Picture a proofing scene. We’re wielding red pens and knee deep in magazine pages for our client Arm. A typical conversation might go a little like this:
“So we’re doing ‘the Internet of Things’, even when abbreviated? ‘The IoT’?”
“That’s right – but not if it’s an adjective”
“What, like – ‘this fridge is Internet of Things-y?’”
“No, like – ‘talking to Arm’s IoT expert we discovered…’ or ‘IoT device’”
‘Shouldn’t that be ‘IoT-enabled device?’ though”
“Just treat it like you would ‘internet’”
“But do we capitalise ‘internet’ though?”
“It’s not a proper noun…is it?”
“Although it is a place…sort of?”
“Is it a thing?”
“What do we mean by thing? – it exists I suppose”
“Wait…does it?
“Do we…?”
We’ll be honest, more than once a tech-induced grammar tangle has threatened to turn full-blown existential crisis.
An avalanche of fresh terms, phrases and acronyms descend almost daily from the mountain of tech, at a rate that’s making us grammar pedants a little queasy. Before our eyes, acronyms become nouns become verbs. It’s on-the-fly grammar, spawned from a need to talk in new ways. The copywriter lives and breathes consistency, but language couldn’t care less about our style guides. As swift changes in the tech landscape impact our habits and methods of communication, it’s starting to feel like English is on fast-forward. Linguist David Crystal backs my hunch, explaining for the BBC: “Language itself changes slowly but the internet has sped up the process of those changes so you notice them more quickly”.
We know words come and go. The Oxford Dictionary annually adds to its ranks (selfie, to google, ransomware, mansplain) while other words are quietly, and tragically, sent packing (scobberlotcher, glebe, purfle). It can be harder to remember that the so-called ‘rules’ – spelling and grammar – shift under our feet too. Dictionaries project a picture of any given language as an almighty list of words. But, there’s nothing like witnessing language growth in double-time to set us right. We’re reminded: language is an activity, a living communication tool – forged in use, recorded later.
The linguistics society of America identifies two key causes of change in a language:
First, “it changes because the needs of its speakers change”. Our habits and routines are drastically different thanks to tech. Bona fide internet philosopher, Lucian Floridi goes so far as to describe humankind as being in the midst of a ‘fourth revolution’ in terms of understanding ourselves in relation to our environment. Should we be surprised if language is running to keep up?
We cut corners with favourite terms. The most common words are often the most irregular –chipped away at with usage, streamlined for purpose. ‘Had’ should be ‘haved’, ‘Made’ should be ‘maked’, and ‘to google’ is a grammatical nightmare, if we remember Google’s a name. The slippery grammar of tech terms is heightened by overhandling. No one’s got time to say ‘to look up on Google’, when it’s happening dozens of times a day.
Doom-mongers decry the English language will have gone full American by 2120.
A grammatical mistake, or misspelling, is only wrong for so long as the majority says so. Crystal claims that in 2006, there were a few hundred instances of the misspelling of ‘rubarb’ on Google, by 2010, 100,000; at his time of writing there were 750,000 uses of ‘rubarb’ (incidentally, the original medieval spelling). As Crystal puts it, ‘people are voting with their fingers’, and the misspelling is on its way to becoming a standard alternative. If language, to run with the linguistics society’s second point, “changes through the meeting of people with different languages/dialect”, then the internet has upped the frequency of these meetings a thousand-fold.
The online archive and search engine algorithms allow misspellings to accumulate and languages to merge. Doom-mongers decry the English language will have gone full American by 2120. And even here, working on international copy, we’ll sheepishly confess that ‘color’ trips off the keys as easily as ‘colour’ these days.
Plus, with IoT (the IoT? I genuinely don’t know anymore) the scope of ‘people with different languages/ dialect’ widens. Some parents are worried that home assistants like Alexa, functioning best as they do on simple commands with basic syntax, are teaching their kids to be rude. Rude and grammatically uncreative. Imagine how language games might continue to adapt to incorporate their newest, robotic, players. (That’s if the new players don’t decide to throw us off the team altogether). With IoT and AI there’s a dizzying multitude of ways we can imagine tech might amplify the effects of causes (1) and (2) – tip of the iceberg is an understatement.
In 2004, Wired released an article entitled ‘It’s just the ‘internet’ now’ declaring – ‘effective with this sentence, Wired News will no longer capitalize the “I” in internet.’ In the opinion of Toby Long, the mag’s copy chief, the internet had by then achieved the kind of ubiquity where it made no sense to capitalise what is simply another communication channel – See radio, or television (small ‘r’, small ‘t’). Decapitalizing ‘internet’ meant ‘giving the medium its proper due’. But, when Wired was bought by Conde Nast two years later, internet became Internet again. The flip-flopping i/I debate even has its own Wikipedia page.
There is a technical distinction, it turns out, between the Internet, and internets (a computer network with smaller networks within it), but common use has – for the most part – ground the capital letter away. It happens. We’re streamlining for use again. Think of it as the hoover or the frisbee effect. Remind yourself that ‘scuba’ is an acronym.
Why do we care? Perhaps it’s something to do with how these debates boil down to discussions about (in this case) the role of the internet from within the context of our own lives. When say ‘common usage’ we’re talking about how a word fits into our way of living. Language change is a game of adaptability, directly reflecting new habits. When we’re talking about changing words, we’re forced to think about our own changing lives. Grammar never got so philosophical.
“Two translators won’t even agree on whether [something] was translated properly or not. Language is kind of the Wild West, in terms of data.”
So, the wheels of language change may be greased by technology, but human behaviour stays in the driving seat. Adapting words to new contexts, creating and playing with new languages, making shortcuts, breaking the rules. It’s what humans do. It’s a byproduct of living. Apps like Grammarly are evidence of technology attempting to keep pace with us, rather than vice versa. Using machine learning to learn from language inputs, the more Grammarly’s used, the better it gets. But as of yet, it still misses the trick. Google translate runs up against the same hurdles.
Machine learning feeds on data. Data processing relies on binary labels: right, or wrong. It ultimately comes down to pre-ordained rules – syntactics, not semantics. The ideal scenario for machine learning involves clear parameters of success or failure. Data science officer Michael Housman explains how language’s adaptability – entangled as it is in human life – is exactly what makes Google Translate and Grammarly fall short:
“Language is almost the opposite of that. There aren’t clearly-cut and defined rules. The conversation can go in an infinite number of different directions. [for machine learning] You need labeled data. You need to tell the machine to do it right or wrong. Two translators won’t even agree on whether [something] was translated properly or not. Language is kind of the Wild West, in terms of data.”
The flexibility of our language, the ways in which we constantly bend the rules to better communicate, may create a change that’s exacerbated or amplified by technology, but this change is necessarily sparked from the very traits that distinguish us from tech in the first place. Arguing over ‘internet’ vs. ‘Internet’ may turn out to be what makes us human after all.
In the hands of our robot overlords, there’s hope for the copywriters yet.
+