Now, The Name Inspector realizes that the “millionth word” story is a ridiculous play for attention that’s not to be taken seriously, and that the folks at the Global Language Monitor know it. But the story has gotten people talking about what a word is, and that’s a topic that The Name Inspector can warm to.

The easiest criticism of the millionth-word story is that Web 2.0 isn’t a word, but a phrase. That’s the main thing that linguist Geoffrey Pullum had to say about the matter on Language Log. And that’s pretty disappointing, actually, because it ignores the fact that the whole enterprise of counting words that precisely is linguistically suspect.

Why would The Name Inspector object to counting words? Believe it or not, it’s not due to a perverse academic refusal to give simple answers to simple questions. The innocent word, which seems to be the very simplest little bit of language to understand, is remarkably hard to pin down. There are very clear examples of words, like dog, but around the edges the word category is fuzzy. That makes it hard to count words with any precision, let alone announce the exact time of day when a word enters the language.

Let’s start with the very dumbest definition of word, the one used by the “word count” function on your word processor: A word is a string of characters (lets say letters) with no spaces. Well, that would mean the following string consists of five words: jjj akjsdhfjkh auygfh tg drqwds.

We can do better than that: A word is a string of letters with no spaces that has a meaning and can be used in a sentence. By this definition, Web 2.0 doesn’t cut it. And many people who’ve weighed in on the issue in blog comments have raised just that objection. Some object to the space, some to the digits, some to the punctuation. Sorry, sorry, and sorry. If inclusion in a dictionary is the ultimate proof of wordhood, then consider this: Even the abridged online version of the Merriam-Webster dictionary includes entries for deep six, 12-step, 20/20, 24-7, 3 D, and even 86ing (a slang term for refusing to serve a customer). All these include numbers, all but one include digits, some have punctuation, and one has a space.

Now about spaces. It’s commonly accepted that English has complex prepositions that consist of parts. In some cases the parts are separated by spaces, and in others they’re not. We write in lieu of as three chunks, and instead of as two, even though their structures are parallel, etymologically speaking. Then there’s notwithstanding.

There are many compound words in the Merriam-Webster dictionary (and others) that are written with spaces. A space is a purely orthographic entity, and it’s silly to define a linguistic unit based on orthography alone. Spoken language is primary. Written language is, ultimately, a representation of spoken language. There are compounds that some people write as “one word” and some people write as “two words”, though the pronunciation remains constant. Website/web site is one example. If you use the no-space criterion, you end up saying that such expressions are sometimes words and sometimes not words, based on orthographic variation. And that just doesn’t make good sense.

Of course, you might appeal to The Language Boss to tell you which version is “correct”. But people, it’s time to wake up and realize that The Language Boss is a fiction, like the Wizard of Oz. There are just different people, sometimes with different opinions, bumbling around behind their curtains. Pay no attention to that language maven behind the curtain!

Lurking behind the orthographic issue, of course, is a deeper linguistic one: If some words are made of pieces that are themselves words, how do we know when a group of words adds up to a complex word as opposed to a phrase or a random stretch of language? Here linguists begin to rely on criteria that distance the definition of word from the pragmatic, what-you-list-in-the-dictionary understanding of what a word is. The linguists might, for example, think about how an expression interacts with the rules of English stress assignment, or about it’s syntactic behavior. In any event, for a group of words to add up to a complex word, it has to be a conventional, cohesive unit.

And here there are no hard and fast rules. Idioms make things especially complicated. Merriam-Webster lists kick ass and kick the bucket under its entry for kick. So these idioms get a sort of honorary word treatment. But notice that idioms don’t always occur in exactly the same form: we can kick a little ass or kick some ass or even kick some Raider ass. In idioms, words begin to blend into grammar, and that’s where things get really tricky.

Some idioms, like kick the bucket and kick ass, are identified mostly by the presence of certain component  words. Others, however, are more like grammatical templates. Consider sentences like There’s only so far a car can go with a flat tire, There’s only so long you can sit before you have to get up and walk around, and There’s only so often you can talk or sneak your way out of a fight. There’s a pattern here that’s something like There’s only so X Y can Z, where X is a scalar measure or property of some kind,  Y is a noun phrase, and Z is a verb phrase. Most people wouldn’t call this pattern a word, but it’s hard to find the exact barrier between this pattern and something like kick ass. (To see lots of patterns like this, you might take a look at the Snowclone Database).

Even when you’re talking about words with simple forms, it can be hard to decide how to count them. That’s because words aren’t just forms–they also have meanings, and it’s often the case that the same form has more than one meaning. If the meanings are very different, we usually think of there being more than one word. For example, bank used in connection with a river is one word, and bank used in reference to a financial institution is another.

But what if the meanings are only a little different? How many “words” are represented by these different uses of the verb see?

Can you see the car?
I see that it’s raining.
I don’t see why you’re so angry.
Let’s go see grandma.
Are you seeing anyone?
I’ll see your twenty and raise you ten.
Let me see you to your door.
See to it that this doesn’t get out.

All these complexities don’t mean it’s impossible in principle to count the number of words in the English language. They do, however, mean that it’s very, very hard, and that you have to know what you mean by word before you start.

This isn’t the first time someone has tried to own a letter. According to MediaLife Magazine, O, The Oprah Magazine, which goes by the nickname of O Magazine, was sued for trademark infringement in 2001 by a German publication called O Magazine. Judge John Koeltl, who dismissed the lawsuit, reportedly said that O, The Oprah Magazine, which features stories about women’s health and well-being, and O Magazine, which features photographs of women in fetish attire, are “devoted to different aspects of women’s lives”. Indeed.

Now the folks at QVC may have had a purely metaphorical, fun, marketing use of the word own in mind. But companies often do get all hot and bothered about trying to own our cultural commons more literally. When The Name Inspector worked at a language technology company in the late 1990s, one of the executives thought we should seek a patent on our code because it would be “like having a patent on the English language”. Such hubris, such naivete.

Ah, now The Name Inspector is taking a trip down memory lane, and is reminded of another story from the late 1990s, which Geoffrey Pullum wrote about in 2004. James Pustejovsky, a famous computational lexical semanticist (at least, about as famous as a computational lexical semanticist can be), started a natural language processing company in 1997. First he called it Lexeme. This is a perfectly nice real-word name. Lexeme is a technical term in linguistics that means, basically, ‘word’. But a large company threatened to sue for infringement upon their trademark. Any guesses about which trademark it was? Lexus. Yep, for the luxury division of Toyota. Somehow they believed they owned the sequence of letters lex, despite the fact that this is an ancient Greek root found in English words like lexicon, lexical, and dyslexia.

As Pullum wrote, “Aren’t the owners of the English language ever going to rise up against greedy corporations like Lexus and Microsoft and Starbucks who lay claim to whole regions of the phonetosphere…?”. The prospect of huge legal fees forced Pustejovsky to change the name of his company, but he found his own way to fight back: he called his company LingoMotors, no doubt intending a dig at Lexus. But The Name Inspector fears that fighting linguistic land grabs with little witticisms is a bit like the idea, lampooned in Woody Allen’s movie “Manhattan”, of countering a Nazi rally with a devastating editorial.

Tags: q, the letter q, qvc, abbreviations, lexeme, the name lexeme, lingomotors, lexus, o magazine, oprah magazine, oprah

We’re really getting a chance to enjoy the graphic possibilities of the letters i and j lately. If you’re a blogger you can put a Lijit Wijit on your blog. And now there’s Kijiji. Actually, there has been Kijiji for a while now in other countries, but eBay has just launched a U.S. version of this local classified ad network, putting them in direct competition with Craigslist.

Just look at all those dots and vertical lines. When the name is written in a serif font, as it is in the company logo, the dots look like heads, the vertical lines look like torsos, and the serifs look like arms reaching out in a welcoming embrace or unbridled enthusiasm or something else great like that. You would be forgiven for thinking that this unusual name was invented just to achieve this graphical effect. But actually kijiji is a Swahili word meaning ‘village’. It’s a diminutive form of the word mji, which means ‘town’. So it’s almost as if in Swahili they call a village a townlet.

Why choose Swahili for the name of an international network of classified ad sites? There are two good reasons. First, Swahili is a lingua franca–a common language used for business by speakers of other languages–in East Africa. Native speakers live in and around Tanzania.

Second, Swahili syllables tend to conform to what phonologists consider the universally preferred syllable structure, which is a single consonant followed by a single vowel, represented CV. (An exception is the first syllable of the word mji, which consists entirely of the nasal [m] sound. That may seem exotic, but in English we actually have syllables consisting only of nasal sounds, like the final syllable of the word button. We just don’t put those syllables at the beginnings of words.)

What does it mean for the CV syllable to be universally preferred? For one thing, it means kids produce this kind of syllable first when acquiring language. It also means that this kind of syllable is found throughout the languages of the world, while other kinds of syllable are more likely not to be allowed in this or that language. Most importantly, it means that in theory just about everybody in the world should find it pretty easy to pronounce this name.

About the first part of Kijiji: Have you studied French or Spanish or German or some other language and struggled with the system of grammatical gender? Well, if you didn’t enjoy that, steer clear of Swahili. It has more than ten genders, or noun classes, as they’re commonly called. Each noun starts with a prefix like ki- or m- showing its class. When a noun is used as a subject, verbs and some other words must be marked for agreement with its class prefix. At least the noun classes are based more on meaning than the random gender systems of other languages are, so it’s often easy to guess which class a word belongs to.

Anyway, The Name Inspector is getting off topic. How does this name do in English? Well, it’s actually a little hard to pronounce, preferred syllable structure notwithstanding. The two affricate sounds followed by high front vowels are kind of awkward to squeeze out. And despite the graphical gimmick of the name (or perhaps because of it), the orthography is a little hard to parse. When you see this name on the page or the screen, it looks like a bunch of scratches–it’s hard to distinguish the letters.

So this name makes perfect sense from a semantic and sociolinguistic point of view, but it suffers from the very orthographic and phonetic properties that make it special.

Tags: kijiji, the name kijiji, swahili, lijit, wijit, lingua franca, syllables