Abercrombie & Fitch v. Hunting World: Judge Friendly's Spectrum and the Architecture of Distinctiveness

Abercrombie & Fitch Co. v. Hunting World, Inc., No. 21, Docket 74-2540, 537 F.2d 4 (2d Cir. 1976), decided January 16, 1976 (with a limited opinion on rehearing dated February 26, 1976), is the case in which Judge Henry Friendly, writing for a unanimous panel, set out the five-category “spectrum of distinctiveness” that organizes American trademark law to this day. The fight was prosaic — two purveyors of outdoor wear quarreling over the word Safari — but the framework Friendly articulated to resolve it has been cited thousands of times and absorbed into the everyday vocabulary of every trademark practitioner.

At a glance

• Case: Abercrombie & Fitch Co. v. Hunting World, Inc., No. 21, Docket 74-2540, 537 F.2d 4 (2d Cir. 1976)
• Court: U.S. Court of Appeals for the Second Circuit, opinion by Judge Henry J. Friendly
• Decided: January 16, 1976; limited opinion on rehearing February 26, 1976
• Holding: Word marks fall along a spectrum — generic, descriptive, suggestive, arbitrary, fanciful — and the category determines whether and on what conditions the term is protectable; “Safari” occupied different points on that spectrum depending on the goods to which it was applied
• Outcome: Several of Abercrombie’s “Safari” registrations were vulnerable, and Hunting World was not enjoined from descriptive and generic uses of the term; the case was remanded for further proceedings consistent with the panel’s category-by-category analysis

The dispute and the difficulty

Abercrombie & Fitch had used “Safari” on a range of apparel and sporting goods for decades and held several federal registrations incorporating the word. Hunting World, a competitor, began selling outdoor goods and using “Safari” and “Minisafari” in describing them. Abercrombie sued for infringement; Hunting World counterclaimed to cancel Abercrombie’s registrations, contending the word was generic or merely descriptive of the goods.

The difficulty Friendly confronted was that “distinctiveness” is not a single switch. A term can be a perfectly valid mark in one setting and an unprotectable common noun in another. “Safari” applied to, say, a particular cut of hat might mean something quite different from “safari” used to describe a type of expedition or a category of bush jacket. To resolve the case rationally, the court needed a vocabulary precise enough to handle the same word doing different work across different goods.

The five categories

Friendly’s enduring contribution was to lay out the categories explicitly and to explain the legal consequence attached to each.

Generic terms are the common name of the genus of goods — “aspirin,” “escalator,” “thermos” after they lost protection. A generic term can never function as a trademark, no matter how much money is spent promoting it, because to grant exclusivity in the common name would hand one seller a monopoly over the language competitors need to describe their own goods. Genericness is fatal and, as later cases confirmed, an existing registration offers no shelter once a term is found generic.

Descriptive terms directly convey an ingredient, quality, characteristic, function, or feature of the goods. They are not inherently distinctive and may be registered or protected only on proof of secondary meaning — that consumers have come to associate the term with a single source. Until that showing is made, a descriptive term remains free for competitors to use descriptively.

Suggestive terms require imagination, thought, or perception to reach a conclusion about the nature of the goods. Because the consumer must take a mental step, suggestive marks are deemed inherently distinctive and are protectable without proof of secondary meaning. The line between suggestive and descriptive is the spectrum’s most litigated seam, and Friendly candidly acknowledged the difficulty of drawing it.

Arbitrary and fanciful marks sit at the strong end. Arbitrary marks are common words applied to goods with which they have no logical connection (the textbook example arrived later: “Apple” for computers). Fanciful marks are coined terms invented to serve as marks (“Kodak,” “Exxon”). Both are inherently distinctive and receive the broadest protection.

Crucially, Friendly stressed that classification is always relative to the goods or services at issue. The same word may be generic as to one product, descriptive as to another, and arbitrary as to a third. That relational quality is precisely why “Safari” could not be assigned a single fixed status.

Applying the spectrum to “Safari”

The court worked through Abercrombie’s uses good by good. Where “safari” had become the generic or common descriptive name for a style of garment or for the activity itself, Abercrombie could claim no exclusivity, and Hunting World was free to use the word to describe its goods. Where Abercrombie’s use was arbitrary or had acquired distinctiveness as a source identifier for particular products, protection could survive. The result was not a clean win for either side but a parsing exercise: some claims fell away as generic or descriptive, others potentially endured, and the registrations were treated according to where each fell on the spectrum.

That granular outcome is itself part of the opinion’s teaching. Distinctiveness is adjudicated mark-by-mark and good-by-good, not brand-by-brand. A company does not “own” a word in the abstract; it owns, at most, the source significance the word has acquired for specified goods.

Open questions

Friendly himself flagged the spectrum’s soft spots, and they remain live. The suggestive/descriptive boundary continues to defy mechanical resolution; courts deploy heuristics — the “imagination” test, the “competitors’ need to use” test, evidence of third-party use — without any of them being dispositive. The opinion also left for the future how acquired distinctiveness interacts with categories beyond the descriptive, a problem the Supreme Court would later confront for trade dress in Two Pesos and Wal-Mart v. Samara. And the treatment of a single word that migrates across categories depending on goods still generates difficulty whenever a registrant asserts broad rights across a sprawling product line.

Implications

• Pick your category before you pick your fight. A mark’s place on the spectrum dictates the entire litigation posture — whether secondary meaning must be proved, how broad the scope of protection is, and whether the registration is vulnerable to cancellation.
• Strength is a spectrum, not a status. Brand owners chasing “descriptive but evocative” names are choosing, by definition, marks that demand secondary-meaning evidence and yield narrow protection. Coined or arbitrary marks cost more to launch but are cheaper to enforce.
• Genericness has no cure. Once a term is the common name for the goods, registration and advertising spend cannot resurrect it; brand owners must police against their own marks becoming the category name.
• Classification is good-specific. Rights are measured against particular goods and services. Sweeping claims to a word “in general” invite the kind of category-by-category dismantling Hunting World achieved.

Frequently asked questions

Why is this case so heavily cited if Abercrombie partly lost? Because the value of the opinion lies in its framework, not its result. Friendly’s five-category taxonomy gave courts a shared language for distinctiveness, and that vocabulary — generic, descriptive, suggestive, arbitrary, fanciful — is now standard in trademark doctrine and registration practice.

What is the difference between suggestive and descriptive? A descriptive term tells the consumer something directly about the goods and needs secondary meaning to be protected; a suggestive term requires a mental leap to connect it to the goods and is protectable immediately. The line is notoriously fuzzy, which is why so much litigation turns on it.

Did Abercrombie lose all rights in “Safari”? No. The court’s analysis was good-by-good. Where the term was generic or merely descriptive of the goods, Abercrombie could not exclude Hunting World; where its use was distinctive of source for particular products, protection could remain. The case was remanded for treatment consistent with that parsing.

Authorities and sources

• Abercrombie & Fitch Co. v. Hunting World, Inc., No. 21, Docket 74-2540, 537 F.2d 4 (2d Cir. Jan. 16, 1976), full opinion at Justia and OpenJurist.
• Case background, docket, decision date, and the five categories summarized at Wikipedia and WIPO Lex.
• Doctrinal treatment of the spectrum of distinctiveness at OpenStax, Introduction to Intellectual Property § 4.6.