[b-hebrew] Unit of Translation
Read, James C
K0434995 at kingston.ac.uk
Fri Jul 29 16:31:36 EDT 2005
In the directory,
,I have placed a series of text files named 1s,2s,3s etc.
In order to view the text files they need to be downloaded and viewed in a
text editor which supports hebrew unicode and has access to the necessary fonts.
The files are the product of software manipulations of a digital form of the
Torah from the Aleppo codex which I scavenged from a module from the JSword project at
The file '1s' is a list of the individual words with their frequency in order
of highest frequency. It would be very difficult to hold a conversation with
any real meaning if we limit ourselves to the first 10 words of this list. As it
would if we were to do the same with any modern language.
The file '2s' is a list of combinations of two words found in the Torah (also
with their frequency). File 3s is the same but with 3 word combinations etc etc.
Examination of file 16s shows that there are 4 occurences of 16 word length phrases
which are repeated 12 times in the Torah. We do well at this point to imagine the
results if we apply the same tests to the whole Aleppo text. We do even better to
imagine the results if we were able to conduct this research using every B-Hebrew
dialogue that has ever been spoken.
It is not long before it becomes transparently clear that the dividing points we
commonly accept as the natural limitations of words is completely arbitrary. i.e
We can just as easily consider getout! to be one word and in fact we would be more
justified to do so from a translative point of view.
Hence, I restate my position (to Rolf especially). I'm not saying that the 'word'
should not be used as a base unit for translation. I am saying that we ought to
reconsider how long a 'natural word' is and use that as the base unit for translation.
NB IT is interesting to note that the ancients put no divisions in their texts to
indicate where one word stops and another word ends.
This email has been scanned for all viruses by the MessageLabs Email
More information about the b-hebrew