What are we even talking about anyway? by Sten Lillieström

What are we even talking about anyway?

by Mr. Sten Lillieström (LinkedIn) 

When UDRP panels decide whether a domain name registration is likely to be an effect of cybersquatting, they rely on legacy trademark law terminology. At the same time, they do not have access to the full array of resources at the disposal of a court of law, should square terminology alone leave something to be desired.

The nature of a device that is supposed to be applicable only to “the relatively narrow class of cases of abusive registrations” is by necessity going to be leaning towards summary speculation. It’s natural. But in the odd case that the subject matter is more nuanced than what the blunt tool is geared to handle, it is probably a good idea to double-check how well the terminology actually aligns with what it purports to measure. Otherwise the terminology may risk living a life of it’s own, and that is not the idea behind any law or legal institution.

Here is therefore an attempt to explain the terminology and to put it into a wider context.

“Generic”

In trademark lingo, if a term is “generic”, it is reason to not allow it to be registered as a trademark. It’s what a language community currently commonly uses to refer to something, and therefore not legitimately protectable. In copyright terms it is “public domain”. In patent terms it would not exhibit an “inventive step”.

Not an option, duh.

Genericness can thus be an argument on behalf of a domain name registrant that the registration is “legitimate”, and it can be the argument from a competitor that wants to cancel your trademark. Conversely, it can also be an argument that something is not “generic”, and should therefore be considered a less legitimate investment.

It’s an important term to understand. It’s part of a handful of interrelated terms that make up the categorization structure for types of names in trademark lore known as the “Abercrombie spectrum” (1976) or “the spectrum of distinctiveness”. So let’s try to understand it.

If we look up the word “generic” in the dictionary, it primarily relates to the ability to denote categories or classes. A “generic expression” is consequently the equivalent of an umbrella term. So, a “butterfly” is a “butterfly” since someone at some point in time categorized all insects with certain traits and gave them a collective name. The specific types of butterflies in turn have even more specific categories all the way down to “cabbage white”, “swallowtail” or “monarch”.

As is often the case with dictionaries it may however not be the whole truth and nothing but the truth as it relates to the way a word is actually used. Ironically perhaps, UDRP jurisprudence has nevertheless evolved into employing precisely dictionary inclusion as a sort of litmus test for when something is to be considered “generic”. Dictionaries are this way regarded as the equivalent of a collection of all recorded “generic” terms. (This shortcut solution has no precedence in actual trademark law statute; it is completely a pragmatic-jurisprudential invention.)

But is this specific shorcut valid? Does everything in the dictionary really refer to a class of things? Common nouns arguably do, but they only comprise around 37 percent of what is called word tokens in all available text data, according to studies.

This hopefully seems to imply that not merely “something that refers to a category” is “generic”, but that there is something else to the equation as well. (Otherwise we would have already destroyed the validity of the “dictionary test” simply by thinking about it for two seconds.)

This something else may relate to yet another possible meaning of “generic”, namely “common”. Dictionaries after all grow primarily based on relative frequency of use. When a term becomes obsolete, it hopefully falls out of the dictionary – when it becomes prevalent, it is hopefully included. Does this sense of “generic” lend more credence to the idea to treat dictionary inclusion as a signal that the domain registration may have alternative motives outside of targeting a rights holder?

As it turns out, word frequency is not the penultimate arbiter on whether a name is generally likely to find a use either. One unwanted side effect of the reasoning that frequency equates to accessibility may be on display in the 2023 decision to transfer the domain name “Sleeptopia․com” to the Complainant (D2023-1317).

In this case, the argument that the “SLEEPTOPIA mark is not a dictionary word, colloquial term, or otherwise descriptive in nature” is instrumental to the decision to allow a domain name registered in 2006, before the existence of the Complainant, to nevertheless be considered registered and used in bad faith. (Circumstantial and unrebutted evidence indicated an ownership change, and so called PPC-links were probably the last nail in the coffin.) As is regretfully often the case, the Respondent did not respond, and there was consequently no alternative explanation to what the Panel decided to infer. It would perhaps not have been unreasonable to infer the inverse; that the name is a meaningful and natural name, constructed in a way that most names are constructed, by combining words and wordparts. In this case “sleep” and “utopia”. This wild idea was however for some reason not examined.

But perhaps to the disappointment of advocates of a quick fix for the issue of cybersquatting, frequency of word of occurence is at the end of the day actually no hard measure of how easily conjured or comprehended a word is either. Occasion and context can trigger words rarely seen or used and we will understand them just as perfectly. Most know what the Cold War is even though it ended long ago. Hippies are not common these days, and therefore not a common word, but we still get it. Context, which last I looked is always present, is also a factor. If we for example see the word “Tamarind” in a recipe, we are likely to activate or deduce what it is, should we not consciously know it already.

All of this confusion notwithstanding, the main issue is however even more fundamentally confusing. It stems from the fact that dictionaries are just not an arbiter of whether or not something is a “real word” or has “meaning” in the first place. This notion is generally considered by experts to be a popular myth with low bearing on actual circumstance. Here’s what the Merriam-Webster dictionary themselves say about words that are not included in their dictionary: “They’re still real words! Many words that aren’t widespread enough to make it into the dictionary—words that are particular to a region or profession or even a family—are perfectly good words; it’s just unlikely that a person outside that area or group would encounter them. For now.”

Irrespective of physical dictionaries and their contents, when a lexical gap emerges in a natural flow of speech, we may sometimes be “at loss for words”. Other times, we freely invent new words on the fly. Sometimes without realizing it. “Outdoorsiness”. “Uncorkable”. “Ghost-like”. “Spring-drunk”. We may also attempt to avoid the trap that relying on one word would set by instead using several. Children may especially excel at forming new words. (Adults may find it a rewarding exercise as well, even the ones that have absorbed the myth of that dictionaries have any authority at all on the subject of language.) The linguist and cognitive psychologist Steven Pinker gives some examples of child-speech creativity in his book “The Stuff Of Thought”.

• The water was *drindling* down the drain.
• I’m going to *sloop* up the gravy.
• Why does the lightbulb *ringle* when you shake it?

These wonderful examples are likely conjured as an effect of something called “phonesthesia”; meaning that lives in sequences of speech sound.

But novel terms and names can be built from a wide array of strategies, too many for an extensive list in this article. Some of them may likely still be undiscovered. Adults may be more prone to use word parts (morphemes) creatively. Here are some examples, also courtesy of Pinker. He quotes the comedian Rich Hall who conjured words that previously may not have seen the light of day:

• Elbonics: The actions of two people maneuvering for the armrest in a movie theater.
• Peppier: The waiter at a fancy restaurant whose sole purpose seems to be walking around asking diners if the want ground pepper.

It seems to simply be the case that when we name things, we do so based on structure and logic. The root of this structure and logic is our knowledge about language. And the basis for our knowledge of language is generic in all of the senses of that word. It is a species-wide ability that everyone possesses. Some may be more artful in their performance than others, but everyone has access to the same internal capacity. (As it turns out, even severely brain damaged individuals can exhibit normal or even supernormal linguistic capabilities.)

It’s plain to see that nobody naming something will do so by letting their cat dance all over the keyboard. Not unless the result by lucky chance happens to be intriguing or suitable enough. Ask a naming agency whether they use randomized letters as a model or whether they look up the “correct” dictionary word for the industry in which their client operates when they mine for viable options. The fundamental question for a naming agency and an IP expert alike is always going to be what the effect on the mind is. There is really no simple answer to that question other than experimentation, in theory as well as practice.

Two co-existing definitions of “generic” would probably have been confusing enough, right? But the crux of the matter of bad faith speculation sometimes also seems to be related to whether the snippet of language in question is literal or figurative. This becomes clear when the playful language experiment of naming is employed in ways that go beyond what trademark law terminology is tooled to cover. Some fringe cases may for example contain “generic dictionary terms” that are combined into novel constructs. A UDRP panel may then for example argue that even though the domain name is *comprised* of “generic” terms, that the end product is not a “common term”, and that the registration is more likely to be cybersquatting for that reason.

This was the case in the UDRP decision regarding “Solverde.com” (D2020-3138).

”The Panel does not consider SOLVERDE to be a dictionary word or phrase. SOLVERDE is a distinctive trademark and not a descriptive or generic term. It is correct that SOLVERDE consists of two dictionary terms in Spanish language (sol, verde). However, the composite term “solverde” is in the Panel’s view not descriptive or generic. There is no “green sun” and to best of the Panel’s knowledge “solverde” is neither a dictionary term nor a common expression.”

This clear opinion of course ignores the fact that language is creative by design, and that naming is pretty much the most obvious example of that creativity. (For good, pragmatic reasons, entirely unrelated to the trademark regime). What you will find exhibited in names is the wonderful language of names, not some textbook example of literal language that purports to only say what it means or limit itself to words or expressions that are only allowed to denote physical objects that “exist in the real world”.

The trademark regime is not a pre-emptive trademark for all viable name identites. That was not the intention.

Names, like words in general, are also regularly formed by utilizing word parts; so called morphemes. Another example where this specific strategy is deemed inherently suspect can be found in yet another categorical rejection of the way that names are formed; the 2024 decision to transfer the domain name “Medaxis․com” to the Complainant (D2024-2027).

“The Panel acknowledges that, as a general rule, a person such as the respondent has the right to acquire a domain name containing a dictionary word or phrase based upon its dictionary meaning and to offer this for sale at a price of its choosing, provided that in so doing it does not target the trademark value of the term. The problem for the Respondent in this case is that the argument that the disputed domain name contains a generic or dictionary term rather than a trademark seems particularly contrived. It requires the Respondent to separate the second level of the disputed domain name into two words, then ascribe a meaning to the first word, then find some kind of dictionary meaning in the two words taken together. For the Panel, these steps are steps too far when the disputed domain name itself has the appearance of being a trademark or coined term rather than having any logical dictionary meaning. The Respondent presented no evidence that “medaxis” is a term in any dictionary or used in any publications or in the media. The fact that other businesses also use “medaxis” as a trademark or brand does not assist the Respondent’s argument on this issue. Further, the Panel does not consider “medaxis” to be a generic word or term. A generic word should be seen by the community or the relevant trade to mean something on its face. The term “medaxis” seems to mean at best something medical on an axis, which does not mean much at all, except perhaps an axis or group of doctors. In these circumstances, the value of the term “medaxis” in the disputed domain name appears to the Panel to derive primarily from the trademark and not from any alleged dictionary meaning.”

An alternate conclusion that would not be all that hard to reach is that “med” is probably short for “medical” (as seen in meds, medpack or medicare) and that “axis” is a natural way to describe the notion of a core basis around which things revolve. “Axis” is an extremely prevalent term in business names because it has relevant descriptive power. It’s virtually a metaphor for business.That the term is abstract or figuratively used is not an argument that the coinage is unnatural. Both terms are natural and meaningful in the context of naming a business in the medical field, and in combination they are also a natural option – trademark existence notwithstanding. This is in fact quite possibly the reason that the Complainant chose this name, and quite possibly also the reason that Respondent did as well. It would not even be a stretch of the imagination to consider that the overlapping coinage by the Complainant and the Respondent could extend even further into other past and future examples of independently reaching the same naming option. It is not shown how the fact that several unrelated entities use the same name is not an argument for it’s natural and general appeal. It is just an opinion presented as fact.

In fact the steps that the “Medaxis” Panel speculates would have to occur in the mind of the Respondent to be able to mentally parse the term “Medaxis” can be called into serious question. In a 2003 paper, Sam Glucksberg, professor in Psycholinguistics, has the following to say about the idea that metaphor for example always takes additional cognitive effort: “We apprehend metaphorical meanings as quickly and as automatically as we apprehend literal meanings.”

TM law however to the contrary traditionally conjectures that literal language requires little to no mental processing and that figurative language requires more. If it is the case that the mental process always refers to literal interpretation first, the speculation that figurative language requires a “mental leap” may be adequate. If the modes of interpretation however are equally accessible, as Glucksberg argues, there may be no reason to assume that metaphor or other naming strategies, are more mentally taxing than retrieving the mental lexicon entry for “car” from memory.

Other studies have further tested this hypothesis. In a 2020 paper, the authors show that even “novel metaphor” (with the assistance of context) does not seem to inherently differ significantly from the processing of literal language. “Our findings demonstrate that context affects neural mechanisms that promote the usefulness of novel language”.

It may be sufficiently clear that the context of a brand name assists in interpretation. The more of a prototypical name that a name is, and the more the context supports it, the stronger the signal to interpret it in a certain way.

According to one theory with wide acclaim called “Career of Metaphor”, the contents of dictionaries are actually to a great extent simply “dead metaphors”. Through normative use, some of them lost their relation to whatever spawned them and they became “literal”. If this theory is correct, and there is no reason to assume it is not, what is called literal language is just a warehouse of staple goods and a mere side-effect of one of the superpowers of language; to convey novel meaning.

In the past couple of decades, our understanding of metaphor from a cognitive perspective has actually changed dramatically. Historically, metaphors were viewed as infrequent and decorative. However, recent studies indicate that metaphors are a fundamental aspect of daily communication and thought, and not some kind of fancy poetry. Trademark law in all it’s forms is built on a very different premise, and it may be useful to recognize this.

In this segment, we saw that the term “generic” seems to have atleast three different plausible definitions:

1. Relating to a category.
2. “Common term”.
3. Literal language.

At the same time, this ambiguity is never made explicit in UDRP decisions. The term must for some reason be considered self-explanatory, even though that is obviously not the case. One could perhaps wonder why.

As noted, the concept of “generic” is also related to other legal concepts that prompt different legal treatment. At the other end of the spectrum of distinctiveness is a collective term that is considered to be the antonym of “generic”.

“Distinctive”

According to the review above, an oversimplified translation of the relation between what is “generic” and what is “distinctive” may be “common VS unique”. (In trademark practice, it seems that the less prevalently used or known a string is, the more distinctive it is.) This however presents a conundrum that is rarely mentioned. Why, if distinctiveness is the key aim, do businesses not just name themselves “PÅN)U#N VÅ” or “#===(VM M” and be done with it? Mission accomplished! As distinctive as it gets!

The answer, again, must of course be that names have structure and logic. Even at the distinctive end.

The question of whether something is distinctive by design (“inherently distinctive”) is important because a term that is considered distinctive carries significantly greater scope as a trademark. Traditionally and contemporarily, the line between what is considered distinctive and what is considered descriptive or generic is drawn at a point in the spectrum of genericness where something called “suggestive” is found.

(“momondo”) (“apple”) (“microsoft”) (“salty pretzels”) (“accounting”)

FANCIFUL -> ARBITRARY -> SUGGESTIVE -> DESCRIPTIVE -> GENERIC

PROTECTABLE <————–> NOT PROTECTABLE

Suggestive names are not considered to literally “describe” the product or service. They therefore require a “leap of the imagination”. As already noted, this may seem like a speculative assumption. Every piece of language that ever existed was once invented. Normative use made some expressions shed their versatility and anchor them to inflexible specifics. But if trademark law is supposed to conjecture about the way that language actually works, it may be prudent to improve on the stale notion that “real language” is literal.

Indeed, delineation between what is “descriptive” and what is “suggestive” is sometimes impossible even to a Court. In a review of this issue, Law Professor Jake Linford lists a set of quotes by courts that have had to decide on similar matters. “Nearly incapable of precise description”; “not always clear”; “shadowy”; “murky”; “wholly lacking . . . in distinctiveness”; “difficult . . . to draw”; “frequently difficult to apply”; “uncertain and impossible to extrapolate from precedent”; and “almost impossible to define in the abstract.” 

And even though this delineation is too commonly a coin-toss, the difference in outcome depending on which category the trademark is chosen to belong to is dramatic. A suggestive name does not require evidence of what is called “acquired distinctiveness” to be able to be registered as a trademark. A descriptive “name” does.

Improving the lens through which these kinds of decisions are perceived could likely be of use to everyone with a stake in intellectual property.

So what is “distinctive”? Is it as fleeting as “generic” in that several definitions are used interchangeably?

Let’s look at the dictionary again.

“Distinctive”: “characteristic of one person or thing, and so serving to distinguish it from others”

Alright, ok, so something unique that renders the person or thing distinguishable is distinctive. This may seem like little more than a tautology. The simple word “set” for example is considered to have 430 different senses and counting. Each one obviously quite distinguishable in context. Even from the perspective of trademark law, a single word mark can have multiple (in theory an endless amount) of unique use cases that do not infringe or impinge on eachother.

The difference when it comes to “prima facie” assessment of distinctiveness in a UDRP context is that this fact is thrown out the window. If a domain is orthographically (textually) identical or similar to a trademark (sometimes one of hundreds of similar marks) – tough luck – this is going to be the first step towards successful appropriation by Complainant.

But what creates distinctiveness? Does creating a new word along the lines of the language capacity we all have always result in something distinctive? Is “Spring-drunk” distinctive because it is rare? Wouldn’t something truly distinctive require that the string has no inherent meaning?

Indeed, some legal sources reference meaning as having a role. A lot of UDRP decisions do. (13,633 WIPO administered UDRP decisions contain the phrase “no meaning”). For legal mechanisms built around “meaning” to be meaningful, it would however probably be required for us to know what “meaning” is in the first place. We have some ideas about this, but consensus is that we pretty much have no idea how meaning works.

It may however be clear enough that “meaning” is not a phenomenon that lives in dictionaries, but in the creatures that coined all of the dictionary entries, all of the non-dictionary-entries, all of the future dictionary entries and all of the future non-dictionary-entries.

In other words, the antonym pair of “meaningful VS meaningless” is certainly not synonymous with the antonym pair “common VS unique” – and there is yet again ambiguity as to what a term that decides the fate of domain registrations and trademark scope alike actually means.

Perhaps “meaningful VS meaningless” would be a better measure of whether something is more or less “distinctive”? Meaning is arguably the fundamental reason for all name selection. It can be meaning that does not previously relate in any apparent way to the source, or it can be lack of apparent meaning, but the subject matter is meaning. And assessment of meaning is universally accessible to a language user. Names are predominantly meaningful in some way, and this meaning does not conveniently disappear because it is applied to a new source. The number of senses and references are just amended and extended. (Other times, the apparent absense of meaning is the aim.)

The idea of “inherent distinctiveness” and that “source significance” can be instant and automatic upon first use sometimes seems does seemt to impede comprehension of the fact that even “distinctive” names have properties that can be more or less appealing. A brandable domain investor is demonstrably in the business of amassing more or less distinctive and appealing name identites, and is arguably legitimately free to do so as long as no targeting of a specific rights holder is the aim. If the brandable domain investor is successful, he will have succeeded in gauging this appeal, without targeting anything but general appeal and likelihood of new adoption.

It’s also relevant to note that trademarks are the most valuable sector in IP, and IP is the most valuable sector in global business. The raw material for brand names and their prospective trademarks are however globally accessible through our shared language competence. This means that the creative output of name selection is going to overlap among different interests in a very natural way.

The shapes and senses of brand name examples in general concert are furthermore a pervasive feature of the public domain. They naturally make an imprint in the minds of consumers (anything less would be a capital destruction scandal). The effect is likely an increased sensibility to signals of “nameness” in general. Nothing exists is in a vacuum; generalizations as to what features are generally present in names will happen, and they will guide any and all coinages of novel options.

The purpose of brands is to make money from consumers. In doing so, they also make it legitimately possible for the audience to make money on the general existence and nature of brand names – without necessarily having to resort to cybersquatting.

That a name is “like a trademark” is not really reason to assume that it inherently targets some specific trademark. Saying that a name is similar in type to how trademarks generally appear is not sufficient to assume bad faith. Trademarks are a very limited special case and they are a subcategory that exist in a much wider context of all available naming options.

I am not arguing that trademarks are impossible, or not useful, or that cybersquatting does not exist.

I am saying that an oversimplified heuristic is going to lead to cases where a legitimate rationale for name selection, completely outside anything “derived primarily from the trademark”, may not be appropriately acknowledged.

 About the Author: As the founder and CEO of Next Venture, a domain acquisition brokerage and naming firm, I help entrepreneurs and businesses find and secure their desired brand names on the internet. As a consequence I also operate a portfolio of domain names that serve as viable naming options in different niches and industries.