Opening UTF-8 file with UTF-16 encoding in gedit results in readable Chinese

Question

I know this isn't really a problem, but I was really curious why this is happening.

How to reproduce this phenomenon

Open gedit and type something in (I used Lorem Ipsum text from lipsum.com). Save the file and close gedit.

Reopen it, and click "Open", then "Other Documents...". Click on the file that you just made. Before you click open, click on Character Encoding and select UTF-16.

Once the file opens, some Chinese shows up in the editor, mixed with some random undisplayable Unicode characters. Now here's where the really weird part begins: Open Google Translate, and paste in the Chinese. Make sure you select "Chinese" as the language.

Depending on the text, you will almost always see some understandable (and even normal) English in there somewhere. In my case, I saw: "For more information, please visit our website at: www.globalcouncil.org", and "This is the first time that you have come to see us in this area. You are welcome to visit our website." See the full text on Pastebin.

Other observations

• When translating the text using the website at freetranslation.com, some of the text comes up reversed, like this: "image ⁥ macro 瑡 桴 礠 樠 with ‮ hunting⁴ 慭 and prices are 礠 楬 挠 macro Ɱ 挠One common phrase
• This text, when reversed, is actually the same as the original text that I typed in when the encoding was UTF-8.
• The reversed text seems to act as if it were actually backwards, too; when I type something at the end of the text, it appears at the beginning.
• Some of the sentences sound so normal (and sometimes funny) that it seems like this can't be a quirk of the translation software, for example: "Soup can not be used in the treatment of acute or chronic obstructive pulmonary disease." This was not in the original text that I put in.
• I found this question on LinuxQuestions, which reproduces the phenomenon when put through Google Translate. Here is some of the understandable English: "This is a place where you can find your favorite bowl and bowl."

The question

Why is this happening? And why do the sentences generated sound so realistic? Is it a quirk of gedit or the translators? (It would be nice if a native Chinese speaker could give me his opinion.)

This is my first question on Super User, so please don't be hard on me. Thank you in advance.

Answer 1

As a Chinese speaker, I can tell you for a fact that, those Chinese character are all invalid and random garbage (sorry for breaking the mystery). The problem occurs here:

Now here's where the really weird part begins: Open Google Translate, and paste in the Chinese. Make sure you select "Chinese" as the language

Google translate for Chinese <> English is not as reliable as it seems. Google not yet have a very accurate Chinese to English / English to Chinese translation just yet, due to the completely different language structure of Chinese and English. To cite your example.

image ⁥ macro 瑡 桴 礠 樠 with ‮ hunting⁴ 慭 and prices are 礠 楬 挠 macro Ɱ 挠One common phrase

All these Chinese words above, doesn't even make any sense. But Google translate thought you actually paste something useful there, so it will just randomly connect words in it's database.

Let's take these two out "挠 楬", which in a long shot might have tiny connection.

"挠" can be watering flower "挠水"

And "楬" can be some type of tree. So maybe "挠 楬" is watering some sort of tree (Even though we never use Chinese words together like above described)

But Google translate was like "挠楬 means Flexibility"

So yeah... The problem is just Google Translate being bad, that's all.

As for this:

"For more information, please visit our website at: www.globalcouncil.org", and "This is the first time that you have come to see us in this area. You are welcome to visit our website."

I suspect you accidentally copy this text along with the Chinese words.

And for the whole backwards words, I think wizzwizz4 already provide that solution that for you