Because McCulloch: A Policy Q&A With the Internet’s Linguist-in-Chief 

Since Montrealer, McGill linguistics grad and WIRED Resident Linguist Gretchen McCulloch published Because Internet: Understanding the New Rules of Language last month, the New York Times rave called it “incisive and entertaining”, it hit the Times Bestseller List and Vox ate it upPolicy Magazine Social Media Editor Grace MacDonald spoke with McCulloch about the politics of internet language, how to avoid being a meme victim and why periods are hostile. 



Policy Magazine: You’ve talked about how the greatest influence on an individual’s online language isn’t necessarily age, but what their first social group was. Have you noticed any political stratification along similar lines? What are the major linguistic differences between political affiliations? 

Gretchen McCulloch: I don’t know if it’s an age thing, but it seems like there are politicians who understand how to make something a meme vs becoming a victim of the meme. I think one thing that’s really interesting is a demonstration of an understanding of how to communicate irony in writing, which I think is very important. So if you’re communicating something that has a particular tone, there are ways to do that. 

You can see this, whether it’s politics or brands or other types of organizations on social media; there are people who are still running their corporate Twitter accounts with simple brand updates and then there’s another wave of people doing big mouthpiece accounts, who are doing that by talking to the internet like itself, with the #relatable content or ironic content. Sometimes those are kind of incongruous brands, like the Museum of English Rural Life, which got famous for the “Absolute Unit” tweet. Because that’s not something you expect to see from a museum. There’s nothing controversial about admiring a photo of a sheep if you’re a museum of rural life, but they could have posted “here’s a nice sheep from our archives” and it wouldn’t have gone viral in the same way.

It works, but it only works to a certain extent, because there’s also been a backlash. Because people are saying “Are brands actually our friends?” and “Do we really want brands to talk to us in this ironic voice, or is it too much?” And I think politicians are just getting clued into that. They’re just kind of beginning to say “What if we did do these tweets in a more edgy or relatable sort of way, in a way that spoke to the internet like itself, spoke like an actual person in the tweets?” without seeing what could go wrong. Somebody doing a good job of this right now is Alexandria Ocasio-Cortez, who has this very casual voice, she goes on livestreams and Instagram. The kind of things that aren’t as scripted, just showing up. The kind of things that politicians do try to show off in smaller scale events, but not so much on social media. And then there’s the question of, if you’re a politician, are you running your own social media or are you getting somebody to run it for you? And if you’re getting someone else to run it for you, what kind of voice are they running it in? Brands have been in this space for a number of years now, it’s less surprising if a company account posts something ironic or funny, but it’s still more surprising if a political account does this. In many ways it may be better to have a totally straight or normcore account than it is to be bad at memes. You may not make headlines for the right reasons, but you won’t be making headlines for the wrong reasons. At least you’re not gonna embarrass yourself.

PM: What advice would you have for our readers — who tend to be senior professionals who aren’t necessarily digital natives — in navigating internet language or emoji when communicating with younger colleagues? Common pitfalls to avoid?

GM: I think it’s really useful to be open to having meta level conversations about what particular things mean. Just because you don’t personally intend to be communicating a particular tone of voice, doesn’t mean you’re aren’t communicating something. So it’s good to be aware of what you’re communicating and if you’re inadvertently making people think that you’re feeling hostile towards them, it can be useful to think okay, it’s not just incumbent on the people that you’re supervising to alter their language to suit you. If everyone else is doing something different and you’re not catching up, then you might be the one that’s being left behind. 

In the nineties, a lot of people had to be told that writing in all caps was shouting. And most people know that these days, but one that people don’t know as much is that if you’re sending an informal message like a text message or something that’s very short, and you put periods at the end of those very short messages, that can come across as hostile or passive aggressive. Especially if it’s a positive message. So if you have younger staff members and they’re not using periods then it may be they’re inferring that you’re more angry than you intended from your periods. It’s a big one. 

And from a younger perspective you can say okay, just get used to the fact that these people aren’t intending to be hostile and this isn’t something to be inferred from them. But if I was to predict where we’ll be in the next twenty years, I expect we’ll be using a lot fewer periods. If something’s widely adopted by a younger group of people, that’s a sign that it’s on the upswing.

PM: Your major focus is obviously English, but has internet language changed how we communicate across borders? How important are things like the flag emoji?

GM: The flag emoji are an interesting space because a lot of people use them to stand for languages when of course they stand for countries. So that gets complicated because if you’re going to say English, are you going to use the American flag or the British flag, and of course there are other English speaking countries. And of course there are languages that don’t have national countries at all. There are many indigenous languages that don’t necessarily have a flag, and definitely don’t have a flag in the emoji set. So if you’re going to use the flag emoji as a way of representing countries, it breaks down very quickly if you start thinking about it, but many people don’t.

It’s been interesting to watch how automatic translation tools have gotten better, but still have issues. A couple years ago, people discovered that if you put weird language into Google translate, like “dog” over and over from English to Yoruba, you would get a sort of doomsday clock message. There are a bunch of these. The languages that generated the strangest results like Somali, Hawai’ian, Maori, are often languages that don’t have as much text translated in them. So languages that have say, the entire proceedings of the European Union translated between them, have a large corpus to draw on, whereas for smaller languages the largest corpus may be the Bible, since that’s been translated into so many languages. So if you put really weird stuff into those languages then the output may resemble the training data more, and if the training data is the Bible, then you may end up getting some strange outcomes. 

Not all languages have been written for the same amount of time or have as large of a literary corpus in the first place. Or if you’re using a corpus of newspapers, not all languages have as many newspapers or as much of a written print tradition. Or if you’re thinking of using Wikipedia, not all languages have as much on Wikipedia or even a Wikipedia in their language at all. So all of these things sort of create an internet that’s more welcoming to some languages than others. One way to try to make the tools better is to say, “We’ll release a machine translator for the language pair even though we know it’s not particularly good because then we can get people to evaluate or correct the results, and maybe that’ll help us improve it”. But in the meantime if you have people relying on a tool where they don’t speak the language then they might think that this is actually producing good results when when in fact all it’s producing is weird biblical hallucinations.

PM: Do you think that symbols like emoji or sarcasm markers have potential in official policy materials? For example, written transcripts of speeches that were presented in person with gestures?

GM: I think we might see some limited use of emoji to transcribe gesture, that’s an interesting idea. But in most cases I think they’re unlikely to cross over into formal varieties. Because think about the exclamation mark, for example. It’s part of the standard repertoire of English. Nobody thinks of it as a new or internet thing, it’s been around for quite a long time. And yet if you try to find an exclamation mark in a formal document like an academic paper or diplomatic treaty or even a book, you’ll generally have to go quite far. And it’s not because it’s not somehow standard or not somehow useful. It’s because the exclamation mark indicates something about the attitude of the writer towards the text. And the convention in formal writing is that the attitude of the writer is removed somewhat from the text, that the writer approaches the text with a level of removal, without making their feelings and their attitudes quite so explicit. It’s kind of like how a newscaster on the radio or on television communicates in a sort of narrow emotional range between cheerful and solemn, where they’re trying to report the news and not be the news. Those moments when a newscaster breaks “character”, those are the moments that usually go viral. 

So if you take a formal written context, you also have this expectation that people communicate in a fairly narrow emotional range, and that includes not overtly signalling their emotions with tools like the exclamation mark. If you can’t even get an exclamation mark into a formal paper, what makes you think we can get emoji in there? Because emoji are also the kind of thing that signal the attitude of the writer to the text. It’s not so much the fact of them being little pictures — because you do have pictures in formal text, you have graphs and diagrams and charts and these kinds of things, if you wanted to display pictures you could do that — but this is a particular type of little picture which indicates the intention or attitude of the speaker. Formal writing doesn’t prioritize the intention or attitudes of the speaker and if they do, they couch it in different types of terms, like “one wonders if” or putting it in the footnotes. Formal writing doesn’t even use “I” statements, so you’re trying to disguise the attitudes of the speaker even for things like that. So if you can’t use the pronoun I, which again, no one thinks is not part of the standard repertoire of formal English, if you’re discouraged from even using the pronoun I in formal writing, what makes you think that emoji are likely to follow? It’s a formality barrier, and the thing that emoji are really good at is adding this extra gestural layer, adding this extra cue towards our intentions, which is the kind of thing we need in informal language, but which we deliberately try to strip out of formal language. 



PM: Can you briefly summarize for our readers how something as innocent as a pixelated frog can become a symbol of fear and hate? How else have you seen internet ‘injokes’ specifically turned into a political locus? Aside from British Parliament having to discuss whether an image of an anime girl counts as hate speech.

GM: I don’t think that the repurposing of existing, potentially innocuous symbols for political effect is a uniquely internet thing. The symbols seem sillier now, but the US parties being a donkey and an elephant are repurposed from a political cartoon, and they’re pretty silly animals, to be honest. Or if you think about the abbreviated names for political parties. In Canada you have the Grits from the True Grits a number of decades ago, or the Dippers for the NDP, which sound pretty silly to me. I think memes are in the continuum of political cartoons, which also try to sum up a particular story or narrative or bring people to a particular point of view through arresting imagery and common visual shorthand. 

If you sum up Canada as a beaver, is that any less silly? They get conventionalized and we stop noticing the political symbols or silly iconographic shortforms that we’re used to.

PM: So do you think that in a few years people will take Pepe or Lily as very serious political markers? 

GM: I think that in another generation or two, just like I had to analyze political cartoons in history class, I think kids will be analyzing political memes in their history class. And they’ll probably think it’s boring.

PM: Well, good luck to them. It’s hard enough to keep up now.

You can get more Gretchen McCulloch on her blog, All Things Linguistic, and her podcast, Lingthusiasm. Follow her @GretchenAmcC.