Glenn West Reopens the “Represents and Warrants” Can of Worms!

You may recall my September 2009 post regarding my correspondence with Glenn West on that deathless subject, the phrase representations and warranties. (If you’re new to this subject, you may want to consult this handy 558-word summary of my analysis.) Well, Glenn couldn’t leave well enough alone—today I received from him an email on the subject, which he sent to me and to Tina Stark, professor at Emory Law School. (Longtime readers of this blog may recall that Tina has offered some thoughts on this subject.)

Here’s Glenn’s email:

In cleaning up my office, after writing my latest article on extra-contractual liability, I was reminded of the attached English case [pdf] (which I originally cited in my M&A Lawyer article on Sandbagging [pdf]). To Tina’s point that the use of the word “representation” to flag a factual assertion may carry with it different remedies than the use of the word “warranty,” this case offers support. To Ken’s point that the word is not as important as making sure that the remedies are specifically limited in either event, this case also offers support (and is certainly the approach I take). As indicated in my article, because I believe convention is a long way from allowing contract drafters to avoid the use of both words, I am not sure the debate matters. But, I couldn’t resist pointing this case out since neither of you has referred to it to my knowledge. In either event, I am now convinced that you are both right (or both wrong). As lawyers, words do matter, but we should not rely simply upon ancient distinctions, we should say what we mean clearly. Note the following quotes from the case:

The first issue to which I turn concerns the nature of the representations themselves. Section 4.1 contained a large number of representations and warranties on the part of Western Star. (Corresponding representations and warranties on the part of MAN in relation to the sale of MAN Australia were contained in section 4.3.) In each case the same statement was expressed in terms of both a representation and a warranty. It was accepted that insofar as these statements were expressed as warranties they involved a promise on the part of the maker that they were correct and that if shown to be incorrect for whatever reason the promisee would be entitled to obtain an indemnity under section 12.1(b) which would be measured primarily by reference to the value of the promise. This was referred to for convenience as a ‘contractual’ measure of damages, though it should be noted that the definition of ‘Damages’ in section 1.1 may enable the promisee to recover a wider range of losses for a breach of warranty than he would be entitled to under the general law. The statements were also expressed as representations, however, and it is inherent in the nature of a representation that it is liable to induce the person to whom it is made to take some action on the faith of it. If the representation is false, therefore, any loss suffered by the representee is likely to have been caused by his acting in reliance on it. In principle, therefore, one would expect him to be entitled to recover what was described as a ‘reliance’ measure of loss rather than a contractual measure of loss. … This reflects the essential distinction between a representation and a warranty.” [At printout page 29.]

By drafting the clauses in question as both representations and warranties the parties have attached different characteristics to the statements they contain which, depending on the circumstances, may give rise to different consequences and different measures of loss. Far from suggesting that the indemnity available to the injured party is restricted in every case to the measure of loss appropriate to a breach of warranty, section 12.1(b) makes it clear that it extends to any loss flowing from the inaccuracy of a representation. [At printout page 41.]

Here’s the email I sent Glenn in response:

Glenn: You sure know how to rattle a guy’s cage! One minute, I’m basking in holidays bonhomie; next minute, I’m gnawing my hind leg over represents and warrants ….

I’m sure you’re feeling pleased with yourself for having been able to point me to the Man Nutzfahrzeuge case, as I’ve previously said that I haven’t been able to find any such caselaw. Unsurprisingly, I find this case both frustrating and irrelevant. I know that you and I agree on this issue, but I’m unable to restrain myself from using this as an opportunity to vent.


Let’s deal with the frustrating bit first. All this case does is confirm that at least one English judge endorses the “magic words” interpretation of the significance, for purposes of remedies, of use of the phrase represents and warrants.

The “magic words” approach to contract language sees in a given word an occult meaning that goes beyond the everyday meaning of the word—even the everyday legal meaning. In this case, the “magic words” approach would have a verb that links a party to a factual assertion—as in Acme represents that the Assets are not subject to any Liens—do double duty by also having implications on the remedies available if that factual assertion turns out to have been inaccurate.

The “magic words” approach to contract language is antithetical to clear drafting. If you ask a random group of U.S. lawyers why they use represents and warrants, I know from experience that few or none would offer the “magic words” explanation. And it’s a safe bet that of those willing to trot out a “magic words” explanation, none of them would have explained it to the person at risk of being charged with having that meaning in mind—namely the client.

Even if the “magic words” approach were to meet with greater acceptance, that doesn’t make it a good idea. It reflects a Rube Goldberg approach to articulating meaning, and as such it will always be inferior to saying directly what you mean to say.

I’m not surprised that this case comes from an English court—many English practitioners and judges appear to be in thrall to the “magic words” approach to contract language, more so than their U.S. counterparts. In particular, English judges periodically offer up opinions that are spectacularly lacking in semantic acuity. (I’m thinking in particular of the English caselaw on best endeavours, a topic discussed here and here in this blog.) As we’ve discussed, U.S. courts haven’t endorsed the “magic words” view of represents and warrants.

I’ve done little research of this issue under English law, but what poking around I’ve done indicates that the “magic words” approach doesn’t hold sway. In MSCD I cite a briefing paper by Jones Day’s London office that comes to a conclusion consistent with my own. (I also mentioned it in this April 2007 blog post.) So even in a jurisdiction sadly amenable to the “magic words” approach, that approach is conducive to confusion.


The Man Nutzfahrzeuge case stands for the proposition that in England, I run the risk of having a semantically-challenged judge see remedies implications in the verbs used with factual assertions.

To that I say, So what! A judge will only have a say in the matter if I create confusion. So if in a contract there’s any risk of confusion regarding verbs used with factual assertions, I’ll go out of my way to avoid it by doing one or both of (1) being explicit as to remedies and (2) including the following provision, which I just dreamed up:

The one or more verbs used to indicate that a party is making any given factual assertion are to have no bearing on whether for purposes of determining remedies that assertion is to be considered a representation, a warranty, neither, or both.

In the U.S., the risk of confusion wouldn’t be sufficient to warrant using this provision unless you aren’t explicit as to remedies and are really risk-averse.


You say that you “believe convention is a long way from allowing contract drafters to avoid the use of both words.” Given that using represents and warrants is pointless and confusing and that the “magic words” approach is so lacking in merit, I’m up for the challenge. Of course, the magic-worders have inertia on their side. But I’m confident of the power of a winning argument if it’s articulated coherently and is readily accessible. In that regard, I’ll keep doing what I can.

About the author

Ken Adams is the leading authority on how to say clearly whatever you want to say in a contract. He’s author of A Manual of Style for Contract Drafting, and he offers online and in-person training around the world. He’s also chief content officer of LegalSifter, Inc., a company that combines artificial intelligence and expertise to assist with review of contracts.

8 thoughts on “Glenn West Reopens the “Represents and Warrants” Can of Worms!”

  1. Ken, it is good to vent!

    One thing I hate, is rhetorical arguments. In your earlier post one of your arguments was that you couldn’t find a single case supporting Tina’s approach. Now that a case has been found, you change your argument and say it is only one case, by a “semantically challenged” judge. Will 10 cases by judges with higher degrees in linguistics close off this point, or will you change the argument again?

    Part of the difficulty with arguing with you is that you set the rules of the game.

    According to my rules, good contract drafting needs to take full account of the approach taken by judges, particularly senior ones, when interpreting contract wording. You can dismiss the case as having been decided by a semantically-challenged judge. I have to take account of it. PLC, the leading UK resource for legal research, began its summary of this case as follows: “In October 2005, Moore-Bick LJ gave one of the most valuable judgments ever delivered by the English courts. ”

    Under English law, there is a distinction between a warranty and a representation, and they do have different legal consequences, both as to the remedies available and the method of calculating damages. When you use one of these magic words in a contract, the court’s starting assumption is that you meant to use the word in its magic sense. However, there are also some legal safety valves, where the legislature or court recognises that the draftsman may not be using the word “correctly”.

    For instance, the UK Sale of Goods Act 1979 distinguishes between conditions and warranties. However, section 11(3) of that Act provides:

    “Whether a stipulation in a contract of sale is a condition, the breach of which may give rise to a right to treat the contract as repudiated, or a warranty, the breach of which may give rise to a claim for damages, but not to a right to reject the goods and treat the contract as repudiated, depends in each case on the construction of the contract; and a stipulation may be a condition, though called a warranty in the contract.”

    Similarly, case law provides:

    “Whether a stipulation in a contract of sale is a condition, the breach of which may give rise to a right to treat the contract as repudiated, or a warranty, the breach of which may give rise to a claim for damages, but not to a right to reject the goods and treat the contract as repudiated, depends in each case on the construction of the contract; and a stipulation may be a condition, though called a warranty in the contract.”(LG Schuller AG v Wickman Machine Tool Sales Ltd [1974] AC 235, House of Lords)

    As a practising lawyer, I need to take account of this stuff. I can’t just dismiss it by saying that the judge is semantically challenged.

  2. You’re all going to kill me, but I’m going to complicate the debate over “represents and warrants.” Once having done so, however, I will propose some solutions.

    At bottom, there is a distinction between a “representation,” which is properly construed as nothing more than a statement or an assertion, and a “warranty,” which encompasses the remedy notion that Glenn discusses. But, I recall an “old-school,” “magic words” focused article suggesting that “represents and warrants” is still not enough! The author of that article claimed that the real magic words were “represents, warrants and covenants.” “Covenants,” the author claimed, was to be read as “stipulates,” thus covering matters as to which a party has made a good faith, yet erroroneous, “representation.”

    In other words: “represents” means “says” or “asserts,” “warrants” has the meaning Glenn suggests, and “covenants” means “agrees” or “stipulates” (without regard to truth).

    Having served as in-house counsel to one of Glenn’s firm’s more cantankerous, nit-picking corporate clients (the CEO was a cranky lawyer), I was sensitized to “magic words.”

    While I agree that all of this is probably the proverbial distiction without a difference, there are still some clients who want to pretend they live in 13th Century England.

    Two suggestions (to the extent they may prove necessary):

    Assuming the contract needs and contains a “Definitions” section, why not simply provide that, “Represents means represents, warrants and covenants.”

    Alternatively, if the client is fixed on encompassing all three concepts, but is not necessarily wedded to particular language, simply say what you mean. Instead of “represent, warrant and covenant,” each party could “assert, warrant and stipulate” or, better yet, “state, warrant and agree.”

    Regards and best wishes to all for 2010,

    /s/ Michael D. Scott

  3. A note on convention from a fourth year attorney in the corporate group at a mid-size (not quite Vault 100) firm. I’ve been looking for opportunities over the past year to take the MSCD approach to “representations and warranties”, and have initiated more than a few drafts in which the parties “represent” (not “represent and warrant”) things to one another. I have only had “and warrants” inserted by partners at the firm on a couple of occasions.

    More often (but still not all that often), attorneys representing smaller counterparties have inserted “and warrants”. When the draft goes to Jones Day, for example, they tend to focus on more substantive issues. I’m most likely to be “corrected” by attorneys at smaller shops for whom (I’m guessing) contract or deal work may make up on a middling part of their practice.

    Of course, the set of contracts that fourth year associates at mid-size firms “own”, however briefly, isn’t representative of (and probably isn’t even a particularly good proxy for) the larger body of agreements being drafted today. But my experience within that set is that convention may not be so far away from allowing contract drafters to avoid the use of both “representations and warranties”.

  4. @michael: contracts often say represents, warrants and undertakes. I will leave it to Ken to comment on whether the third member of the triumvirate is necessary (or, to bring in a pseudo-classical analogy, and having just read the latest Robert Harris novel, whether Crassus adds anything to Caesar and Pompey).

    Even if this is considered necessary, I dislike the word “covenant”. Under English law, a covenant is an undertaking given in an agreement made as a deed, although it is often used in ordinary agreements to make the obligation sound more important. It is therefore magic wording without the magic.

  5. If I had my druthers, I would not use the word represents, warrants, or covenants in any agreement ever — simply because they get you dragged into these kinds of issues and because most business people have no idea what they mean.

    Instead I would have a section in which a party makes factual assertions about current or past states of affairs, introduced by something that is very plain English, like “XYZ confirms that . . . .” By itself, that supports a misrepresentation claim or a breach of warranty claim. If I want to limit the remedy, I would do that separately.

    Then I would put “representations” about future states of affairs as positive promises (using “shall” as Ken requires). There is no misrepresentation claim there, but there is a breach of contract claim.

    I won’t ever get my way on this, so I’ll use what Ken says instead. And, if I am successful in getting a contract drafted to Ken’s standards, I might put in a provision saying that it should be interpreted in light of MSCD. That way, when the question of what “represents” means comes up in a case, we have an interpretive guide. I’d use something like the following: “This agreement is to be interpreted consistently with the second edition of Kenneth A. Adams’ A Manual of Stye of Contract Drafting, which the parties used in drafting this agreement.” I would throw that in right after the standard bit that negates the rule of interpretation against the drafter.


  6. Thoughts prompted by the above comments:

    Mark: So we meet again over this issue!

    My argument isn’t expedient. I hold that the goal in drafting is to articulate the intent of the parties without having to call on the courts to breathe meaning into contract provisions. So my recommendations are primarily grounded in semantics.

    Caselaw is relevant in three ways. First, it provides me with plentiful examples of usages that were sufficiently unclear that the parties ended up in a fight. Second, if courts in some jurisdictions have been irrational enough to insist on certain magic words for a given provision to be enforceable, I obviously take that into account. And third, if the bulk of precedent is consistent with my understanding of a given usage, that suggests that I might be on the right track.

    My understanding of the way language works tells me that the “magic words” approach to represents and warrants is fundamentally unhelpful. And on this issue, I’m not going to automatically defer to a court’s conclusions—a judge is no better equipped than, say, a journalist to tell me how language works.

    And the fact that no U.S. caselaw supports the “magic words” approach, and the fact that the Uniform Commercial Code goes out of its way to refute it, tells me that I won’t have to work too hard to convince lawyers in the U.S. that my approach makes sense. If a case or two to the contrary should surface, would that weaken my argument? Hardly. I might even welcome a few such cases—they’d give me something to throw rotten tomatoes at.

    For the same reason, the reasoning in the Man Nutzfahrzeuge case—I should probably have used the other party’s name, Freightliner!—gives me no pause. Similarly, I’d be a twit to pay attention to what the judge in Rhodia v. Huntsman had to say about best endeavours.

    As for what some PLC functionary had to say, I don’t give a fig. It’s blather of the sort that’s routinely used to introduce a boiled-down version of caselaw. I had a good, hollow, laugh when I saw a similar formulation used in a law-firm newsletter to introduce a summary of Rhodia v. Hunstman.

    I’m aware of the doctrinal distinction between representations and warranties, but that doesn’t have to bear any relation to the verb you use to introduce a factual assertion. You’re certain that the “magic words” approach is clear and understood by all, but the Jones Day briefing paper that I cited suggests otherwise.

    And more to the point, one of the great things about contract drafting is that you’re at liberty to articulate meaning in any way you see fit. Anyone who can improve on time-honored but deficient formulations is perfectly free to do so, and in this case would benefit from doing so.

    By the way, I too have a soft spot for Robert Harris’s historical novels.

    Michael: No killing, I promise! But I have to say that the notion of represent, warrant, and covenant accomplishing anything is, pardon my French, utter bollocks. I discussed it in this June 2009 blog post. Say in standard English whatever you’re trying to accomplish with respect to remedies! If you think I’m missing something, please let me know.

    Aquariumdrinker: It’s great to have your input. I wouldn’t have recommended my approach if I didn’t think it would, for the most part, fly under the radar.

    Chris: I agree that the best way to avoid fights would be to use a verb like asserts, as it’s free of any baggage. But I haven’t recommended that approach—like you, I have little hope of convincing lawyers to use a brand-new word. Instead, I’ve opted to suggest that they drop the more problematic component of the traditional couplet.

    And regarding your notion of adding a provision to the effect that the contract is to be interpreted consistent with MSCD, I’m not sure that would work, as MSCD is geared to drafting contracts, not interpreting them. Maybe you’d achieve the desired effect by including with any draft contract a cover note of the sort I proposed in this January 2009 blog post.


  7. Ken,

    “Utter bollocks!” I love it!

    For the record, I did not begin my subscription to your blog until recently, so I was not aware that you had already addressed this issue. I apologize for raising the dead. Also, to be clear, I wrote to report on what others have argued; I did not necessarily mean to suggest that their arguments were persuasive.

    You struck a chord when you said that caselaw provides examples of usage sufficiently unclear that the parties ended up in litigation. Personally, I’ve deemed every contract I’ve negotiated that ended in litigation as evidence of a failure on my part (and on the part of my adversaries).

    The primary objective of the contract drafter should be to memorialize a complete agreement between parties in language that the parties (and a reasonable trial judge and jury) can understand. Consideration as to how an appellate court might construe clear language should be secondary, because if the language is clear and the agreement complete, litigation will never be necessary.

    The language you “just dreamed up,” however, seems no clearer than the “represents and warrants” language it is designed to replace. [Just as the language of my emails, written “on the fly,” likely is less than optimally clear.] Your proposed provision seems exactly the type of verbiage that a party is not going to understand, at least not easily.

    Why not keep it really simple? In a Party Assertions section [Yes, I’ve read some of the debate over capitalization, and I extend my advance apologies to the offended multitudes.] provide, “Party ‘A’ asserts as follows …” Then, deal with the remedy ramifications of the assertions in the appropriate remedies section [Insert as part of the Indemnity boilerplate, for example, “Party ‘A’ warrants that the assertions made in section (whatever) of this Agreement are true as of the Effective Date and shall be true as of the Closing Date … (continuing with any appropriate materiality limitations).”]. This probably makes “covenant” language unnecessary, but if someone feels otherwise, similar language can be inserted where appropriate. [Party Covenants? Pre-Closing Conditions? I really have no settled view as to the proper place, if any, for such language. It depends on the nature of the assertions and whether they should survive closing.]

    Ken, I don’t think you’re “missing anything.” As I wrote earlier, adding “covenants” is probably surplusage at best. On the other hand, the concept of “breaching” [if that’s the right word] a “representation” is conceptually unclear, while the concepts of breach of warranty and breach of covenant are very clear. In the end, however, it is probably a semantic issue of little significance.

    Regards and renewed best wishes for the new year.

    /s/Michael D. Scott

  8. Michael: The provision I dreamed up simply says that you’re barking up the wrong tree if you try to impute remedies significance to verbs that introduce factual assertions. I doubt anyone will actually use it, but I don’t think it’s that complex, and it certainly bears no resemblance to “magic words.”

    And don’t worry about raising issues that I’ve discussed in months past; it would be pretty sad if I assumed that every reader had read every post. On the other hand, I hope you don’t mind if I critique your suggestions and point you to the relevant analyses.

    So, for example, the idea of having factual assertions separate from remedies makes sense, but I would never use the wording you suggest. This isn’t the appropriate forum for lengthy explanations; you might want to check out MSCD 12.433 regarding use of the verb to warrant. And it’s best not to refer to “breach” of a factual assertion: representations are simply accurate or inaccurate. (See MSCD 12.316.) So I’d state the facts, then I’d state simply, without any magic words, what happens if any of those facts are inaccurate.



Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.