How to Do Things with Word Marks: A Speech-Act Theory of Distinctiveness


For a word or phrase to be legally recognized and protected as a trademark, it must serve as an indicator of source — that is, it must be distinctive. Factfinders tasked with determining whether a mark possesses inherent or acquired distinctiveness essentially must determine whether consumers perceive the mark as a mark, rather than as simply a literal description of the goods or services. This article explores the formal and informal tests for inherent distinctiveness in close cases and concludes that they fail to provide objective guidelines for gauging whether consumers perceive a mark as indicating the source of the goods or services. Instead, the tests invite factfinders to focus on the perceptions of mark selectors over those of consumers; as a result, factfinders reward producer ingenuity by overprotecting descriptive marks, needlessly handicapping competitor speech.

To highlight deficiencies in the classification of descriptive marks, the article relies on speech-act theory, applying to trademark law concepts first articulated by J. L. Austin in HOW TO DO THINGS WITH WORDS. Trademark use is constative in one of two ways. Use is source-constative when consumers perceive the mark as primarily indicating the goods’ source. Use is goods-constative when consumers perceive the mark as primarily describing qualities or characteristics of the product or service itself. When use is source-constative, law renders it simultaneously performative: the trademark use generates enforceable legal rights and places competitors on notice of those rights. The constative speech is directed at consumers; the performative act, at competitors. A producer may intend its use of a mark to be source-constative, but if the mark is descriptive and lacks secondary meaning, consumers will perceive it as merely goods-constative — the use will suffer from an audience uptake error. Just as, per Austin, the performative “I bet you sixpence it will rain tomorrow” is felicitous only if met with assent, so is the performative function of trademarks contingent upon consumer perception.

Existing tests for distinctiveness often lead factfinders to reach the wrong outcome for borderline marks because the tests don’t adequately interrogate whether a proposed mark fulfills its source-constative function. Fanciful and arbitrary marks map neatly onto the speech-act schema: KODAK for cameras and PENGUIN for books are innately source-constative precisely because they are never goods-constative. The adjective “KODAK” has no definition that modifies the noun “camera” other than its source-indicating definition. Conversely, descriptive marks like TASTY TREATS for cookies or AGE ERASER for face cream are, without secondary meaning, goods-constative — consumers have no reason to perceive the component terms as describing anything other than characteristics of the products to which they are affixed.

The US Patent & Trademark Office, Trademark Trial and Appeals Board, and federal courts apply a variety of tests to determine whether a word mark is inherently distinctive or merely descriptive. Such tests include the dictionary, puffery, competitors’ need, competitors’ use, and imagination tests. In addition, factfinders often regard rhetorical devices such as incongruity, musicality, or double entendre as further proxies for distinctiveness. Those tests and proxies lead them to incorrectly deem distinctive such marks as the alliterative CLASSIC COLA for a brand of cola or the ostensibly incongruous SNO-RAKE for a particular snow removal tool. The raison d'être of trademark law is subverted when goods-constative terms are granted protection in registration or infringement proceedings. That protection authorizes the owner of a descriptive, non-distinctive mark to wield the mark offensively against new entrants before the mark has come to serve as a trademark to consumers, the equivalent of a property owner erecting a “no trespassing” sign on a piece of land that he has not yet purchased (and may never come to own). Taking a cue from the statutory fair use defense and asking whether a competitor could use a trademark’s component terms literally and in good faith to describe its own product, based on evidence of non-trademark usage of the terms by consumers as well as competitors, provides a more direct and objective means to evaluate whether the mark is inherently able to signal a product’s source and therefore function performatively.

Publication Date


Journal Title

Alabama Law Review

Document Type


This document is currently not available here.