In MSCD 3.47, I say the following:
For purposes of business contracts, as opposed to statutes, it’s unlikely that anyone could successfully argue that instead of expressing an obligation, a particular shall is “discretionary” and means may or should.
Well, courtesy of @mrsalzwedel I learned of PacifiCorp v. Sempra Energy Trading Corp., No. CIV-04-0701 (E.D. Cal. 2 July 2004), an opinion in which the court accepted that shall can express exactly that meaning. (PDF here.)
Sempra wanted to compel arbitration; PacifiCorp argued that arbitration shouldn’t be compelled because Sempra had failed to satisfy a condition to arbitration. The court granted Sempra’s motion to compel arbitration.
The contract at issue was a 1997 California independent system operator corporation agreement and tariff, which the court referred to as “the Tariff.” Here’s what the court said:
Section 13.2.1 of the Tariff states that the parties “shall” engage in good faith negotiations “prior to” invoking the ADR Procedures. PacifiCorp urges the court to interpret “shall” as “must;” however, there is an equally reasonable alternative interpretation. In context, the word “shall” may reasonably be understood as “should,” a precatory encouragement to the parties to work matters out.” … Because the word “shall” may reasonably be construed to mean either “should” or “must,” the court finds that its meaning is unclear. See B. Garner, A Dictionary of Modern Legal Usage 939–41 (2d ed. 1995) (noting that “shall” has as many as eight different meanings and that its use violates all the basic principles of good legal drafting; “lawyers are not educable on the subject of shall, so the only solution is complete abstinence.”).
This opinion won’t change how I recommend you use shall. But it does have something to tell us about how caselaw relates to contract drafting and how the marketplace of ideas works.
This opinion serves as a reminder that contracts involve risk. You can’t eliminate risk; instead, you mitigate it.
I don’t lose my nerve every time an opinion comes along that’s inconsistent with my recommendations. The court system is a vast enterprise in which courts are endlessly asked variations on a given question, and sometimes courts screw up. This is the only opinion I’ve encountered in which the court endorses the notion that when used in contracts, shall can mean should. Generally, only in the context of statutes do courts hold that shall can mean should. (The one case that the PacifiCorp court cites involves statutory interpretation.)
You could conclude from this case that it would nevertheless be prudent to use must instead of shall to express obligations. But that approach comes at a steep cost, in that it would result in your using must not only to express obligations but also as one way to express conditions. The risk posed by that sort of confusion greatly exceeds the risk of a given shall being interpreted as meaning should.
Similarly, a couple of courts have held that what I call a “backstop” recital of consideration was sufficient to establish that a contract was supported by consideration, but I wouldn’t dream of recommending that you include a backstop recital of consideration in your contracts. (For more on that, see my article Reconsidering the Recital of Consideration.)
Even if you state something clearly in a contract, someone might want to pick a fight over it. If it ends up in litigation, you might end up with a judge who doesn’t have a clue. If you want to eliminate that sort of risk entirely, stay away from contracts. But if you accept that risk is part of contracts, you can take sensible measures to reduce risk. That can require a balancing of the sort I propose with respect to shall.
Prevailing in the Marketplace of Ideas
This opinion also serves as a reminder that when it comes to contract usages, courts generally don’t engage in a grand inquiry to discover the meaning of, say, shall. Instead, they grab whatever authority comes to hand. In this case, that authority consists of one irrelevant case and Bryan Garner.
That the court cites Garner shows me, yet again, why it’s important that I engage in the marketplace of ideas. If the soap box you stand on is big enough, your ideas prevail. That’s why I make a point of saying, repeatedly, that Garner has overstated the risks posed by shall, that he ignores the benefit of using shall, that shall isn’t going away, and that Garner’s argument for using will instead is incoherent.
My most recent summary of my views is this 2014 article. You’ll have to wait for the fourth edition of A Manual of Style for Contract Drafting for my complete analysis. Thanks to Matthew Salzwedel, that analysis will now feature the PacifiCorp case.