Graeme Dinwoodie on Rescuecom v. Google

By Eric Goldman

[Eric's note: As I mentioned, I'm getting a lot of private emails about Rescuecom v. Google, including the email from Margreth Barrett that I blogged last week. Today, I got the following email from Graeme Dinwoodie, a law professor currently at Chicago-Kent Law and soon to be at Oxford. Like Margreth, Graeme has written on the trademark use in commerce doctrine and search engine liability. I've blogged on a few of Graeme's papers before as well; see his SSRN page. Graeme has graciously permitted me to share his email on the blog:]

I think that your bottom-line take on Rescuecom is largely right, though it will not surprise you that I do not regard the decision as “disappointing.” I think the Second Circuit largely accepted the arguments that Mark Janis and I have made in our articles, and so I am pleased with the outcome. I agree that there are some oddities in the reasoning, though these are in large part a product of (1) having to distinguish the very badly reasoned decision in 1-800 Contacts, and (2) the fact that the trademark use requirement does not map well to the concerns that should drive the scope of trademark protection. In fact, looking at where the court appears to want to go, I have to think that – if the 1-800-Contacts decision was not out there — they would have concluded that there was no such thing as a trademark use requirement.

I think that the court largely accepts the critique that some of us have offered of the trademark use requirement: the court recognizes that there is no inevitable symmetry between use sufficient to create rights and use that causes likelihood of confusion; the court recognizes that uses by defendants that fall outside the strict scope of the section 45 definition could be “pernicious”; and the court thinks it important to adopt a rule that allows courts to hold defendants liable when confusion is created (footnote 4 clearly reflects the concern that courts should be able to police this activity).

Of course, because of 1-800-Contacts they could not simply say that Section 45’s definition did not apply to defendants’ uses, which would have been much cleaner. If one accepts, as they do, that the definitions only apply “unless the contrary is plainly apparent from the context,” one could simply have said that almost none of the section 45 definition is intended to constrain what type of activities by a defendant might be actionable. Instead, the appendix proffers a reading of the two sentences in the definition that is truly weird. (Indeed, under one reading, you might even say that they were endorsing — in the last couple of pages — a trademark use requirement linked to the affixation language in what the court called the second sentence of the definition.) But their having to do all this is simply a function of the fact that the Second Circuit had previously applied the second sentence to sections of the act of defining infringement in 1-800 contacts: see fn 12.

Likewise, some of the factual distinctions seem a bit odd (even though they were to some extent predictable given the dicta in 1-800-Contacts). The URL/mark distinction is inconsistent with typical infringement analysis that permits use of a term similar to the mark to be infringing and is functionally ridiculous given the prevalence of Mark.com URLs. I suppose the distinction between an ad triggered by a “product category” in 1-800 Contacts and one triggered by a mark as in Rescuecom (also a predictable distinction given the dicta in the earlier case) might reflect some vague notion of directness or frequency of harm, but the alleged harm that is experienced when the ad appears is surely pretty similar. (It reminds me of the link-counting analysis to determine commercial use?). See Dinwoodie and Janis, Confusion Over Use at 1635.

What I take from the decision is that they really would like to go back and rethink 1-800 Contacts. I agree though they have effectively undermined 1-800-Contacts. The bad news is that the messy way in which they have done it — if they take seriously the details of their analysis rather than the message that they are sending — might generate some silly litigation in the meantime. The good news is that, if courts focus on the message, we might now get greater judicial consideration of the central issues of what types of confusion — if any — are created by this type of advertising, and what types of confusion should be actionable (and I hope that those are separate inquiries). That this is the court’s preferred focus is evident from their alternative explanation for the outcome in I-800-Contacts (see p. 17), their analysis of the product placement analogy, and some of the factual distinctions that they draw between 1-800 Contacts and Rescuecom (at least at the 12(b)(6) stage). It does not seem inevitable to me that search engines or advertisers will lose on the confusion analysis in the cases to come (including this one). And at least I hope we will now have some judicial exploration of whether there is any confusion and whether that should be actionable. To be sure, there are some litigation and compliance costs associated with this, but even those may dissipate over time through accretion of case law. And I hope and expect that defendants will begin to explore the types of uses of marks in ads that might be immunized through defenses such as nominative fair use (we’ve had a couple of lower court cases beginning to move in that direction). The combination of all of this analysis will, I hope, be more helpful to search engines in formulating appropriate policies and responses to trademark owner requests — something they have already given a tremendous amount of thought to – than debate about “trademark use”

In short, although there are some problems with the opinion, the outcome should at least start us talking about types of issues that I believe we should be talking about in this area. To put it (I hope not too) tendentiously, the Second Circuit has decided to opt for analysis of “confusion” over “use”, and has decided that we should litigate the scope of trademark law, with due regard for “context.” See Graeme B. Dinwoodie and Mark D. Janis, Confusion Over Use: Contextualism in Trademark Law, 92 Iowa L. Rev. 1597 (2007). Alternatively stated, I think the law that will be developed in the next few years in the wake of this decision will be heavily driven by factual particular rather than broad legal rules (though rules of sorts may accrue over time).