Chat got your tongue?: How does one regain their voice amongst the victimisation of the em-dash?

ARTS | ANAMATA / THE FUTURE

Written & illustrated by Stella Roper (they/she) | @stellyvision / @dodofrenzy | Arts Editor

There is an unmatched, twisted and complex nature contained in the English language, with its cruel rules, ifs and buts and “i’s before e’s” – you’d wonder how any of the literary greats got anything done. 

I think back to high school, in moments looking over reading materials in class, and remember having thought that the author's writing was leagues beyond me; it didn’t feel as if we were writing in the same language. Yes, all …well, at that time, most of the individual words I knew, and yet, were arranged in a way which read in such a beautifully unfamiliar flow. It was English class that exposed me to the lengths to which words can be manipulated by the human mind, and back then, that introduction to the limitless bounds of what could be crafted, read in these writings, felt so distant from the restrictive sentence structure templates and writing techniques I had learnt.

Despite those initial difficulties that all budding writers face, it’s only once all the specificities are somewhat understood that individual creativity begins to flourish and take hold. To recognise that the methods I was taught in high school were not a strict mandate, but instead, served as guideposts and suggestions, put in place to provide a supportive scaffolding for emerging writers like myself. All that is to say, once the groundwork is laid,  the development of one's authentic voice in textual form is bound to thrive. Consuming work from different people; established authors, classmates, and me from 5 years ago, all undoubtedly assisted in my understanding of what a voice, angle and bias are, as I began to recognise and reflect on my own. When I sent my first contributions to Debate in 2022, it was lovely to hear the thoughts about my work from friends and family, about how “this sounds just like you!" The first time I heard that, it was flattering, of course, to know that I was doing myself justice to how I would speak on a subject; however, it also made me wonder – “what does my voice sound and read like?” and “how did I achieve this?”.

Four years since then, being on the Debate editorial team for half of that time, I think I’ve found the answers to my rabbit hole of questions. A writer’s voice sounds like nothing else – it’s inherently, incomparably, invaluably theirs. Consciously or not, there will be those with voices that you find yourself drawn to consume, whether it’s a niche philosopher I’m out of the loop with, an acclaimed academic, an influencer or a politician. Establishing a voice, present in words, spoken or written, is special. It is one of those unique things that make you human, even in this century, that your voice can transcend physical boundaries and language. 

However dun, dun dun, the world has changed, and will continue to do so at an alarming pace, with innovations implementing artificial intelligence in every facet of our lives, automating education and providing shortcuts to learning (last time I checked, it’s important for humans to learn things). But at this stage in media coverage, and in Debate in particular, I can guarantee you’ve heard about the dangers of AI over and over again. If you’re not in the loop, don’t worry! Just search: “What are the dangers of AI?” in Google, and consume from the AI interface in their system, which simplifies (and sometimes straight up makes up) information into a handy, bite sized answer.

Now we are at a stage where even the AI checker sites have become as undesirable as the AI-assisted writing software itself, with UoA trusting (human) professors over re-subscribing to the programmes. In an interview with RNZ, a president of the Massey Tertiary Education Union branch; Dr Angela Feekery, stated: "I've been teaching for 25 years. I've been marking student writing for years. I know what it looks like, and it's not what they are submitting now. In many of the cases, when you've got students who can write better than I can in first year, there is an issue." 

In a way, I can agree with Dr Feekery. As an editor, I recognise contributors writing styles and the natural emotions that come through, and have had moments where I question whether there may have been AI-assistance. However, morals, ethics and journalistic integrity aside, it’s never a conversation of “you did this wrong”, but to communicate how important the presence of their voice is within the piece. It’s counterintuitive to steamroll over your authentic imperfections with generic phrasing, all for the sake of ensuring you’re understood, when it’s not even your words. When you go beyond auto correcting grammar, and ask an AI assisted programme to “make this paragraph simpler” or “easier to read”, it’s not helping, but instead delegitimises your impact and lessens reader appeal. Once “edited” by the programme, your voice and work are set to become another victim of AI, assisting in stripping down that authentic sense felt while reading it previously, which you, after the changes, may be led to assume are forgettable imperfections.  

However, it’s not like AI turns your emotions into 1s and 0s, it just has a particular way about rewording text for what it deems to be maximum coherency. That being said, if everyone is using it, will distinct voices and their writing styles be replaced by a generic, universal blob? Will the way in which people organically communicate be tainted by the AI-appropriated adaptation of the human language? 

Well, for better or worse, it seems that this is already happening. One study, published in scientific journal Acta Psychologica this September, “explored the impact of AI-powered speaking tasks on EFL (English as a foreign language) learners' speaking performance and anxiety”, which resulted in positive results, indicating how AI could prove to be a “powerful mediational artifact” within Activity Theory. Additionally, in the last few months, publications like NewsWeek, Verve and Vice (which is basically ripped off of the Verve piece), have jumped to write their uncreatively titled pieces on “How AI is changing/shaping the way we speak”, and without ripping them all off myself, they’re right. Words which weren’t as commonly said pre-AI are having a comeback, while others have become outlawed, labelled “AI speak”. 

As my first witness… I would like to call the em-dash to the stand, your honor.

If you haven’t heard already, yes, even my beloved (—) is on the AI-accusatory blacklist. While I admit my confusion between the single dash versus the em-dash only faded once I started university, that fact shouldn’t devalue my passion for the punctuation. The em-dash was created in the 15th century, which last I checked, was way earlier than the introduction of Chat GPT. Fellow em-dash users should not be at fault for the AI-assisted writing software appropriating the overuse of this device, despite it already being rebranded as an unsafe option for avoiding accusations of AI assistance.

With certain words, sentence structures, and now punctuation on the chopping block, I am reminded of how artists with styles stolen by AI programmes were also given the same treatment. While I might feel a moment of hesitation before I use an em-dash in my writing, I can recognise that AI cannot replicate the unique experiences and people which have formed my background, perspectives, biases and personality – unless you use a hell of a good prompt. As users continue to point out patterns of what many sites dub as  “GPT-isms”, perhaps AI will continue to be finetuned, in attempts to perfect human mimicry. Whatever the future does hold in this space, consider placing creativity above coherency, and build on your own voice and abilities, rather than giving data toward an attempted replication. 

Artificial Intelligence has its own voice too – and it’s built by stripping yours.