Language, Dialect, Translation

Back to the Syllabus

There's an unintentionally funny moment in the 1996 film Phenomenon with John Travolta. As you recall, Travolta in this film magically acquires super powers. He can move pencils without touching them, and memorize books, and such. In one scene, a little boy has gone missing on a nearby farm. The farm community, a mixed group of Anglo- and Mexican-Americans, does not know where the boy is--but an old bedridden man does. But the man speaks only Portuguese! So the farmers call up Travolta, who grabs a Portuguese phrasebook and starts to memorize it on his way over in a car. Big suspense. Travolta enters the sickroom, drops to his knees beside the bed, says "bom dia, senhor" and then asks breathlessly: "Onde esta o menino?" . . . as if none of the surrounding Spanish speakers could possibly have thought of asking something like that :-)

This silly movie moment points to a popular misconception about languages. Many people think of a language as an impenetrable Other, a form of expression that is sealed off unless one studies it in a formal way. Many people, too, think of a language as a kind of code, a set of grammar rules plus coded vocabulary items that one acquires by memorizing a book.

These misconceptions are the products of several factors, including: the rise of national standard languages and formal methods of teaching them, the prestige of correctness in standard dialects of languages, the great difficulty of learning a new language after childhood.

In Europe at the time of the origins of the English language, about 1,500 years ago, the concept of national languages was very weak. Even Latin, which was obviously the national language of the (just fallen) Roman Empire, and had a great international currency as the language of the Catholic Church, was spoken in many different local dialects across its range (from Iberia to the Balkans). As one moved across the range of spoken Latin, or across the range of the spoken Germanic languages, each tribe or village could understand its immediate neighbors pretty well, but mutual intelligibility decreased with distance. Which variety of Latin, or of Germanic, was a "language"? Which was a "dialect"? Where did one stop and another begin?

In such a situation, the language/ dialect distinction makes little sense. Languages as distinct modern entities developed later in Western Europe. And earlier--one might in fact say that Latin was the only "language" in Europe 1,500 years ago, and then only in its written form, which was essentially a preservation of the way it had been written by literary authors 500 years before that. Latin was the only language that you could learn formally--indeed you had to learn written Latin formally, then as now. All other "languages" of Europe were spoken dialects of Celtic, Germanic, or other language families, and they were acquired by hearing and speaking, in childhood or afterwards.

The growth of the national languages (examples of some will follow) is a growth not just of linguistic difference but of nation-states, central educational authorities, courts and legal systems, national churches, national publishing industries. It is very important to understand (and very difficult for Americans to grasp, since we live in such a homogenous nation in terms of language) that what we call "French" or "Spanish" or "German" are single "dialects" of larger and quite diverse speaking communities that still today have many features of the linguistic world that existed 1,500 years ago. In Italy, for instance, what we call "Italian" and learn as a foreign language in classes is a single dialect--a dominant one, because it is the literary dialect, the dialect of courts and publications and government and literature. But "Italian" is really a rather artificial version of the dialect of one part of Italy (Tuscany, specifically Florence) as it was written down and accepted as a national standard over many centuries. Even spoken Tuscan today differs from spoken "Italian," and the great range of Italian "dialects"--Milanese, Roman, Neapolitan, Calabrian, Sicilian, Sardinian--might well be considered as separate languages, though they grade into one another across the territory of modern Italy. Most Italian citizens today are fluent in two "languages": "Italian" and their own local "dialect." Both serve the purpose of communication perfectly well, but Italian is more prestigious, and serves to link the whole nation as a common medium. (It links it rather poorly, in fact--Italy has only been a single nation for 130 years, and has great internal diversity and a very weak central government.)

Even today, after centuries of national standardization, the world's languages form a gradation of intelligibility with English. This is because English shares origins with close family neighbors among the Western European languages, and because English, as the language of travel, technology, and the Internet, continues to have an impact on the vocabulary of many languages around the world. The closest "relatives" of English share a large amount of basic vocabulary with English. For example, here's a news story in French. Even if you don't read French, you will be able to identify a large number of words in the passage, and most of them will have the same meaning in French that they have in English. In fact, of all other world languages, you'd probably have the best chance of puzzling out the meaning of a French text, if you knew no language other than English.

Now turn to stories in Dutch and German. You will still be able to identify lots of words in each language, but the pattern of word similarity has changed. It's now the "little" function words that are easy, and the words that carry most of the noun and adjective meaning of the story that may be mysterious. The pattern is almost the inverse of the pattern that connects English to French.

If we move to Italian, fewer words are familiar, but there's a basic pattern of common words, probably fewer than in the preceding examples, but enough to establish a similarity between the languages. In honor of John Travolta, let's also try Portuguese--here too we have many similarities, but perhaps not quite so many as with French or Dutch.

How about Czech? Here we have a lot of words that have probably been borrowed into both English and Czech from Latin, as well as many phrases that have been more recently borrowed into Czech from English. The original cognate family relationships are very difficult to determine, though Czech is in fact a distant Indo-European cousin of English.

Here's an example from Finnish. Aside from the occasional obvious loan-word from English (like seksiskandaalissa) there are no words that I can connect to English words. Finnish and English are unrelated. But there is clearly an influence of present-day English on present-day Finnish, as those loan-words indicate.

We can see the world's languages not as separate compartments, but as part of a gradation of family relationships--as a space shared by several large families who have frequent contact, not as a set of cells each inhabited by an individual.


Languages are inexact and idiomatic media. Translation is not a simple decoding and re-encoding but an art that is sensitive both to general concepts and to particular expressions of those concepts in different languages. There are two principles at work here, and they are sometimes at odds.

One is the "principle of effability," which means that anything you can express in one language you can express in another. You might have to borrow some vocabulary and provide some context, but you can say anything in any world language.

Another is sometimes called the Sapir-Whorf hypothesis, after its joint originators Edward Sapir and Benjamin Whorf. This hypothesis posits that there are great differences in the way people think, depending on what language they acquire. Severe forms of this hypothesis would actually contradict the principle of effability, because there might be some concepts in some language that you really couldn't get across in another.

Even the simplest translation exercises point to the tension between these two linguistic ideas. Most languages have a simple phrase you say to someone when you will see them again soon. "Good bye," we say, which is a contraction of "God be with you," though most of us say instead "See you," "so long," "later." In French? "A bientot" literally means "To well soon." "Au revoir" is "To seeing-again." In Spanish, "hasta luego" or just "luego"--"till then," "till later." Nothing terribly remarkable here; it's all quite effable; but try saying "to seeing again" or even (think about it) "till later" to someone in English. It just sounds funny. "To well soon" or even "till soon" sound crazy. Are we all really saying "good-bye"? And what about Italian, where "ciao" means "hello" or "good-bye" (and etymologically means "slave," though that's another story altogether).

Or try saying "his wife" and "her husband" in a gendered language like French. The usual phrases are, respectively, "sa femme" and "son mari". Where did the "his" and "her" go? A Frenchwoman calls her husband "mon mari" and her car "ma voiture." She uses two different words for "my," and she actually cannot say "his" or "her." Does the French language have a concept for "his or "her"? Are the ideas expressed in those two words "effable" in French? Or--more subtle--does French have the same idea for "my" as English has?

Reflect for a moment on the simplicity of those phrases we've just gone over, and you'll see how hard this issue becomes.

Let's take one of the stories we looked at briefly above--the Italian story might be good--and run it through AltaVista's machine translator. Do this a few times, going from one language to English to another language and back to English.

Back to the top of this page.