In this blog, I promised stories about the history of language technology, and the promise still stands: stories will come. But the things that happen in science and technology do not happen in themselves or by themselves, and when you tell stories about events in technology, it helps if you are conscious about how and why they happen.
Humankind created amazing things since the dawn of history, which brought about unprecedented wealth and an almost twofold increase in the life expectancy of those lucky enough to be born in the so-called “developed” world. This makes us talk about the development – or progress – of science and technology, and the improvement of the quality of life along with it.
Today’s world takes progress for granted, and it’s generally viewed as a desirable course of human life. This wasn’t always like this – it’s legitimate to ask if progress exists, and if it does, what its nature is. Some historians even talk about a progress trap, which implies that progress has limits.
We are continually creating new things that we expect, in some way, to improve the quality of life of those who use it. Apart from the very basic items of sustenance, the newer a product is, the easier it is to sell it on today’s global market.
When someone releases a new technology product, the market will demand a so-called unique selling point (or ‘unique selling proposition’), something new that none of the other, similar products have. (Marketing gurus refine this and introduce ‘positioning’, which means it is enough to provide the novelty to the group of people you want to sell to.) When this novelty is finally created and it gets to the people, we talk about innovation.
Humankind craves and dreads innovation at the same time. For example, I, as a customer, am really happy that finally I found a simple and affordable widget that tracks my exercise and my sleep. On the other hand, I, as a co-owner of a software company, am frightened that in secret, someone else is working on a new thing that will make my product obsolete and worthless in a moment, and put me out of business.
We see progress when a lot of innovation is accumulated. Innovation can bring unprecedented wealth and unprecedented destruction, sometimes both of them.
New things often come in small steps. But every now and then, new technology is created that makes other, more conventional technologies useless. Today’s buzzword for that event is disruptive innovation. This is happening to CDs and DVDs as online streaming makes them useless. This is happening to music publishers and filmmakers who don’t endorse the pay-on-demand or subscription models, and insist on demanding substantial prices for music and movies on physical media. This happened to old-fashioned CRT screens when various types of flat screens became cheap and reliable enough. This is happening in lighting where low-consumption LEDs seem to phase out all other types of lamps. This is happening in manufacturing that is being revolutionized by 3D printing. This is happening in translation, where… wait, what is happening in translation?
The hype about disruptive innovation has a dark side, too: there is considerable scaremongering going on about it. Developers of technology often try to make their product more attractive by advertising them as disruptive technologies. But even before ‘disruptive’ became a buzzword, “experts” used to work the fears of others and mystify the technology they deal with – so that their services were sold easier and at a higher price.
A few months ago, I heard a presentation about how translation and translation services will cease to exist as we know them, and how translation companies should become ‘language solution partners’, whatever that means. [I will not include a reference to this presentation, out of consideration for the presenters.] We also heard how important it is to involve the customers in the process of translating complex content. The presenters suggested that translators and translation companies involve their customers by sharing linguistic resources [translation memories, terminology databases (glossaries)] over the cloud using top-notch technology. And how it’s nothing like what translators and translation companies do today.
Up until the last sentence (in italics), this is more or less fair. The falsehood – no doubt added to elicit fear – is in that one last sentence. The predecessors and founders of my company dreamed up technology for precisely the same thing as early as in 1998, and actually implemented it by 2005 (ten years ago!). And even in the 90s, we have seen a good number of committed translators and translation companies who did not just take up translation jobs but partnered with their customers, provided them with consultancy, and integrated with the clients’ organizations when it was necessary. We might talk more about this, or find new terms for it – that won’t make the approach or that technology any more new or disruptive.
That said, it’s worth spending a word or two on what disruption might really be, because it has a lot to do with my approach to the up-and-coming stories in this blog.
I am with Thomas Kuhn when he says that scientific and technological revolutions take place in the form of paradigm shifts. In science – the focus of Kuhn in his work The Structure of Scientific Revolutions –, this happens when one paradigm is replaced by another. According to Kuhn, a paradigm is the collection of “universally recognized scientific achievements that for a time provide model problems and solutions to a community of practitioners”. A paradigm shift happens when a new set of achievements and problems emerges, which is not compatible with the old set of problems, or is incomprehensible for the practitioners of the old paradigm. This happened, for example, when Newton’s laws were discovered.
For quite some centuries now, technology has been derived from science. To my mind, this means we can witness changes in technology that resemble – or can be considered as – paradigm shifts. Thus in technology, a paradigm shift happens when an old framework or technology is made irrelevant (unnecessary) by something new, and that new thing cannot even be conceived or understood if you stay within the old framework.
Let me give you an example. I am a great fan of Mr. Elon Musk and his Tesla cars. But they are still only cars, however beautiful or electric they are. There’s nothing disruptive about them. The disruption – the paradigm shift – will happen when someone comes up with a safe teleportation technology that makes cars irrelevant. (Electric cars can also become disruptive once the electricity used to charge them is mainly drawn from renewable sources rather than fossil or nuclear fuels.)
Disruptive innovation often comes from a completely different place, brought about by discoveries outside the field. Modern digital telecommunication protocols are fundamentally based on finite-state machines, a mathematical model that coincides with Chomsky’s theory of generative linguistics – because Chomsky’s hierarchy of formal languages includes Type 3, or regular, languages, which are mathematically equivalent of finite-state machines. The generative approach might no longer be a big thing in natural language research, but it’s a fundamental piece of modern telecommunications.
How does this translate into the world of language technology, or into my narrower subject, translation? When we want to communicate to people in another language, we translate. No matter how we do it, what technology or process we use, we still translate. Disruption will happen when there will be no more need for translation. I don’t think that fully unattended machine translation will ever be possible – when that happens, it will be something completely different, totally foreign to the framework of translation. For now, we have examples of dreams about this: in the Bible, at the Pentecost, the apostles made themselves heard in the multitude of native tongues of the crowd. Or there is the Babel fish that transmits brainwaves. Only it doesn’t exist, or at least we don’t know about it.
People at various levels of respectability often make predictions about what new technologies to expect. Sometimes they get carried away, and predict the advent of, say, human-equivalent artificial intelligence in just a few years (see here, here, here, and here). As to myself, I’m deeply skeptical about artificial intelligence (for the most part, I don’t question its value, but I question its feasibility) – there will be another post on that sometime soon. For now, let’s just be really careful about predictions. I have good reason to say this: because we can almost be sure that human beings are not entirely rational (see also here), even the future actions of a single human being are unpredictable. Always. And we know that paradigm shifts happen – which means that there will be new things that we won’t understand with our current thinking.
Predictions can often be very accurate – but they fall outside of structured reasoning. They have more to do with dreams – in the good sense I mentioned in the introduction of this blog. In fact, I’d call them dreams rather than predictions: not because I think they’re unrealistic, but because we work to make dreams true – while we just passively wait for predictions to be fulfilled. That said, being extra careful about technology predictions isn’t just a fatalist acceptance of our powerlessness, but an active openness to new things that might come. This is probably the personality treat that makes some people innovators or early adaptors, who are always ready to learn about new things, rather than run away from them.