Five critical patch updates for the English language

In which I identify five vulnerabilities that keep causing conversations to crash.

Hello again. For anyone new here, I’m a writer of poetry, downbeat diaries, branding and advertising projects, articles for Creative Review and The Guardian, and books about design. My plan is to send something once a month and make it a substantial read. The broad theme: using language as a way into wider cultural and political issues. The good news is I’m getting away from the word ‘purpose’ this month to talk about something else.

Preamble: Let’s all speak Lojban
I’ve been thinking lately about a language called Lojban. The red and blue symbol above is its logo – it’s good when languages have logos.

Lojban is a constructed language that was first developed by the Logical Language Group in 1987, building on an earlier language called Loglan. The idea was to create a language that eliminates, or at least minimises, ambiguity – in both written and spoken form. So no two words have the same sound with different meanings (like your and you’re). And no single word has multiple meanings (like bank). All spellings are phonetic and the grammar has a consistent internal logic.

Such a language would allow computers and humans to communicate with fewer misunderstandings, so the theory went. But Lojban has also sparked its own lively subculture within humans: people have written Lojban novels and poems.

Since discovering this new language, I’ve often wondered how Twitter would work if we all spoke it. It’s not that Lojban would eliminate all arguments. It’s still possible to say things that are incomplete or vague, or express a point of view with which the other person simply disagrees. But it would reduce the possibility of certain kinds of argument – those that rely on a misunderstanding (wilful or otherwise) of what the other person is saying.

When I see arguments like that playing out on Twitter, I find myself staring at the words on the screen in the same way that I imagine a developer looks at code. As the conversation spirals out of control, I often get the sense that a particular word is ‘glitching’ and causing the conversation to crash. It’s not that language is exacerbating the problem – it is the problem.

Preamble continued: Ambiguity kills

The textbook example of language ‘glitching’ is the case of Derek Bentley, who was convicted and hanged as a ‘joint party’ in the murder of a policeman during a burglary gone wrong in 1952. When the policeman cornered his 16-year-old accomplice (Christopher Craig) and ordered him to hand over the gun, Derek Bentley shouted ‘Let him have it, Chris’. The case hinged on whether you believe he was telling Chris to let the officer have the gun, or colloquially inciting him to shoot. The jury convicted Bentley, but he was posthumously pardoned in 1993.

That case would never have happened in a world that spoke Lojban.

But then a lot of other things wouldn’t happen in a world that spoke Lojban.

I’d struggle to write most of the poems in Realtime Notes (three and a half years of poems that used chronic wordplay as a way to chronicle events). And my job as co-editor on A Smile in the Mind would have been trickier. Given that so much of the work relies on verbal double meanings, the Lojban edition would have been much thinner (to some people’s relief, no doubt).

So rather than recommending we all switch to Lojban, I’m proposing a rolling programme of patch updates on the most serious vulnerabilities that afflict the English language. I’ll list five of them, then reveal the patch at the end.

Vulnerability #1: COMPARE (Disambiguation)

The word ‘compare’ is a culpably negligent piece of verbal design, in that it can mean at least two entirely different things:

- to note limited similarities/contrasts between two things (I’m a writer like Shakespeare but I’m slightly shit compared to him)

- to put on a par with / equate (Nick compares himself to Shakespeare!)

These meanings are so often elided, either wilfully or cluelessly, that the word should be accompanied by a clown emoji in all newspaper headlines and Twitter hot takes.

You can search Google News for the word ‘compares’ and find new examples every day. A few came up while I was writing this article:

Daily Mirror headline: Sarah Ferguson compares herself to Nelson Mandela (implied second meaning of ‘compare’ – to put on a par with)

Actual story: In a longer speech, she said “One day I got up and thought, Mandela forgave his persecutors; surely I can forgive and move forward.” (obvious first meaning of compare – to point out one limited parallel between)

Fox News headline: Stephen Colbert compares the Taliban to Americans in Capitol Riot (implied second meaning of ‘compare’ – suggesting equivalence between)

Actual story: Colbert said “Why should our soldiers be fighting radicals in a civil war in Afghanistan? We’ve got our own on Capitol Hill.” (using the word ‘radicals’ to describe both groups – barely even the first meaning of ‘compare’)

Jewish News headline: MP compares footballers taking knee to Nazi salute (implied second meaning of ‘compare’, suggesting equivalence)

Story: Tory MP makes a clumsy point about mixing sport and politics (an argument that it’s possible to refute without having to deliberately take the worst version of it)

I include the last example because all ‘sides’ do it. People continually slide between the various meanings of compare in whichever way suits them. “Someone compares X to Y” should be a near-meaningless headline. It tells you nothing about whether they’re saying X is wholly different to Y, or pretty much the same. This one little word is a glaring vulnerability that has been generating more heat than light for years.

Vulnerability #2: EXPLAIN / EXCUSE (Disambiguation)

I am borrowing this from US author and political commentator Robert Wright, but I hope he won’t mind as it’s in the spirit of averting the apocalypse.

As with ‘compare’, the ambiguity is built into the word ‘explain’, but in a subtler way.

The key point should be clear enough: to explain something is not to excuse it. For example, if we seek to explain China’s motives in its treatment of the Uyghurs, that is not to excuse those motives. But by seeking to elucidate and pick apart those motives, we might get a better handle on how to view the situation.

But any politician or commentator who tries to ‘explain’ the motives of another party is often accused of seeking to ‘excuse’ them. By providing a rationale for the actions in question, you are seen to be justifying them.

This confusion is more understandable than the ‘compare’ vulnerability. If you walk in on your partner in bed with someone else and he shouts ‘Wait, I can explain!’ then the implication is that the explanation will somehow exonerate him. We often use ‘explain’ in this sense, so a sense of ‘justifying’ is built into the word. But in an ideal Lojban-speaking world, it wouldn’t be. Attempting to see the world from another person’s point of view is an existentially valuable exercise that shouldn’t be shut down by a language error.

Vulnerability #3: APOLOGIST / APOLOGISE (Error in root folder)

This leads directly to the third vulnerability: apologist. Again, I’m getting this from Robert Wright, who points out how any commentator who seeks to ‘explain’ the motives of Putin (for example) will be labelled a Putin apologist. By seeking to understand his motives, you are seen to be justifying them.

This vulnerability can be traced to the Ancient Greek source code. The original Greek sense of apologia is a ‘defence’ and this meaning has persisted through the ages. When Elizabethan poet Philip Sidney published his Apology for Poetry in 1595, he wasn’t saying sorry for all the rhyming – he was mounting a defence of the art form. In that sense, apologia means the opposite of apology. Rather than “I can’t defend it, I was in the wrong”, the original sense of apologia means “I can defend it, I am in the right”.

Thanks to this etymological loophole, anyone constructing a rational analysis of Putin’s motives can be falsely labelled an ‘apologist’ in the Ancient Greek sense of the word, while smuggling in all its more recent overtones of not only defending someone but doing so in a vaguely subservient and ‘apologetic’ way.

Vulnerability #4: SORRY (Unknown value)

This in turn links to the fourth vulnerability, which is a general design flaw in the word ‘Sorry’. What a profoundly dysfunctional word it is.

When we demand that someone say ‘Sorry’, we do it because the word implies an admission of guilt and remorse. Yet in so many situations we use the word ‘Sorry’ to mean nothing like that – it’s simply an expression of sorrow. If you tell someone your mother has died and they say “I’m so sorry,” we don’t assume they’re confessing to a murder – we know they’re using ‘sorry’ in the purely “I’m sad it happened” sense.

This wormhole in language allows politicians to worm their way out of anything.

“I’m very sorry about what happened and any distress caused” – can be enough to answer the demands for an apology without accepting blame.

“I’m very sorry for what happened and any distress caused” – nudges you further up the scale of blame-accepting, but with plenty of room for manoeuvre.

“I’m very sorry if anyone was offended by my words” – the classic formulation that can mean EITHER ‘I’m sorry I said a shitty thing’ OR ‘I’m sad that you’re so batshit crazy as to claim offence at my words’.

From one perspective, it’s strange that language evolved to include such a fuzzy word at the heart of all moral discourse. But even putting it like that makes you realise it’s not that strange at all – maybe it’s useful to have a word that allows disputes to fizzle out in a haze of face-saving ambiguity. In that sense, the word ‘sorry’ may be a feature of the English language, rather than a bug. I have hedged this by calling it an ‘Unknown value’ – more updates to follow.

Vulnerability #5: PENDULUMS AND SLOPES (New analogy plug-in)

Finally, I want to move away from the individual-word-unit flaw to a higher level of language error: bad analogies.

This one could be a whole post in itself, but one thing I notice repeatedly in political arguments is the persistence of analogies that feel true but aren’t when you think about them. One is the slippery slope – the idea that a small step in one direction will inevitably lead to a headlong slide to the worst extreme of that direction. (See also: thin end of the wedge). One thing that rarely gets mentioned is that not all slopes are slippery. In fact, it’s a crucial aspect of life that most slopes aren’t slippery. Almost every law and convention in human history has been constructed on a notional slope, and most of them don’t slide any further.

The second popular analogy is the pendulum theory of progress – the idea that we proceed by dramatic swings between right and left and you actually have to overshoot your target in order to arrive eventually in the right place. My sense is that this analogy runs deep on both sides, but I notice it more in my general tribe of mainly left-leaning people.

I sense many are aware of the excesses of the left – the crazier extremes of what gets called ‘wokeness’ (a word that is much maligned for not having a clear definition, but then it can be equally hard to define other broad terms like conservatism). When you point out these excesses, you feel the background ‘pendulum’ analogy kicking in. Many people agree that certain things are excessive and over-reaching, but (they reason) that’s how progress works. You have to push too far in your favoured direction because that’s the only way you’ll get anywhere. Soon the pendulum will swing back and it’ll be fine.

There’s something seductive about this analogy – particularly the way it relies on a natural and inevitable force like gravity (likewise slippery slope). But it’s also not a great parallel. One thing about pendulums is that they don’t progress anywhere. And if they do finally come to rest in the middle, it’s because they’re broken.

I once heard a better analogy with which to replace it, but it was on a podcast that I’m struggling to relocate. I know it featured the writer, podcaster and musician Coleman Hughes.

Briefly, it relied on the fact that a significant proportion of fatal car accidents (4%) are caused by the driver over-correcting in response to a hazard – whether it’s an animal in the road, or the driver nodding off and drifting to his side of the road, then waking with a start and wildly steering back in the opposite direction, either straight into oncoming traffic or at such a sharp angle that the car overturns on an otherwise empty road. This over-reaction happens more when the driver is tired and/or responding to a mild but sudden danger (bumping the driver side kerb, then oversteering into the opposite lane).

You can see where this is going as an analogy for political and social progress. I fear a lot of excesses are not inevitable over-corrections that naturally even out, but avoidable errors that, uncorrected, prove more dangerous than the original hazard.

The analogy gets particularly apt when you look at the advice for avoiding accidents through over-correction, which is to avoid the common mistake of steering away from the object you see as the danger (which causes you to steer too enthusiastically) and instead focus on steering towards where you want to go. Easing off the gas is also advisable, but not slamming on the brakes.

As an analogy for social progress, I like the idea of steering towards where you want to go. Broadly speaking, I would say both the American constitution and Martin Luther King’s dream are examples of defining a target in order to steer towards it. And that usually involves basing the target on values that will be as morally defensible and desirable in 500 years’ time as they are now. On the dangerously slick road of the early 21st century, I think it’s useful to think about most moral questions in terms of where we want to steer towards.

Postamble: Patch installation notes

So those are five vulnerabilities I’ve been pondering – five words and pieces of language that I believe cause a lot of failed communication in the world.

It would be great if there was a magic patch to solve them – maybe some new coinages or an ingenious piece of software.

I once worked with Google on a project called Sideways Dictionary that aimed to populate the tech world with analogies to explain technological concepts to the public, press and policy-makers. With Morissette-like irony, the website is currently down (having fallen between the two stools of project partners Google and the Washington Post). But the image above gives an idea of how the Chrome plug-in worked – highlighting techy words in any article and suggesting everyday analogies. It would be tempting to do a similar project for words like compare, explain, apologist, sorry, pendulum and slippery slope.

But however ingenious the plug-in, language ultimately doesn’t work that neatly. The only way to change it is to add your voice to the system and hope it connects with a few nodes elsewhere in the network. So for now, this post is the patch – and hopefully you’ve just installed it by reading it.

Thanks for subscribing – please like, share, comment, send flowers.

PS: I’d be interested to hear new examples of words like ‘compare’ causing mischief, or if you have identified any similar glitches in this software we’re all running.