Oracle and Google Spar Over Whether Programming Languages Can Be Copyrighted

pcritter writes "With the Oracle v. Google trial date set for next Monday, the Judge has asked Google and Oracle to take a position on whether a programming language is copyrightable. This presumably relates to whether Google violated copyright by using a variant of the Java language and its APIs in the Android framework. Oracle, who thinks it can be, has used J.R.R. Tolkein's Elvish language as an examples (PDF) of a language that can be copyrighted. Google disagrees (PDF)."

Comments

Sure.

[JustAnotherIdiot]

Just don't expect anyone to take interest in your language if you copyright it.

Re:Sure.

[Desler]

Being open source does not mean it isn't copyrighted...

Re:Sure.

[nedlohs]

No, it's a specific way of licensing your creation.

Re:Sure.

[Barefoot Monkey]

No, copyright is there with or without licensing. Copyright means that nobody can use it (for certain definitions of "use") without your permission. Licensing is giving that permission, usually with certain conditions attached. The two concepts are related in that there would be far less need for licensing without copyright, but nonetheless are very distinct.

Re:Sure.

[Chrisq]

Sad that in 2012 people are still confused by this.

Well if the Mayans were right they won't be in 2013

Re:Sure.

[DragonWriter]

While you're 100% correct about copyright being automatic (at least in the US) with respect to the implementation of the language, the question here is if that also applies to the language itself.

Copyright is automatic in US law for anything that is a copyrightable work in US law. The only question is whether a programming language, as such, is a copyrightable work -- if it is, every programming language is copyrighted at the instant it is first created (a work being "created", in US copyright law, "when it is fixed in a copy or phonorecord for the first time" -- 17 USC Sec. 101.)

Of course, the fact that the language -- as opposed to a description or implementation of the language -- is an abstraction that cannot be set in a fixed form makes it impossible for it to be "created" as defined in copyright law, but also makes it outside the scope of what is subject to copyright in the first place.

This would be a fairly new thing. Langauges like C and Pascal, for example, had zillions of different implementations each the copyright of their respective creators, but no one "owned" the language itself. (i.e. UCSD didn't pay Wirth for Pascal, Borland didn't pay Ritchie for C)

If programming languages were copyrightable works, then somebody has owned each of those languages from the moment they were created, even if, to date, they've been fairly lax in enforcing their rights against others who use the languages without permission, either directly or by creating derivative languages of their own.

(If the courts were to accept Oracle's position, I suspect that a lot of that laxness would end swiftly.)

Re:Sure.

[gtall]

A programming language has a syntax and a semantics. The semantics is generally considered mathematics. The language itself is not math. There are no reduction rules telling you how to compute in the language as opposed to, say, the lambda calculus (note the name calculus).

Consider the type system for a compute language. Where do the types exists? In the language? Nope, they are in a mathematical system into which the language must be interpreted.

That said, Oracle can suck eggs, Java isn't copyrightable because languages shouldn't be copyrightable. Their implementation may be, but language is an abstract thing. Being an abstraction does not make something mathematics, i.e., love.

Re:Sure.

[Jonner]

Just don't expect anyone to take interest in your language if you copyright it.

I'm pretty sure Sun never made any ludicrous claim that they held a copyright on the Java language rather than specific implementations of it. There's a long history of independent implementations (as opposed to Microsoft's derivative of Sun's code) based on the good public specifications from Sun that they never attacked. The fact that Oracle is completely reversing this stance is not only illogical, but extremely anti-competitive. We've been worried about Microsoft patents on DotNet stuff for a long time, but it seems we were looking at the wrong anti-competitive behemoth. Perhaps the deepest irony is that while Sun sued Microsoft because Microsoft's version wasn't compatible with theirs, Oracle is suing Google because Google's implementation is compatible with theirs at a source level.

What About Machine Language and Assembly?

[eldavojohn]

If they were copyrighted, wouldn't the assembly and machine language folks get the last laugh? I mean, copyright lasts nearly forever now in the United States and the first guy who thought up what to call registers or how to represent them in a language or what shift left should look like or even the people who came up with RISC, CISC, etc would be laughing all the way to the bank ... right? I mean, even though it might just be a handful of instructions that interact with hardware, wouldn't this position make it just as copyrighted as the higher level languages which, in the end, depend on this stuff to interact with the hardware?

Re:What About Machine Language and Assembly?

[Coffeesloth]

I agree. You can copyright a derivative work of the language but not the language itself. For instance, you can copyright a book written in English but not the English language.

Re:What About Machine Language and Assembly?

[Anne Thwacks]

In the 1960's there were a bunch of law suits over this, which is why many modern assemblers have Really Stupid (TM) mnemonics. I remember TI as deliberately using stupid mnemonics so they could copyright them. This strategy was assisted by very poorly designed documentation:

A - this instruction affects some registers

B - This instruction does not affect some registers

etc

Hint A = Add, B = Branch (which means jump) but only real gurus knew this, because the documentation did not bother to tell you.

Re:What About Machine Language and Assembly?

[maxwell demon]

A programming language is not an idea; it is a creative product, like a novel or a song or a software program.

Not every creative product is copyrightable. For example, algorithms are often a creative product, but they are not copyrightable (they may be patentable, but that's something different). Of course a specific implementation of an algorithm can (and usually will) be copyrighted.

Programming languages are more like algorithms than like programs. They correspond not to a song, but to the note system in which the music of the song may be written, and the natural language in which the text of the song is written. A completely different thing.

It has rules and symbols.

Which makes it an idea, not an expression of an idea. Because the expression of an idea has no rules, it is a specific combination of symbols, sounds, graphic elements etc. following (or sometimes not following) certain rules.

You cannot copyright the idea of a programming language, but the invention of a programming language is a specific manifestation or expression of that idea.

No. A specific implementation is an expression of the idea. The language itself cannot be an expression because it has no form. There is no way to write down a programming language, or to perform it. You can describe a programming language (either in the form of a specification, or in the form of a compiler/interpreter implementation), and those descriptions are of course copyrightable. But there is no way you can write down the language just in the same way you cannot write down an idea, but only a description of an idea (or the implementation of it, if it is an implementable idea).

Re:What About Machine Language and Assembly?

[msauve]

A programming language is a system or method of operation, and hence specifically not copyrightable in the US. A language would be a subject for patent protection.

In no case does copyright protection for an original work of authorship extend to any idea, procedure, process, system, method of operation, concept, principle, or discovery, regardless of the form in which it is described, explained, illustrated, or embodied in such work.

- 17 USC S 102

You can copyright a program expressed in a language, a book describing the language, a compiler or interpreter which implements the language, but not the language itself. Although, if Oracle can get a ruling that a language is copyrightable, then IBM may want to exert a copyright over SQL.

Re:What About Machine Language and Assembly?

[maxwell demon]

The language is defined by (a) rules to decide whether a given source is a valid program, and (b) rules to define the behaviour of any valid program.

For example, I could define a language by stating that a text is a valid program if it doesn't contain any 'e', and the behaviour of the program is to print the number of lines of the text.

That's not a very useful language, but a language. Note however that the sentence above is not the language, but a description of the language. A valid program in the language would be

bhfjdks sdfjk bvcnxbvxc nm,vc.xyn v,yxcb fjkdg ff
gfds gfjdksl mvc,xyäa dvsqwzt z cv ödf hgfdö

Patents Versus Copyright

[eldavojohn]

Except that the title of this article clearly says "copyright" and the 20 year limit is what patents are in the states (although extensions and whatnot make that mean jack anyway). Enjoy your "one generation" graph.

Derivative work of C/C++

[Luthair]

Since Java borrows a lot of syntax from from C/C++ wouldn't that make it a derivative work ;)

What is Java?

[swm]

The technical definition of the Java language is "the set of all Java programs".
This is an infinite set.
Therefore, it cannot be fixed in a tangible medium.
Therefore, it it is ineligible for copyright protection.

Re:What is Java?

[jd]

An interesting argument, there are others that are perhaps better-suited to the courtroom. Java is an API, the implementation is in the JVM and not the language. APIs are not automatically copyright, since they are a meta-description. The -implementation- of those APIs is copyright, because it is a specific case. It's the same reason that a dictionary can be copyright but the words within it cannot.

Sun successfully argued in court that to call something Java it has to implement the Java API. That's fair enough. It would be deceptive if you implemented Basic and called it C and it would be maliciously deceptive if you did so for the purpose of damaging C (whether or not it had much impact in the end), so Microsoft's deliberate attempt to destroy Java by violating the standards was definitely in the wrong and enforcing those standards in court was an impressive feat by Sun.

Sun did NOT argue Microsoft could not implement Java, merely that if they wrote something they promoted as Java then it damn well should be Java.

Oracle's argument is significantly different. The enforcement of a standard by trademark is very different from the enforcement of using a specific implementation by saying that the standard is the work.

I can respect the former. I would have no difficulty with Oracle arguing that all trademarks and licenses involving Java require adherence to the official standard, that said licenses stipulate that there should be no fragmentation of that standard as a condition of use, and that the Android implementation cannot violate the licenses involved (for the trademarks or anything else) as per the Microsoft case.

I cannot, however, respect the idea that the standard IS the implementation. It patently isn't (pun not intended), since the standard implements nothing. To allow that argument would be extremely dangerous, as it would mean that Oracle could circumvent all analysis on the legality of patents on intangibles like business methods by simply claiming they're copyright instead. Indeed, because of international copyright agreements, it would mean that things which are NOT protected (for good reason) in the EU would suddenly become protected despite contrary rulings by the EU's courts. ie: it would allow Oracle to impose a law on Europe that Europe has rejected as unlawful. Complain about the EU being undemocratic all you like, but I sure as hell didn't elect Oracle as president of anything.

Would Oracle's PL/SQL Suffer the Same?

[eldavojohn]

Oracle doesn't care about that stuff. They will say just about anything so that Larry Ellison can buy another yacht.

Uhhh well, they should. I mean Oracle's PL/SQL is an extension of SQL which, would be copyrighted by someone from the long long ago. And if that person wanted to, they could basically say "Yeah, you know that language that your bread and butter runs on? It's infringing on my copyrights so you owe me ... gosh I don't know ... a hundred billion trillion dollars?"

And, like every other language, PL/SQL has to be turned into machine language at some point ...

Re:Would Oracle's PL/SQL Suffer the Same?

[Anonymous Coward]

That someone would be...oh... IBM.

goto: Elbereth ?

[fahrbot-bot]

Oracle, who thinks it can be, has used J.R.R. Tolkein's Elvish language as an example of a language that can be copyrighted.

Sure, but I wasn't aware that was a *programming* language.

Re:goto: Elbereth ?

[tepples]

to the extent that it might be copyrightable, it would be the non-Latin alphabet

I was under the impression that alphabets, typography, and calligraphy weren't copyrightable in the home country of Oracle, Google, and Slashdot. Code of Federal Regulations, Ch 37, Sec. 202.1(e); Eltra Corp. v. Ringer; Copyright Office Practices 503.02(a). See Wikipedia:Public domain#Fonts. Fonts can be subject to a design patent, but unlike a copyright, a design patent has to be applied for, and a design patent expires.

Re:Embrace Extend?

[ppanon]

Thing is, while some third-party books do claim to "Teach ... Java for Android", as far as I know Google has never claimed that Dalvik is Java. Whereas for a typical embrace and extend, be it Kerberos or Java, Microsoft did claim that their extended fork was still Java, and marketed it as such. The Microsoft lawsuit was really about the use of the Java trademark (hence why Microsoft renamed their product J++), and not a copyright infringement case. So that case doesn't really apply as precedent here and it would seem that Oracle is overreaching.

Any lawyers here?

[nine-times]

I'm totally baffled by this and would like an explanation of how a language could possibly be copyrighted. Is Tolkein's Elvish language copyrighted, and if so, what does that mean? I can understand specific phrases from his books being copyrighted, but if I translated this post into Elvish, does Tolkein's estate suddenly own the copyright to this post? Or what?

Sorry, but the idea of owning the copyright to a language seems silly. I might understand patenting a use of a language or patenting a method of translation, but the language itself? Doesn't copyright need to apply to a specific expression? Like... I can copyright the image of a painting, but I can't copyright paint.

Re:Any lawyers here?

[mrchaotica]

The words are under copyright because Tolkien created them and defined their associated meaning. You would be perfectly fine with using the script (typefaces are not subject to copyright in the US, they are in the UK) but because all of the meanings for the words come from Tolkien having defined a combination of sounds/letters as "meaning" something a translation would be a derivative of Tolkien's creative work.

I don't buy that argument at all. If that were true, then people would also be able to copyright new English words they coined.

Re:My ass

[idontgno]

I suspect engraving "Elbereth" on the floor wouldn't prevent Oracle from attacking me. More's the shame.

Re:Yes.

[Anonymous Coward]

In the US only purely creative work can be copyrighted. Basically you can only copyright something if it has no practical application. This means that literature movies and music can be copyrighted but clothing, recipes and industrial design cannot. A fictional language (such as Klingon) could be argued to fall under the realm of copyright because it is not really practical. Google has a fairly strong case that a programming language is not copyrightable. The libraries written using that programming language in both compiled and uncompiled form would fall under copyright though.

Oracle silliness

[swillden]

I suppose lawyers have an ethical duty to argue every point that reasonably could help their client, but this is silly. I really like the comments in a couple of the footnotes in Google's response:

Similarly, fictional languages such as Na’vi and Dothraki cannot be copyrighted. While the film Avatar and the television series Game of Thrones are copyrightable (including the portions in the fictional Na’vi and Dothraki languages), and while, for example, a dictionary or grammar textbook for either language would be copyrightable, the languages themselves are not. Oracle asks why copyright should not protect such languages, see Oracle 4/5/12 Br. [Dkt. 859] at 9; the answer is that Section 102(b) says that they are not protected. Moreover, there is no reason to believe that allowing copyright owners to control who can express themselves in these languages would further the aims of copyright law.

Umm, duh. Do they really want to argue that, say, stating "Your mother is ugly" or some other random sentence in, say, Dothraki is a public performance of a copyrighted work? That's what it would have to be if the language itself were copyrightable.

Even clearer, though:

Oracle also argues that a computer language can be “original, text-based, and capable of fixation,” and thus that it must be copyrightable. See Oracle 4/5/12 Br. [Dkt. 859] at 9. First, Section 102(b) bars copyright protection for “original works of authorship” that fall within its enumerated classes of exclusion. See 17 U.S.C. 102(b). Thus, the fact that a system is original, text-based and fixed does not mean that Section 102(b) does not apply. Second, a language cannot be fixed. Certainly, a description of a language (e.g., a specification) can be fixed. A computer program written using the language (e.g., the Gmail application on Android phones) or an implementation of a language (e.g., a compiler or interpreter) can be fixed. But none of those things is “the language,” any more than a dictionary “is” English, Das Boot “is” German, or a C compiler “is” the C programming language. See Baker, 101 U.S. at 102 (“But' there is a clear distinction between the book, as such, and the art which it is intended to illustrate. The mere statement of the proposition is so evident, that it requires hardly any argument to support it.”); cf. René Magritte, La trahison des images.

Exactly. How can you write down a "language"? You can describe it. You can list the allowable words and explain how they can and cannot be put together, but such a description isn't a language. You can use it. You can create prose, poetry, or, computer programs with it, expressing and fixing your own ideas in the constructs provided by the language, but an expression in a language isn't the language.

A language is about as pure an abstract idea as I can imagine, and ideas are not copyrightable, only expressions.

(Disclaimer: I work for Google, but have nothing to do with any of this Java folderol, other than using the language occasionally. I'm a programmer but not a language expert and would not be qualified to offer expert testimony on this topic, even if I were asked to, which I haven't been. Other than the quotes, the above is my own opinions, nothing more. In this case it appears that they align closely with Google's officially-stated opinions, however!)

Re:Oracle silliness

[Derekloffin]

It goes beyond just that silliness too. Let's just give for a second that a programming language is copyrightable. Well, derivative works are also automatically copyrighted too, and Java is derivative of C++ and C before it, and probably something before that, and none of those have expired. Plus, forward facing, all use of Java of would be copyrighted too. So, suddenly, you'd have every business using Java pissed at Oracle for claiming copyright on their work, but not only that, Oracle has thrown the door wide open for being sued themselves for copyright infringement on a vast scale. Yeah, this is an argument you want to go down in flames, and really even Oracle, they may not realize it, but they too want it to go down in flames as well.

Interesting question for the judge to ask

[SkOink]

I have to admit, I'm impressed with the judge's question. I'd agree that this is really what's at the heart of the matter, and I'm glad that the judge is asking it. It certainly seems like he's taken the time to do his homework into programming languages and computing.

If Your Language Can Be Copyrighted

[Greyfox]

Then every program in that language must therefore be a derivative work. If a judge finds that a language can be copyrighted, there would be no reason why Oracle could not claim every single java program ever written is a derivative work based on their copyright. I'm pretty sure their license does not have an exclusion for that, either.

What about the language should be considered copyrightable? The keywords? Because Java looks a lot like many other languages. Perhaps the creators of those languages should sue Oracle for Copyright Infringement, then. Oracle's class and function layout looks a LOT like C and C++.

Is it the standard library? Would a clean room implementation of a published API be considered copyright infringement? I think there are precedents that it would not, at least going back to the IBM PC BIOS reverse-engineering.

Is it the idea of object orientation? That was around long before Sun released Java. I have a LISP textbook I got in the '80s that showed a lisp program doing object-oriented kinds of things with lisp data structures.

Attempting to copyright any of these things would run you afoul of the people who actually invented them. In theory you could patent aspects of your language (And they probably did) but doesn't last nearly as long as copyright would.

This case suddenly became a lot more important

[ilsaloving]

And in one fel swoop, this case has gone from simple money grabbing to downright surreal. To decide whether a language itself can be copyright-able is going to be incredibly significant, regardless of which way the final decision goes, and I'm believe it was unwise for Oracle to even raise this issue.

If a programming language cannot be copyrighted, then their whole argument goes down the tubes and they can potentially lose a lot more than just this lawsuit. They could conceivably lose control of java.

If a programming language CAN be copyrighted, I expect to see a flurry of lawsuits as different language authors start suing one another for "stealing" parts of "their" language. The vast majority, if not all, of java syntax is directly lifted from other languages. There is absolutely nothing unique about Java's grammar.

I am going to be very interested now in the outcome of this case.

Re:This case suddenly became a lot more important

[JamesP]

Yes

I sincerely think Google should quietly buy the copyright to C/C++ before this

Then let Oracle go ahead with this.

As soon as Oracle gets copyright protection on languages google does 2 things:

1 - (motion to) Block the sale/usage of anything Java since Java comes from C/C++
2 - (motion to) Block the use of anything written in C/C++: Oracle products, JVM, etc, etc

Losing Android at this point is merely an annoyance.

Congratulation Oracle, if that's what you want, that's what you get.

Re:Haven't we seen this?

[js_sebastian]

Didn't Sun already win this case against Microsoft?

AFAIK they won a trademark case, not a copyright case: Microsoft could not fork their own bastardization of Java and still call it Java.

Re:My ass

[interval1066]

No, but a circle of protection engraved with Postgresql specific charms might.

Derivative Works?

[Bob9113]

Copyright includes the notion of derivative works. Java is an excellent example of a derivative work, borrowing most of its core syntax from C. Oracle is effectively proposing that Dennis Ritchie's estate owns a huge swath of the language space; Objective C, ECMAScript, Java, C++, C# -- a big chunk of commercial programming is done in languages that are not even distant derivations, but nearly direct copies.

Re:Yes.

[Anne Thwacks]

Elvish is not "functional", and thus is copyrightable.

IMHE it is every bit as functional as Lisp.

Re:My ass

[SteelKidney]

Tolkein developed a fairly in-depth grammar, vocabulary, pronunciation guide, and even name patterns for Elves. He wasn't *just* making up words, he was creating a language.

Re:My ass

[mrchaotica]

And who says Tolkein's Elvish is copyrighted anyway? His description of it could be copyrighted, sure, but that's not the same as copyrighting the language itself. Don't tell me some dumbass judge granted an injunction on other people writing elvish grammar books...

Re:Haven't we seen this?

[pak9rabid]

Right....they called it C#.

Re:Specifications themselves are allegedly copyrig

[mrchaotica]

But the specification is not the language. The specification is a description of the language, but that's not the same as the language itself. Anybody else is free to write their own specification that's functionally equivalent but described differently.

Of course, I'd argue that it's unreasonable and unethical to charge for access to an official standard (in the same way it would be unethical to charge for access to the text of a law), but that's a different discussion.

Oracle's Elvish example undermines their own point

[JDG1980]

Oracle's reference to the Tolkien Elvish language is exceptionally weak. They note that Tolkien's estate claims that Elvish is copyrighted, but they provide no case law to substantiate that claim. Instead, they cite a link to the personal Web site of a lawyer who is also a Tolkien fan. And the page includes the following:

I'm not the first to look at this issue. Back in 1999, some Tolkien scholars wanted to publish a journal entitled Tyalià Tyelelliéva, which was to include original poetry written using Tolkien's languages, but they didn't want to run afoul of copyright law. So they solicited a legal opinion from Robert P. Wade, former General Counsel, National Endowment for the Arts. This opinion is available online. He researched the law and concluded that this use would not violate any Tolkien copyrights, for a few different reasons. I won't have time to go into all of them, but here's one: The copyright law specifically says it does not protect any procedure, process, or system.12 A language is a system.13 In fact, there have been cases saying that computer languages are not copyrightable.14 Computer programs are copyrightable, but not the languages used to write them. Would a court find that a language intended for use by humans -- or elves -- falls into the same category? I can't say for sure, but I think there's a good chance of it.

So not only does their own citation say that it is likely Elvish is not copyrightable, it also says that computer languages are not copyrightable – directly undermining Oracle's own specific case! Did the attorney who included this citation assume no one would check it? I expect this kind of practice from desperate college students, not high-priced corporate lawyers.

Re:My ass

[shutdown -p now]

And who says Tolkein's Elvish is copyrighted anyway? His description of it could be copyrighted, sure, but that's not the same as copyrighting the language itself. Don't tell me some dumbass judge granted an injunction on other people writing elvish grammar books

Tolkien Estate says it's copyrighted. And, yes, they have been going around and suing people publishing linguistic works on Sindarin and such based on that interpretation. It's pretty well known in the community that deals with those things.

That said, I don't think it ever came to a court injunction. Generally speaking, the enthusiasts don't have the time, the money and the inclination to go to court over such things, so they just back off when threatened with a C&D.