Here’s the “after” version of a provision I used in my class to illustrate verb use in language of policy relating to a contingent future event:

This agreement will terminate if the Market Price falls below $1.00.

One of my students suggested that the verb be changed to terminates, that using will terminate leaves it unclear when exactly the contract terminates—right when the Market Price falls below $1.00, or at some point afterwards?

I responded that changing the verb to terminates would have no bearing on that issue. And that in standard English, an if … then structure with will indicates cause and effect, with the cause triggering the effect.

It follows that I wouldn’t use immediately this context. As I say in MSCD 12.23 with respect to analogous examples, “A drafter might be inclined to use instead immediately … , but even without immediately the reasonable reader would conclude that the result in question would occur at once.” Any anyway, immediately doesn’t quite capture cause-and-effect.

And I don’t use automatically either. Besides being unnecessary, it conveys that no intervention by the parties is required, but it has no bearing on timing.

But those phrases, and analogues using “merely,” make me uneasy, particularly when used to compare conditions and obligations. Here are a couple of examples of that, selected at random, the first referring to a “mere condition,” the second referring to “merely covenants”:

“We are clear, also, that this lack of identity, taken in connection with the whole record, at least tended to show that the provision about ‘giving us all your cotton’ was not a mere condition, but was a covenant on the part of the defendant.” Phoenix Cotton Oil Co. v. Churchill, 274 F. 53, 55 (6th Cir. Tenn. 1921).

“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.” Jacobsen v. Katzer, 535 F.3d 1373, 1380 (Fed. Cir. 2008).

Using the word “mere” (or “merely”) in such comparisons suggests that one category of provision trumps the other, but in fact we’re dealing with apples and oranges: Failure of a contract party to satisfy a condition will relieve the other party of any related obligation. Failure to a contract party to comply with an obligation will entitle the other party to claim damages and, in an M&A context, might relieve the other party of any obligation to close the transaction. Because conditions and obligations serve different functions, it doesn’t make sense to describe one category as inherently trumping the other.

That much is suggested by the fact that on Lexis, instances of use of “mere condition” in court opinions are roughly equal in number to the aggregate instances of “mere convenant” and “mere obligation.” If one kind of provision inherently trumps the other, you wouldn’t expect those references to balance each other out. The fact that they do indicates that the relative significance of one kind of provision over the other is entirely a function of context.

And use of the rather fusty “mere” in this context suggests that use of these phrases is primarly a matter of parroting jargon.

So I suggest that you not use “mere” or “merely” when comparing categories of contract language.

A big part of my recent article on the AAA standard arbitration clause (pdf here) was a discussion of categories of contract language. The boilerplate governing how contracts are amended raises a similar issue.

Here’s what a text on contract boilerplate offers as a “basic” no-oral-modification provision (optional language omitted):

Amendments. The parties may not amend this Agreement, except by written agreement of the parties.

That’s language of prohibition. It follows that if the parties amend other than by written agreement, they’re in breach. That doesn’t make much sense.

Here’s my version:

Amendment. No amendment to this agreement will be effective unless it is in writing and signed by both parties.

That’s language of policy, which is used to state the groundrules for how the contract is to operate. I think it states more clearly what’s going on.

As is often the case, you could use a yet another category of contract language to express the same meaning. Here it is expressed as a condition using language of obligation:

Amendment. To amend this agreement, the parties must express the amendment in a writing signed by both parties.

This version exhibits the elements of this kind of condition: must used instead of shall, because you’re not stating a duty, and an introductory phrase that makes it clear that you’re dealing with a condition. But using a condition to express this meaning is way too cumbersome—I’d stick with language of policy.

At this point you might be asking yourself why anyone should care about this. After all, no one would ever find themselves in a dispute over whether the “amendment” section articulates prohibition or policy. I can think of three reasons:

First, it might be that not much is at stake with respect to any one provision, but the cumulative effect of minding your categories of contract language is dramatic.

Second, mishandling the categories of contract language can give rise to disputes. A classic example of this is confusion over whether a given provision states a condition or an obligation (see this January 2007 blog post).

And third, minding your categories of contract language throughout a contract enforces discipline on the drafter. Spotting a categories-of-contract-language glitch, such as a shall that fails the “has a duty” test, can result in your picking apart the provision and restructuring it. That’s what I did with the AAA standard arbitration clause.

SunPower hereby guarantees that, subject to Section 22, it shall supply and deliver each of the Products to the delivery point specified in a given Purchase Order (each, a Delivery Point ) on or prior to the scheduled delivery date therefor specified in such Purchase Order (with respect to such Products and such Delivery Point, a Guaranteed Delivery Date ).

When in the MSCD-approved manner you’re using shall to impose an obligation on the subject of the sentence—in this case, SunPower—nothing is accomplished by having the party under the obligation guarantee that it will perform the obligation. The obligation itself is all that’s necessary.

And in this context, the verb to guarantee doesn’t serve to convey its legal meaning, namely “To assume a suretyship obligation; to agree to answer for a debt or default.” Instead, to guarantee that is a form of rhetorical emphasis that simply reinforces the speaker’s passion, as in I guarantee that Accrington Stanley will defeat Scunthorpe on Saturday! Obviously, that isn’t what the drafter of the SunPower contract intended.

It also follows that, more generally, to guarantee that has no place in a contract, whatever the category of contract language. You wouldn’t even say Parent hereby guarantees that Sub will repay the Sub Loan. Instead, you’d say Parent hereby guarantees repayment of the Sub Loan—you’d use guarantees without a that-clause.

By the way, while we’re on the topic, you might find of interest this October 2006 blog post on guaranty or guarantee.

This issue arose in connection with a trademark license agreement I’ve been redrafting. The original version says that the license is “nonassignable,” and I’ve been contemplating deleting that adjective on the grounds that the question of assignment of the license would be covered by the general nonassignment provision included in the contract boilerplate.

If this all sounds familiar, it’s because it cropped up in this August 2008 post on license-agreement granting language. So I revisited the issue with Michael Fleming, who in the comments to that post was the principal advocate of the position that a license is distinct from the contract granting that license. Yesterday he repeated in an email to me the gist of his argument:

I’m still of a mind that the license arises independently from the contract, much like a purchase agreement for a house is a different thing than the house itself. The license is the widget. Thus, rules that apply to the license should be clearly attached to the license separate from the rules that apply to the agreement, much as you would attach a covenant that goes with the land in a real estate transaction.

I’m not crazy about the house analogy. For one thing, my selling you a house is very different from my letting you use my trademarks—in the former transaction you get the widget, whereas in the latter you just get the right to use the widget.

But more importantly, any discussion of the meaning of contract language has to start with semantics. I suggest that in this case, the semantics indicate that granting language in a license agreement is simply a variant form of language of discretion and as such is simply a creature of the contract, like any other provision.

Consider the follow provisions:

• Acme hereby grants Widgetco a license to use the Marks in …
• Acme hereby grants Widgetco the right to use the Marks in …
• Widgetco may use the Marks in …

They all convey the same meaning. What distinguishes the first from the other two is that it seeks to flag that discretion is being accorded with respect to something that Acme controls. But such control wouldn’t be inconsistent with the other two variants. (In the absence of such control, the first two variants would be preposterous.)

When you can convey the same meaning in different ways yet you accord a special significance to one of those ways, then—Hey presto!—you have magic-wordery, the enemy of rational drafting. In terms of the semantics, granting a license is just one way to accord a party discretion under a contract. Debating whether a license arises independently of the contract, or instead is covenant not to sue and therefore part of the contract, is to miss the point.

I haven’t yet researched how this plays out in the caselaw—as long as I’m clear in my drafting, I needn’t concern myself too much with other people’s messes. But it would be rash of me to expect the rest of the world to think like I do: Although I think it’s only logical to treat license-granting language as just another contract provision, I’ll make it clear, for purposes of the trademark license agreement I’m redrafting, that the general nonassignment provision applies to the license-granting language.

I imagine some people will have strong views on this subject. I’d be happy to hear them.

In license agreements I see the following two alternative constructions:

Acme hereby grants Widgetco a license to …

Acme hereby grants to Widgetco a license to …

The second alternative represents the inferior choice. Consider the following sentences:

I gave John a book.

I gave a book to John.

*I gave to John a book.

The first sentence matches the structure of Acme hereby grants Widgetco a license to … For any grammar junkies among us, it’s a ditransitive clause, with a direct and indirect object. The second sentence, by contrast, is a monotransitive clause—it has just one object—plus a prepositional phrase using to. See The Cambridge Grammar of the English Language 248 (2002).

When you use a monotransitive structure, the prepositional phrase conventionally comes at the end. If you have a simple direct object, it sounds odd to put the prepositional phrase before the direct object, as in the third sentence.

But if the direct object is lengthy, that can preclude putting the prepositional phrase after the direct object. That would be the case with a direct object beginning a license to …  Your only choice would be to put the prepositional phrase before the direct object, as in Acme hereby grants to Widgetco a license to … But not only is the to oddly positioned, it’s also superfluous, as without it you’d have a conventional ditransitive structure—Acme hereby grants Widgetco a license to … That’s why the ditransitive structure is your best alternative.

Why am I bothering with this? Because I want to be able to offer a reasoned basis for selecting among alternative usages, no matter how modest the difference between them.

Note that this post applies to other verbs, including to pay, and applies to other categories of contract language in addition to language of performance.

Consider the following example:

If any Person brings a proceeding to compel the Recipient to disclose any Confidential Information …

In this context, the active voice is wordy. It’s obvious that a proceeding would have to be brought by someone. Because it doesn’t matter who brings a proceeding, we don’t care about the subject in the above sentence. [Updated Oct. 1 11:30 a.m. EDT: I flirted with the idea of saying "a nonparty" rather than "any Person," but Alex's comment persuaded me to abandon that approach.] You could eliminate it by using the passive voice. That would turn the subject into the passive agent, and you could simply omit it:

If a proceeding is brought to compel the Recipient to disclose any Confidential Information …

[Updated Oct. 1 11:30 a.m. EDT: I'm back to thinking this is the better option.]

Depending on the context, a more economical active-voice structure can offer the best solution, and that might be the case here:

If a proceeding seeks to compel the Recipient to disclose any Confidential Information …

Any preference?

[Updated 8:25 p.m. EDT: On Twitter, @gideonalper said "Economical active version would be incorrect. Only one party in proceeding seeks compulsion, not proceeding itself." I think Gideon's right: I hereby wave the white flag. I'm being schooled by commenters this evening.]

A case in point is Managment Strategies v. Hous. Auth. of New Haven, 2009 Conn. Super. LEXIS 1550 (Conn. Super. Ct. June 2, 2009). In that case, the following release language was at issue:

Upon completion of the Completion Work, [FNICA] and [the plaintiff] shall, immediately upon full payment by [the defendant] of all amounts due hereunder, release [the defendant] from any and all liability hereunder, under the Contract, and with respect to any action taken by [the defendant] with respect to the Contract, the [Takeover] Agreement, and the Project.

The plaintiff never issued the release, even though the defendant had satisfied the conditions. Then in the lawsuit, the plaintiff objected to the defendant’s motion for summary judgment, claiming that because the plaintiff hadn’t released the defendant, the defendant wasn’t entitled to summary judgment. Sensibly enough, the court held that the plaintiff couldn’t base it its claim on its failure to issue a release that it had been required to issue under the contract.

The plaintiff would have been precluded from making this lame argument if the release had been automatic. And more generally, making a release automatic should simplify matters. Chasing after Acme to get it to issue a release could be a nuisance—if Acme’s not inclined to do you favors, it might be willing to drag its heels.

How do you make a release automatic? Instead of saying If X, Acme shall release Widgetco, you say If X, Acme will be deemed to have released Widgetco. (The former is language of obligation; the latter is language of policy, using deem; see MSCD 12.65.) The release happens automatically, without requiring Acme to do anything.

Making a release automatic would work only if the conditions to the release can easily be established and if the party being released doesn’t need to show someone (for example, a lender) a piece of paper called a release.

The court in Managment Strategies referred to the notion of a release being automatic, but it also expressed the same concept by using the term “self-executing.” Black’s Law Dictioary defines self-executing as “(Of an instrument) effective immediately without the need of any type of implementing action.” But the term is applied primarily to treaties, and I don’t see that anything would be gained trying to apply it to contracts too. I think it makes more sense to refer to an automatic release.
.

Here’s D.C.’s summary:

Wal-Mart requires its suppliers to agree to a code of conduct. It was sued, in California, by employees of suppliers in various Third World countries, on grounds that • the retail giant had failed to police the suppliers’ compliance with the code’s standards, and • the employees were allegedly third-party beneficiaries of those standards.

Wal-Mart succeeded in having the case dismissed, on grounds that it did not undertake to police compliance, and therefore had no duty to do so; consequently, the foreign employees were not third-party beneficiaries.

Here’s the sentence from the code of conduct that was at issue (emphasis added):

To further assure proper implementation of and compliance with the standards set forth herein, Wal-Mart or a third party designated by Wal-Mart will undertake affirmative measures, such as on-site inspection of production facilities, to implement and monitor said standards.

According to the Ninth Circuit, “The language and structure of the agreement show that Wal-Mart reserved the right to inspect the suppliers, but did not adopt a duty to inspect them.” That makes sense, as it would have been odd for the suppliers to want to impose on Wal-Mart a duty to inspect them, and for Wal-Mart to have assumed such a duty.

I agree with D.C. that the language at issue could have been clearer. In the MSCD scheme of things, the clause “Wal-Mart or a third party designated by Wal-Mart will undertake affirmative measures” sounds like language of policy with respect to a contingent future event—in other words, a ground rule of the contract with respect to something that’s to happen automatically in the future under certain circumstances.

But in this context, that doesn’t work. Instead, the sentence in question would have made more sense as language of discretion. This is where I part ways with D.C. Tracking the court’s analysis, he would have used reserves the right to instead of “will.” But I say that reserves the right to is one of many pointless alternatives to may, which is what I’d have used instead. See MSCD 2.87.

I’m preparing a partnership agreement and have been given precedent to work with. Using MSCD, I have spent some time trying to rework the precedent’s archaic language to make it more readable. It says, several times, “Each of the Partners severally represents, warrants, covenants and agrees with each other Partner that such Partner (a) has the capacity to enter into the agreement, (b) shall ensure that its status shall not be modified,” etc.

I’m not sure how to tackle the wording in bold italics. I’ve looked at four other partnership agreements and the same quadruplet wording appears in various places in each agreement. Is it used for belt and suspenders purposes, or just because no one has any idea what to say? I’ve looked at your book, and I see that you have tackled covenants and agrees to, but it is a bit daunting to read two couplets side by side. And the drafter obviously thought this language was important. What do you recommend I do with it?

Here’s my reply:

The initial question is, what one or more categories of contract language follow “represents, warrants, covenants and agrees”? The example you cite consists of (1) language of representation and (2) language of obligation imposing a duty on the subject of the sentence. In an ideal world, for the former you’d say “Acme represents” and for the latter you’d say “Acme shall.” Your language is clumsy because it jams the two concepts together and also adds redundancy.

Regarding the redundancy, you’ve noted MSCD‘s discussion of covenants and agrees to. MSCD also discusses represents and warrants, but click here for a recent blog post that looks at it from a Canadian perspective.

And look at what my book says about joint and several. Several is a liability concept; it doesn’t make sense to have anyone “represent severally.”

I hope this helps!