March 22, 2004

Homesteading the phonetosphere

Following up on a sequence of posts about organizations laying claim to portions of what Geoff Pullum has taken to calling the "phonetosphere", I have a small collection of stories about similar cases. Here's one, which shows that it's not always big companies that are the ones laying claim to chunks of sound space. Sometimes they're on the receiving end.

In 1993, the British conglomerate ICI "demerged" its pharmaceuticals interests into a new entity named Zeneca. As I understand it, this was a completely made-up name, supplied by one of the consulting firms specializing in such things. Shortly afterwords, I got a phone call from a lawyer representing Zeneca (or so he said), asking if I would be willing to provide advice and perhaps expert witness testimony. According to my caller, the issue was a threat from the Seneca Nation of Indians to sue for intellectual property infringement over the company's new name. I declined to serve, but offered for free my opinion that /s/ and /z/ are generally regarded as distinct phonemes in English, and therefore capable of distinguishing one word from another. In fact, I continued, this capability is not merely theoretical but is realized in practice, in minimal pairs such as sue and zoo, sip and zip, or peace and peas. Getting carried away with the story, I noted that these different words denote entirely unrelated concepts.

To my surprise, the lawyer seemed to find this informative. He asked me to spell "phoneme", and to repeat the minimal pairs slowly, so that he could write them down. After a bit more Q & A, he asked again if I'd like to be a consultant. I again declined, mainly on the grounds of hassle avoidance. I once agreed to testify in an intellectual property dispute involving a pronouncing dictionary, and the lawyers for a certain large electronics firm subpoenaed me to provide them with a copy of all notes, papers, correspondence and other records in my possession dealing with word pronunciation, speech technology and related topics. I learned that it is possible to "quash" a subpoena when it's obviously just a form of harrassment, as this one was; but you have to pay a lawyer to ask the court in the right way, and the courts put the boundary between legitimate discovery and totally ridiculous harrassment in a different place than I had imagined. I have friends who do quite a bit of expert witnessing, including one who does it for a living, so I'm sure it's possible to adapt to that universe, but as far as I'm concerned, life is too short to deal with hostile lawyers unless some worthwhile principle is at stake.

Anyhow, Zeneca didn't have to change its name, and survived to merge in 1999 with Astra into AstraZeneca, which seems still to be doing business. I don't know if they beat back the challenge from the Seneca, or settled with them, or what. "Zeneca" is a little more similar to "Seneca" than "Star Bock" is to "Starbucks", but then pharmaceuticals are perhaps more different from Indian nations than beer is from coffee, I don't know.

You'd think that an intellectual property lawyer working for Big Pharma on a naming case would be familiar with a bit of basic practical phonology, but apparently not. This may be because phonology itself is less relevant than psycholinguistics, according to my limited understanding of the law of trademarks. The U.S. Lanham act, for example, defines "infringement" as the "use in commerce [of] any reproduction, counterfeit, copy, or colorable imitation of a registered mark in connection with the sale, offering for sale, distribution, or advertising of any goods or services on or in connection with which such use is likely to cause confusion, or to cause mistake, or to deceive."

I don't think that U.S. federal intellectual property law would apply to the Zeneca-Seneca issue, because it requires names to be registered to be protected. If the Seneca Nation have registered their name with the USPTO, searching for "Seneca" in the "Full Mark" field of TESS did not turn it up among the 46 registrations, though some of these are owned by relevant Indians, such as a wholesale cigarette company formed in 1999 by the Sac and Fox Nation. There are also quite a few gambling-industry trademarks, such as "Seneca Alleghany Casino" registered to the Seneca Nation of Indians in 2003. The rest of the registrations show that non-Indian companies have been using the "Seneca" name for nearly a hundred years. I suppose that the "goods and services" aspect of the Lanham law makes it inappropriate for a group like the Seneca to register, and therefore offers them no protection independent of particular commercial activities that they might carry out under their name. However, state law in the U.S. depends on a common-law tradition that doesn't require registration, and this may have been the premise for the challenge allegedly raised to Zeneca.

I wonder if any psycholinguists have gotten involved in quantifying the degree to which particular names are "likely to cause confusion, or to cause mistake". There seem to be a few examples of psychologists doing brand-confusion experiments in a general way, but the ones that I've heard about aren't focused on linguistic aspects of the problem. And legal practice still seems mainly to depend on dueling assertions by lawyers: there don't seem to have been any psycholinguistics experiments in the Star Bock or Lexeme cases. Maybe it's time for a bit of pro bono psycholinguistics in this general area, to protect the phonetosphere from inappropriate encroachment? Of course, this would do no good without parallel legal efforts.

Surely the sound space of human languages is the most crucial single part of the public domain. The right to name new enterprises without excessive constraint is essential to individual freedom, and is just as important to civil society as the right to own such names. If trademark lawyers trying to expand their clients' claims are the only voices in the debate, we risk having the whole sound space of English taken over by the penumbra of owned names, as Geoff Pullum has warned.

[Update: James Gleick has an article on this topic entitled "Get out of my namespace" in yesterday's NYT magazine.(via Rosanne at the X-bar)]

Posted by Mark Liberman at March 22, 2004 09:27 AM