Column – In Love with Language

By Dianne Loyet

Why Can’t I Haz Grammar?

As a future crazy cat lady, I can’t get enough of the felines of the Internet. Cat memes, cat videos, cat comic strips—you post a link to it, I click on it.

Although I enjoy Internet kitties, I’ve long had one question: Why do the posters of cat memes think that cats cannot communicate in grammatical English?

Well, it turns out that the cats in the memes may not be speaking ungrammatical English, per se; they may be speaking a new dialect.

The language often used in cat memes is sometimes referred to as ‘Lolcat’. Apparently, Lolcat originated on MeowChat Village (not to be mistaken for the MeowChat app!) in the 90s. On MeowChat Village, users could role-play as cats, among other activities. Some users insisted that cats would speak in dignified English, as shown in this quote from Christa Keith’s 2007 article on SFGate:

“We are highly intelligent. Some of us have difficulty spelling because our paws don’t fit the human keyboard or because we telepathically send our posts to our humans, who type them for us, and something gets lost in the translation.”

Other users, however, delighted in voicing their little kitty avatars much differently. This language is referred to as “MeowChat.” Keith provides this quote from the MeowChat historian’s story of the First Thanksgiving:

“we know dat da pillbugs camed here frum da old comfy, but we not know why and how, aftpurr dey landed at plywood rock, did dey get to meowchatvillage?”

MeowChat also appeared as the text of cat memes posted on 4Chan in in the early 2000s. As Lev Grossman recounted in a 2008 Time article, “there used to be a tradition on 4chan that every Saturday people would post pictures of cats. It was called Caturday. People added captions representing what the cat would say if cats could talk. One day somebody posted a shot of a fat gray cat looking at the camera and saying, “I CAN HAS CHEEZBURGER?” It seems that this viral explosion of cat memes spread the language far and wide, and it has since been referred to as Lolcat.

As Anil Dash described in a 2007 post on his blog, one way you can recognize that Lolcat is in fact a dialect is that you can tell incorrect utterances when you see them. As an example, Dash provided the following examples of unsuccessful attempts to write about Dune in lolcat: “The spice must flow,” and “I are dunecat I controls the spice I controls the universe.” “The spice must flow,” is wrong because it has no differences from standard English in morphology, syntax, or spelling. “I are Dunecat,” and, “I controls the spice,” differ morphologically from standard English but are not consistent; in lolcat, the first person singular of the copula appears to be “is,” so in lolcat, the text of the meme should read, “I is Dunecat.”

Dash goes on to speculate about whether Lolcat is a pidgin or a creole. My own conclusion is that it is neither. Pidgins arise when two groups of people who have no common languages create a rudimentary language with some elements from two or more of their native languages; creoles are spoken by the descendants of pidgin speakers and are linguistically distinct (although similar) to the pidgin spoken by their parents and the native languages of both their parents and the other group. Lolcat doesn’t involve two actual groups, considering that the cats aren’t really talking, so it is not a pidgin or creole.

I concur with Mark Liebermann’s 2007 Language Log post that Lolcat is, or at least began as, a form of pet caretaker talk or kitty baby talk. As such it has any number of simplifications, such as fewer person-number distinctions in verb forms. As the MeowChat excerpt shows, it has mispronunciations, for example, ‘pilbugs’ for ‘pilgrims’. One distinct feature of Lolcat is that it is written more than it is spoken. The fact that it’s written makes a whole other set of features possible. Many words, for example, are standard text abbreviations, such as “Y” for “why”. Others, such as “cheezeburger” for “cheeseburger,” mimic common spelling errors among beginning writers of English.

Lauren Gawne and Jill Vaughan conclude in their 2011 presentation at the Australian Linguistic Society conference, “I can haz language play?” that Lolcat is language play—an activity rather than a discreet entity. They propose other sources for some of the features mentioned above, such as the mistakes that appear to mimic the writing of language learners. Gawne and Vaughan also provide impressively detailed explanations of the likely sources of many Lolcat features, such as misspellings. One drawback of their research is that it is limited to one Lolcat text, The LOLcat Bible. This Bible was written by multiple authors, and final drafts were decided upon by consensus. However, this one text and the number of authors involved can’t possibly compare to the number of extant Lolcat texts and individual Lolcat text producers. Gawne and Vaughan based their research on this Bible because of their own constraints. Analysis of Lolcat based on a larger more representative corpus of authentic Lolcats might result in different conclusions.

The impossibility of severing language from interaction is fundamental to post-modern linguistics. Nevertheless, Gawne and Vaughan’s focus on Lolcat as activity is too narrow. Although Lolcat users are clearly engaged in linguistic interaction, the aggregate of their interactions is a rule-based system which includes aspects of discourse, syntax, morphology, phonology and orthography. To a certain extent the system is consistent with the intuition of a typical native speaker of English. To focus on Lolcat as play would be the same as focusing on writing to the exclusion of text.

If you would like to try out Lolcat for yourself, several websites purport to translate English into Lolcat. I preferred which provided me with the following translation: