Now we will do just what we talked about doing on the previous page, except that this time, we will actually submit the Spanish text to Altavista's BabelFish program, obtain their translation of it, and save that translation in a file of our own.

The point of this is that once we have learned this process, we will be able to do something quite similar to what we did on the previous page with any text inside any word processor, using the side-by-side or two-column command of that word processor. That way you will be able to compare the two versions more easily.

STEP 1: Here is the Spanish text you are about to ask the computer to translate for you. Highlight it. And when you have it highlighted, go to the Edit command and click on "Copy" or "Copy Text."

El Sistema de Mapas Verdes es un marco global, y al mismo tiempo modificable a escala local, para lograr comunidades sustentables. Los Mapas verdes muestran sitios ambientales de importancia en las ciudades alrededor del mundo. Cada mapa es creado localmente en forma única por grupos de todas las edades y experiencias. Cada uno de ellos, usan los iconos del Mapa Verde para identificar, promover, y connectar los recursos ecológicos locales, junto con los lugares que hacen del ambiente un lugar especial. Visite la página web: www.greenmap.org donde podrá ver la variedad de mapas creados y de como participar en este esfuerzo colaborativo de interconnectar naturaleza y cultura a través de los Mapas Verdes.

STEP 2: Finish reading all these instructions before you do anything else. Make sure you have them memorized, or simply write them down.

STEP 3: Go to the website of Altavista's BabelFish program simply by

STEP 4: Once you are on the Altavista BabelFish website, go to the space for entering text to be translated. You can't miss it. It is a big white box.

STEP 5: With your mouse put your cursor inside the box and click. Now go to the Edit command and click on "Paste." Your Spanish paragraph will appear inside the box.

STEP 6: Now go to the BabelFish command below the box and choose "Spanish to English" from the pull-down list of languages. Once you have done this, click on "Translate."

STEP 7: The program will ask you to confirm your request. Do so.

STEP 8: The English translation will appear in another box above the first one. You can save it by highlighting it with the cursor and saving it to a file of your choice. You save it by going to the File command and clicking on "Save As." The system will ask you to specify a file name, which you now need to do.

STEP 9: Later, once again using the Copy and Paste commands, you can take that file and load it in your word processor. If you know how to use your word processor's Side-by-Side command, you can display it and the original English text side by side for the purposes of comparison.

STEP 10: Now that you know how to do this, you can do something similar with any on-line text that appears in two or more languages. Or you can enter text manually from a foreign book or magazine, though this takes longer. Or you can scan it into the computer, but you should try to be sure that the scanning process does not produce any errors.

And now that you've tested the Spanish version, you can follow exactly the same procedure for the French one. Here it is:

Tout en étant ouvert sur le monde, le système de cartographie vert est un cadre de travail localement adaptable pour un développement durable des communautés. Les cartes vertes identifient les sites environnementaux significatifs situés à l'intérieur d'un cadre urbain. Chaque carte se crée localement et de façon unique par des équipes pluridisciplinaires de tous âges. En utilisant les icônes collectives, on identifie, suscite et lie les ressources écologiques du milieu et des endroits qui rendent notre environnement remarquable. Regardez chacune des cartes vertes sur notre site web, www.greenmap.org. Initiez-vous au système cartographique vert et à la manière de participer à cet effort collectif afin de découvrir le lien entre la nature et la culture.

The makers of Babelfish advise against repeatedly retranslating a text back and forth between two languages. And they are making a perfectly good point here. Unlike numbers, words are not truly precise, and the more often you try to retranslate them, the further you will get from the original text and the original meaning. It is something like the game children play while sitting in a circle, with each child whispering a sentence to the next child in the circle. By the time the sentence has made it all the way around the circle, it may bear no resemblance at all to the original sentence.

But there is nothing wrong with comparing the machine translation of an original text with a good translation of that text in another language. One fairly astute observer has complained that he translated the famous example "The spirit is willing, but the flesh is weak" into Russian, and then retranslated that version back into English. The text he got back was just as silly as the famous blooper of several decades ago that went: "The vodka is strong, but the meat is rotten." He therefore concluded that computer translation has made absolutely no progress over the past few decades.

He is fairly close to being correct, but this still isn't quite fair. Even a human translator working from Russian to English could make a mistake with this sentence.

"How can this possibly be?" you may ask, "isn't the meaning of this saying perfectly obvious in any language."

And there you have put your finger right on the real problem of translation, especially computer translation, but even translation by human beings.

"No, the meaning is not obvious," comes back the sad but totally true answer. The meaning of this saying may not be totally obvious to a Russian translator. Or to any translator working from any language into English.

Even though most other languages may have a way of saying this very same thing, it most probably won't be the same way we say it. So even if the translator into Russian has used the precise equivalent saying in that language, it most probably won't read the same way as the English version, especially if he translates it back literally. And there is no way of making sure that the translator working from Russian knows enough to translate the saying back the way we think it should be phrased.

Besides which, if the Russian translator simply translates our saying word for word into Russian, it may 1) make little sense; 2) make no sense at all; 3) sound more or less alright but not be very clear; or 4) just get passed over in reading without being noticed.

And there you have one of the truly major problems of translation in a nutshell. If humans can't be sure of doing it right, how can we program computers to do so? How can we even be sure of all the solutions various translators might devise to deal with this saying?

Furthermore, it is rare that any two human translators will ever translate the same text in the same way. So how are we to tell our computers precisely which one may be the correct way? Or precisely what latitude between different but equally correct translations may be allowed? And how can the computer be expected to determine this all by itself in deciding how to translate even the simplest of phrases?

So now you may want to read some of the other articles about translation—or language or linguistics—on this website. With a bit of luck they may make a bit more sense than they would have if you hadn't visited these pages first. Good luck!

This page is not copyright in any way, so feel free to copy it and show it to your friends or incorporate it into your own website.