Loading ...
Sorry, an error occurred while loading the content.

Comparative word counts

Expand Messages
  • James Kirchner
    Someone has contacted me (and probably some of you) to find out if there s any typical ratio that can be used as a rule of thumb for estimating the probable
    Message 1 of 14 , Jul 31, 2006
    • 0 Attachment
      Someone has contacted me (and probably some of you) to find out if
      there's any typical ratio that can be used as a rule of thumb for
      estimating the probable word count in a Czech-to-English translation
      and an English-to-Czech translation.

      Does anybody know of something like that? All I could tell her was
      that my English usually comes out about 25 percent shorter than the
      original Czech, IF the Czech formality level is inappropriate for
      English.

      Does anybody have any generally accepted numbers?

      Thanks.

      Jamie
    • Martin Janda
      Not sure about generally accepted figures but I guess these two things may vary a lot - most of my target Czech texts are about 10% shorter than the source
      Message 2 of 14 , Aug 1, 2006
      • 0 Attachment
        Not sure about generally accepted figures but I guess these two things
        may vary a lot - most of my target Czech texts are about 10% shorter
        than the source text, but I vaguely recall that in times when I was
        working into English, the resulting texts were at least the same size
        but mostly a bit longer than the source. I guess it has often much to
        do with language/writing skills - those who are less gifted in creative
        writing (even if writing in their mother tongue) tend to need longer
        phrases to tell their message. (Now, this was a long phrase, wasn't it?
        :-)

        But yes, there is a myth shared by at least some Czech agencies that a
        translation has to be about 10% longer (probably because they are
        working with poor translators only).

        Martin


        > 1. Comparative word counts
        > Posted by: "James Kirchner" jpklists@... kirchnerjk
        > Date: Mon Jul 31, 2006 2:27 pm (PDT)
        >
        > Someone has contacted me (and probably some of you) to find out if
        > there's any typical ratio that can be used as a rule of thumb for
        > estimating the probable word count in a Czech-to-English translation
        > and an English-to-Czech translation.
        >
        > Does anybody know of something like that? All I could tell her was
        > that my English usually comes out about 25 percent shorter than the
        > original Czech, IF the Czech formality level is inappropriate for
        > English.
        >
        > Does anybody have any generally accepted numbers?
        >
        > Thanks.
        >
        > Jamie
        >
        >
        >
      • James Kirchner
        Martin, your remarks make me think of a time I was working on publications for General Motors that had to be put into French for Quebec. The French
        Message 3 of 14 , Aug 1, 2006
        • 0 Attachment
          Martin, your remarks make me think of a time I was working on
          publications for General Motors that had to be put into French for
          Quebec. The French translations done in house at GM tended to be 25
          to 33 percent longer, and we had horrible trouble with the layouts,
          especially when a five-word English headline would cover THREE LINES
          in French. The department head at GM finally decided to farm the
          work out to an independent translation company near Toronto. That
          agency's French translations came out the same length -- or even
          shorter -- than the English original, which is amazing, since here
          they always tell designers to allow an extra 25 percent more layout
          space for French. Their French also sounded less labored.

          Jamie

          On Aug 1, 2006, at 7:00 PM, Martin Janda wrote:

          > Not sure about generally accepted figures but I guess these two
          > things
          > may vary a lot - most of my target Czech texts are about 10% shorter
          > than the source text, but I vaguely recall that in times when I was
          > working into English, the resulting texts were at least the same size
          > but mostly a bit longer than the source. I guess it has often
          > much to
          > do with language/writing skills - those who are less gifted in
          > creative
          > writing (even if writing in their mother tongue) tend to need longer
          > phrases to tell their message. (Now, this was a long phrase,
          > wasn't it?
          > :-)
          >
          > But yes, there is a myth shared by at least some Czech agencies that a
          > translation has to be about 10% longer (probably because they are
          > working with poor translators only).
          >
          > Martin
          >
          >
          >> 1. Comparative word counts
          >> Posted by: "James Kirchner" jpklists@... kirchnerjk
          >> Date: Mon Jul 31, 2006 2:27 pm (PDT)
          >>
          >> Someone has contacted me (and probably some of you) to find out if
          >> there's any typical ratio that can be used as a rule of thumb for
          >> estimating the probable word count in a Czech-to-English translation
          >> and an English-to-Czech translation.
          >>
          >> Does anybody know of something like that? All I could tell her was
          >> that my English usually comes out about 25 percent shorter than the
          >> original Czech, IF the Czech formality level is inappropriate for
          >> English.
          >>
          >> Does anybody have any generally accepted numbers?
          >>
          >> Thanks.
          >>
          >> Jamie
          >>
          >>
          >>
          >
          >
          >
          >
          >
          > Yahoo! Groups Links
          >
          >
          >
          >
          >
          >
        • Michael Trittipo
          Martin s invocation of the direction as important (because of overtranslation, sometimes) bears investigating. One English translation of Valka s mloky has
          Message 4 of 14 , Aug 1, 2006
          • 0 Attachment
            Martin's invocation of the direction as important (because of
            overtranslation, sometimes) bears investigating. One English
            translation of Valka s mloky has 87607 words and 499353 characters
            according to OOo Writer, but that includes some extra "credit"
            paragraphs extraneous to the work and a table of contents and so on that
            I didn't delete. A Czech download seems to have 349683 characters and
            67083 words, according to Textmaker. So that expansion was in the Cz>En
            direction (even discounting a bit for what I should have deleted).
            There must be downloadable, countable versions of other texts, going in
            both directions: say, The Hobbit, or R.U.R., or The Oxbow Incident
            (which I have in Czech), or some Dickens, etc., that would provide some
            useful corpora for the comparison, with direction of translation known.
          • Terminus Technicus
            I would normally say an Eng translation (of a technical or other professional text, not fiction) should be a bit shorter, English being more to the point ,
            Message 5 of 14 , Aug 2, 2006
            • 0 Attachment
              I would normally say an Eng translation (of a technical or other
              professional text, not fiction) should be a bit shorter, English being more
              "to the point", but it totally depends on how the original was written. You
              know those wordy Czech text with passive sentences, often repeating the same
              thing three times.. if you translate what's there, it'll never sound
              English, if you translate what should be there (i.e. the meaning), it's 1/3
              shorter, but much clearer and easy to read.

              On other occassions, though, a Czech tech. text will be full of very
              specific abbreviations and tech terms which need to be explained/written in
              full, or the sentences will be compacted to the point where saying the same
              thing in English requires and extra word or two to make sense - and then the
              resulting ENG file could be 10-15% longer.

              There's big differences in the words to characters ratio depending on the
              style and professional area of a text as well, so I don't think we can come
              up with an universal rule regarding lengths in either of these languages ..

              I would say that the bigger the difference, the more probable it is that
              either the original or the translation are way off (too verbose, repetitive,
              etc.). Poor translators often think that everything must be translated,
              good translators will edit their sentences over and over until they are
              perfect - and that usually means shorter but clearer..

              In other language pairs, the difference is more demonstrable, Czech/English
              to German, for example, as Helga could no doubt tell us, or Czech/English to
              Russian (that could be up to 40% longer)..

              Matej
            • Jan Vanek jr.
              ... Right, especially if you look in the pirate underground. (I doubt if anybody has bothered to bother The Oxbow Incident or R.U.R. in English, but of course,
              Message 6 of 14 , Aug 3, 2006
              • 0 Attachment
                --- In Czechlist@yahoogroups.com, Michael Trittipo <tritt002@...> wrote:

                > Martin's invocation of the direction as important (because of
                > overtranslation, sometimes) bears investigating. One English
                > translation of Valka s mloky has 87607 words and 499353 characters
                > according to OOo Writer, but that includes some extra "credit"
                > paragraphs extraneous to the work and a table of contents and so on that
                > I didn't delete. A Czech download seems to have 349683 characters and
                > 67083 words, according to Textmaker. So that expansion was in the Cz>En
                > direction (even discounting a bit for what I should have deleted).
                > There must be downloadable, countable versions of other texts, going in
                > both directions: say, The Hobbit, or R.U.R., or The Oxbow Incident
                > (which I have in Czech), or some Dickens, etc., that would provide some
                > useful corpora for the comparison, with direction of translation known.

                Right, especially if you look in the pirate underground. (I doubt if
                anybody has bothered to bother The Oxbow Incident or R.U.R. in
                English, but of course, weird things happen.) Just another datapoint
                based on a quick search for an overlap of both languages in my ebookz
                directory, rounded to a meaningful number of significant digits:

                Pratchett's Wyrd Sisters 85 300 words, 470 000 chars incl. spaces
                Kanturkovy Soudne sestry (with a few "pozn. prekl." untrimmed, but
                OTOH he has a tendency to leave out a thing or two sometimes) 81 600
                slov, 493 000 znaku s mezerami.

                Hmmm... weird; my experience with fiction (in shorter lengths) lately
                was that Czech target tends to be some 8 % shorter than the original;
                or is Kanturek so loose? Let me try some more:

                John Varley's novella Press Enter 23 300 / 122 000; Czech translation
                (abominable but mostly word-for-word faithful) 19 200 / 115 000.

                Ha: Frank Herbert, Dune 190 000 / 1 061 000; Duna (without various
                Prilohy not represented in the original plaintext) 170k / 1 067 000.

                Weird, weird, weird.

                --
                Jan Vanek jr.
              • Josef Hlavac
                Well, it seems that the rule is - There s no rule , doesn t it? Joe
                Message 7 of 14 , Aug 3, 2006
                • 0 Attachment
                  Well, it seems that the rule is - "There's no rule", doesn't it?

                  Joe

                  Jan Vanek jr. wrote:
                  > --- In Czechlist@yahoogroups.com, Michael Trittipo <tritt002@...> wrote:
                  >
                  > Ha: Frank Herbert, Dune 190 000 / 1 061 000; Duna (without various
                  > Prilohy not represented in the original plaintext) 170k / 1 067 000.
                  >
                  > Weird, weird, weird.
                  >
                Your message has been successfully submitted and would be delivered to recipients shortly.