new expressions

P&T notes – generalities

Down here we’ll write some “Parsing & Translation”, or “P&T”, notes that describe what we are doing.　The “Parsing” has to do with how we parse the Japanese dialog, while the “translation” has to do with how we move from either “parsed Japanese” to “literal English”, or “from literal English” to “English”.

parsing　

The dialogs have three columns. The first column, “Parsed Japanese” is the original dialog with parsing aids. We try to make the parsing using spaces, color codes and groups.

The first parsing aid is the color codes of some words. We try to follow this scheme:

key

factual particles: post-positions like に, と (with), で (by means of), まで, から, etc. (easy to translate)

emotional particles and words: honorifics (お, ご, titles), softeners (です, んですが), agreement seekers (ね, な) (unlikely to translate)

structural particles: sentence part markers, i.e., が, は, を, と and って. (don’t translate well)

verb and adjective sufixes: okurigana-like

expressions: translate well but the resulting translation might have no word-to-word relationship to the original

The actual color of a word might change inexplicably depending on how new we are to Japanese. One reason might be that we made a mistake and colored the word wrong. Another is that the word itself changes its function. For example, consider the word “fax”; if we were coloring it depending on whether it is a noun or a verb, then “fax” would change its color in “John sent me a fax”, where we use it as noun, and in, “fax it to me”, where we use it as a verb. Japanese is the same, e.g.,

In the first two sentences above, です follows a noun and a na-adjective, so it plays the role of the verb “to be” and closes the sentence; it is not optional. In the last sentence, です follows an i-adjective; since i-adjectives also work as verbs, they can close sentences, so the です is optional and it only plays a decoration role that raises the politeness of the sentence. Hence, we code it in red, because it works as an emotional marker.

group parsing　

We are trying to do some phrase parsing but it is difficult to find a middle ground between keeping the parsed Japanese lean, and making it useful. In English, we attach prepositions to specific words forming sets which we can move around without changing the meaning of the sentence, e.g.,

on Monday, the boyfriend of Mary went to Tokyo by subway

We can group-parse this sentence to make clear what prepositions are attached to what words:

(on Monday), (the boyfriend of Mary) went (to Tokyo) (by subway)

As long as we preserve the subjects and direct object phrases, we are free to move around these groups without changing the meaning of the clause:

(the boyfriend of Mary) went (to Tokyo) (by subway) (on Monday)
(the boyfriend of Mary) went (on Monday) (to Tokyo) (by subway)
(to Tokyo) (by subway), (the boyfriend of Mary) went (on Monday)

These clauses might sound strange but they are grammatically correct. Another way to show these groups is to join them with dashes and separate them with spaces:

on-Monday, the-boyfriend-of-Mary went to-Tokyo by-subway

Japanese works the same way, except that the role of prepositions is carried out with ‘postpositional’ particles. In this case, we can just attach the particle to the word that preceeds it:

Monday-on, Mary’s-boyfriend-of Tokyo-to train-by went
日曜日に, マアリのかれしは 東京へ 電車で 行きました

The only issue with this approach is that sometimes these groups can become large and it is not clear whether showing the grouping actually clarifies anything, e.g.,

although sometimes group-parsing, either with parentheses or spaces, does help:

Hence, we will try to group-parse when the translation might be ambiguous.

translation

Many of the feelings transmitted in Japanese get lost in translation, e.g., from the Japanese dialog of the anime scene above, we can get that both Matoko and Kei are friendly, but Makoto is polite while Kei is casual. English simply doesn’t have a way to express these nuances. In chats, we supplement this lack of expression with emojis, but we are not going to clutter the translation with emojis. Still, we highlight in red the particles or words that are a sign of politeness, e.g., honorifics like さん, くん, ちゃん, お, and ご, or that express an emotion or to seek agreement, e.g., particles like な, ね and -んです. The translation to English would be identical if we were to remove them all; clearly there is something in the Japanese original that we lose in the English translation.

In English, we indicate punctuation marks with voice inflections except, maybe, quote marks, which we sometimes ‘write’ in the air; Japanese also uses voice inflections to indicate punctuation marks but, in addition, it also uses spoken particles. We highlight in blue the particles when they work as spoken punctuation marks, e.g., か and の for ?; よ, ぜ, and ぞ as !, etc. These spoken punctuation marks are likely to survive translation.

We color the particles that work as sentence markers in green. These particles cannot be translated easily because English doesn’t have words that perform the same functions of many of them, e.g.,

• が: subject maker
• は: topic marker
• を: direct object
• に: indirect object
• と: formal quotation marker
• って: casual quotation marker