Here’s a generic bit of granting language from a software license agreement:
Acme hereby grants Widgetco a nonexclusive, perpetual, irrevocable, royalty-free, fully paid-up, worldwide license to the Software (that license, the “License“).
I’m not a licensing guy, so it is with some trepidation that I ask the following question: Are such adjective-heavy formulations really the best alternative?
As with any such string, you run the risk of redundancy. For example, “fully paid-up” would seem unnecessary, given that the consideration for the license would be stated in the contract.
You also run the risk of confusion. Describing a license as “irrevocable” could mean that it can only be termined if the licensee breaches the terms of the license. Or it could mean that the only remedy is damages, even in the event of breach. It would seem a best idea to address this explicitly in a separate provision.
And doesn’t one grant a license to do something, rather than grant a license to a thing?
So how about saying something along the following lines:
Acme hereby grants Widgetco a nonexclusive license to use the Software in perpetuity without restrictions as to territory (that license, the “License“).
Widgetco is not required to pay Acme royalties in connection with its use of the Software under the License.
Acme may terminate the License only in the following circumstances: …
And might there be any benefit to being explicit as to what “nonexclusive” means? Is that something anyone ever fights over? How about “exclusive”?
30 thoughts on “Granting Language in a Software License Agreement”
I’m glad you’re raising this issue, but I don’t think this is your typical target for cleanup. Most of the license grant adjectives have some heavy interpretative history, so using alternatives may not be the best solution. There are also cases distinguishing between words that delimit the scope of an IP license (creating IP infringement claims for breach) and words that set contractual covenants/conditions (which only create breach of contract remedies at most), so there may be value of having the words in the license grant rather than elsewhere in the document. So I don’t think it’s an easy project to clean up the license grant language.
I do agree with you that the word “exclusive” is inherently ambiguous and usually should not be used without a definition.
Oh, boy. Where to start?
Irrevocable means (notwithstanding 17 USC 203) that the licensee can’t revoke the license, during the term or otherwise, even for breach (although the licensor might be able to terminate the agreement, which would allow revocation of the license). Perpetual means the license lasts as long as the contract. This is a big deal for licensees. You kind of have to understand how licenses work – term vs perpetual, etc. Basically, you don’t want a licensor revoking the license but still expecting the subscription payment. They mean two different things in copyright law.
Fully paid and royalty free go to the payment obligations. If it’s fully paid, you’ve paid up front everything you need to get started (or it’s zero). If it’s royalty free, you’ll have no recurring payment obligations. These are also terms that have distinct meanings from each other.
Yes, one should be granted a license to “do” something. Usually it’s the list from the Copyright Act; i.e., to “use” plus reproduce, prepare derivative works, distribute/reproduce, perform, etc. (Which also demonstrates that one of the flaws in your proposed grant is that it leaves out a number of distinct rights that are highly and separately negotiated)
Why say “without restrictions to Territory” when “worldwide” suffices?
Yes, people fight over exclusive licenses all the time. They’re significantly more valuable than non-exclusive licenses. Someone else can get into the whole host of procedural issues as to who can sue to enforce an exclusive license, but they’re quite different from each other. And one can negotiate for a license that is exclusive as to a particular field, as well.
I’m just scratching the surface on this stuff. But in general, I think you might be tilting at a windmill with this one. Between the Copyright Act and the Patent Act plus the long line of cases, there actually is a technical reason for each of the phrases you questioned, even if most practitioners don’t understand them as well as they should.
I’m pretty sure Eric Goldman reads this blog and will comment far better than I could, but check out his Manual. I put the link in the “website” field above.
The string of adjectives are the filling in of the blanks that the IP laws provide for us. Think of them as you would a metes-and-bounds description of real-estate — They define the boundaries of the grant, and what’s inside the grant is sometimes (colloquially) referred to as the scope of the license. (I know some judicial opinions rely on the word ‘scope’ but I don’t think the term has any legal significance except to the extent a judge creates it within an opinion.)
A license grant is not itself a contractual undertaking — It’s the widget that’s being sold in the contract. (I think some people don’t agree with that. Congress got itself in a bundle over that question when it created an exception in the US bankruptcy code to say that licenses in IP that were granted by the debtor are (more or less) not rejectable as executory obligations in bankruptcy — There were many who said this was not necessary given that licenses aren’t executory obligations in the first place. I think that latter group is right. Canada has not chosen to follow the US’s lead on this exemption, although it’s not clear if that’s a conscious choice on the Canadian’s part.)
I believe that if you start to mix the grant with the executory obligations of the contract, you’re exposing yourself to arguments that the breach of the other language is not an act outside of the scope of the license (thus exposing the bad guy to IP infringement remedies), but rather just a breach of contract (and its much less interesting set of remedies).
Most drafters of license grants don’t take much pain to clearly segregate the ideas of the grant and the executory obligations, and sometimes that leads to disputes over the nature of a breach. (Can I sue for copyright infringement? OR, am I limited to contract remedies?)
Rather than do what you suggest, I’d rather develop a style that takes the ‘license grant’ and segregates it from the contract flow itself and clearly defines it as the widget that is the subject of the transaction. Until now, all we’ve really had is the adjective-laden stream that you (rightly…) object to. But, since it’s not really ‘contract’ interpretation we’re dealing with here (if you buy into the idea that the grant is not a contractual undertaking but rather a description of the object of the transaction), I think many are afraid to step out here.
10803 and Michael: This topic raises two questions: what does one need to say, and what’s the best way to say it. Outsider though I am, I think there might be room for improvement.
For example, is it really clear what “irrevocable” means? In particular, can you count on parties being in agreement as to what it means? I’m not convinced that you can.
And Michael, as a matter of semantics I have a hard time with the idea that a license isn’t a creature of contract.
The Court of Appeals for the Federal Circuit just disagreed with you, Ken!
Jacobsen v. Katzer was decided last week and came down the other way. I’m grossly oversimplifying here, but the holding was that breach of an open-source license grant was a breach of copyright, not contract, and the rightholder was entitled to all of the remedies available under copyright, not mere contractual damages.
I’m not going to look for a direct link, I’m sure you’ll find it easily enough. I’ve got a harcopy on my desk…
“But Michael, as a matter of semantics I have a hard time with the idea that a license isn’t a creature of contract.”
It’s a creature of contract in that you use contract to make the transfer of the grant from party A to party B, but that doesn’t change the nature of the grant itself into a contract. Just because it uses contract-like drafting doesn’t make it a contract. Would you argue that a real estate deed is a contract? Is a negotiable instrument a contract? I think by definition neither is a contract, yet they have words and phrases on them that have contract-like aspects (and either of those fields of practice could well use Adams-like drafting skills even though they aren’t contracts).
I think the fact that we have tended to meld the license grant within the contract without all that much fanfare (and often with very clumsy drafting), rather than doing this like real estate practitioners and banking lawyers have always done — make it a separate document — has been a disservice to the users of license documents. Maybe we should start writing separate deeds of IP licensure in order to put the license grants back where they belong in our legal framework.
Michael Fleming writes: “A license grant is not itself a contractual undertaking ….”
Some would take the view that a license grant is simply a covenant not to sue, which (if made in exchange for consideration) is precisely a contractual undertaking.
We have to jump to similar hoops in other jurisdictions (in my case Denmark) to override a number of the less practical defaults that IPR laws (e.g. copyright) provide. I do agree that it’s ugly and I regularly see drafters try to get around this for instance by breaking such sentences into bulleted lists or similar typographical constructs.
I also think that most of the terms in the string of advjectives are pretty well defined (if not in the statute then in well-fleshed out case law). Some of the terminology that is used in a metes-and-bounds real estate description seems esoteric, but I’d daresay that many of them are terms of art that are not subject to reasonable arguments of ambiguity.
Your example of irrevocable is a good one. It’s one that is used in defining the scope of a license — the giving over of one of the sticks in the bundle if you want to resort to the popular property analogy. (I’m not arguing that I like the bundle of sticks analogy in IP — I don’t in fact — but for sake of argument today I’ll let it slide.)
If you tried to place ‘irrevocable’ in the context of a contract, one could make arguments under contract law about why irrevocable is ambiguous, or is nonetheless not what it says because so-and-so did such-and-such, etc. In the context of a license grant, it means just that — irrevocable, no matter what. Licensees in certain circumstances might insist on that, even while at the same time they’re acknowledging that certain contract rememdies might flow from a breach of the contract, one of them won’t be the revocation of the license grant.
I say that inside of the license-grant box, when reviewed as a grant and not as a contract term, it is unambiguous and necessary (if its part oft the deal). Outside of the grant box and as part of the contract, it’s arguably neither, or could be stated in different terms.
This is a pretty complex topic, and I dare not wade too deeply with such a short period of time. But, first, “use” is not an exclusive right of the copyright owner. See, 17 USC 106. More specifically, the right to “use” a copy for the purpose of a computer program is already protected by law. 17 USC 117.
Second, I’m not sure how much help elaborating on non-exclusive is. Non-exclusive simply classifies the license into the type that is basically a promise not to sue. An exclusive license–at least in copyright–is closer to an actual ownership of some interest in the actual property. They don’t really say much beyond who might also have rights to the same grant.* In any event, you actually have to look at the actual rights granted to fully appreciate the scope of your right to “use.”
In both cases, elaboration is definitely important when you divide up the rights such as by territory and by field of use. Explicitly stating how you plan to divide up those rights is important and usually very contentious.
But the designation as “exclusive” or “non-exclusive” is usually debated BEFORE you get to the drafting stage.
Third, I have so far assumed software was protected by copyright, but software is frequently hard to classify since it’s likely protected by copyright and patents. In that case, there is some questions about how to structure your granting clause.
Finally, for reasons highlighted by a recent Fed. Cir. case, many take pains to make sure that certain elements of the “contract” make it into the granting clause to avoid being sued for infringement, or, on the flip side, to enable suits for infringement, when the contract is breached.
To DC: I acknowledge some say that licenses are merely a covenant not to sue — I just happen to (respectfully) disagree with them!
For those who haven’t already read the Jacobsen decision referenced above, it can be found at http://www.cafc.uscourts.gov/opinions/08-1001.pdf
My practice spans numerous international jurisdictions and this adjective string is a regular cast member.
As to “irrevocable”, as I see it, it = I won’t terminate at will or at any time without cause.
As to “exclusive”/”nonexclusive”: I’ve never seen it misunderstood.
At minimum, “exclusive” is inherently ambiguous because it isn’t clear whether it’s exclusive as to the licensor or except for the licensor. Eric.
When counseling licensor clients, I ask what they want to do if the licensee breaches the restrictions (say by decompiling the code or distributing it without a license agreement possibly rendering that transfer a sale). If they say, I want to terminate the license, then I don’t accept irrevocable or perpetual language, although I have been forced to accept the words followed by the paranthetical “(except as provided in the termination provision)” to express the intent that it’s perpetual until its not. I think some of these magic words evolved from pure IP licenses (e.g., patent license that was fully paid up, makes sense to say perpetual and irrevocalbe, even though in essence if it’s only one patent, the license only lasts as long as the patent is in force) and have been carried over to the “products” licensing, like software.
Fully paid up has some applicability when the agreement has a term, but the business deal is that no further consideration will be paid upon renewal of the term, as the license was fully paid up. Begs the question as to whether it was only fully paid for the term, hence why effect of termination becomes a critical provision in the license.
It all depends on the context of the business deal, but I’d say that this language is adequate for the vast majority of licenses in the US. These are business transactions where the licensee pays a one-time fee to buy a license to the software, and the software has a market-limited lifespan because the licensee is not entitled to free upgrades to future versions. The only big option that — as a licensor — I’d want to address is whether the license is revocable for breach of its terms, or I am satisfied with my right to sue for infringement.
Now, this might not be language that we lawyers see and argue over most frequently, because our purchasing departments (sensibly) don’t usually bring us standardized, non-negotiable, commercial-off-the-shelf license agreements for review. So, much of the discussion so far has really been about the exceptions: licenses with terms (a growing exception), territorial restrictions, contingent future payments, royalties, exclusivity, etc. That’s not to say they are not important, but most actual business gets done under something like the language that Ken cites.
I don’t really see any need to change it when our purchasing and IT departments have all been trained to look for those words and know what they mean.
Having now read the Jacobsen case, I am thinking that the meaning of the case for drafters is that the phrase “provided that” is an invitation to dispute. In the case, the alleged infringer claimed that “provided that” means (essentially) “but you also promise that,” while the copyright owner claimed that it meant “but only on the condition that.”
The whole problem that the court dealt with (which the district court got wrong, according to the appellate court) would have been avoided by using a better phrase than “provided that.” The appellate court said that the meaning of “provided that” as a condition was perfectly clear under California law, but (a) what licensee wants to have to research state law opinions for every open source license it has and (b) who wants to spend the time and money fighting over the meaning of “provided that”? This case has been through the district court for a preliminary injunction hearing and an appeal, and is now going back to the district court. I can’t see how the defendant could have spent less than $100k by now.
If I had been the appellate court, I would have been less enthusiastic about finding that California law clearly makes “provided that” a condition. Here’s why:
First, a contract is the creature of the parties. If we want to define “provided that” to mean “the sky is green,” we can. So what some other set of parties defined it to mean seems irrelevant to me.
Second, the licensor is in the position of the drafter, so ought to have responsibility to eliminating ambiguity, especially where the contract is one of adhesion.
I might still come out on the side the the appellate court did. It cited a purpose clause in the license that said that the purpose was to state the conditions of the license. I find that to be a persuasive indiction of the parties’ intent that helps define the phrase “provided that.” I’d be curious to know whether it also have the kind of lead-in of “the parties agree as follows” that might lend weight to the other side.
Anyway, the lesson for us practitioners is to strike out all those instances of “provided that.” And now we have a case to show our clients and counter-parties for why we do it.
Re MF:”A license grant is not itself a contractual undertaking — It’s the widget that’s being sold in the contract. ”
“Thus, if the terms of the Artistic License allegedly violated are both covenants and conditions, they may serve to limit the scope of the license and are governed by copyright law. If they are merely covenants, by contrast, they are governed by contract law.”
As for exclusive vs. non-exclusive, I suggest those who believe the distinction is straight forward examine “exclusive” licenses and their relation to anti-trust laws in Europe and the US.
10803 wrong again.
“Perpetual means the license lasts as long as the contract.”
As Roger correctly states “If they say, I want to terminate the license, then I don’t accept irrevocable or perpetual language, although I have been forced to accept the words followed by the paranthetical “(except as provided in the termination provision)” to express the intent that it’s perpetual until its not.”
Perpetual means, well, perpetual (yes, this definition is for Mr. Adams). If the contract does not explicitly terminate the license on contract termination then the license will not be terminated.
To Eric’s point, exclusive can be ambiguous. I in fact meant to include a footnote where the * in my previous post is.
Exclusive can mean “you’re the only one,” or it can mean “you’re the last one that can ever have this license, but there are others out there that may have the license already,” or it can be “only you and the licensor.”
In some contexts, I’ve seen “co-exclusive licenses” that are essentially “non-exclusive” licenses with a prohibition on future licenses (aka standstill).
To make matters worse, if you license from a university, there’s inevitably a qualification in an exclusive patent license agreement for government rights from Bayh-Dole.
I’ll also point out that exclusive grants have other implications like in antitrust, taxation, standing to sue, etc. The point being, there’s a lot within the definition and grants are definitely a trap for the unwary.
Finally, there are definite differences between patent, copyright, and trademark. So, you need to check the right rules for determining what applies when.
“Why say “without restrictions to Territory” when “worldwide” suffices?”
Because in a patent license license “worldwide” creates an implication that rights are granted, or will be sought, in countries in which no patent has been issued or granted and in which no one in their right mind would seek patent protection. Particularly where you have a pending PCT or EPO application, you don’t as the licensor want any language implying an obligation to nationalize the application in each country in the world. “without geographic restriction” is the modern and more precise way to address the concept.
This is a fascinating discussion. I’ve been working with software license agreements for over 30 years and I’ve seen and negotiated both long and short versions of these grants with strings of license qualifiers.
Software licenses evolved when software began to be thought of as a contract subject separate from hardware (because initially, hardware and software were inseparable and were provided at the same time by the same vendor). Lawyers did what lawyers always do when something new comes along: work from something familiar and make adaptations where the status quo doesn’t fit. In the case of software, the typical contracts at hand we looked at as templates, were patent and trade secret/know-how licenses (developed by industry), and copyright licenses (coming from publishing and the music and movie industries). We simply spliced these contract forms together. Why did we splice them together? It completely reflected back on the intellectual property rights underlying (or not underlying) the subject of the contract.
Software was not expressly protected by copyright until 1980, and even after that, it was not patentable for several years. And it was still an open question in 1985 whether patent and copyright protection could both apply to any one intellectual property, and at that time, the courts had not yet completely interpreted how copyright law would interact with contract law and whether contract law would be preempted (it is preempted, where contract law seeks to affect the same rights as are provided for and covered by copyright law. 17 USC section 301). But this was when the IBM PC and clones were exploding all over the place and mass-market software was becoming the norm. We were covering all the bases.
I agree with the line of thought expressed in the earlier comments that while these long strings of adjectives may or may not be appropriate to a particular situation and may or may not be duplicative — they go on being used on a daily basis because of a. familiarity and comfort level of purchasing and procurement departments with pattern language, and b. caselaw interpreting particular contract terms, terms of art, and legal jargon. (As a young lawyer I heard several times that there is apparently a Michigan case which discusses the meaning of “witnesseth” at length).
For Eric Goldman: I for one do not believe that “exclusive” is ever ambiguous in a software license — can you provide an example? The parties may be making a mutual mistake in using the term “exclusive license” because they do not mean to exclude the licensor, but that’s a separate issue from its meaning as a legal term.
I was taught in law school as a general principle that an exclusive license always excludes the grantor. In the context of US copyright law, the “exclusive rights” in software being referred to are the section 106 rights of the owner. Surely in this context the owner’s rights exclude the exercise of those rights by all others (subject to the balancing factors of the copyright act such as “fair use.”) The patent law uses the word “exclusive” in a similar manner.
When you look at copyright cases like the Franklin Mint case (the ‘cardinal on a branch’ case) and progeny, there is plenty of caselaw in the copyright context saying that an owner who transfers exclusive rights by license excludes himself from subsequently enjoying those rights. (In the Franklin Mint case, which is widely used in casebooks as a leading case, the artist excludes himself from the right to make derivative works by transferring all rights in his painting of a cardinal on a branch. The court said he was an infringer by making other paintings of cardinals on branches similar to the one he sold the rights to — these were derivative works and though he was the original creator, he had effectively excluded himself from making these kinds of paintings).
This is a meaning of “exclusive” coming out of copyright law, rather than a contract law interpretation.
Patent, trade secret and know how license grants are highly formulaic as well. It is typical in these to grant an exclusive license to the licensee in one subparagraph, and have a “grantback” of certain rights to the licensor in the next subparagraph, and these are simultaneous so that the licensor is never excluded. This drafting technique obviously implies the belief that the licensor would otherwise be excluded.
I haven’t looked into it, but how does a court construe the meaning of “exclusive” in a requirements, exclusive output, or exclusive dealing contract? I’d be interested in knowing if “exclusive” becomes ambiguous out of the intellectual property law context.
I’m not taking away from Ken’s thoughtful dissection of the utility of these strings of words by merely saying “always has been” — I’m just pointing out that it’s more than a ‘plain English’ question to revise how these grants are made.
I often see language whereby someone “grants a license.” A grant is, by definition, not a license nor is a license, by defintion, a grant. Thus the use of the 2 words in the clause makes for lack of clarity. I have a number of articles on my site dealing with precision in contract drafting and how it almost always affects the outcome. Click on “Articles About Being an Entrepreneur.”
On the exclusive front, I use the words “exclusive license” when the licensor will not have any residual rights to the licensed property (and typically couple that with an express statement to that fact, so there is no confusion), and then use the words “sole licensee” when the licensor will continue to practice what is licensed, but is agreeing that there will be only one licensee.
As noted by others in this thread, the string of licensing terms is fairly common in software licenses. I agree with Ken and others that careful drafting reduces ambiguity, with the additional benefit that our clients can better understand what they’ve agreed to.
Jacobsen v. Katzer is an interesting case in many respects. There’s the lengthy discussion of “provided that” as used in the draft (which Ken has separately blogged about). There’s the Brandeis brief on which open source is a good thing. But I’m personally find interesting the discussion of covenants vs. conditions.
The Federal Circuit finds that all of the “provided that” language were conditions of the copyright grant. Uses outside of the grant were held to be infringing uses. However, the court does not provide much guidance on how one drafts a *condition* in a license as opposed to drafting a *covenant*.
Certainly, when drafting a license I could (and sometimes do) say,
“Any uses of the Software in violation of Sections through of this Agreement are violations of the license granted by Licensor to Licensee, and thus are infringing uses under the U.S. Copyright Act.”
Not surprisingly, opposing counsel occasionally objects to such language. ;-)
So, if I write or receive a software license that is one page long or 50 pages long, how do I determine whether a provision agreed to by the parties in writing is a condition of the license grant or a covenant of the contract?
What do you think?