Who Owns the Internet? You and i Do at the NYT looks at Joseph Turow’s campaign to encourage journalists to, er, decapitate the word “Internet” when used in copy.

Much to my frowniness, the article educated me on the widespread misuse of the capitalization of the word. That’s thanks in part, as the story notes, to Word’s insistent auto-capitalization of “Internet”. That usage reflects dictionary-based references to the word.

This reminds me of a quibble I have with AP style, the basis of the Cinescape style guide, which insists that “website” is more properly written “web site.”

Both judgments are based on reasonable analytic examinations of the meanings, derivations, and uses of the words: “the Internet” is something like a country; it’s not a naturally-occurring environment (like the forest or the ocean), and so forth.

“Web site” is best justified on the grounds that “web page” is the preferred usage, and furthermore, that similar digital constructs such as an “ftp site” are best rendered as two words.

Unfortunately, the capitalization of “internet” strikes me as just wrong, and I can’t recall ever thinking differently. In fact, I still recoil from the usage, which seems to me to broadcast, in big red blinking neon letters, “It is the editorial policy of this publication to promote ignorance and poor usage.”

I think that this is probably because I learned my usage of these words in the context of digital rather than print culture. Digitally oriented writing has long lower-cased even clearly derived acronyms, such as “ftp”, following the unix practice of coining words from acronyms (in this case, ‘file transfer protocol’) and treating the words as verbs or common nouns within the context of a programming command.

That is, an ftp program will perform ftp and may also be named ftp (pronounced “eff-tee-pee”). Similarly, but distinct, a gif is neither a Gif nor a GIF.

Thus, taking a step back, we can see that I’m a linguistic relativist who certainly prefers that definition and grammatical application follow usage, and in this case I prefer the usage of the coining dialect.

Of course, marketing in the computer industry has also employed goofy capitalization (iMac, CorelDRAW!) in ways that are intended to emulate this, but, (ahem) IMHO don’t.

I suspect that the idea of common property is the underlying assumption which drives the practice of coining common nouns and verbs from the acronym-derived names of a given program. Not Xerox but ftp.

As I recall, it’s common on Wintel to use ImpactedCaptalization in the context of crafting a program (naming variables and handlers and so forth); and of course DOS eschewed lower-case altogether for a time. Is it possible that the lower-case world of the unix progeny and the UPPPERCASE or UpperCase world of Wintel are direct expressions of ideology?

That’s a start to understanding what provides my certainty that the internet is L/C; but why do I think “website” is more correct than “web site”?

17 thoughts on “lower-case i

  1. What about “Web site” then? Speaking as a student of such matters, it’s pretty clear that we’re in an early transition period. Bill Walsh, copy editor at The Washington Post and writer of “Lapsing into a Comma,” and theslot.com practically has a conniption over the lower-case “internet.” I think he’s more worried about a slippery slope eventually legitimizing things like “email.”

  2. Capitalization, to me, seems like an anthropomorphizing activity. Words which have attained some immortal characteristics, some essence which gives them some intangible weight are capitalized. Names of people, titles of people, cities and countries of people, beginnings of thought, all these things are capitalized. I deplore (and mock) the ancient habit of capitalizing every other word in titles and dedications. Excessive capitalization devalues the currency, as it were.

    in some respects, i think that the move towards removing all capitals is an attempt to level the playing field, to remove the undue emphasis that capitals place on certain words, whether they deserve emphasis or not.

    whether it works is rather a moot point. lack of capitalization seems invariably to be accompanied by appalling lack of spelling as well…

    I am relatively indiferent on the subject of word jointure or non-jointure. I prefer website, but if all other things are equal (spelling, wit, wisdom, etc.), then I will accept web site. Personally, I prefer internet. I am with Mike on this. I do not yet attribute godly (or Godly) qualities to the connection of myriad networks, no matter how cool…

  3. Paul’s right. Pulling out my slowly disintegrating Tanenbaum (1996), I read: “…A collection of interconnected networks is called an internetwork or just internet.

    “A common form of internet is a collection of LANs connected by a WAN….

    “To avoid confusion, please note that the word ‘internet’ will always be used in this book in the a generic sense. In contrast, the Internet (note uppercase I) means a specific worldwide internet that is widely used to connect [various entities].” (16)

    Note that this is before “commercialization of the web” started going into full gear. Elsewhere Tanenbaum states that he will use two senses of the word “subnet” because it will be clear from context which sense he means. He never says this about the internet/Internet distinction, which means that if we start mixing up the use of internet with the Internet, the distinct terms may be confused in real contexts. Like you said, you prefer to preserve the originating dialect’s usage: use the Internet–here’s a link to a paper by several of the people who defined what the Internet is: http://www.isoc.org/internet/history/brief.shtml. Also, http://www.itrd.gov/fnc/Internet_res.html defines “Internet.”

    As an aside, the whole idea of URIs started out with the concept of different internets being used: between the two slashes after the protocol, the specific internet you were addressing was supposed to be placed– http:/arpanet/foo.domain.net/index.html; ARPANET, THEORY, SERCnet, BITNET or whatever (my first contact with Usenet was through a BBS connected to BITNET, I think, in 87). I don’t remember where I learned this.

  4. As an old-school print editor, I tend to think of the Internet and the Web as proper names and hence capitalized.

    There’s no reason for “Website” or “website” either, if you ask me. Should be two words.

  5. See, Anne, I don’t think that these words are proper nouns or names. “World Wide Web” and “www” (note L/C: see “ftp” above) are distinct, but related things: one’s a collection of documents and communications protocols. The other is a server-naming convention, which is why it’s lower case; I suppose if it’s uses as an unpunctuated acronym a la USA or FBI, WWW is tolerable, but still jarring, to me.

    The original capitalized usage of World Wide Web served, if I understand where my impulses are on this matter, not to signify proper noun status but rather novelty: the Next Big Thing.

    It’s not novel, now.

    The NYT article alludes to this line of reasoning, citing the Telephone and the Phonograph as precursors, I think strongly.

    (There’s an interesting side-thread on this stuff: English has such unclear practices for this because, as I understand it, our grammar is a mishmash of Germanic and Latin practices: in German, hypercompoundwords and Noun Capitalization are Standard Linguistic Practices. Perhaps my preference for “website” reflects this tradition.

    I have some recollection of one Mark Twain penning an essay on just this subject, actually.)

    Paul’s take on the matter is interesting. It inverts my prior hypothesized relationship of Capitalization and property: if it’s a LAN, and can be controlled and owned by a single entity, well then, it’s “an internet”, while if it’s an indefinable, an entity which escapes into the commons, it’s “the Internet.”

    Interestingly, this links up with felicity’s observation. Paul’s preference appears to reflect an appeal to the Universal and the Ideal; the Internet, than, is much like the Chair, or the Infinite. This sounds somehow facetious, it’s not intended to be; I’m genuinely trying to work through the usage.

    I suspect that this is the root of my preference for “the internet;” I don’t buy the idea of a Platonian ideal for networking. Not that that’s specifically what Paul was describing, it’s just the meaning I’ve made of his preference.

    Allan, OTOH (“otoh,” perhaps?) notes the historical evidence for “Internet” as used by the originators of the technology, a reasonable tack, but not persuasive to me, certainly; see Phonograph and Telephone, above.

    I think an additional argument could be constructed on the evidence of Allan’s citations; namely, the originators of the technology were capitalizing “Internet” in reference to the exact sort of limited, specific deployment Paul cites.

    An additional point on this matter is that I strongly discount the need to adhere to a specific originator’s prescriptive intent in adopting the coiner’s word; for me, the dialect is the authority, not the parser or the coiner.

    Here’s an amusing, jokey thought experiment: Bill Quick’s “blogosphere” or “Blogosphere” ? I know my preference on the matter; what’s yours?

    Wow, this comments thread is getting out of hand!

  6. Actually, Mike, I think that Paul and my argument turns on the fact that “internet” and “internets” is a generic term that refers to things networkers create, while the Internet describes a specific internet that evolved out of the ARPANET (the Defense Advanced Research Projects Agency internetwork of universities and military installations), distinct from other internets, such as BITNET and THEORY. As the name of an internet, it ought to be capitalized.

  7. So, Allan, if evolutionary heritance is the key, what about large scale internets whch are isolate from or gated to the public internet, such as the rumored military internet or, even, kinda sorta, the Chinese internet?

    (Other zaibatsu or globocracy maintained proprietary WANs, insert your fave here)

    These are WANs using TCP/IP, presumably, yet they’re also isolate and under specific, finite control. In my thinking, they look like candidates for capitalization specifically because of these qualities.

    (sidenote: were BITNET or THEORY, or, God knows, PLATO an internet if it didn’t use TCP/IP?)

    If I’m following, your preferences and reasoning (and maybe Paul’s) they are probably lower case internets as privately-maintained instances (please forgive my use of “privately”, I believe you follow).

    They might also be captialized Internets because of their proprietary nature. However, your identification of the public internet as a specific network which evolved from ARPANET might exclude them from this usage.

    I think, maybe, we share the view that internets are generic things. I think I diverge from the evolutionary inheritance argument partly because the theoretical isloate internets above can be argued to be evolutionary inheritors of ARPANET as well, although clearly not using the same infrastructure (at least not visibly or unless gated).

    I do grant the chain of reasoning, though, Alan. I think it’s a better argument than the generally cited “it’s like a country” analogy, since, well, it’s not like a country.

    I’m still a lower-case man, though. I suspect the power transmission infrastructure (now upper case that mentally, just for fun) bears me out.

    Maybe, just maybe, I want “the internet” and not “the Internet” in reflection of my perception that the system we use today has become something different from that whch evolved from ARPANET; I think this was covered in my novelty thesis in the initial post, though not in depth.

    And what ABOUT “email” or “e-mail” or, may it never come to pass, “e mail”, “e.mail”, and other such discomforts? I guess I use “email” but don’t have a thesis other than the dash is mildy more difficult to type than not typing it, which is no excuse.

  8. One more thing on Allan’s contributions:

    Perhaps the Internet is a restrcted, historically documented subset of the internet (in my usage) which is constituted of the originating servers and infrastructure at the time of wide educational, but noncommercial implementation; i.e., the end of BITNET et cetera?

    That constituency resolves and harmonizes both of our views, possibly. Thus if you’re at umich.edu or u.washington.edu or suchlike, maybe you are in fact using the Internet.

    (and I just realized I have been misspelling Allan’s name, like, FOREVER. Sorry, sir!)

  9. I love reading you educated cats…

    About the only thing I could possibly add is that it kills me to see acronyms not capitalized, but that’s my mainframe background, replete with years of IBM brainwa-, uh, training, doing the talking.

    And, if you care, in the commercial enterprise world, everybody just calls it a network (sometimes a LAN) now. Interconnections are assumed, whether subnetted, flat, routed, WAN, MAN, or CAN. Traffic — whether it’s email, web, voice, video, storage or whatnot — has all been bent to will of IP, and his dilligent workhorse brother TCP.

    [See? I automatically capitalized all of that without even thinking.]

  10. Ah, but my A. P. style manual sez letter-period-repeat unless its a pronoumced acronym (by “sez” i mean I think that’s what it says, being too lazy right now to lean allll the way over t’other side of the desk right this secont).

    Nonetheless, them usages are the way I’d write that stuff, too, Scott.

    But what’s your position on the burning “ftp” issue?

  11. Ah, but in the IBM mainframe world, we did what IBM told us, not what some fancypants style guide said. Who’s running this show, anyway? IBM or A.P.? Durn right.

    Sad to say, I would actually use FTP if it was a document someone would read, and ftp when I gotsta do it. That would be the unceasing urge to clarify for my beloved readers, versus the lazy-man refusal to capitalize (you know – lifting that right little finger) unnecessarily. Somewhat like telling the kids to clean their room from the middle of the rubbish tip that is my office.

    This, of course, does not address the midWord capitalization issue. For the record, it drives me around the bend unless it’s a frikkin’ variable in the midst of a pile of them. Even then, I just think it’s too precious for words, and I eschew them in favor of manly 8-byte globals.

  12. Mike, you’re thinking too hard. Allan’s right.

    As for ftp, I would go with uncapitalized, as it comes from the glorious land of UNIX (BSD specifically), where everything is lowercase (except, of course, for UNIX and BSD). After all, we don’t write about the VI or the ED editors, do we?

  13. Of course, today while working, it was all I could do to concentrate on getting things done while ftp’ing crap all over the place trying to get the damn network up. This frikking post loomed large. Style guides, indeed.

  14. Mike, concerning “Alan,” so many people do that or “Allen,” and yet I cringe inwardly every time I see it. But you’re forgiven.

    Scott, I am surprised that IBM, a Java and Smalltalk vendor, has not drummed it into your head that intercapping descriptive variable and method names is the Right Thing to do.

    Hm. I would use FTP to describe the protocol, ftp to describe the Unix program implementing the protocol, ftp’ed and ftp’ing to describe having used the program implementing the protocol, and uploaded/downloaded to describe what I did when using the program implementing the protocol. E.g.: Today I uploaded a copy of my C file to the OS X box, grep’ped and sed’ded the thing all to heck and back, then downloaded it using Fetch’s FTP GUI. Bones?

  15. I have no technical reason for my response unlike our techie, Scott. I believe internet should be lowercase. It not only feels right, it is! I feel it fits in with the lower case spelling of television and telephone. Throw in library and school,also.

    Web site – two words.

    My grip is email or e-mail. My work manual on correct usage states “e-mail” and always follow with the word message. That seems to imply it is an adjective? Therefore, I e-mail a message is wrong? I sent an e-mail message is correct. My preference is email and use it as a noun or verb.

Comments are now closed.