Japanese "word" count

1,145 views
Skip to first unread message

Warren Smith

unread,
Sep 22, 2014, 3:33:30 PM9/22/14
to hon...@googlegroups.com

A client of a client of a client (that is, I don’t have the ability to make an inquiry directly) has quoted a job is XX cents per “Japanese word.”

 

When the second-level client was asked if he meant “character,” the answer was, “No - ‘Japanese word’.”

 

What do you suppose was meant by this? How does one come up with this count?

 

The second-level client was asking for a concession in price, but if this was a price per character, it is substantially more than I am charging. If it was a price per English word, it was tight (considering that there are two levels of people between me and that client taking their respective cuts), but not horrible.

 

Has anybody else run across this? While I typically quote using 2 characters per English word (unless I am doing chemical patents, in which case I use 2.3), what conversion would one use for “Japanese word”?

 

Thanks in advance.

 

Warren Smith

---------------------------------

Dr. Warren Smith

Japanese-English Technology Services

 

 

 

Charlie Milroy

unread,
Sep 22, 2014, 3:38:06 PM9/22/14
to hon...@googlegroups.com
Can’t you ask for an example of a Japanese word?
 
Charlie Milroy
--
You received this message because you are subscribed to the Google Groups "Honyaku E<>J translation list" group.
To unsubscribe from this group and stop receiving emails from it, send an email to honyaku+u...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Matthew Schlecht

unread,
Sep 22, 2014, 4:02:11 PM9/22/14
to Honyaku
On Mon, Sep 22, 2014 at 3:33 PM, Warren Smith <warren...@comcast.net> wrote:

A client of a client of a client (that is, I don’t have the ability to make an inquiry directly) has quoted a job is XX cents per “Japanese word.”

 

When the second-level client was asked if he meant “character,” the answer was, “No - ‘Japanese word’.”

 

What do you suppose was meant by this? How does one come up with this count?

 

The second-level client was asking for a concession in price, but if this was a price per character, it is substantially more than I am charging. If it was a price per English word, it was tight (considering that there are two levels of people between me and that client taking their respective cuts), but not horrible.

 

Has anybody else run across this? While I typically quote using 2 characters per English word (unless I am doing chemical patents, in which case I use 2.3), what conversion would one use for “Japanese word”?


     I have run into this before, and I have always understood "Japanese word" to be equivalent to "Japanese character" (kanji + kana).
     Anyone who asks about counts in Japanese words probably isn't very well acquainted with the Japanese language.  Take the example:

好きなだけ食べさせなさい

     What is the verb "word"?  "食べ"?  "食べさせ"?  "食べさせなさい"?

    The folks who speak of Japanese words are probably accustomed to European word counts, and don't want no funny business with them Asian languages.
     If it comes down to it, let them give you their "Japanese word" count and compare it with your analysis of the document using, for example, Ryan Ginstrom's CountAnything, which reports in Asian words (kanji + kana) the same as characters.  If you have the same count (or close), then pricing should be unambiguous.  If it's off by a factor of ~2, then that should also be helpful.
     Alternatively, you can quote by your predicted target count, or if they don't mind waiting, the actual target count.

Matthew Schlecht, PhD
Word Alchemy
Newark, DE, USA
wordalchemytranslation.com

Warren Smith

unread,
Sep 22, 2014, 4:03:47 PM9/22/14
to hon...@googlegroups.com

I’d love to – but my direct client (with whom I have discussed this) is kinda shy about asking his client to the next level of client, for fear that someone will think he lacks critical professional knowledge. I have tried to reassure my client that this is NOT a “stupid question,” but he is still quite hesitant to ask.  That being said, I have suggested that before he bid on this project (including a very large number of patent apps), that no one would think him unprofessional if he were to request a copy of a sample of work done in the past, including the Japanese, English, and “Japanese word count” for reference.

 

W

 

------

Charlie asked: Can’t you ask for an example of a Japanese word?

Warren Smith

unread,
Sep 22, 2014, 4:05:05 PM9/22/14
to hon...@googlegroups.com

You are the best!

 

Thanks, Matthew.

 

Warren

 

 


From: hon...@googlegroups.com [mailto:hon...@googlegroups.com] On Behalf Of Matthew Schlecht
Sent: Monday, September 22, 2014 1:02 PM
To: Honyaku
Subject: Re: Japanese "word" count

On Mon, Sep 22, 2014 at 3:33 PM, Warren Smith <warren...@comcast.net> wrote:

A client of a client of a client (that is, I don’t have the ability to make an inquiry directly) has quoted a job is XX cents per “Japanese word.”

 

When the second-level client was asked if he meant “character,” the answer was, “No - ‘Japanese word’.”

 

What do you suppose was meant by this? How does one come up with this count?

 

The second-level client was asking for a concession in price, but if this was a price per character, it is substantially more than I am charging. If it was a price per English word, it was tight (considering that there are two levels of people between me and that client taking their respective cuts), but not horrible.

 

Has anybody else run across this? While I typically quote using 2 characters per English word (unless I am doing chemical patents, in which case I use 2.3), what conversion would one use for “Japanese word”?

     I have run into this before, and I have always understood "Japanese word" to be equivalent to "Japanese character" (kanji + kana).
     Anyone who asks about counts in Japanese words probably isn't very well acquainted with the Japanese language.  Take the example:

好きなだけ食べさせなさい

     What is the verb "word"?  "食べ"?  "食べさせ"?  "食べさせなさい"?


    The folks who speak of Japanese words are probably accustomed to European word counts, and don't want no funny business with them Asian languages.

     If it comes down to it, let them give you their "Japanese word" count and compare it with your analysis of the document using, for example, Ryan Ginstrom's CountAnything, which reports in Asian words (kanji + kana) the same as characters.  If you have the same count (or close), then pricing should be unambiguous.  If it's off by a factor of ‾2, then that should also be helpful.

     Alternatively, you can quote by your predicted target count, or if they don't mind waiting, the actual target count.


Matthew Schlecht, PhD
Word Alchemy
Newark, DE, USA
wordalchemytranslation.com

--

Warren Smith

unread,
Sep 22, 2014, 4:20:36 PM9/22/14
to hon...@googlegroups.com

Matthew:

 

I just downloaded Ryan’s Count Anything, but it just gives me the same EXACT counts as in MS Word (“Words,” “Characters,” “Characters (no spaces),” “Asian”, and “Non-Asian”).

 

On a Japanese test file (from my archives, not from this client), Ryan’s program gives me, for example, 11753 “Characters (without spaces),” (the same as MS Word gave me). From this, I would anticipate the English to be about 5880 words. Ryan’s program (and MS Word) count the Japanese doc as “11615 Words.’ Do you suppose that is what the client is referring to?

 

Warren

 

Matthew Schlecht

unread,
Sep 22, 2014, 4:27:00 PM9/22/14
to Honyaku
     I think so.
     The programs they use report counts in "words" as the default.  They're just going with what they know.
     Some agencies use Trados or their favorite other CAT tool to determine counts.  These frequently ignore text squirreled away in text boxes, and of course they ignore dead text that is present in images.  (I won't start on repeats/fuzzy matches!).
     I like CountAnything because, with a few exceptions (newer formats) it does count anything.

Chris Blakeslee

unread,
Sep 23, 2014, 10:42:24 AM9/23/14
to hon...@googlegroups.com
I can confirm that the word count in a Japanese document can be the same as or much less than the character count without spaces, depending on the number of numerals and Roman letters in the document. The difference is that the number 1,000, for example,  is counted as a single word but five characters, and USA, for example, is counted as a single word but three characters. So the bottom line is, if you have a lot of numbers and romaji in your document, the word count will be considerably less than the character count. Worst case is the Asian word count, which does not count numerals or Roman letters at all.

********************
Chris Blakeslee
--

Jon Johanning

unread,
Sep 23, 2014, 5:40:29 PM9/23/14
to hon...@googlegroups.com
It’s enough to drive you back to estimating character counts the old fashioned way — guessing an average number of characters per line and laboriously counting the number of lines. At least that’s the way I used to do it. Now I use Word’s character count and say to myself, “Well, it’s good enough for government work,” even if it’s not government work.

Jon Johanning // jjoha...@igc.org

Warren Smith

unread,
Sep 23, 2014, 11:13:35 PM9/23/14
to hon...@googlegroups.com
I have recently discovered the joys of ABBYY FineReader 12. Its amazing in
its ability to take graphic documents and turn them into word files,
enabling the characters to be counted by word.

Kind of saved me this morning -- A client dropped 105 pages of patents (some
old enough that they aren't available in editable format from the JPO), and
asked, "Can you have these done by Friday." Within minutes FineReader had
converted them (as a batch) into a single Word file, which I was able to
count (with a counting macro I had written to catch words in text boxes,
etc.), and tell the client that 72,000 words will take a LOT longer than
Friday!

It would have taken a LONG time to do this the old-fashioned way that Jon
described!

Warren


a2ztranslate

unread,
Sep 24, 2014, 5:30:19 PM9/24/14
to hon...@googlegroups.com
We have hit this before; of course many agencies in Japan actually quote by target English word, so be careful you don't get caught out there. Otherwise in the translation world there really is no such thing as a Japanese word count, there is only a character count.

christopher blakeslee

unread,
Sep 25, 2014, 2:34:51 AM9/25/14
to hon...@googlegroups.com
There may be no such thing, but I have several clients who insists on calling it a word count instead of a character count, even though we both use MS Word's character count excluding spaces, because that is what I quoted, that is what I bill for, and that is what the client pays. They just can't seem to bring themselves to call it a character count (all Japanese clients working in foreign companies and communicating in English). The key I think when quoting is just to spell it out as the character count excluding spaces as computed by Microsoft Word (if that is what you want), after which you can let them nickname it whatever they want.

Chris Blakeslee

Jon Johanning

unread,
Sep 26, 2014, 11:58:00 AM9/26/14
to hon...@googlegroups.com
On Sep 25, 2014, at 2:34 AM, christopher blakeslee <cpbl...@gmail.com> wrote:

> There may be no such thing, but I have several clients who insists on calling it a word count instead of a character count, even though we both use MS Word's character count excluding spaces, because that is what I quoted, that is what I bill for, and that is what the client pays. They just can't seem to bring themselves to call it a character count (all Japanese clients working in foreign companies and communicating in English).

Could this be because they are really thinking of 単語 or even just 語 (which have pretty vague meanings, I think), and interpreting it as “word”?

Jon Johanning // jjoha...@igc.org

Reply all
Reply to author
Forward
0 new messages