English is a weird language. It is globally known, and to various extents, are spoken to varying proficiencies. With it, also comes its mingling with other languages, and more interestingly, geographical isolation. English has become an ubiquitous language — spoken almost everywhere on Earth, probably except the uninhabited islands that dot the Earth’s crust. The far extremities of human settlement, like Tristan da Cunha and St Helena, are so isolated from the rest of the world, that they have developed their own English accents or even dialects. But observing the creation of an accent has never been empirically observed.

Enter Antarctica. A continent in the deep south, in a landscape so frigid and arid, and where barely any permanent settlements exist. Research bases are maintained year-round, where scientists conduct sampling or run experiments in various fields in this icy continent. As close as it is to countries like Argentina, Chile and New Zealand, Antarctica is essentially cut off from even these countries for most of the year. The extreme winter months make it impossible for aircraft to safely land and take off, leaving the summer months the only time period where people can fly in and out of the continent.

Scientists and other staff who live on the continent during the winter months mostly only have the people within the research base to communicate, and pretty much that is it. Required continued maintenance (and survival) of the research base would mean that these people would have to cooperate, and communicate with one another. Perhaps this is one of the situations where a new accent or a dialect could arise.

And to Harrington et al., that was what they tried to assess, and their findings published in The Journal of the Acoustical Society of America in 2019. Here, the main vowels they focused on were the /ɪː, u, ou/ vowels, from which they formulated a word list for the participants to pronounce. The other sound they also focused on was the /ju/ sound, which in some varieties of English, is found in words like “few”. With this, they aimed to see if there were any changes in how these sounds were articulated in the participants living on Antarctica over the winter — the winterers.

Before going onto how these changes in articulations were assessed, we should introduce the winterers. There were only 11 participants, with mixed backgrounds of English accents, with most of them from the UK, one from the Pacific Northwest of the US, one with Icelandic as a first language, and one with German.

The researchers expected the assessed sounds to be articulated further front of the mouth should there be a change, a fronting of sorts. To see if there was indeed such a change, the researchers first took a baseline measurement from the participants, which was conducted before their deployment to the Antarctic. This meant that the baseline measurements of each participant was done in relative isolation from other participants, and their accents would be more influenced by their background or place of residence. And on Antarctica, another four recording sessions were done, at six-week intervals. If a word selected from the predefined word list was pronounced wrongly by more that one participant, that word would be excluded from the study. This gave the researchers a list of 28 words to analyse.

With these recordings, the researchers had to analyse the sounds through the use of formants. These are basically the frequencies where the human vocal tract resonates. These can be visualised as spectrograms for individual morphemes or sounds, while formant plots are used to visualise and compare a bunch of vowels. The latter does so on two axes, using the lowest frequency F1, and second lowest F2. This is widely regarded as the most important formants used in distinguishing vowel quality. And with agent-based modeling, the researchers assessed if there was indeed a vowel change in the Antarctic winterers.

And here is what they found. During the participants’ time in Antarctica, the researchers found that there was indeed vowel fronting in the /ou/ sound, but negligible change in the /ɪː/ sound. Another finding was that the participants sounded more similar to one another for those sounds, as the differences recorded between the participants in Antarctica were noticeably lower than that before their time in Antarctica. They also noted particular factors that could contribute to this change, including background, socialising, and various demographic factors. With more factors not included in the model, it is possible that these could have contributed to vowel change in some form. Remember the German speaker? They would have demonstrated a vowel shift as they learned English and became a more native-like speaker of English through practice. As such, language learning could have been an unmeasured factor here.

But is this really an observation of the Antarctic accent in real time? The authors pointed out not quite. After all, only 11 participants were assessed, a small study population selected from an already small cohort of winterers to begin with. Furthermore, this study likely only took place in one research station run by the British Antarctic Survey — several more research stations run by the BAS exist, and other stations like those run by the United States Antarctic Program. If all of these stations were staffed or manned during the winter months, it would be likely that each station could have had their own phonetic changes occurring, be it by demographics (English accents or linguistic background), social factors, randomness, or any other reason. As a result, the authors noted the rather limited generalisability of this study, as it cannot really represent the wintering population of Antarctica. Nevertheless, the authors concluded that they observed the very early stages of accent formation in Antarctica, just within a subset of a wintering population, in one research station among several others dispersed throughout the continent. As such, calling this the Antarctic accent would be a little too misrepresentative.

Perhaps if more phonetic studies were conducted in other research stations, linguists could sooner or later put together a map containing sound changes in certain English sounds, and properly ascertain if a rise of the Antarctic English varieties is indeed happening, with each variant being different by research station.

Overall, I find it a fascinating article as this is probably one of the first, if not, the first study actually tracking changes in English sounds in real time in Antarctica. With Antarctica being known for its remoteness and isolation, observing such changes could shed light (empirically, at least) on how accents form, something that we have observed in Englishes all over the world, but never quite tracking the early stages of it. Extending this over time, perhaps a “sound change log” could be established, and track the development of accents in real time. Of course, this would require more resources, made even more difficult considering the challenges of logistics in managing research projects primarily based in Antarctica. But nevertheless, this paper does indeed bring to light the conditions and factors that have to be considered for the birth of a new accent.