Language Lounge

A Monthly Column for Word Lovers

Machine Translation Dreams

"A serious artist doesn't start with a kitschy piece of error-ridden bilgewater and then patch it up here and there to produce a work of high art. That's not the nature of art. And translation is an art."   —Douglas Hofstadter

The noted author Douglas Hofstadter wrote a piece for the Atlantic last month called The Shallowness of Google Translate, about the general topic of machine translation (that is, translation of human languages by computer). Hofstadter is not a linguist by training, but he's a cognitive scientist, a polyglot, and a brilliant scholar. The article is an interesting read, and recommended to all. Of the many great points that Hofstadter hits on, one struck a chord that continues to reverberate in the Lounge, and that is the persistent but not very productive idea that more data gives computers the capacity to do better translation.

Accurate and dependable machine translation has long been a dream among linguists and computer scientists alike. It is the subject an extraordinary amount of research (Google Scholar has more than three million hits on the term) and that research has certainly paid off in improvements that are now available to all. Probably everyone reading this has used either Google Translate or some other bilingual or multilingual translation service online to help them past a stumble in a foreign language. For that simple task—let's say, for example, "how do you say ice skate in Swedish?"—machine translation is usually dependable. But if you step up the game—that is, if you want to translate something so complicated as an actual sentence, or heaven forbid, a longer passage held together by the rules of logic, grammar and syntax, machine translation is an odd, baffling, and sometimes hilarious mashup of hits and misses, as some of Hofstadter's examples show.

We live in the era of Big Data, and because massive amounts of formatted data are available to anyone with a computer and an internet connection, it's tempting to look to Big Data as the solution to any of a number of problems, including machine translation. The great advantage of Big Data is the evidential support it provides. Most things that happen have happened sometime before, and we now seem to have enough data about everything that has ever happened to get a statistical picture of how likely it is that a particular thing will happen again, given similar conditions. But that simple scenario also points up the weakness of throwing a problem at Big Data. The fact that something happens most of the time is certainly no assurance that it happens all the time, or that it happens at the particular time that you are interested in learning about.

Here's an example. If you look at hundreds of examples of vetted, gold-standard human translations of the consecutive words red tape from English into German, there are a handful of variations, but more often than not, red tape is translated as Bürokratie. Armed with evidence like this, Google Translate feels confident to translate for you. Ask it to translate the sentence "We sealed the package with red tape" and you get:

Wir haben das Paket mit Bürokratie versiegelt.

And if you dutifully return to Google Translate to render this sentence back into English, you get:

We have sealed the package with bureaucracy.

To characterize the failing generally here, Google Translate takes a statistical, Bayesian approach to a situational problem, and that's sensible on the face of it: look at the statistical record of instances of a phenomenon and on that basis, make an informed guess about possible outcomes when a similar phenomenon arises. Or in more technical talk, consider "how the conditional probability of a set of possible causes for a given observed event can be computed from knowledge of the probability of each cause and the conditional probability of the outcome of each cause." That's the restatement of Bayes' Theorem in the VT. Here's what goes wrong: Google Translate doesn't look beyond the simple juxtaposition of the words red and tape to characterize the situation. The translation algorithm simply homes in on the collocation, has the computational equivalent of an aha moment, and plugs in the translation that works most of the time. Except it doesn't work here. Why not?

As Hofstadter puts it, "Google Translate isn't familiar with situations, period. It's familiar solely with strings composed of words composed of letters." It's not as if the "situations" that clue up a reader about the correct meaning of red tape in this sentence are absent: there are two excellent cues. One is the presence of the verb seal, which collocates more than randomly with tape; the other is the presence of the noun package, which is also found with greater than random frequency in the vicinity of tape. Could a computer not make these small inferential leaps and "realize" that the red tape in this sentence is literally red tape?

The challenge that computers (and their human programmers) face is that analyzing the situational aspects of "strings composed of words"—in other words, the pragmatics of such strings—is something extremely easy for humans to do, and nigh impossible for computers to do. Humans are natural language machines. We invented and have evolved natural languages and they are therefore ideally suited for communication between minds embedded in the sensate skin envelopes we are all born into. All of the components that enable us to interpret natural language are in-built and highly developed. Today's computers of even the greatest sophistication, on the other hand, simply do not have ways to emulate this complexity of what is very literally the human condition.

Will computers get there eventually? Possibly, but probably not after going down many other algorithmic rabbit holes. Because a computer is not a mind, much less an embodied mind, it will never understand language; it will not equate words with the physical or abstract things they represent, because it has no notion that relationship, or of meaning. The most computers can attempt is to decode language—that is, map its coded message into its own form of coding, or the coding of another language. Computers do not read text (which suggests some understanding of it); they process text. Understanding and translation require access to material that is partly encoded in language, but also largely encoded in the minds of those who invented language—namely, humans—and computers don't yet have a way of gaining access to this.

Rate this article:

Click here to read more articles from Language Lounge.

Orin Hargraves is an independent lexicographer and contributor to numerous dictionaries published in the US, the UK, and Europe. He is also the author of Mighty Fine Words and Smashing Expressions (Oxford), the definitive guide to British and American differences, and Slang Rules! (Merriam-Webster), a practical guide for English learners. In addition to writing the Language Lounge column, Orin also writes for the Macmillan Dictionary Blog. Click here to visit his website. Click here to read more articles by Orin Hargraves.

Join the conversation

Comments from our users:

Friday March 2nd, 5:27 PM
Comment by: Graeme Roberts (Pittsford, NY)
Surely enough instances of red tape for sealing packages will train the engines and increase sensitivity to context.
Saturday March 3rd, 6:50 AM
Comment by: Nathan L.
Wonderful piece! Thanks a million!

It is probably worth noting that Google often provides options. And of course it is also worth noting that with its "suggest an edit" Google invites all the free labor it can get from users like me who go to it for rough drafts of all kinds of texts in languages we already know. Those free "edits" obviously get vetted, but the good suggestions do get used.

In your example, at least now after you pointed it out, the drop down menu offers, among others, this option that gets the "red tape=red sticky ribbon" idea across: wir versiegelt, um das Paket mit rotem Band.

And the Russian option offers the "red ribbon" translation without the "bureaucracy" option at all. The Russian does, however, offer an unexpected interpretation of the English that results from an inexact match between the uses of the English preposition "with" and the Russian "dictionary" equivalent of "with".

The first Google interpretation is what most of us English speakers understand, something that means "We used red tape to close up the box". But the drop down box offers another Russian sentence that means "We sealed the box with ["with"="that had"] red tape" meaning maybe red tape was being delivered in the box or there was red tape already decorating the box that we sealed. It certainly isn't the "common sense" interpretation, but given the right "situation" it is possible to imagine "the box with red tape" in the English sentence to mean "the box that had red tape."
This all, of course, goes to support the idea that computers can't do a better job of "contextualizing" a text than a human can, but even a human, without a broader context, is going to rely on a "best guess" option.
Thanks again!
Wednesday March 28th, 12:49 PM
Comment by: Richard W.
Computers will get there eventually for two reasons. The first is that eventually someone will put more pragmatics behind machine translation, which will do better at understanding these situations.

The second is that machine learning is based on looking at the statistics of features in the sample. There are many combination of words that can constitute a feature. It's amazing that Google Translate can get meaningful statistics given the large number of features that occur when you start to combine words. Your sentence has 7! combinations of words. Imagine generating all those combinations, looking up their statistics, and getting the probability of them occurring, then fusing the various results into a coherent sentence. (They are going to throw out the stop words, etc., but for an order of magnitude, think n!.) And, they propose to do that for any passage of text you throw at the site.

However, processors and memory used to perform machine translation keep improving. At a point, they will simply be powerful enough to process the combination of red, tape, seal, and package, and calculate that the probability of these words occurring together means a color of tape, not a governmental process.

Do you have a comment?

Share it with the Visual Thesaurus community.

Your comments:

Sign in to post a comment!

We're sorry, you must be a subscriber to comment.

Click here to subscribe today.

Already a subscriber? Click here to login.