There is a saying in philosophy that if you want to say something true, say something vague. The aphorism is particularly effective in what philosophers call ordinary language, where in the spirit of cooperation, vague utterances invite charitable interpretation, thus fostering successful communication between speaker and hearer.

Within UX, successful communication of content through a navigation scheme is a proverbial challenge for designers. The words we choose, despite our best intentions, may fail for a variety of reasons, what philosophers call infelicities. Contributing to the challenge are warranted incentives for brevity, which as a result invite the failure we endeavor to avoid.

With this essay, I discuss the tension between the risk as well as utility of vagueness and draw lessons from observations in philosophy of language.

The challenge may be reduced, I argue, by adopting an approach to defining nomenclature that considers the user as hearer. The approach encourages application of conversational maxims that adjudicate between word choices, favoring those that at least increase the likelihood of successful communication.

A Heap of Names

The phenomenon of vagueness is of particular interest to philosophers for its relevance to what is called “the sorites paradox.” The proverbial example is a heap of sand. Since a grain of sand does not constitute a heap, the addition of one more grain makes no difference, but with additional grains, a heap results. When? The answer is not obvious and although the topic may strike some as esoteric, for others there are moral implications. Our interest, however, is elsewhere.

Vagueness attracts philosophers, as well as linguists, for its role in speech and communication. As speakers, we often say vague things that successfully communicate our intentions. The observation that San Francisco is a small city, for example, is generally met with agreement whether one’s audience is from New York or elsewhere. Yet, what exactly constitutes small or a small city is not obvious. Nonetheless, in the spirit of cooperation agreement is typically found and communication successful.

Within UX, we rely on such cooperation when we choose words to denote the content we define, but such reliance is taken for granted and consequently neglected. My original design for Facebook Studio, for example, consisted of “Case Studies” as a top-level navigation item but despite almost universal familiarity with the term, internal preference for “Showcase” won out. The result said too little to be readily understood.

Other sites consist of navigation schemes that say too much while some say things unclearly. Appliance Replacement Parts, for example, is arguably verbose under the heading Appliance Parts & Services within a main navigation menu labeled Appliances on Sears.com. There may be, of course, SEO considerations for such copy but SEO is not equivalent to usability nor does it necessarily translate into a good user experience.

At the same time, Bloomberg’s main navigation is unclear given a multi-colored set of links within an offscreen menu, competing with a default set of utility links prominently displayed, sometimes hidden by a sticky navigation consisting of blue copy on a blue background that unsurprisingly fails Web Content Accessibility Guidelines, not to mention a row of links preceding the set described. Bloomberg’s navigation constitutes an extreme violation of presentation rules.

A Familiar Culprit

The challenge is not unfamiliar and warnings abound. Hoa Loranger of the Nielsen Norman Group, for example, advises us to stay away from cool names, despite temptation, because they suck (See “Avoid Category Names That Suck”).

Among other reasons, clever names suck because they cause doubt and and hinder site exploration. Such names are symptomatic of all kinds of things that, as Steve Krug puts it, “can make us stop and think unnecessarily” (Steve Krug, Don’t Make Me Think Revisited, 2014). The typical culprit, Krug laments, “are cute or clever names, marketing-induced names, company-specific names, and unfamiliar technical names” (Krug). Junos comes to mind. Central to the core offering of network giant Juniper, Junos is the company’s operating system but as my research has shown it is typically thought of as the company name or something vaguely related to Juniper.

There are, of course, exceptions. Google, for example, is derived from a child’s made-up word, but is obviously well known as a brand and as a verb. Nonetheless, the exceptions are rare and the presence of unfamiliar technical names in one’s interface may not only be ignored, but also end a user’s session.

Whatever the characteristic, such names may all be considered vague in the context we find them, but this only reveals that it is not the words themselves that stop us so much as the context. Put differently, use is the culprit. As the philosopher J. L. Austin remarked, it is usually the “uses of words, not words in themselves that are properly called vague” (J. L. Austin, Sense and Sensibilia, 1962). After all, every word has its place and an unfamiliar technical name is familiar somewhere and only unfamiliar elsewhere.

The culprit is most always in us, at least those of us who choose words that present more of an obstacle than a meaningful pathway.

I plead guilty. When redesigning The Walt Disney Family Museum website, I became attached to vocabulary that extended from the museum’s brand strategy. My initial navigation scheme led with “Imagine,” followed by “Explore.” Yet, the context from which these terms arose was lost on even the most enthusiastic Disney fan. The terms readily invited the questions Imagine what? and Explore what? The uncertainty, if not confusion, was confirmed in user testing. We went with Walt Disney and Galleries instead.

Brevity and Vagueness

Motivation is not entirely emotional. There is an incentive for not vagueness but brevity as there is only so much space within any interface and none of the advances in screen resolution allow us to ignore tenets of usability. There is, in some sense, a “Goldilocks” zone for consuming content on a screen, as a page too wide or copy too small will upset any intended communication.

A further incentive for brevity is expressed in the I-Principle, a law of the economics of language: Produce the minimal linguistic information sufficient to achieve your communicational ends I (Stephen Levinson, Presumptive Meanings – The Theory of Generalized Conversational Implicature, 2000). Formulated by linguist Stephen Levinson, the principle is an extension of philosopher Paul Grice’s conversational maxims and in particular the sub-maxim Be brief. Essentially, there is utility in describing some things in fewer words than more.

I present the relevance of Grice’s work later but here I want only to point out that real estate, usability, and economics motivate designers to not only minimize the number of items within a navigation scheme, but also minimize the number of words in such a scheme. To be sure, some items should not be given top-level status, but even with a short list of candidates, the competition remains fierce. We strive to distill the chosen categories of content or directives into as fewest words as possible. The challenge mixes an imperative with pragmatics. If we can say things in one word, why use two?

Yet, the problem with the forgoing question is that it is not as simple as saying anything. To be sure, what we say pertains to meaning, but the mere act of speech does not necessarily equate to anything meaningful. Communication does not work that way, particularly in the absence of any of its core components namely speakers, hearers, words, and context (One may wonder if someone says something in a forest and there is no one else around to hear her, is what is said meaningful?). If the mere act of speech entailed meaning, then our challenge with nomenclature would be limited to real estate only. Vagueness would be a non-issue and being brief would be a matter of course.

Still, despite the utility brevity is a double edged sword and while there are contexts in which a brief description may be as meaningful as its precise counterpart, the chance of misinterpretation increases with fewer words. Manfred Krifka, for example, argues that brevity is a worthier goal to aim for than precision in the context of distance descriptions (See “Be Brief and Vague!”), but brevity can only get you so far. Compare, for example, ‘The gas station is at 1200 19th Avenue on the corner of 19th and Lincoln Way’ with ‘The gas station is south of the park’. Though the latter statement may be true, the location is relatively vague given its brevity, and as directions, practically unhelpful.

Ontology and Deliberation

The risk of vagueness begins with our deliberation over content. Whether a given word or phrase is better than others when describing content is a question that only follows confirmation of the content in question. That content constitutes what is called a product’s ontology.

Within philosophy, ontology is roughly theory of being, of what is and of the hierarchy of what is. For many, such talk evokes thoughts of deities and demigods, but ontology is not strictly a metaphysical inquiry. On the contrary, ontology is very much part of our corporeal experience, a world in which we find among other things cats and dogs and whose relative import translates into the difference between sleeping inside or out.

Within UX design, ontology is not dissimilar. The apps or sites we design are essentially sets of objects with relation to one another, some equal in import, some relatively so. The status of objects means the difference between, for example, a place in the main navigation and a place in the footer.

A product’s ontology also includes its feature set. From this perspective, conversion paths and search functionality are as much a part of a product’s ontology as are calls-to-action and search results, which are metaphorically more tangible objects in the form of buttons, copy, images, links, and videos (For the philosophically curious, an analogy in metaphysics is the Greek goddess Charis and her acts of kindness; both are objects, the latter dependent on the former).

Whatever the set, the names and relation we give to these objects, a product’s taxonomy, are presented in some fashion or other to users and as UX designers know, a site’s navigation scheme is a pathway to that content, one which we endeavor to make meaningful. Therein lies the challenge.

Our familiar and internal description of content requires translation to external and meaningful expressions referencing that very content. The factors of real estate, usability, and economics arise in this case and are followed by temptations of brevity and consequently, the risk of vagueness. Yet, the language we use internally, if matched with the language used externally (conversationally), may avoid the ambiguity we endeavor to avoid.

Semantic Calibration

The starting point for a meaningful navigation scheme is ironically with our deliberation over content, that is, our casual and informal conversations about content. Whatever the form of communication, verbal or written, deliberation over content for the UX designer is not merely a hashing out of ideas but a design technique. As UX researcher David Peter Simon emphasizes, developing an ontological perspective is fundamental to the work of UX designers and in particular Information Architects (See “What We Talk About When We Talk About Ontology”). We ask, (1) What things exist?, (2) What is the hierarchy of the things that exist? and as it pertains to our concerns, (3) What do we name the these things?

The answers to these questions do not come easily, for it is not merely a matter of conducting an audit or taking inventory. These questions may give rise to others more difficult to answer, particularly when those questions touch on a company’s identity. “Are we a service organization?” “A solutions provider?” “What are we?” And although the existential crisis may be solved, one’s answers lay a foundation for a vocabulary that may or may not resonate with one’s intended audience. The possible tension between a company’s self-conception and conception of that company from the outside is perpetual. The dictum Choose your words wisely applies as much here as anywhere. If one describes oneself in vague terms, then although the description may be true, it is also open to interpretation.

The challenge extends into deliberation over nomenclature but that deliberation is essential to our objective. Dan Klyn of the Understanding Group refers to the technique as ontological tuning, an activity to ensure cognitive consistency across a system (See “5 SEO Questions For Dan Klyn – SES Toronto”). I prefer semantic calibration, to borrow from some programming efforts, for the issue is fundamentally a matter of semantics and as I teach my students, semantics is everything.

Whatever words we we choose to describe our content internally, we want only those words that translate into a navigation scheme calibrated to our audience. Only then will we ensure meaningful direction.

The Informal Context

Methods of testing the resonance and efficacy of the vocabulary that emerges from our ontological investigation are, of course, established. Card sorting is a standout and worthwhile example but what is more telling are the casual and informal conversations we have with the very persons we may in another context refer to as subjects.

The informal context, the ordinary context, is one that mitigates the potential risks that come with testing or even formal interviewing. Such risks include, among other things, The Hawthorne Effect: the tendency for those being observed to be influenced by knowledge of the observation. For example, it has been shown that users behave differently in the laboratory than on the Web, and one possible explanation is that test subjects are more inclined to make an effort given the presence of the experimenter, or, authority figure as indirect motivator (See Michael Schulte-Mecklenbeck and Oswald Huber, “Information search in the laboratory and on the Web: With or without an experimenter”).

Here, the Web is the informal context I mention above but the informality begins with our deliberation over content. Essentially, we want to transcribe that conversation to the navigation schemes we present. We want to leverage the water cooler talk we have about the content, the domains, and the actions we identify. We want to then extend that vocabulary to our audience as if they are a part of the conversation and even as if our navigation schemes are an extension of the conversation. Put differently, we want to semantically calibrate to the informal context and vocabulary emerging from that environment.

Grice’s Cooperative Principle

The tact suggested here is fundamentally an extension of the philosophy of Paul Grice, whose work on meaning and theory of implicature is well known in linguistics as well as philosophy.

Grice was the first to observe that conversations are typically cooperative in nature. Our conversations are, in a Wittgensteinian sense, a game where participants play in a way so as to preserve the game. We neither talk or listen with indifference to one another. On the contrary, we endeavor to say and interpret each other so as to ensure meaningful communication.

As speakers, we most always utter what we mean but what we say is not necessarily what we mean. Within the philosophy of language, specifically speech act theory, Grice explained this ordinary and familiar phenomenon as implicature: the act of meaning something other than what is said, often by convention but also within unconventional contexts. I say, for example, I have $10.00 when $20.00 is requested but I imply that I have no more than $10.00 since saying I have $10.00 does not mean I do not have $20.00. The implication is readily understood in a conventional context and Grice teased out the workings of such communication.

As hearers, we interpret what is said in a way that gives speakers the benefit of the doubt. If I ask for $20.00 and you say that you have $10.00, I take it that you do not have more than $10.00. Similarly, we casually pass over vagueness with interpretations that complete communicatory exchanges, that renders such interaction successful. We dine together, play a game, converse, and when one of us remarks that this is fun, the other nods in agreement despite the vagueness almost inherent with indexical, in our case our use of the word this. What was fun? The dining, the playing, the conversing? One might think it is all of the above and perhaps that is what was intended, but what is intended is not equivalent to what is said. That is why ‘this’ may refer to any or all of the moments in question.

The contribution we give in interpretation extends itself to the written word. We accept and even champion our unalienable rights, for example, “Life, Liberty and the pursuit of Happiness,” though use of these terms even in the context given is vague. This is particularly true of the word happiness but as hearers, that is readers, we interpret the word as meaningful, contributing to the context in which it is presented variables that validate its use. We essentially cooperate with the speaker, or writer, to understand if not identify with the message.

Grice formulated his observation philosophically, in what is called The Cooperative Principle:

Make your contribution such as it is required, at the stage at which it occurs, by the accepted purpose or direction of the talk exchange in which you are engaged.

We can think of “talk exchange” as any form of communication and on this account we can extend the principle from speech to writing, from, say, manuals to instructions, and consequently, from directions to navigation schemes.

Conversational Maxims

The Cooperative Principle consists of four maxims, namely Quantity, Quality, Relation, and Manner, which we may roughly understood as Be informative, Be truthful, Be relevant, and Be perspicuous.

Grice regarded these maxims as commonly adhered to rules for successful communication but his contribution to the philosophy of language stems from The Cooperative Principle to an account of meaningful violations of the principle’s maxims, conventional and unconventional.

For us, flouting the maxims may come into play, but not before understanding the benefit of following the rules. We may think of the maxims as follows:

  • Quantity in terms of the adage less is more, unless more is required
  • Quality in terms of transparency and trust
  • Relation in terms of relevance of content in context
  • Manner in terms of design

For example, if we are designing a conversion point, we should present a button and not only copy (quantity), we should point to the expected page and not an interstitial (quality), we should solicit conversion after providing incentive and not before (relation), and we should avoid a green CTA on a green background (manner).

As principles of design, the maxims are straightforward and perhaps even obvious but the import of Grice’s work to UX design is the emphasis on the ordinary language aspect of conversation, an aspect that consists of violations of the very rules conducive to successful communication. Ordinarily, we do not always speak literally nor do we always speak figuratively but we rely on the literal (rules) to impart meaning through the figurative (violations). Our research should be anything but formal, if by research a lab coat is entailed. On the contrary, we should investigate casually. Our approach should be fundamentally ethnographic in nature, with subtle but profound emphasis on informality. We do not want to interview subjects, at least not heavy handedly—we want to converse with them and we want to do so in the most casual, unassuming, and comfortable setting as possible for the vocabulary that emerges from that conversing is the vocabulary that we want as nomenclature, appropriately placed.

Solutions

By adopting a Gricean vantage point, one where we consider our design an extension of conversation and our users as hearers in that conversation, we can define a navigation scheme and directional copy in general that mitigates the session ending risk of vagueness but at the same time usurps that very risk into meaningful nomenclature. We may use two or even one word to indicate a pathway but only if we have first established a context with our users where that vocabulary resonates.

Take the word solutions, for example. As one client of mine remarked, it is a pearl of a word, but if it is a pearl in any sense it is for what we can imagine it referring to since like pearls, it hasn’t any intrinsic value; out of context, a pearl is like anything else. Besides, as anyone familiar with Steinbeck can attest, a pearl may cause more harm than benefit and for my client, the word was not a cause of harm so much as it was problematic, intensified by internal descriptions and even quasi-branding of solutions. The resulting vocabulary was generally unfamiliar to the target audience and as testing revealed, the word was regularly dismissed as being vague and out of context.

Therein lies the key. The pearl of a category term like solutions is vague precisely for the reason that in preceding content it is out of context. Similar to the reaction I experienced with “Imagine” and “Explore,” solutions invoked the question, Solutions to what? For my client, the revelation was clear. Solutions are as such to problems but without first presenting those problems, the would-be pearl was indistinguishable from any other thing. The navigation scheme I was tasked to wrangle with failed to take into account its audience in a way that treats that audiences as participant in a conversation where the navigation scheme itself is an extension of the conversation.

The paradigm to facilitate such understanding, I have argued, can be found through Grice’s Cooperative Principle and the maxims within.

Add new comment

Comments

Yeah, Google is not a made up word: "The name 'Google' is a play on the word 'googol,' coined by Milton Sirotta, nephew of American mathematician Edward Kasner. A 'googol' refers to the number represented by a 1 followed by 100 zeros. It's a very large number. In fact, there isn't a googol of anything in the universe -- not stars, not dust particles, not atoms. Google's use of the term reflects our mission to organize the world's immense (seemingly infinite) amount of information and make it universally accessible and useful," Google tried to explain the origins of the company's name.

all words are made up but that point aside the name is as i say dervied, and as you describe from the word googol (which is made up). so, we are in agreement in this respect. as a side, one could argue that there is a googol of objects in the universe and in particular atoms. such an estimate would take into account not only the observable universe that underlies the rough estimate of 10^8 atoms. additionally, if we posit a multiverse, then there is most certainly a googol of atoms. and one could also argue that there is a googol of metaphysical objects, such as the total number of (human) thoughts to date.

It's really valuable article. I think vagueness is underestimated.

This is so perfect for me right now. I needed those tips!