Monday, October 30, 2006

A niche of millions

Clay Shirky, Web and computer industry pundit, flexes some neologistical chops:

I define a meganiche as a thin slice of the Web that nonetheless represents roughly a million users. The meganiche is something new, and it will have a lasting impact on online business and culture.

This would appear to be an oxymoronic concept, but I predict this one will stay, if perhaps not with as much currency as terms like the long tail. Shirky's example is a forum devoted to cell phones that gets 250 million page views a year for topics as obscure as modding the firmware in a cell phone, etc.

Sunday, October 29, 2006

Irony and evolvery

On polyglot conspiracy, Lauren observes:

I think the shift to talking about internets rather than The Internet might actually be happening. A couple years after Bush called it the internets, we all laughed, and it became a kind of joke to pluralize: the internets, internets, interwebs. But then just this week [...]


I’d be interested in hearing whether people notice usage of internets in a not-tongue-in-cheek way, or if this still seems to call attention to itself as an ironic formulation.

What do we think?

Elsewhere in the Ist-a-verse

Here'a blog post (note their clever title, which I lift wholesale) that I can practically just to link to and say "Look, English on the march." The city blog Seattlest, which comments wryly on the goings-on-about-town[1] (town here being Seattle, I prolly should not need to add), does a nice roundup of similarly themed blogs from other metropolises, linking to the following:

Austinist, Torontoist, DCist, Parisist, Phillyist, Londonist, Chicagoist, LAist, Gothamist, Bostonist, SFist, and Sampaist.

Although I don't think it's really needed, I'll provide some thin value-add for y'all, to wit:

  • What's the exact semantic of -ist? Perusing the menu of meanings offered by AHD, we might select:
    2. A specialist in a specified art, science, or skill: biologist.
    The art, science, or skill here being "place where I reside," I guess. Or maybe:
    4. One that is characterized by a specified trait or quality: romanticist.
    Wherein the respective authors are characterized by their place of residence. Hmm.
  • Of the blogs in the ist-a-verse, only Seattle's ends in -est; the others all in -ist. What's the rule here? Is it phonological? (Doesn't seem like it.) An orthographical thing? (Maybe Seattlist looks too much like a site you could browse for used tools, jobs, and alternative mating options.)
  • Pronunciation would not appear to be the primary focus of at least one of these. How would you say SFist, anyway?
  • And hey, how about that ist-a-verse term, anyway? A bonus neologism that exercises the -verse morpheme for us.

Should I have named this blog EvolvingEnglishist?

[Update 7 Nov 2006 I emailed the folks at to ask them whence their name. Turns out that the names and were already taken. Dan Gonsiorowski told me of Seattlest "Whatever, we love it. We're the est of the ists." There you go: mostly a commercial issue, only incidentally a linguistic one. Probably not the first time that's happened, eh?

[1] Or as your mother-in-laws and those attorney-generals might say, the going-on-about-towns, haha.

Tuesday, October 24, 2006

developing -ability

Surely a very productive suffix we can slap willy-nilly onto things is -ability. In one of our older topics from work, we talked about (and I quote) availability, manageability, reliability, scalability, and secureability.[1] In its days as a draft, the topic was referred to as the "-abilities topic," as in "Hey, who's writing the -abilities topic?"

Our company style guide (elsewhere mentioned in this post) makes a point of saying that words with the -able suffix (hence the -ability suffix) "take their meaning from the passive sense of the stem verb from which they are formed." Their example is forgettable (forgettability), which they define as "susceptible to, capable of, or worthy of being forgotten," to which they carefully add "... not of forgetting."

Their counterexample is bootable (bootability), which is proscribed because it does not carry this passive sense. Thus a bootable disk is not a disk that's "suspectible to or capable of being booted"; nah, it's a disk that is "capable of booting." So, like, verboten. Welp, Google gets 196,000 hits for bootable disk, including (paging Alanis Morissette!) articles on the Microsoft Web site.

Nonetheless, the definition of -able/-ability is, I think, generally true (the -abilities topic confirms). At least, in my nearly minute of trolling for examples, I find none that obviously do not conform, except the damnable bootability of that disk. (Although: would you say that a particular brand of paint has great paintability or that the surface to which it is applied has that paintability? Hmmm, probably both, depending on what you mean.)

I am thinking about all of this today because I ran across a blog post discussing ... well, I'll just quote and you can see:

When you're designing for Users, you do a usability study. When you're designing for Developers, you need do a a developability study.

This conforms to the definition just fine. I was pretty sure this was a singleton, a one-off coined by the blog author, but nope, you can find several thousand uses already. Moreover, although Random House does not give the word its own entry, it lists it without comment at the bottom of the definition for develop. The OED has no cite for developability, but does have several for developable, starting in 1816.

I don't have the tools for this, but it would be interesting to see whether use of -ability has increased over time; my instincts say yes, but without the numbers, that remains a hunch. But I do opine with some confidence that -ability is a productive suffix that, as I noted at the beginning, we can probably add to many (transitive-ish?) verbs (a song with excellent singability, a program with promising podcastability). Which is to say, you are free to develop your own -abilities.

[1] A source of common discussion is whether to include or drop the -e- in words like manageability. Our company style guide says that you keep the -e- after -ce or -ge (manageability), drop it after -e (scalability). Obviously, this is purely orthography and has nothing to do with the ability to whack the suffix onto a word. Which would be the suffixability of the, um, suffix.

I created this entry with Windows Live Writer!

Monday, October 02, 2006

Didst troubleshoot

One of our folks here sent around a query asking "What's the past of troubleshoot?" That's one of those "No, wait ..." questions. Whichever past you initially come up with -- troubleshot, troubleshooted -- you do a mental double take, because neither of them sounds right. AHD declares troubleshot, not surprisingly, but when's the last time you ever heard someone say that? Confusion seems to be common; a Googlefight reports about a 3:8 ration for troubleshooted:troubleshot. At least 37% of speakers are willing to actually write the former.

There's undoubtedly a name for this phenomenon, and if there isn't, the Language Log folks can come up with one in a jiffy. What's happening is that a word with an uncontroversial past tense (verbs) or plural (nouns) is used in a new context. The new context can be a new definition (e.g. a computer mouse) or with some sort of morphological twist (trouble+shoot). The new context is just sufficiently different to cause speakers to think of the word as new, or at least, to not intuitively connect it to its related form.

Some examples that I've noted here before (I think):

  • Plural of (computer) mouse: mice or mouses? Contemplated at leisure, it's easy to be confident that you know. But there is that "No, wait ..." factor, and Google lists nearly 2 million hits for mouses.
  • In baseball, today a hitter flies out; yesterday he flied out.
  • The past of cast is cast, but the past of podcast and broadcast is very often -casted.
There are many more, not that I can think of any.

Update 10/17/06: Saw the past of to cheerlead in the New Yorker recently: they cheerlead. Again, correct per the stem, but still a "No, wait ..." moment.

This phenomenon really only occurs when the original word has some sort of irregularity to it -- for example, the past of shoot is shot (irregular), not shooted (regular). But in the new context, folks apply the rules for formation of new words, which are overwhelmingly to use regular inflections and declensions and conjugations. New nouns are pluralized by adding -(e)s -- whatever your classics teacher might have told you, the common plural in English of octopus is octopuses. New verbs form the past tense with -d/-ed -- if we make up a new verb to bim, its past tense is going to be bimmed.[1] (A form of not particularly hilarious humor is to apply faux irregular rules to regular verbs, e.g., squeeze-squoze, think-thunk, bring-brang, status-stati, etc.)[2]

There is a certain, mmm, class of people who look down on this type of formation, but I don't see any particular reason why that should be. When little kids do it, we think it's cute, although the more appropriate sentiment might be astonishment at how quickly and thoroughly small children deduce morphological rules. And anyway, did you have a "No, wait ..." moment when you thought about troubleshoot? All right, then.

[1] A pattern that can throw people is a verb whose root contains -ing, like ring, sing, or fling. Make up a verb with -ing (e.g., fring), and a certain number of people will intuitvely use the irregular past. See Pinker in Rules and Words (or equivalent).

[2] A few times these forms have scrabbled their way into acceptance, the commonly cited example always being snuck in place of the (nominally) historically correct sneaked. Also: quit (vs. quitted), knelt (vs. kneeled), drug (vs. dragged).