Imagine two bakeries. The first wants to register the name "Fresh Bread" for its loaves. The second wants to register "Aerie" for the same thing. They open on the same street, sell the same sourdough, spend the same money on signage. Yet the U.S. Patent and Trademark Office will hand one of them a registration certificate and slam the door on the other, and it will do so for reasons that have nothing to do with who is the better baker. Why?

Because trademark protection is not a reward for effort or quality. It is a set of substantive standards a designation must satisfy before the law will treat it as property at all--and then a gauntlet of statutory bars it must run before the federal government will register it. "Fresh Bread" fails the first test: it merely describes the product, so it is not distinctive, and even if the baker could nurse it to distinctiveness over years it would still have to clear the descriptiveness bar of the Lanham Act. "Aerie" sails through: it is arbitrary, inherently distinctive, and nothing in the statute stands in its way. The difference between them is the subject of this article.

This is Part 2 of our four-part Trademark Overview series. Part 1 explained what kinds of things can be trademarks--words, logos, slogans, colors, sounds, product shapes (see Trademark Overview--The Subject Matter of Trademark Law). Part 2 explains the substantive standards a designation must meet to become a protected, registrable trademark. There are two affirmative requirements--the mark must be distinctive and it must be used in commerce--and a longer list of statutory bars to registration in Section 2 of the Lanham Act (15 U.S.C. § 1052) that can sink an otherwise distinctive, properly used mark. Part 3 covers how you convert those rights into a federal registration and how you license them (see Trademark Overview--Obtaining Protection and Licensing); Part 4 covers what happens when someone steps on your rights (see Trademark Overview--Infringement and Related Rights).

We will work through the whole apparatus: the Abercrombie spectrum that ranks marks from unprotectable to bulletproof; the secondary-meaning doctrine that rescues weak marks; and then, one by one, the bars of Section 2--mere descriptiveness, deceptive misdescriptiveness, geographic descriptiveness, surnames, functionality, likelihood of confusion, deception, and the disparagement and scandalousness bars that the Supreme Court demolished on First Amendment grounds in Matal v. Tam and Iancu v. Brunetti. We will give you the tests examiners actually apply, the cases that supply them, and worked hypotheticals that turn doctrine into intuition. Let us start where every trademark starts: with the question of whether it can identify a source at all.

The First Pillar: Distinctiveness

Distinctiveness is the ability of a designation to identify a single source of goods or services. That is the whole game. The question is never "is this a clever name?" or "did they think of it first?" It is whether, in the mind of an ordinary consumer, the designation functions as a brand--an indicator of who stands behind the product--rather than as a description of what the product is. A term is protectable as a trademark only if it is "capable of identifying a single source of goods or services, even if that source is unknown" to the consumer.

The reason the law insists on distinctiveness becomes obvious once you remember why trademarks exist at all. A mark is a tool that serves two audiences. For consumers, it is a shortcut: when you see the swoosh, you know who made the shoe and roughly what to expect, and you need not audit the supply chain every time you shop. For producers, it is the way to capture goodwill--the reason it pays to make a good product is that customers can find you again. Both functions collapse if a mark cannot point reliably to one source. A word that everyone uses to name the product itself ("bread," "shoe," "stapler") points to no one in particular and so cannot do a trademark's job. Worse, handing one seller exclusive rights in the common name for a product would let that seller monopolize the language competitors need to sell their own goods--a concern the Federal Circuit's predecessor court flagged in In re Abcor Development Corp., 588 F.2d 811, 813 (C.C.P.A. 1978): the descriptiveness restriction exists to prevent a trademark holder "from inhibiting competition" and to maintain everyone's "freedom to use the language" without fear of harassing infringement suits.

Distinctiveness comes in two flavors. A mark can be inherently distinctive, born ready to identify a source the moment it is used. Or it can lack inherent distinctiveness but acquire it over time, through use and promotion, until consumers come to associate it with a single source--this is acquired distinctiveness, also called secondary meaning. A third category sits outside the system entirely: generic terms, which can never be trademarks under any circumstances. Figuring out where a designation falls is the single most important skill in trademark selection and clearance, and lawyers do it with a famous tool: the spectrum of distinctiveness.

The Abercrombie Spectrum

In 1976, Judge Henry Friendly, writing for the Second Circuit in Abercrombie & Fitch Co. v. Hunting World, Inc., 537 F.2d 4 (2d Cir. 1976), sorted trademarks into a spectrum that has governed the field ever since. From weakest to strongest, marks fall into five categories: generic, descriptive, suggestive, arbitrary, and fanciful. The Supreme Court adopted the framework in Two Pesos, Inc. v. Taco Cabana, Inc., 505 U.S. 763, 768 (1992), so it is not a regional rule of thumb; it is the law of the land. Climb the ladder.

Generic terms are the common name of the product or service itself. "Aspirin" for the pain reliever, "thermos" for the vacuum flask, "escalator" for the moving staircase--each was once a brand and each tumbled into the public domain when it came to name the thing rather than the source. (Trademark lawyers call this fate "genericide," and it stalks every famous brand; Xerox, Band-Aid, and Google all spend real money fighting to keep their names from becoming the everyday word for the product.) A generic term gets no protection, ever, no matter how strongly the public associates it with one company, because the alternative is monopoly over the product's only common name. That is why "Fresh Bread" from our opening can never be a trademark for bread. The Supreme Court added a wrinkle in USPTO v. Booking.com B.V., 591 U.S. 549 (2020): a "generic.com" term is not automatically generic and can register if consumers in fact perceive it as a brand--a point we develop in Part 1.

Descriptive terms describe a characteristic, quality, ingredient, function, or feature of the product--"Cold and Creamy" for ice cream, "Quick Print" for a copy shop, "WHITE" for milk, "HANDY" for repair services. Descriptive marks are not inherently distinctive, because on first encounter they tell the consumer about the product, not about who makes it. But--and this is the crucial mercy in the doctrine--descriptive terms can become protectable if they acquire secondary meaning. We devote a full section to that below, and another to the descriptiveness bar that descriptive marks must overcome to register, because that is where most real-world fights happen.

Suggestive terms suggest a quality of the product but require the consumer to use some imagination, thought, or perception to make the leap. "Coppertone" for suntan lotion, "Netflix" for streaming, "Greyhound" for a bus line, "Jaguar" for a fast car--none literally describes the product, but each hints at it. Suggestive marks are inherently distinctive and protectable immediately on use, with no proof of secondary meaning. The line between suggestive and descriptive is the most litigated boundary in this entire field; it is fuzzy, fact-bound, and worth fortunes, because landing on the suggestive side means protection from day one, while landing on the descriptive side means you own nothing until you prove secondary meaning.

Arbitrary terms are real words used in a way utterly unrelated to their dictionary meaning. "Apple" for computers, "Amazon" for an online store, "Camel" for cigarettes, "Shell" for gasoline. The word exists and means something--it just has nothing to do with the product. Arbitrary marks are inherently distinctive and strong.

Fanciful terms are invented words coined to be trademarks and nothing else: "Kodak," "Exxon," "Verizon," "Pepsi," "Häagen-Dazs." Because they mean nothing else, they are the strongest marks of all--maximally distinctive, broadest in scope, easiest to enforce. "Aerie" from our opening sits near the arbitrary/fanciful end (a real but obscure word with no connection to bread), which is exactly why it registers while "Fresh Bread" cannot.

A useful way to internalize the spectrum: strength and marketing-convenience run in opposite directions. Fanciful and arbitrary marks are legally strong but tell the customer nothing, so they must be advertised into meaning. Descriptive marks are marketing-convenient--"Best Buy" instantly says what the store sells--but legally weak, because the law resists letting anyone fence off ordinary descriptive language. Brand owners are perpetually tempted toward the descriptive end for the free advertising; trademark lawyers are perpetually pulling them back toward the suggestive-and-stronger end for the protection. When you choose a brand name, you are choosing a position on this spectrum whether you realize it or not. (For the practical art of choosing and clearing a strong mark, see How to Conduct a Comprehensive Trademark Clearance Search and Trademark Basics.)

A Worked Example of the Spectrum

Suppose our hypothetical client, Acme Corp., wants to launch energy bars and has five candidate names. "Energy Bar" is generic--the name of the product, off the table no matter what. "Quick Energy" is descriptive--it describes what the bar does, so Acme could use it but would own no enforceable rights and could not register on the Principal Register until it proved secondary meaning, which for a brand-new product means years of vulnerability. "Summit" is suggestive--it hints at peak performance without describing the bar, so it is inherently distinctive and protectable from the first sale. "Anvil" is arbitrary--a real word with no connection to energy bars, inherently distinctive and strong. "Zylo" (a coined word) is fanciful--strongest of all, though Acme must spend to teach consumers what it means. For most clients the right answer lives in the suggestive-to-fanciful range: strong enough to enforce, distinctive enough to register without a fight, and not so descriptive that a competitor can crowd up next to it and credibly say, "I'm just describing my product."

Acquired Distinctiveness (Secondary Meaning): How a Weak Mark Grows Strong

Here is the doctrine that gives hope to descriptive marks, and it is one of the most useful concepts in all of trademark law. A term that is not inherently distinctive--a descriptive term, a surname, a geographic name, a product design--can still become a protected trademark if it acquires distinctiveness through use and promotion. The technical phrase is acquired distinctiveness; the older synonym, used interchangeably, is secondary meaning. The "secondary" meaning is the source-identifying significance that grows up alongside the term's ordinary (primary) descriptive meaning. When enough consumers, encountering the term, think first of the brand rather than the description, the term has acquired secondary meaning and becomes protectable and registrable under 15 U.S.C. § 1052(f).

Think about "American Airlines." Its primary meaning is descriptive: an airline that is American. But decades of use have layered a secondary meaning on top, so that to the flying public the phrase now identifies one specific carrier. The same is true of "Best Buy," "Sharp" televisions, "Bank of America," "The History Channel," and--per the Supreme Court--"Booking.com." Each started life describing something. Each earned trademark protection by teaching the public to hear, in those ordinary words, the name of one company.

The Five-Year Presumption

The Lanham Act offers a generous shortcut. Under 15 U.S.C. § 1052(f), a mark that has been in substantially exclusive and continuous use in commerce for five years before distinctiveness is claimed is entitled to a presumption of acquired distinctiveness. In practice, an applicant whose descriptive mark has five clean years of use can submit a simple sworn declaration to that effect, and the examining attorney will generally accept it as prima facie evidence of secondary meaning. Two cautions. First, it is only a presumption--an examiner or opponent can rebut it, and a highly descriptive mark may need more than the bare declaration. Second, the use must be substantially exclusive: if competitors have used the same descriptive term all along, five years of your own use does not show consumers associate the term with you alone.

Proving Secondary Meaning Without the Presumption

When five years is not available--because the mark is younger, or the owner must prove secondary meaning in litigation rather than at the USPTO--courts and the Trademark Trial and Appeal Board (TTAB) weigh a familiar cluster of factors. No single factor controls; the inquiry is holistic and asks one ultimate question: in the minds of consumers, does this designation now identify a single source? The factors typically include:

  • Consumer surveys. Direct evidence of the consumer association is the gold standard. A properly designed survey showing that a substantial share of relevant consumers connect the term with a single (even anonymous) source is powerful and sometimes decisive.
  • The amount and manner of advertising. Large, sustained, brand-focused spending suggests effort to build a source association--especially "look-for" advertising that expressly tells consumers to recognize the term or feature as a brand.
  • Volume and length of sales. Long, high-volume sales under the mark support an inference that the public has come to recognize it; courts weigh duration of use heavily.
  • Unsolicited media coverage. Articles and third-party references using the term as a brand name (rather than a generic description) show the marketplace already treats it as a source identifier.
  • Intentional copying by competitors. Deliberate copying is treated as circumstantial proof that the term had acquired secondary meaning worth copying--why imitate it otherwise?
  • Direct consumer testimony and evidence of actual confusion, where available.

These are the proofs a brand owner assembles when its mark is descriptive, when its mark is a product design under Wal-Mart Stores, Inc. v. Samara Bros., 529 U.S. 205 (2000), or when it is defending a registration against a claim that the mark never deserved protection. The practical lesson is to document as you go. Smart owners keep records of advertising spend, sales figures, press coverage, and any evidence of copying from day one, because the day you need to prove secondary meaning is the day you will wish you had kept the receipts. (For how these standards play out in disputes, see Navigating the Maze of Trademark Confusion.)

When an applicant cannot yet prove secondary meaning, a descriptive (or surname, or geographically descriptive) mark can still be registered on the Supplemental Register--a kind of waiting room that confers fewer benefits but lets the owner use the ® symbol and builds a record of use that can later support a § 2(f) claim on the Principal Register. We cover the two registers in Part 3.

The Second Pillar: Use in Commerce

Distinctiveness alone is not enough; in the United States, federal trademark rights arise through use of a distinctive mark in commerce. You do not get rights by being the first to think of a name or the first to register it. With one important exception--the intent-to-use system below--you get rights by being the first to use the mark in the marketplace. This "first to use" rule stands in sharp contrast to the "first to file" systems that govern most of Europe, China, and much of the world, where whoever races to the trademark office first generally wins.

The Lanham Act defines "use in commerce" in 15 U.S.C. § 1127 as the bona fide use of a mark in the ordinary course of trade, and not use made merely to reserve a right in the mark. That "bona fide... ordinary course of trade" language is the law's filter against fake use. For goods, the mark must appear on the goods, their containers, or associated displays, and the goods must be sold or transported in commerce. For services, the mark must be used in selling or advertising the services and the services must actually be rendered--advertising a service you have not begun to perform is not use, the trap that voided the application in Aycock Engineering, Inc. v. Airflite, Inc., 560 F.3d 1350 (Fed. Cir. 2009), and the registration in Couture v. Playdom, Inc., 778 F.3d 1379 (Fed. Cir. 2015). The bar for "commerce," though, is low: a single bona fide interstate sale of caps was enough in Christian Faith Fellowship Church v. Adidas AG, 841 F.3d 986 (Fed. Cir. 2016).

The 1988 Trademark Law Revision Act killed the old practice of token use (a single contrived shipment to manufacture "use") and replaced it with the intent-to-use (ITU) application, 15 U.S.C. § 1051(b). An ITU application lets you apply before using the mark, provided you have a bona fide intention to use it, and--if it matures into registration--gives you nationwide constructive use priority as of your filing date under 15 U.S.C. § 1057(c). The catch, per M.Z. Berger & Co. v. Swatch AG, 787 F.3d 1368 (Fed. Cir. 2015), is that "bona fide intent" is an objective standard: the intent must be firm and documented, and the absence of business records can by itself defeat it. We treat use, priority, the ITU system, and the Tea Rose-Rectanus good-faith-remote-use doctrine (see Stone Creek v. Omnia) in depth in Part 1 and Part 3; here we keep our focus on the registrability bars, because that is the heart of the substantive-standards inquiry and where this article does its real work. (For the mechanics of filing, see How to File a Trademark Application with the USPTO and Federal Trademark Application Checklists.)

The Statutory Gauntlet: The Section 2 Bars to Registration

Suppose your mark is distinctive and you are using it in commerce. You still are not home. To obtain (and keep) a federal registration on the Principal Register, the mark must clear Section 2 of the Lanham Act, 15 U.S.C. § 1052, a list of bars that an examining attorney can invoke to refuse registration--and that opponents can later use to oppose or cancel. These bars are the substance of most real prosecution fights. A surprisingly large share of applications die here, not at the distinctiveness or use stage. We will take the most important bars in turn: the § 2(e) family (descriptiveness, deceptive misdescriptiveness, geographic descriptiveness, surnames), functionality, § 2(d) likelihood of confusion, and the § 2(a) bars (deception, false association, and the once-mighty disparagement and scandalousness clauses now reshaped by the Supreme Court).

Section 2(e)(1): Merely Descriptive and Deceptively Misdescriptive Marks

Section 2(e)(1) bars registration on the Principal Register of a mark that is "merely descriptive" of the goods or services--the statutory backstop to the Abercrombie distinctiveness rules. The test, settled by the Federal Circuit, asks whether the mark immediately conveys knowledge of an ingredient, quality, characteristic, feature, function, purpose, or use of the goods or services. In re Chamber of Commerce of the U.S., 675 F.3d 1297, 1300 (Fed. Cir. 2012); In re Gyulay, 820 F.2d 1216, 1217 (Fed. Cir. 1987). The examiner cannot judge descriptiveness in the abstract; she must consider the particular goods or services, the context of use, and the term's likely significance to the average purchaser, In re Stereotaxis, Inc., 429 F.3d 1039, 1042-43 (Fed. Cir. 2005).

Two refinements are worth knowing because they drive outcomes. First, the question is not whether a consumer could guess the product from the mark; it is whether someone who already knows the product would understand the mark to convey information about it. DuoProSS Meditech Corp. v. Inviro Medical Devices, Ltd., 695 F.3d 1247, 1254 (Fed. Cir. 2012). That asymmetry is exactly what separates descriptive marks ("you'd understand it describes the product") from suggestive ones ("you'd need imagination to connect it"). Second, the mark must be judged as a whole. An examiner may not dissect a composite into individual descriptive words and stop there; descriptiveness of the parts does not doom the whole if the combination has source-identifying significance. Estate of P.D. Beckwith, Inc. v. Commissioner of Patents, 252 U.S. 538, 545-46 (1920); In re Oppedahl & Larson LLP, 373 F.3d 1171, 1174-75 (Fed. Cir. 2004). A mark also need not describe every feature to be descriptive--describing one significant attribute is enough (TMEP § 1209.01(b)).

A few practical doctrines round this out. Because the suggestive/descriptive line is genuinely murky, doubt is resolved in favor of the applicant and the mark is published for opposition, leaving competitors free to oppose later (In re Fat Boys Water Sports LLC, 118 U.S.P.Q.2d 1511 (T.T.A.B. 2016)). Under the doctrine of foreign equivalents, examiners translate words from common modern languages into English before testing for descriptiveness, so a foreign word that is descriptive in translation is usually treated as descriptive--unless the ordinary American purchaser would not stop and translate it (Palm Bay Imports, Inc. v. Veuve Clicquot Ponsardin, 396 F.3d 1369, 1377 (Fed. Cir. 2005), declining to translate VEUVE). And a descriptive mark refused under § 2(e)(1) has three escape routes: argue it is actually suggestive, claim acquired distinctiveness under § 2(f), or amend to the Supplemental Register. (For the broader strategy of responding to refusals, see Part 3 and Trademark Registration Guide.)

The same subsection also bars deceptively misdescriptive marks: marks that misdescribe the goods in a way consumers would plausibly believe, but where the misdescription is not material to the purchase. (If the misdescription is material--if it would actually affect the buying decision--the mark crosses into outright deceptive territory under § 2(a), discussed below, which is an absolute bar.) The classic illustration: "Glass Wax" for a glass cleaner that contains no wax is deceptively misdescriptive--consumers might believe the misrepresentation, but it does not drive the purchase. A deceptively misdescriptive mark, like a merely descriptive one, can be saved by acquired distinctiveness. A deceptive mark cannot be saved at all.

A Worked Example: Descriptive Versus Suggestive

Return to Acme. It applies to register "Quick Energy" for energy bars. The examiner refuses under § 2(e)(1): "Quick" plus "Energy" immediately conveys that the bar delivers fast energy--it describes a function. Acme's best arguments are that the composite, taken as a whole, requires a mental step (weak here) or that the mark has acquired distinctiveness (no, it is too new). Realistically, Acme's options are to amend to the Supplemental Register and build a § 2(f) record over time, or--better--to have picked "Summit" in the first place, where the examiner would have had to explain what imaginative leap a consumer must make, and the doubt-favors-applicant rule would have tilted toward publication. That is the whole strategic point of climbing the spectrum: it moves you from a refusal you must dig out of to a publication you receive almost automatically.

Section 2(e)(2) and (3): Geographic Marks

Geography gets its own pair of bars because place names raise the same monopoly worry as descriptive words--no one baker should fence off "Vermont" for everyone else's Vermont cheese--plus a deception worry when the place name lies.

Section 2(e)(2) bars marks that are primarily geographically descriptive. The examiner makes a prima facie case by showing three things: (1) the primary significance of the mark is a generally known geographic location; (2) the goods or services originate in that place; and (3) purchasers would be likely to believe the goods come from that place--a goods/place association that is presumed when the place is known for the goods (TMEP § 1210.01(a)). "Minnesota Cigar Company" for cigars made in Minnesota is the paradigm (In re JT Tobacconists, 59 U.S.P.Q.2d 1080 (T.T.A.B. 2001)). Like other descriptive marks, a primarily geographically descriptive mark can be rescued by acquired distinctiveness under § 2(f) or parked on the Supplemental Register.

Section 2(e)(3) bars marks that are primarily geographically deceptively misdescriptive--and this one is harsher, because after the NAFTA implementing legislation it is an absolute bar that no amount of secondary meaning can cure (for marks acquiring distinctiveness after December 8, 1993). The test adds a materiality element to the geographic test: (1) primary significance is a known place; (2) the goods do not originate there; (3) purchasers would likely believe they do; and (4) the misrepresentation is a material factor in a substantial portion of consumers' purchasing decisions. In re California Innovations, Inc., 329 F.3d 1334, 1339-41 (Fed. Cir. 2003). The Federal Circuit there explained that § 2 "no longer treats geographically deceptively misdescriptive marks differently from geographically deceptive marks," collapsing the § 2(e)(3) and § 2(a) geographic analyses into one materiality-driven inquiry. "Old Havana" for rum made in Florida flunked it (In re Compania de Licores Internacionales S.A., 102 U.S.P.Q.2d 1841 (T.T.A.B. 2012)): consumers would believe the rum is Cuban, and Cuban origin materially drives rum purchases.

The "primary significance is geographic" threshold is low and forgiving to examiners. The place can be a neighborhood or a fuzzy region; nicknames, misspellings, and adjectival forms count (so "Yosemite" reads as the national park, In re Spirits of New Merced, LLC, 85 U.S.P.Q.2d 1614 (T.T.A.B. 2007); "Baikalskaya" reads as Lake Baikal, In re Joint-Stock Co. "Baik", 80 U.S.P.Q.2d 1305 (T.T.A.B. 2006)). The escape hatches are to attack the goods/place association, to show the term has a dominant non-geographic meaning (think "Amazon" for books, where the river is not the primary significance), or, for § 2(e)(2) only, to claim acquired distinctiveness.

A Worked Example: The Geographic Bar

Acme launches a premium line and wants "Napa Valley" for wine vinegar. If the vinegar is actually made from Napa grapes, the mark is primarily geographically descriptive under § 2(e)(2)--registrable only with § 2(f) secondary meaning or on the Supplemental Register, because Napa is famous for grape products and the goods come from there. If the vinegar is made in New Jersey, the mark is primarily geographically deceptively misdescriptive under § 2(e)(3): consumers will believe a "Napa Valley" vinegar is from Napa, that belief is material (Napa origin commands a premium), and the mark is dead on arrival--no secondary meaning, no Supplemental Register, no rescue. The lesson for clients: a place name only helps you if you are actually from the place, and even then it is a weak mark you will spend years strengthening.

Section 2(e)(4): Primarily Merely a Surname

Section 2(e)(4) bars marks that are primarily merely a surname, registrable on the Principal Register only on a showing of acquired distinctiveness. The policy is intuitive: the law wants to keep surnames available for the many people who share them and may want to use their own names in business. (For the closely related problem of building a brand around your own last name, see Trademarking Your Own Name.)

The test asks the primary significance of the mark to the purchasing public: when consumers encounter the term, do they perceive it primarily as a surname or as something else? In re Kahan & Weisz Jewelry Mfg. Corp., 508 F.2d 831, 832 (C.C.P.A. 1975). The TTAB's In re Benthin Management GmbH, 37 U.S.P.Q.2d 1332 (T.T.A.B. 1995), supplies the working factors (guidelines, not a rigid test): (1) how rare the surname is; (2) whether anyone connected with the applicant bears it; (3) whether the term has a recognized non-surname meaning; (4) whether it has the look and feel (structure and pronunciation) of a surname; and (5) the effect of any stylization or design elements. As with descriptiveness, doubt is resolved in favor of the applicant (TMEP § 1211.01).

Two counterintuitive points reward attention. First, rarity does not save you if the public still perceives the term as a surname. Even a vanishingly rare name can be barred when a famous bearer gives it surname significance--the TTAB affirmed a refusal of "Belushi's" for restaurant and travel services despite only five U.S. residents named Belushi, because John and Jim Belushi made the surname meaning unmistakable (In re Beds & Bars Ltd., 122 U.S.P.Q.2d 1546 (T.T.A.B. 2017)). Second, the quality of the examiner's evidence matters enormously, and a surname refusal is often beatable by attacking it. Telephone-directory listings alone are not determinative (Kahan & Weisz); foreign directories and listings of non-U.S. residents are irrelevant to U.S. consumer perception; and references buried in obscure or limited-circulation publications do not show the broad public exposure a refusal requires (In re Joint-Stock Company "Baik", 84 U.S.P.Q.2d 1921 (T.T.A.B. 2007)). The applicant's strongest moves are to show a recognized ordinary-word meaning ("Bird," "King," "Rose") and to attack the rarity evidence as duplicative, conclusory, or insufficient.

Functionality: The Bar That Keeps Useful Features Out of Trademark Law

Functionality is the most conceptually important bar for anyone trying to protect a product shape, feature, or look--trade dress. The rule is blunt: a functional product feature can never be a protected trademark or trade dress, no matter how distinctive it is and no matter how much secondary meaning it has acquired. Functionality is an absolute defense and an absolute bar.

The reason is the boundary between trademark law and patent law. Patents grant a limited-term monopoly on useful inventions in exchange for public disclosure; when the patent expires, the public is supposed to be free to copy the invention. Trademark protection is potentially perpetual. If a company could trademark a useful feature, it could monopolize that feature forever and make an end-run around the patent bargain. The functionality doctrine slams that door. The Supreme Court set the standard in Inwood Laboratories, Inc. v. Ives Laboratories, Inc., 456 U.S. 844, 850 n.10 (1982): a feature is functional, and unprotectable, if it is essential to the use or purpose of the article or if it affects the cost or quality of the article. Qualitex Co. v. Jacobson Products Co., 514 U.S. 159 (1995), confirmed that even color can serve as a mark only when it is non-functional. And TrafFix Devices, Inc. v. Marketing Displays, Inc., 532 U.S. 23, 29-30 (2001), added a powerful evidentiary lever: a prior utility patent disclosing the feature's advantages is "strong evidence" of functionality, putting a heavy burden on the would-be trademark owner to prove the feature is merely ornamental.

There are two species. Utilitarian functionality is the Inwood/TrafFix core--the feature works better, costs less, or performs the product's job. Aesthetic functionality is the subtler cousin: a feature that is not mechanically useful but whose protection would put competitors at a significant non-reputation-related disadvantage--say, the color black for outboard motors, desirable because it coordinates with any boat and makes engines look smaller (Brunswick Corp. v. British Seagull Ltd., 35 F.3d 1527 (Fed. Cir. 1994)). In assessing functionality, courts weigh the existence of related patents, advertising that touts the feature's utilitarian advantages, and the availability of alternative designs.

Crucially, functionality also fixes the dividing line that decides whether secondary meaning is even possible. Recall Wal-Mart v. Samara Bros.: product packaging trade dress can be inherently distinctive, but product design (configuration) can never be inherently distinctive and is protectable only with secondary meaning. Stack the two doctrines together and product-shape protection becomes the hardest corner in trademark law: you must prove the design is non-functional and prove it has acquired secondary meaning. We treat this in depth in The Intricate World of Trade Dress Protection, The Evolution of Trade Dress Protection for Product Design and Configuration, and the patent-vs.-trade-dress boundary in Design Patents vs. Trade Dress Protection for Product Configurations.

A Worked Example: Functionality

Acme designs an energy bar with a distinctive double-twist wrapper that customers come to recognize on sight. Can Acme trademark the wrapper shape? It depends entirely on why the twist exists. If the double twist is purely decorative and Acme can show non-functionality plus secondary meaning, it may protect the packaging trade dress. But if Acme's own marketing brags that the double twist "keeps your bar fresher longer," or if Acme holds a utility patent on the twist-seal, the feature is functional under Inwood and TrafFix, and no amount of consumer recognition can save it. The takeaway is uncomfortable for marketers but iron-clad in law: the better a feature works, the harder it is to own as a trademark.

Section 2(d): Likelihood of Confusion

Even a distinctive, non-descriptive, non-functional mark can be refused if it collides with someone else's prior rights. Section 2(d) bars registration of a mark that "so resembles" a mark already registered, or previously used and not abandoned, "as to be likely to cause confusion, or to cause mistake, or to deceive." 15 U.S.C. § 1052(d). This is the single most common substantive refusal at the USPTO, and it is the same question that decides infringement in court--which is why clearance searching before you adopt a mark is not optional (see How to Conduct a Comprehensive Trademark Clearance Search).

Likelihood of confusion is a multifactor inquiry. At the USPTO and Federal Circuit it runs on the DuPont factors from In re E.I. du Pont de Nemours & Co., 476 F.2d 1357 (C.C.P.A. 1973): the similarity of the marks in appearance, sound, connotation, and commercial impression; the relatedness of the goods or services; the similarity of trade channels and classes of purchasers; the conditions of purchase (impulse buy versus careful, expensive decision); the fame and strength of the prior mark; the number and nature of similar marks on similar goods; the existence of actual confusion; and more. In the federal courts the same idea travels under circuit-specific names--the Polaroid factors in the Second Circuit (Polaroid Corp. v. Polarad Electronics Corp., 287 F.2d 492 (2d Cir. 1961)), the Sleekcraft factors in the Ninth (AMF Inc. v. Sleekcraft Boats, 599 F.2d 341 (9th Cir. 1979)), and parallel tests elsewhere. No factor is dispositive; the two that usually do the heavy lifting are the similarity of the marks and the relatedness of the goods. Because confusion is the spine of infringement, we develop the full analysis in Part 4 and in Navigating the Maze of Trademark Confusion. For purposes of this article, the point is that § 2(d) is the screen through which your distinctive mark must still pass: distinctiveness gets you to the door; non-conflict gets you through it.

Section 2(a): Deception, False Association, and the First Amendment Earthquake

Section 2(a) is a grab-bag of moral and consumer-protection bars, and it has been the most constitutionally turbulent provision in the statute.

The durable, uncontroversial part of § 2(a) bars deceptive marks and marks that falsely suggest a connection with persons, institutions, beliefs, or national symbols. A deceptive mark--unlike a merely deceptively misdescriptive one under § 2(e)(1)--is an absolute bar: it cannot be saved by secondary meaning, because the law will not let anyone build a brand on a lie that materially affects purchasing. The test (from In re Budge Manufacturing Co., 857 F.2d 773 (Fed. Cir. 1988)) asks whether the mark misdescribes the goods, whether consumers are likely to believe the misdescription, and whether the misdescription is material to the purchasing decision. "Lovee Lamb" for synthetic seat covers (no lambskin) flunked it. The false-association prong is what stops you from registering "Pope" merchandise that implies Vatican endorsement, or a mark that appropriates a living person's identity without consent.

Then there is the part that blew up. For most of the twentieth century, § 2(a) also barred marks that were immoral, scandalous, or disparaging. Two Supreme Court decisions struck those clauses down as unconstitutional viewpoint discrimination under the First Amendment. In Matal v. Tam, 582 U.S. 218 (2017), the Court invalidated the disparagement clause in a case brought by an Asian-American dance-rock band that wanted to register "THE SLANTS" as a defiant reappropriation of a slur; the government, the Court held, may not refuse registration because it disapproves of the message. Tam also doomed the long-running cancellation of the Washington NFL team's marks, a saga we chronicle in Washington Redskins Trademark Troubles--Disparagement and the Aftermath. Two years later, Iancu v. Brunetti, 588 U.S. 388 (2019), struck down the immoral or scandalous clause in the case of the clothing brand "FUCT," holding that bar, too, was unconstitutional viewpoint discrimination.

But the demolition was not total, and the Court drew a careful line in Vidal v. Elster, 602 U.S. 286 (2024). There it upheld the "names" clause of § 2(c)--which bars registering a living person's name without consent--against a First Amendment challenge by an applicant who wanted to register "TRUMP TOO SMALL." The Court reasoned that the names bar is viewpoint-neutral and content-based but not viewpoint-based, and that content-based trademark conditions grounded in a long history of restricting the trademarking of others' names are compatible with the First Amendment. The upshot after Tam, Brunetti, and Elster: the government may not deny registration because a mark is offensive or expresses a disfavored viewpoint, but it may impose viewpoint-neutral conditions, including the deception bar, the false-association bar, and the consent requirement for using a living person's name. These subject-matter questions are the heart of Part 1, so we leave the fuller treatment there.

Putting the Standards Together

Step back and the architecture is clean. A designation becomes a federally protected, registrable trademark when it satisfies two affirmative requirements and survives the statutory bars:

  1. It is distinctive--inherently (suggestive, arbitrary, fanciful) or through acquired secondary meaning (descriptive terms, surnames, geographically descriptive terms, product designs). Generic terms never qualify.
  2. It is used in commerce in the ordinary course of trade (or is the subject of a documented bona fide intent to use).
  3. It clears the § 2(e) bars--it is not merely descriptive or deceptively misdescriptive, not primarily geographically descriptive (without secondary meaning) or deceptively misdescriptive (ever), and not primarily merely a surname (without secondary meaning).
  4. It is non-functional if it is a product feature or trade dress.
  5. It does not create a § 2(d) likelihood of confusion with a prior mark.
  6. It does not fall under the § 2(a)/(c) bars on deception, false association, or use of a living person's name without consent.

Notice how the bars track the affirmative requirements. The § 2(e) family is the statutory enforcement arm of the distinctiveness pillar: descriptive, geographic, and surname marks are exactly the marks that lack inherent distinctiveness, so the statute makes them prove secondary meaning before it will register them. Functionality polices the boundary with patent law. Section 2(d) polices the boundary with everyone else's prior rights. And § 2(a) polices the boundaries of deception and--after the First Amendment cases--no longer polices taste. Understand why each bar exists and you will rarely be surprised by how it applies.

Keeping Protection: The Standards Sustain Rights, Too

The same standards that create protection also sustain it. A registered mark can be deemed abandoned after three consecutive years of nonuse with intent not to resume (15 U.S.C. § 1127), and registrants must file periodic declarations of continued use to keep a registration alive. On the distinctiveness side, a mark can be lost to genericide if the public turns it into the common name for the product, and it can be lost through naked licensing if the owner licenses the mark without policing the quality of the licensee's goods, because the mark no longer reliably signals a consistent source. We cover use-based loss of rights and the licensing-and-quality-control rules in Part 3. The symmetry is the lesson: a trademark lives on distinctiveness and use, and it dies when either one fails.

Key Takeaways

  • Distinctiveness is the threshold. Marks run the Abercrombie spectrum from generic (never protectable) through descriptive (protectable only with secondary meaning) to suggestive, arbitrary, and fanciful (inherently distinctive). Choose your name knowing where it sits.
  • Weak marks can grow strong. Descriptive terms, surnames, geographically descriptive terms, and product designs can acquire distinctiveness. Five years of substantially exclusive use earns a § 2(f) presumption; short of that, courts weigh surveys, advertising, sales, media, and copying. Document as you go.
  • The Section 2(e) bars enforce distinctiveness at the USPTO. Merely descriptive and deceptively misdescriptive marks (§ 2(e)(1)), primarily geographically descriptive marks (§ 2(e)(2)), and primarily-merely-a-surname marks (§ 2(e)(4)) are refused absent secondary meaning. Geographically deceptively misdescriptive marks (§ 2(e)(3)) and outright deceptive marks (§ 2(a)) are absolute bars that no secondary meaning can cure.
  • Functionality is the hardest bar. A feature that is essential to the product's use or purpose, or that affects its cost or quality, can never be a trademark (Inwood; Qualitex; TrafFix)--and product designs need secondary meaning on top of non-functionality (Wal-Mart).
  • Likelihood of confusion (§ 2(d)) is the gatekeeper against prior rights. Analyzed under the DuPont/Polaroid/Sleekcraft factors, it is the most common refusal and the spine of infringement. Clear before you adopt.
  • Section 2(a) no longer polices taste. After Tam and Brunetti, the disparagement and scandalousness bars are gone; after Elster, the viewpoint-neutral conditions (deception, false association, consent to use a living person's name) survive.
  • The standards sustain rights, not just create them. Abandonment, genericide, and naked licensing can undo a mark you fought to register.

Master these standards and you understand the engine room of trademark law. Everything else in this series--how to register and license, and how to enforce--builds on the foundation laid here.

Frequently Asked Questions

My business name describes exactly what I do. Can I register it? Not right away, and maybe never on the Principal Register without help. A merely descriptive mark is refused under Lanham Act § 2(e)(1) until it acquires secondary meaning--proof that consumers associate it with you as a single source--which the statute presumes after five years of substantially exclusive use under § 2(f). In the meantime you can register on the Supplemental Register to use the ® symbol and build a record. A stronger choice from the start is a suggestive, arbitrary, or fanciful name, which is protectable and registrable immediately. See Trademark Basics and When to Trademark Your Brand.

What's the difference between "descriptive" and "suggestive"? A descriptive mark immediately conveys information about the product to someone who knows the product; a suggestive mark requires a mental leap, some imagination, to connect the name to the product. "Quick Print" for a copy shop is descriptive; "Coppertone" for suntan lotion is suggestive. The line is genuinely blurry, which is why the USPTO resolves doubt in the applicant's favor and publishes the mark for opposition (In re Fat Boys Water Sports). Suggestive marks are protectable on first use; descriptive marks are not.

Can I trademark a place name, like my city or region? Only carefully. If your goods actually come from the place and consumers associate the place with the goods, the mark is primarily geographically descriptive under § 2(e)(2)--registrable only with secondary meaning or on the Supplemental Register. If your goods do not come from the place but consumers would believe they do, and that belief is material to the purchase, the mark is primarily geographically deceptively misdescriptive under § 2(e)(3)--an absolute bar no secondary meaning can fix (In re California Innovations). A place name with a dominant non-geographic meaning (like "Amazon" for books) may avoid the bar entirely.

My last name is my brand. Can I register it? A mark that is "primarily merely a surname" is refused under § 2(e)(4) absent acquired distinctiveness, because the law keeps surnames available for everyone who shares them. The USPTO weighs the Benthin factors--rarity, whether anyone connected with the applicant bears the name, any ordinary-word meaning, the look-and-feel of a surname, and stylization (In re Benthin). A surname with a recognized non-surname meaning ("King," "Rose") fares better, and surname refusals are often beatable by attacking the examiner's evidence. See Trademarking Your Own Name.

Why can't I trademark the shape of my product? You sometimes can, but it is the hardest case in trademark law. A functional feature--one essential to the product's use or purpose, or that affects its cost or quality--can never be a trademark, because that would let you monopolize a useful feature forever and bypass the patent system (Inwood; TrafFix). And even a non-functional product design (as opposed to packaging) is protectable only with secondary meaning (Wal-Mart v. Samara Bros.). A prior utility patent on the feature is strong evidence it is functional. See The Intricate World of Trade Dress Protection.

The USPTO refused my mark because it's confusingly similar to another. What now? A § 2(d) refusal turns on the DuPont likelihood-of-confusion factors--mark similarity, relatedness of goods, trade channels, purchaser sophistication, fame of the prior mark, and more. You can argue the marks or goods are distinguishable, narrow your identification of goods, seek a consent agreement from the prior owner, or appeal to the TTAB. Because this is the most common substantive refusal, the best fix is upstream: a real clearance search before you adopt. See How to Conduct a Comprehensive Trademark Clearance Search and Part 4.

Can I register an edgy or offensive brand name? After Matal v. Tam (2017) and Iancu v. Brunetti (2019), yes--the Supreme Court struck down the bars on disparaging, immoral, and scandalous marks as unconstitutional viewpoint discrimination, so the USPTO can no longer refuse a mark because it is offensive. What survives is viewpoint-neutral: the deception bar, the false-association bar, and the rule that you cannot register a living person's name without consent, which the Court upheld in Vidal v. Elster (2024). See Part 1.

Where does this fit in your trademark series? This is Part 2. Part 1 covers what kinds of things can be marks (Subject Matter of Trademark Law); Part 3 covers obtaining registration and licensing (Obtaining Protection and Licensing); Part 4 covers infringement and related rights (Infringement and Related Rights). For a plain-English starting point, see Trademark Basics and The Trademark Process.

Related Articles

This article provides general information about U.S. trademark law and is not legal advice. Trademark questions are fact-specific and the law continues to develop; consult qualified counsel about your particular situation before acting.