Talking to me? Getting language right matters, even for bots

By Hugh Cadman, Specialist Content Creator

I was brought up to regard America as brash. US slang was banned and my parents were so traditional I had to play football in a tweed jacket (all right, I made that bit up). But when punk rock erupted, I loved it. British youth culture started expressing itself in British accents, singing (shouting actually) about what mattered here, not over in the US. The Clash summed it up: “I’m so bored with the USA.”

Image credit: Etsy

US tech and the force for good

Now of course, I readily acknowledge the dominant role of the US in changing everyone’s lives (mostly for the better) through technology. But the language problem remains. As a content creator in B2B technology PR you quickly learn to de-Americanize (yes -ize as recommended by the Oxford English Dictionary) copy to make it relevant to the UK audience. It is not a trivial matter. While Britons can understand most American English, any heavily North Atlantic text gives the impression of remoteness – that the company responsible for the copy doesn’t really understand the UK market or regards potential British customers as an afterthought.

Where do voice bots feature in language battles?

This is, however, not another gripe about tech jargon. What intrigues me is where developments in Natural Language Processing, virtual assistants and voice-bots will take us in the journey across world English.

According to a Zion Market Research report published earlier this year, the global intelligent virtual assistant market accounted for US$2.3 billion in 2018 and is expected to reach US$19.6 billion globally by 2025, at a CAGR of 35 per cent between 2019 and 2025.

The global sale of smart speakers in 2018 was about 98 million units and is expected to reach 164 million units this year. Growth will be further driven by the autonomous vehicles market, using artificial intelligence and smart voice assistance to make the experience more reliable. In Europe, Germany and UK are projected to be the prime revenue contributors.

What kind of English?

But what sort of language will these assistants speak and who will do the talking? While voice bots will of course be in the native language or languages outside the non-Anglophone sphere, what kind of English will be used within it?

The tech giants who deploy voice bots will doubtless employ native speakers. (A recent BBC radio programme focused on this topic, presented by Jon Briggs – the first voice employed by Apple to present of Siri in the UK) But we can expect to hear American grammar and idioms wherever we go outside the UK. Does it really matter when billions around the globe are completely steeped in US culture through streaming services from Netflix to YouTube? We all know language is a living entity and it is not the job of business preserve its purity. We just need to effective communicators.

Do we need to learn languages anymore?

Similarly, does anyone really need to learn a foreign language anymore? Google Assistant, Amazon Alexa (powered by Microsoft Translator) and Siri all emerged pretty well from complex translation tests this year. Google Assistant will soon be able to act as a real-life translator in 27 different languages as it obtains a new “interpreter mode” that can translate in real time, enabling conversations with someone who speaks a different language.

Is there any point in the hand-wringing that followed a BBC investigation which found declines of between 30 and 50 per cent since 2013 in the numbers taking language GCSEs in some areas of England? Perhaps nobody can stop these trends any more than they halt the tides, because they arise from collective appreciation of how the world is developing.

We still need to pay full attention to the nuances of language

In time, the AI family of technologies behind voice assistants and bots will be capable of picking up and using all the tics, nuances and idiosyncrasies of spoken language. But we are not there yet, and we are not even there yet with the use of NLP to produce written text. In fact, whatever the method, whether through automation or painstaking keyboard-bashing and hand-crafting, we should all concentrate on using the most effective, relevant, direct and respectful language possible when communicating – especially in tech PR.