Technology and language make an unlikely couple. The 1s and 0s on our screens promote peculiar and fascinating developments within our ever-changing languages. Before the 1s and 0s, technology was connecting cultures and promoting new words and styles of communication for ages.
Among the fascinating developments is the emergence of technology-related words. Like all things, technology has facilitated the coining of new words because of simple necessity. These new words are often portmanteaus of existing words. Take the now well-known ‘vlog’: a portmanteau of video and blog, its meaning lying therein. ‘Blog’ is also a shortening of another portmanteau of the early 90s: weblog, making vlog a double portmanteau. Some other recent portmanteaus include infomania (information + mania, as in maniac) and clickbait (from click and bait).
Naturally, portmanteaus are not the only emerging words. Online, thanks to social media, there is an increasing desire to communicate with the least amount of words possible. People use emojis as a form of pictorial communication or abbreviations to shorten the time needed to write a word. This “regressive tendency,” as the IJSTM describes it, harkens back to the communication of ancient societies.
On top of this, technological advancements have historically been a way for cultures to contact and communicate with each other. This is what causes loanwords, a borrowed word from a foreign language. English is no stranger to loanwords, with almost 29% of English coming from French, but recently more international words have infiltrated the language. These include cheetah, jungle, karma, avatar, and more. These seemingly unsuspecting words are all from Hindi or Sanskrit.
Of course, technology does not just spur the use of more words but allows for intercultural communication. Translation devices and machines are all very common in modern society, with the likes of Google Translate exceeding 1 billion active monthly users, 95% being outside of the U.S. The rise of globalization and the international desire for traveling has exponentially increased the use of these devices.
In a perfect world, these devices would take the words from an input language and spit out the exact meaning and context perfectly in the output language. This does not happen and it is all too common to hear about errors these devices make. Yet, these devices are still important to our current global linguistic climate. Intercultural understanding has never been greater, and the benefits that businesses and travel-obsessed people get are numerous.
Nevertheless, current translator devices usually only offer services for just under 100 languages, out of the 7,000 spoken languages in the world. While businesses and individuals can benefit from them, a machine translator handling medical or political affairs promotes more likely mistakes. There was already an incident where Chinese President Xi Jinping’s name was mistranslated as a curse word on Facebook. In 2017, the University of California-San Francisco noted that based on a Google machine learning algorithm, accuracy between English and Spanish in the medical environment was 92%, while English to Chinese was 81%. This unreliability will be an enormous issue for patients and doctors alike. Translation devices offer hope for an opportunity for interlingual understanding, but time will tell if they are to achieve this goal.
Intercultural understanding is not just the goal of machine translators but has also become an aspiration for many people, so language apps cater to those needs. Language apps run on the promise of efficient and easy language learning and offer an array of customization options that serve each individual. These options can range from a ‘slang slider’ like that presented in TripLingo, an option that controls the formality of the phrase or word learned, or just simple vocabulary practice. The more successful and popular apps usually feature some gamification of language learning to add a fun factor, and there is none more prominent than Duolingo in this aspect.
Duolingo needs no overview. It not only has its courses reviewed by native speakers, but its courses are also written by them. This allows for the so-called quirky sentences that appear in their courses and acts as a way to keep the courses accurate.
From one popular language app to another, Rosetta Stone features its ‘TruAccent’ technology that “provides real-time pronunciation feedback,” according to their website. ‘TruAccent’ is a speech recognition system that tracks your voice and compares it to that of a native to critique it. If popular apps like Rosetta Stone use a speech recognition system as one of their major selling points, are speech recognition systems really that effective?
Well, IBM defines speech recognition as the transcription of verbal speech to a text format. The accuracy of speech recognition is partly determined by its Word Error Rate (or WER for short) and speed. Word errors usually result from any amalgamation of pronunciation, pitch, volume, accents, or loud background noise, among other factors.
Most speech recognition algorithms are run by natural language processing or NLP. Most likely, your phone has an NLP speech recognition system installed, like Siri or other phone assistants. NLPs are naturally designed to alleviate the process of speech recognition systems. They disambiguate words that may have multiple meanings within a context, as well as disambiguate the part of speech and grammar of a text. On top of phone assistants, Google Translate uses NLP as well.
Even through all these efforts to increase the accuracy of these systems, Rosetta Stone’s TruAccent is certainly not perfect. Speech recognition technology is a breakthrough in language translation and transcription, but its accuracy has been flawed from inception.
The peculiar pair of technology and language goes from new emerging words to upcoming translation devices, bustling language mobile apps, speech recognition, and NLPs. Technology has always impacted language and will continue to have one for the foreseeable future.