Google makes their translation system, based on machine learning techniques, available to everyone for browser web use for free, and for other commercial type uses for a minimal fee. Their free translation system is great for anyone learning a language. They use crowd sourcing to help improve the system, whereby anyone who sees something incorrect can help provide a better translation. Be aware that some phases do not translate perfectly. Idioms may not translate well. And the case, upper, lower, etc., may not match as the translations are based on machine learning matching.

Google translate is very good at using statistical pattern matching from source and target texts. It has been shown to be clearly better than using rule-based systems. That is, it does not use specific grammar rules, spelling rules, etc., just statistical pattern matching (i.e., machine learning). However, translation issues remain. This translation my not work well. Google translate has some issues when it translates from, say, Russian to Greek and needs to go through, say, English.

Consider translating Greek to Russian.

• The modern Greek word "πηγάδι" (pee-GA-thee) ≈ "well" as in a "hole in the ground from which one gets water".
• The modern Greek word "καλά" (ka-LA) ≈ "well" as in "very well".
• The English word "well" can mean a "hole in the ground from which one gets water" or it can mean "very good".
• The Russian word "хорошо" (ha-ra-sho) ≈ "well" as in "very good".
• The Russian word "колодец" (ka-la-dets) ≈ "well" as in a "hole in the ground from which one gets water".

Whenever Google does not have enough source text translated into target text from which to get patterns to do the translation, it uses an intermediate language such as English. This is where some translation issues arise. Note: For some cases such as this, one would need to recognize the problem and then do a more specific search for "well" in English into Russian and then study the (sometimes) long list of alternatives in order to find the correct translation.

This requires one to realize that the initial answer provided is not correct!

This page looks at translation issues by way of a somewhat silly self-referential sentence. The issues identified happen in real translations and show the difficulty of creating accurate language translations (in the presence of errors or ambiguities).

Here is a self-referential sentence.

No problem here. How about the following self-referential sentence?

How might this be "true" or "false"?

Here is the puzzle in text form.

Version #1:

First them. Then ask yourself the following question. How easy is it to fix them?

The following is a sequence of changes to fix identified errors. There are many possible sequences. One is chosen here. The underlined parts are the parts that have been changed from the previous sentence.

First, let us change "errers" to "errors".

Version #2:

There are now only two errors in the sentence. Next, let us change "is" to "are".

Version #3:

There is now only one error in the sentence. Next, let us change "three" to "one".

Version #4:

We have fixed that error, but introduced two more errors in the sentence. Let us change "are" to "is". Note that this was in the original sentence.

Version #5:

There are now only one error in the sentence. Let us change "errors" to "error".

Version #6:

Notice that there are no errors in the sentence. Let us change "one" to "zero".

Version #7:

There are now two errors in the sentence. Let us change "error" to "errors".

Version #8:

There is now one error in the sentence. Let us change "is" to "are", a change previously made and undone.

Version #9:

The sentence is now correct. But what have we changed in the meaning and sentence itself?

The preceding sequence may seem like a silly little example, but it exemplifies the issues of translating one text into another text and trying to address apparent mistakes in the original text. Let us look at a summary of the progression.

Have we lost anything in trying to "fix" the errors? That is, in trying to find a "fixed point".

Translation can be very non-obvious and difficult if not impossible (to retain the original meaning). Consider the following word puzzle.

There is three errers in this sentence.

In translation, rather than trying to find and fix the errors, a better way might be to let the original sentence as it is and add another sentence (as a gloss, etc.) that refers to the erroneous sentence, something like the following.

The previous sentence contains three errors.

When translating Bible passages, what should be done when the meaning of the passage is unclear? How does one know what intended meaning in order do the translation.

Many Bible translators (and translator organizations) make assumptions that they know what the passages mean. This can be done by implicitly trusting theologians as to the meaning of passages.

Some principles of contemporary software project management can be useful in understanding the trade-offs.

Identifying parts of a Bible passage that are unknown can, at times, make people feel uncomfortable. Many would rather be in ignorance than to discuss what might be unknown and what the possibilities might be.

In many of his discourses, Jesus talks in code whereby what is said can be taken in multiple ways. What is interesting is that, on close inspection, each of the many ways can be true (in one or more senses).

Aside: Polished politicians (or University presidents) use double-talk to do this so that everyone hears what it is that they want to hear.

The problem with text that has multiple meanings is that those multiple meanings can be hard to translate accurately - especially when there may be additional meanings that have not yet been discovered.

We see this often in what Jesus says when going from Matthew to Luke.

• Matthew, a Jew, was a tax collector, needed to remember details, would have a shorthand to take notes, was an eye witness, etc., and not much later wrote from his memory and notes what had happened.
• Luke, a Greek, interviewed people (and likely documents) many years later and had to piece together what had happened from memories, etc.

Luke has some interesting details not present in Matthew, but has many recollections that influence the passages written to follow one of the many possible ways to interpret what was said. Usually this is the more literal interpretation, which is embellished with more details. Example: In the houses founded on rock on sand, there are no floods. The rivers beat on the houses. The waters rising and the floods are added in the Luke account. This obscures some of the double meanings that could otherwise be inferred.

Many contemporary Bible translators make many assumptions on the meaning of passages when they translate them into other languages. Arguments often go along the lines as the following.

Well, the color white as snow that appears to mean good does not mean good in that language, so we will change it to black as coal.

What happens if there are hidden meanings or implications of the original text?

The ESV (English Standard Version) Bible (2001) tends to translate ambiguity in the original language into the target language (English).

In doing research in any area, it is important to go back to the original sources. In order to convince other people at the time, the original sources will go to great lengths to make analogies and intuitive arguments about why they are proposing what they are proposing.

This context gets sanitized over time and one is just presented with those sanitized results.

It is my understanding that a translation of the Islamic Koran, to be considered official, needs to include the original (or approved) Arabic with the target language into which it was translated.

This is rarely done with, say, the GNT (Greek New Testament).

The Monks of the Middle Ages tended to keep the original while adding a gloss to explain any parts of the text that might not be clear or that might appear to contain errors.

It is my understanding that, to be official, any translation of the original Islamic Quran (or Koran) must contain the original Arabic. (There may be other requirements; this is just one). Some translations contain glosses that have become somewhat official, so that, in some cases, there are glosses to the glosses.

Discussion question: What types of issues arise when a Christian relies on (one or more) English translations of the original Hebrew and/or Greek texts, all of which contain variants and parts that may not be perfectly clear?

Some translation issues arise due to the "whisper game".

Whenever I hosted a meeting on which software to use in a class, school, etc., I did not want people on the committee who had only used one software package of that type, or one programming language, one hardware type (e.g., Windows, Mac, Linux) etc.

If you have only ever used one brand of software (hardware, etc.) how can you compare what you have used to what else might be used.

If you have only ever attended one church, one denomination , etc., how can you make valid comparisons?

From Shakespeare (1564-1616):

• "It is a tale told by an idiot, full of sound and fury signifying nothing." (Macbeth, Act 5, Scene 5, lines 26- 28)

The Google person in charge of translation, at a NLP (Natural Language Processing) and CL (Computational Linguistics) conference, said that there was not enough text to do the machine translation satisfactorily.

Enough people asked him, including myself, that a few weeks later it appeared on Google Translate.

Google makes their translation system, based on machine learning techniques, available to everyone for browser web use for free, and for other commercial type uses for a minimal fee. Their free translation system is great for anyone learning a language.

They use crowd sourcing to help improve the system, whereby anyone who sees something incorrect can help provide a better translation. Be aware that some phases do not translate perfectly. Idioms may not translate well. And the case, upper, lower, etc., may not match as the translations are based on machine learning matching.

A related area of topic modeling, but requiring more natural language semantic analysis, is that of sentiment analysis - trying to determine if comments are positive, negative, or neutral - or some other semantic grouping. I started to look at this in terms of 30,000+ comments (most in German) from a company with a presence in Germany, but then the company decided they wanted me to move on to something they considered more important. Note that sentiment analysis is related to but very different than topic modeling. Topic modeling is an unsupervised machine learning method that groups together topics of words in documents without direction and without knowing what the words or topics actually mean. Sentiment analysis machine learning approach whereby the computer needs to have some idea of whether a word or group of words refer to something positive or negative. There are usually only two or three decisions for a sentiment analysis decision, "yes" or "no" or "maybe", "positive" or "negative" or "neutral", etc. What does "awfully good" mean?