I’ve recently become acquainted with a specialized form of contract language—architectural specifications, which are attached to construction contracts and define the requirements for products, materials, and workmanship on which the contract is based and requirements for project administration and performance.
My entrée to this field was Andrew Wilson, whose business, AWC West, prepares custom-tailored architectural specifications for leading design firms. (You might want to check out his new blog.)
Andrew is particularly energetic and innovative. He wants his model specifications to be clear and concise, and he decided that since specifications constitute a form of contract language, he should run his specifications by me. He found our discussions sufficiently helpful that I’m now on his advisory board. That’s the sort of ongoing involvement that law firms and company law departments might find useful—it’s amazing that the first business to take advantage of it is a small but feisty architectural-specifications firm.
Andrew introduced me to the Construction Specifications Institute and its Project Resource Manual, which, according to CSI, is the “authoritative resource for the organization, preparation, use and interpretation of construction documents.” The Project Resource Manual provides some guidance on language, and this bit of advice leapt out at me:
5.8.4 SENTENCE STRUCTURE
Two basic grammatical sentence moods can be used to clearly convey specification requirements:
• Imperative mood
• Indicative mood.5.8.4.1 Imperative Mood: The imperative mood is the recommended method for instructions covering the installation of products and equipment. The verb that clearly defines the action becomes the first word in the sentence. The imperative sentence is concise and readily understandable.
• Spread adhesive with notched trowel.
• Install equipment plumb and level.
• Apply two coats of paint to each exposed surface.5.8.4.2 Indicative Mood: The indicative mood, passive voice requires the use of shall in nearly every statement. This sentence structure can cause unnecessary wordiness and monotony.
• Adhesive shall be spread with notched trowel.
• Equipment shall be installed plumb and level.
• Two coats of paint shall be applied to each exposed surface.
This is the first time I’ve ever encountered anyone recommending that one use the imperative mood in contract language.
Here’s a bit background: A verb may be in one of three moods: the indicative mood, the imperative mood, and the subjunctive mood. (For our purposes, we can ignore the subjunctive.) The indicative mood is the most common and is used to express facts and opinions or to make inquiries; as a general matter, contract prose uses the indicative mood.
The imperative mood is used to instruct or request that someone do or not do something. The imperative is identical in form to the second-person indicative: Feed the dog. Don’t eat the pizza. Stop!
CSI’s recommendation that one use the imperative is evidently motivated by two factors:
- First, that using the active-voice indicative is too wordy, in that you’d have to say, with respect to obligations, Contractor shall.
- Second, that the passive-voice indicative is wordy and monotonous, in that it requires that one use shall constantly.
In the MSCD scheme of things one would use must to state obligations using the passive voice, to the extent that’s ever appropriate. But the approach CSI has adopted presents bigger problems:
Whose Obligations and Prohibitions? Stating obligations and prohibitions in the imperative mood is equivalent to using the second person—you. What happens if in the specifications you want to state obligations imposed on someone other than the contractor? (That happens when a nonparty that’s under the control of one or more parties is required to behave in a certain way.) You can’t have you apply to more than one person, so you’d have to use the indicative for any such additional obligations, and that would make for an odd mix.
What About Language of Discretion? Language of obligation and language of prohibition represent only two of the categories of contract language. What if you want to provide that the contractor may do, or isn’t required to do, certain things? You can’t accomplish that with the imperative. If a set of specifications were to use the imperative for obligations (Notify manufacturer’s representative to schedule a final inspection date) and prohibitions, then switch to the indicative for discretion (Contractor may use the following materials), it would read very oddly. To mask that, specifications tend to express discretion in the passive voice (The following materials may be included) or turn it into language of policy (Authorized materials include the following). That’s not ideal.
The Project Resource Manual is being inventive in relying on the imperative. If the language of specifications were more limited—if, for example, specifications consisted of a simple list of do’s and don’ts—that approach might work fine. But that approach falls short, because specifications constitute contract language, albeit of a slightly narrower sort than that used in construction contracts proper.
So I recommend sticking with the indicative mood, more particularly the MSCD categories of contract language.
Why not do what the legislature does and use imperative for the opener and make a bulleted list, as in the following:
The contractor shall perform the following jobs by July 15, 2009:
– Spread adhesive with notched trowel
– Paint all exposed surfaces
That is the indicative mood.
Maureen: That’s exactly what I had in mind. Ken
In the context of the traditional owner-contractor agreement, there are two parties – the owner, whose documents describe what is to be built, and the contractor, who is solely responsible for constructing the building. The design professional (architect or engineer), who designed the building, acts sometimes as an agent of the owner and sometimes as supposedly independent interpreter of the construction documents.
The general conditions of the contract describe the relationships between the owner, contractor, and design professional, and their responsibilities to each other. Because they are written from the perspective of an outside observer, they are in the indicative mood.
In contrast, the specifications are directions to the contractor. As such, there is no need to use the indicative mood. Imagine a conversation between the owner and the contractor. Would the owner say, “The contractor shall paint the fence”? No. The owner would say, “Paint the fence.”
Because there is only one party, the contractor, who is responsible for construction, the specifications can and should be written as directions to the contractor. It doesn’t matter if the contractor’s own people or subcontractors do the work, all of them, as employees of the contractor, are responsible to the contractor, and the contractor is responsible for what they do. One party to the contract, the owner, is telling the other party, the contractor, what to do. In general, the owner does not care who does what or how it gets done; that is the contractor’s responsibility.
There is good logic to support this approach. Not only does the owner have no direct relationship with a subcontractor or other employee of the contractor, but trade agreements and contractor-subcontractor relationships are beyond the knowledge of the owner or design professional, while the contractor works with them daily.
The argument that you can’t specify obligations of persons other than the contractor is specious; it is because there are only two parties to the contract that you cannot specify obligations of those who are not parties. In terms of the contract, it doesn’t matter who paints the fence, as the contractor is the party responsible for doing it. Looking in the other direction, the contract does not specify where the owner’s money comes from, only that the owner must pay the contractor.
In those cases where the contractor has discretion, there is no conflict in allowing options. The perception that it “would read very oddly” is hardly a convincing argument. Is it so difficult to understand “Paint the fence, using one of the following materials”? The average contract written in legalese reads much more oddly.
Yes, the specifications do constitute contract language, but, properly written, they are refreshing in the lack of ambiguity and complexity that seems to be required of documents written by attorneys. Writing in “abbreviated form”, i.e., short, concise imperative statements, is less likely to be misunderstood, and less susceptible to creative interpretation.
The Project Resource Manual, which incorporates CSI’s old Manual of Practice, is not “inventive”, rather, it goes back to basic rules of writing, as espoused in Strunk & White’s “Elements of Style” for nearly a century, and, perhaps more importantly, the “keep it simple, stupid” principle. The MOP has been around for decades; it is curious that “[t]his is the first time” the author has encountered the recommendation to use imperative, rather than indicative mood. Why we accept – and expect – “legal” documents to be written in the most complex way possible is a mystery.
Comments from an attorney I know shed some light on this subject. In essence, he said the more words you use, the greater the opportunity he has to interpret them in favor of his client. An architect I know said the same thing in a different way. He said that if he specifies simply that a coating be “waterproof” there is no room for misunderstanding or interpretation. The more he adds to make it “easier to understand” the weaker the specification becomes. For example, if he specifies that the membrane must prevent passage of a certain amount of water at a certain pressure over a certain period of time, he has a weaker specification. (I know, there is no such thing as waterproof, just as there is no such thing as bulletproof; it’s only an illustration to make the point.)
And for goodness sake, Maureen, please don’t use what the legislature does as an example! If anything, our legislators and attorneys should be required to understand and use “Elements of Style” in their work.
Sheldon: In everyday English, the imperative is used sparingly—Do the dishes!. Specifications do seem to contain less variety than actual contracts, but those few that I’ve looked at nevertheless consist of more than a simple list of A’s instructions to B. The result is that anything that isn’t a simple instruction has to be twisted to fit in an imperative world.
Regarding the comments by your attorney friend, comprehension isn’t, unfortunately, entirely a function of how few words you use.
As for Maureen’s suggestion, you’re throwing the baby out with the bathwater. The approach Maureen suggested is also applied to good effect in contracts.
(By the way, readers, Sheldon is a specifier who was previously institute director and president of the North Central region of the Construction Specifications Institute.)
Ken
I’ll be the first to admit I had never heard of the “indicative mood,” but really it seems to me that Sheldon’s point is simply: indicative mood is not as useful in a specification because the obligations are really only the contractor’s.
But assuming this is even true, it would be a quirk of the setting in architectural contracts. Because the moment the specification applies more than one person (e.g., to delineate responsibility between more than one contractor), then it falls flat. As we computer scientists like to say: it doesn’t scale.
But really, as pointed out by Maureen though, a well-written imperative mood clause does not change overall wordiness or level of ambiguity:
[Subject] shall [verb] …
I think the CSI argument is actually less persuasive if, as Ken points out, the examples weren’t written in the passive voice, which increases the overall awkwardness.
As to Sheldon’s final point, regarding use of fewer words, the fact that something has more or fewer words does not really change the amount of misunderstanding.
The type and amount of information you need in a specification is more a function of the specificity needed. And the length of most legal contracts has more to do with figuring out all the things that can happen and resolving them now rather than later. Nothing is worse than a client coming to me with a one page contract with nothing but the business terms and stating, “I can’t get them to do this, now what?
Bob: I don’t have a background in specifications, but I’ve spent quite a few years immersed in language that seeks to regulate conduct. A broader perspective can be useful.
I did read the CSI Project Manual sections you point to. They’re necessarily simplistic, as they cover in a few pages what I cover at much greater length in my book.
I understand your concerns about making specifications comprehensible for people who aren’t used to legal language. There’s a whole industry geared to such writing, for example for purposes of insurance policies and credit-card documentation. It would be interesting to see whether they make use of the imperative mood. On the other hand, such documents tend to use the second person, which makes an indicative-imperative mix more feasible.
Ultimately, the clearest way to examine this issue would be to present an analysis of some actual specifications language. I may do that at some point.
Ken
A specification is not a contract- It is an Instrument of Service, just like a drawing. Because it uses words that look like a contract does not inherently make it so. Why is it not a contract? Because it does not contain all of the elements of a contract- parties, obligations, money, and time as a cohesive whole.
The function of the Contract is to define responsibilities, schedule and cost. If you follow Maureen’s observation then we would add schedule clauses into each sentence? In fact, we already do that but in separate locations- the Contract provides schedule i.e. complete the Work by July 15 2009, while the Project Manual (note that I did not say specification) provides the action i.e. Paint the wall white.
As the Instrument of Service, the Project Manual has a specific purpose-to provide instruction/direction to the Contractor. It includes drawings which show what goes where, and specifications which state what and how. Both are inherently, though not absolutely, imperative.
Examine the other part of the Project Manual – drawings. Are drawings indicative or imperative- they are imperative. Consider a dimension of 12 feet long wall- should the dimension be written as 12′-0″, arguably imperative, or should it be written indicative “The Contractor shall build this wall to be twelve feet long”? If you argue specifications are Contracts then you must accept that Drawings are as well and by extension of the argument posited in this commentary then the drawing s should be developed in an indicative voice.
In specifications as there are many thousands of discrete simple actions that are required to be performed by a single entity, the Contractor, not the relative few of a contract that has complex actions to be performed by both parties.
Brevity has value- more words do not change the simple action.
I recommend the Construction Specifications Institute Construction Document Technologist coursework as a means to understand the full relationship of all components of a project and the relationship of the Contract to the Project Manual.
Keep the contract language where it belongs- in the Contract.
First, I apologize for leaving a comment on this article almost fifteen years after it was first posted. I do so because, having just started consciously asking this question myself, I found the discussion here relevant and engaging.
Having reviewed both the original post and the subsequent replies, I’m back to where I myself started on the subject: Why not simply allow the mix of both moods?
Mr. Adams, the post’s author, presents a single argument against mixing moods: the result would “read very oddly.” But I must agree with one other commenter in discounting it., because its aesthetic concern does not necessarily rise to the level of a functional obstacle.
However, using the indicative throughout as Mr. Adams suggests would be neither odd nor non-functional, so I must advocate for the imperative’s place in contract documents. With CSI, we understand that this mood can be more “concise and readily understandable” than the indicative; and its interpretation in traditional owner-contractor agreements is nearly unambiguous. To that, I would add that even in multi-contract project delivery methods, most good-faith interpreters “naturally” construe imperatives as being addressed to “you who have been tasked by the structure of the contracts with doing the actual work, whomever you are.”
(Granted, I use the word “naturally” advisedly, acknowledging the potential differences between “natural” interpretations and “legally enforceable” ones. Good-faith interpreters can have legitimate misunderstandings, and the damage bad-faith interpreters can do is readily understood.)
That said, Mr. Adams does identify two examples that the imperative mood doesn’t fit. One involves a third party, and the other grants discretion in meeting a contract requirement. Teh indicative mood is a clear choice for these examples in addition to the general and supplemental conditions.
So I think contracts really want both moods: the imperative where a provision can be expressed in a clear and brief instruction, and the indicative where a provision’s complexity requires more space and flexibility for its airing out. And I believe it’s possible for a good writer to keep any oddness produced by the mood mix from rising to unendurable levels.
This has been both helpful and, dare I say it, fun. A thank-you both to Mr. Adams and to every commentator who got here fifteen years before I did.