 4), a leading
caseon the scope of protection in computer programs,
the court held that "the efficient organization of a dental laboratory"
was an idea, and that all lower levels came within the scope
of copyright protection in the program (code). Otherwise stated, the court
broadly, interpreted the scope of expression in a computer program to
extend beyond literal code to structure, sequence and organization. Whelan has been followed by a series of decisions broadly interpreting the scope of protectable expression in a computer program by applying the "look and feel" test to the issue of substantial similarity between a program and the allegedly infringing work.
On the
other hand, other decisions have refused to extend protection to such levels of abstraction as structure, sequence,
organization, or look and feel, and have limited the scope of
protection by examining constraints in specific fact situations.
In the Plains Cotton decision of 1987, 5) it
was held that similarities between programs were dictated by the externalities
of the cottonmarket, and that such similarities were at a level
beyond copyright protection. Again, in the NEC v. Intel decision of February
1989, 6)the court took a narrow view of the scope of protection
did not find infringement.In considering substantial similarity between two
sets of micro-code, the court took into account the fact that micro-code
is technically conditioned and restricted by hardware (microprocessor)
and the microinstruction set.
Thus,discussion of the legal protection of programs in the United States has revolved mainly around the idea/expression
dichotomy (i.e., whether the defendant's program is substantially
similar to the plaintiff's,and whether the similar portions are idea or expression).
In West Germany, in contrast, the Supreme Court in 1985 rendered a decision to the effect that "copy-right law does
not protect programs which are on a level within easy reach of programmers
of average skill."7) It appears from this decision that the Court will
afford extensive copyright protection to scientific and technological
works like computer programs, which are clearly based on factual observations
and technological ideas, on the condition that such works evidence
originality or creativity of an exceedingly high level.
From the above decisions we may derive three approaches to the legal protection of programs :
The first is to recognize copyrightability in works of low-level originality and to extend program protection beyond literal
code to structure, sequence, organization, and look and feel.
The secondis to recognize copyrightability in works of low-level originality but limit the scope of protection by examining
constraints in the expression of specific programs.
The third is to limit copyrightability to works of high-level originality or creativity but to afford a broad scope of
protection for such works.
The firstapproach, by offering copyright protection to (technological) ideas essentially devoid of novelty or advancement,
runs the danger of becoming a detriment to the technology protection
system. The second and the third lead to much the same conclusion.
Both deny copyright protection to (technological) ideas, the only difference
being that the one does so at the entrance, the other at the
exit.
IV. Approaches to infringement in the Japanese cases
Several infringement cases concerning computer programs have been reported from Japanese courts. However the video game cases 8) and the Microsoft v. Shuwa case 9)
relating to a BASIC interpreter simply involved what amounted to dead copies, and did not
give rise to any controversy concerning copyright infringement by the
copiers. There was no case in which the creativity of the plaintiff's program
or its substantial similarity to the defendant's program was
an issue. (NEC v. Seiko-Epson involved the uses of whether the BIOS and interpreter
of a compatible machine infringed NEC's copyright, but the
case was settled.)
After the 1985 Amendment to the Copyright Law, there was great interest in how the Japanese courts would decide questions
like a computer program creativity, the idea/expression dichotomy,
and substantial similarity between programs, and in how they would
choose among the three approaches outlined above. They waited with keen
interest for a judicial decision. An important then came in the form of
the Tokyo High Court's judgment rendered on June 20, 1989. See supra
I. Obviously the first approach was not adopted, and it appears that perhaps
the third approach was. (But the conclusion concerning "flow
of processing" suggests that there may be some room left for the second
approach.)
Thereis no imediate need to decide this question. Rather, we will review in order the decisions of the Tokyo District
Court and the Tokyo High Court.
V. Outline of the System Science case
1. Facts of the case
The facts of the case are as follows:
The System Science Co., Ltd. (petitioner and appellant, hereinafter
referred to as "System Science") manufactures
measuring devices and control systems used mainly in the area of biotechnology.
The representative director and another director of System Science
produced the following programs in the course of performing their duties:
(1) the MIC Program (for use in the device for measuring the image processing minimum growth-inhibiting concentration (MIC)), by around March 1981 ;
(2) the ZA-FM II Provisional-Version Program (for the provisional- version Zone Analyzer system ZA-FM II), by around September 1985;
(3) the ZA-FX II Provisional-Version Program (for the provisional- version Zone Analyzer system ZA-FX II), by around September 1985; and
(4) the CA-7 II Program (for the Colony Analyzer System CA-7 II), by around March 1986.
System Science brought to market and published under its corporate name the above-listed devices (1) to (3), incorporating the respective corresponding programs (1) to (3) into a ROM mounted on a circuit board -- (1) around December 1981 and (2) and (3) around February 1986. (Program (4) was developed after the abolition of the statutory requirement for "publication under the corporate name" when the author is a corporation). The copyright in Programs (1) to (4), therefore, belongedto System Science. (However, whether the copyright was assigned became an issue in the case, as discussed below.)
Toyo Measuring Co., Ltd. (respondent and appellee, hereinafter referred to as "Toyo") and Nihon Technat. Ltd. (respondent and appellee, hereinafter referred to as"Technat" distributed, advertised, or displayed for distribution devices (1) to (4), each of which included a circuit board with attached ROM in which the corresponding programs (1) to (4) were stored.
Toyo and Technat also distributed, advertised, or displayed for distribution, devices including a circuit board with attached ROM in which the CA-9 Program was stored. (Whether CA-9 was an adaptation of Program (4), i.e. , the CA-7 Program, became an issue.)
System Science, as owner of Programs (1) to (4), petitioned the court that Toyo and Technat be preliminarily enjoined from copying or adapting those programs, and from distributing, advertising, or displaying for distribution, devices in which those programs (including CA-9) were stored.
Although it is not clear from the written opinions, may be added that Toyo seems to have hired System Science to develop devices storing Programs (1) to (4) and paid System Science for them upon delivery.
2. Issues
The issues in this case were as follows:
i. To whom did the copyright in Programs (1) to (4) belong?
ii. Was the CA-9 Program an adaptation of the CA-7 II
iii. Was an injunctive relief an appropriate remedy? (Otherwise stated, was there a likelihood that serious injury or irreparable change of circumstances would result unless the right were preserved by preliminary injunction? If not, no such writ can issue.)
It need hardly be said that issue ii is of particular interest from the viewpoint of program creativity and scope of protection. The CA-9 Program, though not an exact copy of the CA-7 II Program, had similar functions. The court had rule on the creativity of CA-7 II and on substantial similarity between the two programs (and, if similar, was the similarity in idea or expression?).
3. Decisions
i. Tokyo District Court
The Tokyo District Court denied System Science's petition on all counts, giving the following reasons:
(1) Issue iii (need for injunctive relief) (regarding Toyo and Technat's claim that they were no longer copying Programs(1) to (4), and that they had no intention of doing so in the future) --
Allegedly as part of its plan for changing Programs (1) to (4) into new programs, Toyo created the CA-9 Program to replace the CA-7 II Program, and is manufacturing and distributing the CA-9 device employing the new CA-9 Program. Toyo entered into a development contract with Technat for creating new programs to replace Program (2), i.e., ZA-FM II Provisional-Version Program, and Program (3), i.e., ZA-FX II Provisional-Version Program. Toyo claimed it was also contemplating doing the same with Program (1) i.e., MIC Program. Such being the case, the court could not find that Toyo and Technat were copying or were likely to copy Programs (1) to (4). (Under these circumstances, there was no need for injunctive relief , but the issue of adaptation remained to be decided.)
(2) Issue ii (was the CA-9 Program an adaptation of the CA-7 II Program?)
The court formulated this test: "A finding of infringement of the adaptation right in a program requires, at a minimum, access by the producer of the allegedly infringing program to the infringed program, and similarity of expression in the creative portions between the allegedly infringing program and the allegedly infringed program (identity in internal expressive form). In the absence of similarity of expression in the creative portions, there is no infringement of the adaptation right, even if there is similarity or even identity of expression in non-creative portions". Copying a copyrighted work constitutes a copyright infringement. However, because direct copying is hard to prove, American law presumes copying when it is established (1) that the alleged copier had access to a program, and (2) that his program is substantially similar to it. The Tokyo District Court seems to have adopted much the same test, on the assumption that the original program is creative.
Based on this test, the court considered the following points:
(1) Expressions in the flow of processing, the main routines and the corresponding subroutines in the so-called "processing routines after input of data from the mainframe";
(2) Expression in the flow of processing, the registers used for time monitoring, and the individual instructions in the so-called " processing routines used when the printer is inoperative"; and
(3) The stack area used for subroutine call.
The court found that the two programs were practically identical to each other in these points, although the CA-9 Program contained less than one-tenth the number of bytes, of the CA-7II Program, as it consisted solely of sub-programs comprising the most basic parts. As to point (1), the CA-9 Program was a commonplace program comprised of common forms of expression that for the most part cannot vary with the programmer because of hardware constraints. As to point (2), the CA-9 Program was such a commonplace program that any average programmer would produce a similar one. As to point (3), address-setting for the stack. is, in itself, not a portion that has creativity as expression The CA-9 Program is similar to the CA-7 II Program in some portions, but this is insufficient for a finding similarity to creative portions of CA-7 II Program. Thus, petitioner has failed to present prima facie evidence that the CA-9 Program is an adaptation of CA-7 II Program.
On the grounds that Programs (1) to (4) were not being copied or were not likely to be copied, and further the CA-9 Program could not be considered an adaptation of the CA-7 II Program, the court denied System Science is petition on all counts, seeing no need to address issue i (copyright ownership).
In determining creativity and substantial similarity in the present case, the court took the following factors into account: (1) common place expression, (2) hardware constraints on forms of expression, (3) a commonplace program that would be much the same regardless of who made it, and (4) things like address-setting are not portions that have creativity as expression. On this basis, the court concluded that there was similarity in some portions but not in creative portions.
Credit is due the court for illustrating element not to be considered in determining creativity or in comparing for substantial similarity between programs (not in function or idea, but in expression as copyrightable works). (They will serve as tools if the second or third approach discussed in part III supra is adopted.) By taking up creativity and similarity together, the court made its holding somewhat unclear, but logically it appear that the court's holding is, "Petitioner holds a copyright in its program. Respondents' program is not similar to expressive portions of petitioner's Program that contain creativity.
In this case the court denied creativity of the allegedly similar portions but did not positively identify creative portions of petitioner's program. The decision may be said to rest on the assumption that there is no question of copyright infringement if the program under consideration is not similar to expressive portions of petitioner's program that contain creativity, or, in other words, if ideas and expression are not similar or if similar but noncreative ideas or expression are copied.
From this decision System Science took an appeal to the Tokyo High Court.
ii. Tokyo High Court
The Tokyo High Court granted a preliminary injunction with regard to Programs (1) to (3) only. It decided the three issues outlined above as follows:
(1) Issue i (copyright ownership )--
The copyright in Programs (1) to (4) belong to System Science.Toyo contended that those programs were produced on its initiative by System Science's engineers in the course of performing their duties while working on Toyo's business. The court was not persuaded, however, that System Science's representative director and another director produced the programs as part of Toyo's business. Toyo also argued that the devices in question were developed at Toyo's expense of Toyo, that they were manufactured by System Science in accordance with Toyo's instructions, that Programs (1) to (4) were produced solely for those devices, and that by transferring the devices the copyright in the programs naturally passed to Toyo. Again, the court was not persuaded that there was any such assignment agreement, whether implied or otherwise.
A work cannot have a corporate author unless made by "persons engaged in corporate business". Contractor and persons working to order have not traditionally been held to fall within this category. Moreover, in the absence of a contract, a copyright does not pass to the heaver of the expenses. In view of these rules of law, the decision may be called correct.
(2) Issue iii (need for injunctive relief) --
Toyo and Technat claim that they have discontinued copying Programs (1) to (4), and that they are contemplating changing them intonew programs. Indeed, they have entered into a development contract for that purpose. But the court found it premature to draw from this the conclusion that the possibility of their copying or adapting Programs (1) to (3) had been eliminated.
(3) Issue ii (was the CA-9 Program an adaptation of the CA-7 II Program?)--
The court set the necessary conditions for finding infringement of the copyright in a program work as follows: "There are combinations of instructions in the program work that can be found creative, and the combinations of instructions in the later created program are similar to portions of the program work that can be found creative." The court then outlined precautions to be taken in examining similarity between program works:
"Because the symbols expressing programs are extremely limited and the relevant system (grammar) is rigorous, any attempt to make a computer function so as to achieve a certain result in a more effective way will necessarily lead to many combinations of instructions that cannot escape being similar. Therefore, we must be careful in finding copyright infringement regarding program works."
The infringement test and the precaution are well stated. "Similarity of expression (i.e., combinations of instructions in the case of programs) in creative portions" is the traditional test that has been applied in copyright infringement cases in general10), and is the test adopted in this case by the court below. Moreover, the precaution follows from restrictions like limited symbols, rigorous system (grammar), and functional goals, which leave only a narrow scope for selection and interchange of expression.
From these premises the court proceeded to consider the issues of creativity and similarity. It begins with the following findings:
(1) Both programs govern only the printer portion (to read measurements and other data upon entry into common memory and to convert them into a code for the printer). Combinations of instructions here will necessarily be inherently similar because of hardware constraints. (2) As a processing routine to be used when the printer is inoperative, both programs adopt a very common combination of instructions.
(3) It is a matter of common sense to choose the round and convenient figure of 4100H to set the subroutine stack area.
It is difficult to find creativity in the portions of instruction combinations pointed out by System Science. Moreover, with the CA-7 II Program containing 12 kilobytes and the CA-9 Program 763 bytes, the portions that System Science alleges to be similar cover only a very small number of bytes. Taken together, petitioner's claim that CA-9 is an adoption of CA-7 II is far from persuasive.
Like the Tokyo District Court, the Tokyo High Court considered creativity and similarity together. The court creativity in the portions pointed out by System Science on the grounds that the combinations of instructions (i.e., expression) in both programs were constrained by the hardware and so could not escape being inherently similar and were general combinations. Further, the court denied that the mere presence of similar portions is enough to constitute copyright infringement. This treatment of creativity and similarity together makes the holding unclear, but logically we can say that it comes to the following; "Appellant's program is copyright protected. Appellees' program is not similar to creativity." expressive portions of appellant's program containing.
In a copyright infringement case, the defendant is likely to winin the following fact situations (see the figure below) --
1. The plaintiff's work has no (or little) creativity, so it is not eligible for copyright protection.
2. The defendant did not have access to the plaintiff's work.
3. There are no similar (or identical) parts, whether in ideas or in expression.
4. There is similarity (or identity) only in non-creative expressive parts (combinations of instructions in the case of a program) .
5. There is similarity (or identity) only in ideas.
{ EMBED Word.Picture.6 , }
The decisions of both the courts seem to come under no. 4. (Certainly they do not fall under no. 1, because nowhere is the creativity of appellants' program denied). The decisions, however, might have fallen under no. 5 had the courts chosen it. The Tokyo High Court held that the possibility of appellees' copying, adapting, distributing, advertising or displaying the CA-7, II Program was eliminated by the fact that they were distributing advertising, and displaying for distribution, devices containing the CA-9 Program.
In final analysis the court declared that the copyright in Programs (1) to (4) belonged to System Science, but recognized the need for injunctive relief only for Programs (1) to (3). rejecting appellant's claim that the CA-9 Program infringed the adaptation right in the CA-7 II Program. Consequently, the court granted a preliminary injunction accordingly.
After discussing creativity and similarity, the high court made an additional observation: "The program's 'flow of processing' itself, being an algorithm, i.e., a 'solution' as provided for in Art. 10 (3) (iii) of the Copyright Law, is a portion that does not receive protection as a work and therefore has no relevance to program creativity." Thus, the court made it clear that the "flow of processing," being a "solution", is not eligible for protection.
As stated in II above, two points have to be considered in examining the scope of copyright protection in a work. One is creativity, fully discussed in this case. The other is the idea/expression dichotomy. An idea itself, however creative or original, is not protectable.
In its decision the court concluded that the only similarities were in non-creative portions of expression (combinations of instructions). Therefore, the court did not have to specify whether the similar portion were ideas or expression, nor in particular, did it have to determine whether "flow of processing", if creative or original, was protectable as expression (combinations of instructions). Presumably this is why the high court added the obiter dictum quoted above as an "observation". When program a "flow of processing" is an issue (for example, if flow of processing is creative or original, whether it is idea or expression is a significant question) that is to say, even if processing flow is creative, it will not be protected if it is only an idea. Indeed, characterization of processing flow may be disputed in connection with the scope of copyright protection. It is also necessary to consider further whether flow of processing is in the same class as structure, sequence and organization.
The conclusion that "the program's 'flow of processing' itself, being an algorithm, i.e., a 'solution , . . . is a portion that does not receive protection as a word and therefore has no relevance to program creativity" is directly opposed to that of the Whelan decision above quoted; though an obiter dictum, it has no small significance.
Vl. Sources of Constraints on expression in programs
The Tokyo District Court and the Tokyo High Court gave some examples of matters to be excluded from consideration in determining the creativity of programs and substantial similarity of expression between programs. We could expand the list of such excluded elements as follows:
1. Expression dependent on restrictive conditions for matching interfaces with hardware, operating systems or other pieces of software to be interconnected;
2. Expression dependent on restrictive conditions for matching protocols;
3. Ideas;
4. Expression that merges with idea;
5. Common expression that does not vary with the programmer; and
(6. Flow of processing, structure, sequence and organization?)
With regard to item 6. copyrightability may depend upon the level of details.
In considering the legal protection of computer programs, the test to be applied to creativity and substantial similarity of expression is one of the most important problems in Japan as well as elsewhere. The siguificance of the decisions of the two courts lies in the contribution they have made to the formation of the test.
Notes
1) Japanese Copy Right Law (hereinafter referred
to as "JCL") Art. 10 (3):
(3) The protection granted by this Law to works
mentioned in paragraph (1), item (ix) shall not extend to any programming
language, rule or algorithm used for making such works. In this
case, the following terms shall have the meaning hereby assigned to
them respectively:
(i) "programming language" means letters and other
symbols as well as their systems for use as means of expressing
a program;
(ii) "rule" means a special rule on how to use
in a particular program a programming language mentioned in the
preceding item;
(iii) "algorithm" means methods of combining, in
a program, instructions given to a computer.
2) See Note 1)
3)JCL Art. 2 (1) (i):
(i) "work" means a production in which thoughts
or sentiments are expressed in a creative way and which falls within
the literary,scientifrc, artistic or musical domain;
4) Whelan Associates v. Jaslow Dental Laboratory, 797 F. 2d 1222 (1986), cert. denied 107 S. Ct. 877 (1987).
5) Plains Cotton Co-op v. Goodpasture computer Serv. 807 F. 2d 1256 (1987).
6) NEC Corp. v. INTEL Corp. No. C-84-20799-WPG, (N. D. Cal. Feb. 6, 1989).
7) BGH 9. Mai 1985, BGHZ 94, 276 = GRUR 1985, 1041.
8) e.g., "Space Invader Case", Tokyo District Court, decided on December 6, 1982 (Intellectual Property Law Report (Mutaizaisan-kankei Minji-Gyosei Saiban-reishu), Vol. 14, No. 3, 796, Hanrei Jiho, No. 1060, 18, Hanrei Times, No. 482, 65).
9) Tokyo District Court, decision on January 30, 1987, Industrial Property Research (Kogyoshoyukenho Kenkyu) Vol. 33, No. 1, 27.
10) e.g., "Light Emitting Diode Paper Case", OsakaDistrict Court, decided on September 25, 1979 (Hanrei Times, No.397, 152).