From: Bulent Murtezaoglu
Subject: help us nail the nuance, lexical->textual ?
Date:
Message-ID: <87zmu4zrpg.fsf@p4.internal>
Folks, I am working with a talented bunch of young computer people to
entice enthusiastic Turkish-speakers to look into lisp by translating
documents. We are finding that the technical jargon created by
well-meaning translators in the CS field is lacking and is blurring
the meaning of some of the terms we'd like to translate. The latest
instance of this is the IMHO misguided direct translation of 'lexical'
into the Turkish 'sozcuksel' which which is a plain derivation from
'word' (sozcuk) that's readily apparent to any native-speaker. It
would translate back to English roughly as "word-wise" or, umm,
wordic(?) I think this obscures the technical meaning and causes the
novice learner of the concepts (as in lexical scope or lexical
closures) to mistakenly concentrate on individual words. I think we
can safely translate lexical as 'metinsel' (textual) w/o being unduly
inaccurate and capture the right connotation (as in the programmer
and/or the compiler being able to tell what is going on by looking at
the text of the code rather than the dynamic execution contours). I
also think we can do his and still leave 'sozcuksel' as the
translation of lexical in the context of lexical analysis for
compilers etc. so our usage is not in conflict with the existing
dictionaries.
I am in a position to bully the translators (ill-deserved authority)
into adopting this translation scheme. But would I be wrong to do
this? We're confusing enough people as it is.
I'd appreciate comments. Think of it as refactoring (no! revising,
nonono cleaning up ...) a natural language.
cheers,
BM
Bulent Murtezaoglu wrote:
> ... We are finding that the technical jargon created by
> well-meaning translators in the CS field is lacking and is blurring
> the meaning of some of the terms we'd like to translate. The latest
> instance of this is the IMHO misguided direct translation of 'lexical'
> into the Turkish 'sozcuksel' which which is a plain derivation from
> 'word' (sozcuk) that's readily apparent to any native-speaker. It
> would translate back to English roughly as "word-wise" or, umm,
> wordic(?) I think this obscures the technical meaning and causes the
> novice learner of the concepts (as in lexical scope or lexical
> closures) to mistakenly concentrate on individual words.
Perhaps you should consider not translating the word at all, and just
steal it unchanged from english. English just stole (inherited?) it from
Latin, after all. If you steal it, students can google for it and not
have to remember that [new word] is equivalent to 'lexical'.
Are there typesetting considerations which would make this impractical?
--
Lex mea lux, but IANAL
Cameron MacKinnon
Toronto, Canada
From: Bulent Murtezaoglu
Subject: Re: help us nail the nuance, lexical->textual ?
Date:
Message-ID: <87vf4sznqq.fsf@p4.internal>
>>>>> "CMK" == Cameron MacKinnon <··········@clearspot.net> writes:
[...]
CMK> Perhaps you should consider not translating the word at all,
CMK> and just steal it unchanged from english. English just stole
CMK> (inherited?) it from Latin, after all. If you steal it,
CMK> students can google for it and not have to remember that [new
CMK> word] is equivalent to 'lexical'.
This is an excellent argument actually, but there is considerable
political/cultural aversion to borrowing words here. So if we leave
lexical as-is, my fear would be that someone else will translate it
and perhaps will not have the same concerns with capturing the meaning
precisely (I shudder to think what would happen if MS shook hands with
some bureaucrats/politicians here about some huge gov't purchase and
rushed a translation of parts of their MSDN just to say "lookie lookie
our crap is in _pure_ Turkish.").
On the other hand, people we did manage to interest in Lisp here
appear to read English just fine and had no trouble digging up the
usual on-line resources. So perhaps I am being overly cautious about
the potential for damage by not getting things 'just right' in
translation. Maybe the translation effort helps only by getting
'enough' Turks _talking_ about lisp among themselves and as such the
translations just need to be interesting enough for now. When a
sufficient number of people are knowledgeable in the concepts, the
natural language evolution mechanisms will work.
CMK> Are there typesetting considerations which would make this
CMK> impractical?
Nope. If this were 30 years ago maybe we'd subtitute ks for x, but
nowadays x's and q's and w's are no problem.
Thanks for your input.
cheers,
BM
Cameron MacKinnon <··········@clearspot.net> writes:
> Perhaps you should consider not translating the word at all, and just
> steal it unchanged from english. English just stole (inherited?) it
> from Latin, after all. If you steal it, students can google for it and
> not have to remember that [new word] is equivalent to 'lexical'.
Yeah right! Go tell Turks to speak Latin!
;-)
--
__Pascal Bourguignon__ http://www.informatimago.com/
Until real software engineering is developed, the next best practice
is to develop with a dynamic system that has extreme late binding in
all aspects. The first system to really do this in an important way
is Lisp. -- Alan Kay
textual is good. Cameron's idea of not translating would be the ideal,
though, methinks. You gotta teach the concept anyway, may as well
introduce the word. Give them a head start on reading comp.lang.lisp.
(When do you introduce students to that Savagery?)
:)
--
Kenny
Why Lisp? http://lisp.tech.coop/RtL%20Highlight%20Film
"If you plan to enter text which our system might consider to be
obscene, check here to certify that you are old enough to hear the
resulting output." -- Bell Labs text-to-speech interactive Web page
From: Bulent Murtezaoglu
Subject: Re: help us nail the nuance, lexical->textual ?
Date:
Message-ID: <87r7ffzmcv.fsf@p4.internal>
>>>>> "KT" == Kenny Tilton <·······@nyc.rr.com> writes:
KT> ... Give them a head start
KT> on reading comp.lang.lisp. (When do you introduce students to
KT> that Savagery?) ...
From day 1! If I had any actual say in what people read, I'd much
rather hand them over to their uncle Kenny here than spending time
on the Turkish equivalent of /. I do notice Turkish names here
occasionally, but I don't know if our efforts had anything to do with
it. Rest assured, though, that _hordes_ will show up _eventually_.
Turkey isn't exactly a tiny country and we probably are reaching Turkic
people from points East (did I say hordes?).
God help me, I am getting your optimism.
cheers,
BM
Bulent Murtezaoglu <··@acm.org> writes:
> Folks, I am working with a talented bunch of young computer people to
> entice enthusiastic Turkish-speakers to look into lisp by translating
> documents. We are finding that the technical jargon created by
> well-meaning translators in the CS field is lacking and is blurring
> the meaning of some of the terms we'd like to translate. The latest
> instance of this is the IMHO misguided direct translation of 'lexical'
> into the Turkish 'sozcuksel' which which is a plain derivation from
> 'word' (sozcuk) that's readily apparent to any native-speaker. It
> would translate back to English roughly as "word-wise" or, umm,
> wordic(?) I think this obscures the technical meaning and causes the
> novice learner of the concepts (as in lexical scope or lexical
> closures) to mistakenly concentrate on individual words. I think we
> can safely translate lexical as 'metinsel' (textual) w/o being unduly
> inaccurate and capture the right connotation (as in the programmer
> and/or the compiler being able to tell what is going on by looking at
> the text of the code rather than the dynamic execution contours). I
> also think we can do his and still leave 'sozcuksel' as the
> translation of lexical in the context of lexical analysis for
> compilers etc. so our usage is not in conflict with the existing
> dictionaries.
As usual I defend the argument that even though we Turks deserve
nice and to-the-point Turkish technical terms, since they
are technical they'll have their "special" meaning. "Metinsel"
(textual) also sounds fine, since it is going to appear in
the context of programming, coding, source code, I hope Turkish
readers will think the "text" refers to the text, the source
code of the program. However I must admit "metinsel" can
be confusing, too, just as "s�zc�ksel" because Turkish speaking
programmers never use the word "metin" (text) to refer to the
actual source code. Maybe the problem is that when we talk
about sentences we just use the term "word" and not every lexical
item is a word, right?
So maybe we must think harder, go and bang our heads to Latin
dictionaries :) Or think more about the term "lexical" by
looking at meanings again and again:
"lexical a meaningful atom of a language, eg. word."
(http://infolab.uvt.nl/~fjl/fin4/node44.html)
"In linguistics, a lexicon has a slightly more specialized
definition, as it includes the lexemes used to actualize words.
Lexemes are formed according to morpho-syntactic rules and
express sememes. In this sense, a lexicon organizes the mental
vocabulary in a speaker's mind: First, it organizes the vocabulary
of a language according to certain principles (for instance,
all verbs of motion may be linked in a lexical network) and,
second, it contains a generative device producing (new)
simple and complex words according to certain lexical rules.
For example, the suffix '-able' can be added to transitive
verbs only such that we get 'read-able' but not '*cry-able'."
(http://en.wikipedia.org/wiki/Lexicon)
I believe "... (as in the programmer and/or the compiler
being able to tell what is going on by looking at the text
of the code rather than the dynamic execution contours)."
to be the crucial point but unfortunately I cannot think
of a single word to stress that (something better than "metinsel"
[textual]) :(
--
Emre Sevinc
eMBA Software Developer Actively engaged in:
http:www.bilgi.edu.tr http://ileriseviye.org
http://www.bilgi.edu.tr http://fazlamesai.net
Cognitive Science Student http://cazci.com
http://www.cogsci.boun.edu.tr
Bulent Murtezaoglu schrieb:
> I'd appreciate comments. Think of it as refactoring (no! revising,
> nonono cleaning up ...) a natural language.
And what about explain the meaning in a footnote?
stefan
Bulent Murtezaoglu <··@acm.org> writes:
> Folks, I am working with a talented bunch of young computer people to
> entice enthusiastic Turkish-speakers to look into lisp by translating
> documents. We are finding that the technical jargon created by
> well-meaning translators in the CS field is lacking and is blurring
> the meaning of some of the terms we'd like to translate. The latest
> instance of this is the IMHO misguided direct translation of 'lexical'
> into the Turkish 'sozcuksel' which which is a plain derivation from
> 'word' (sozcuk) that's readily apparent to any native-speaker. It
> would translate back to English roughly as "word-wise" or, umm,
> wordic(?) I think this obscures the technical meaning and causes the
> novice learner of the concepts (as in lexical scope or lexical
> closures) to mistakenly concentrate on individual words. I think we
> can safely translate lexical as 'metinsel' (textual) w/o being unduly
> inaccurate and capture the right connotation (as in the programmer
> and/or the compiler being able to tell what is going on by looking at
> the text of the code rather than the dynamic execution contours). I
> also think we can do his and still leave 'sozcuksel' as the
> translation of lexical in the context of lexical analysis for
> compilers etc. so our usage is not in conflict with the existing
> dictionaries.
Here lexical is taken in its compiler jargon acception (which comes
from the linguistics). Lexical analysis is the job implemented by
compiler scanners, producing a token flow sent to the parser which
analyse the grammar and builds the syntactic tree.
Check if there are already good translations to Turk of the compiler
domain, and if so you'll see what term is used. Otherwise, a "lexical
scope" is indeed equivalent to "textual scope", so in this acception
the translation of textual would be good.
--
__Pascal Bourguignon__ http://www.informatimago.com/
Until real software engineering is developed, the next best practice
is to develop with a dynamic system that has extreme late binding in
all aspects. The first system to really do this in an important way
is Lisp. -- Alan Kay
Bulent Murtezaoglu wrote:
> The latest instance of this is the IMHO misguided direct
> translation of 'lexical' into the Turkish 'sozcuksel'
> which which is a plain derivation from 'word' (sozcuk)
> that's readily apparent to any native-speaker. It would
> translate back to English roughly as "word-wise" or, umm,
> wordic(?)
The word "lexical" never really conveyed much to me. I am a
native speaker of English. "Lexical" is not a common word
nor does it resonate strongly with meaning of the CS
concept.
First suggestion: Feel free to pick a term with emotional
content. For example "ink" variables. "Ink" directs the
reader to find the value of his free variables in the
written text. Also ink contrasts with pencil in suggesting
fixity and permanence.
If a student is trying to work out what a piece of code does
he might draw a line from a free variable to the binding it
refers to. If it is a lexical variable he can do this in ink
because it will not change. If it is a dynamic variable he
can use pencil and rub it out and draw in a new line when
required.
Hmm. That is more of a challenge than a suggestion,
challenging you to come up with a fresh and vivid term in
Turkish.
Second suggestion: Focus on nesting. I found the concepts
became much clearer when I attended to the distinction
between scope and extent.
Lexical variables = lexical scope & indefinite extent
Dynamic variables = indefinite scope & dynamic extent
Steele writes
In the case of dynamic extent, if the time intervals of
two entities overlap, then one interval will necessarily
be nested within the other one. This is a property of
the design of COMMON LISP.
He could have added that in the case of lexical scope, if
the texts of two entities overlap, then one text will
necessarily nest within the other.
One could contrast lexical variables and dynamic variables
by saying that one nests scopes while the other nests
extents. Ah but how are you translating scope, extent,
shadowing, and nesting. Can one find Turkish phrase, meaning
nested-scope variable, that trips off the tongue?
Usually, I would say that one should either not translate or
go for banal, straight forward translations.
However the English terminology of lexical, dynamic,
special, etc. works poorly. There is an opportunity to do
better in Turkish.
Alan Crowe
Edinburgh
Scotland
From: Robert Uhl
Subject: Re: help us nail the nuance, lexical->textual ?
Date:
Message-ID: <m3hdgar8xf.fsf@4dv.net>
IIRC 'muscle' means 'little mouse' and 'dura mater' means 'hard mother';
I'd just take the term from English. Latin and Greek had their day;
English is having its. And someday no doubt Urdu will have its.
--
Robert Uhl <http://public.xdi.org/=ruhl>
Pascha the beautiful, Pascha, the Lord's Pascha, the Pascha
all-venerable hath dawned upon us. Pascha, with joy let us embrace one
another. O Pascha! Ransom from sorrow, for from the tomb today, as
from a bridal chamber hath Christ shone forth, and hath filled the women
with joy, saying: proclaim unto the apostles.
From: Kent M Pitman
Subject: Re: help us nail the nuance, lexical->textual ?
Date:
Message-ID: <uhdg9v23x.fsf@nhplace.com>
Bulent Murtezaoglu <··@acm.org> writes:
> Folks, I am working with a talented bunch of young computer people to
> entice enthusiastic Turkish-speakers to look into lisp by translating
> documents.
As they helpfully did for my Slashdot interview! (Anyone interested can
find pointers to the interview and the translation through my site at
http://www.nhplace.com/kent/misc-postings.html
> We are finding that the technical jargon created by
> well-meaning translators in the CS field is lacking and is blurring
> the meaning of some of the terms we'd like to translate. The latest
> instance of this is the IMHO misguided direct translation of 'lexical'
> into the Turkish 'sozcuksel' which which is a plain derivation from
> 'word' (sozcuk) that's readily apparent to any native-speaker.
In some programming languages, there are two terms of which the Lisp
terminology uses only one. One of the terms is "scope" and not directly
applicable here, since this is the term to be modified by "lexical".
The other is "region". A region, if I understand it, is almost like a
lexical scope, except that (a) regions can overlap and (b) regions are
subprimitive to the definition of lexical scope. So the region of the
outer LET's x in:
(let ((x 3)) (+ (let ((x 4)) x) x))
is the whole body of the LET and the lexical scope is defined upon the
term region as the region minus any region of any contained binding.
Why do I raise this? Well, because it gives you a way of understanding
what the term lexical is really getting at. In a sense, regions are
textual and lexical is different in precisely the way you say: they are
about the way that words 'nudge' you toward and away from certain subregions.
So in
(dotimes (x x) (print x))
the thing that makes you know that the second x is not part of the scope
of the x the dotimes binds is that a _word_ (that is, the word dotimes)
has semantics that guides you through the odd-shaped scope.
Whether you think "textual" is a synonym then, depends on whether you
define text as "a sequence of characters" (having no semantics capable
of dictating such choices) or "a sequence of words and punctuation marks"
(which might have semantics powerful enough to force such shapes).
And then there's the old Lisp standard term "unspecial". :)
- - -
The other analysis of this situation that comes to mind for me is that
the term "static" (or "statically apparent") might work since it's the
opposite of dynamic, and captures the kind of reasoning that you do when
thinking about these.
The term "transparent" (as in "referentially transparent") also seems to
have some relevance here.
And "private" comes to mind as well, since lexical variables are private
variables, not susceptible to dynamic variable bindings.
> It
> would translate back to English roughly as "word-wise" or, umm,
> wordic(?) I think this obscures the technical meaning and causes the
> novice learner of the concepts (as in lexical scope or lexical
> closures) to mistakenly concentrate on individual words.
For the reasons I described above, this might not be so bad if instead
of shying from this interpretation, you clarified it.
> I think we can safely translate lexical as 'metinsel' (textual) w/o
> being unduly inaccurate and capture the right connotation (as in the
> programmer and/or the compiler being able to tell what is going on
> by looking at the text of the code rather than the dynamic execution
> contours). I also think we can do his and still leave 'sozcuksel'
> as the translation of lexical in the context of lexical analysis for
> compilers etc. so our usage is not in conflict with the existing
> dictionaries. I am in a position to bully the translators
> (ill-deserved authority) into adopting this translation scheme. But
> would I be wrong to do this? We're confusing enough people as it
> is.
My experience with learning and translating languages is that you're
better off picking a word that is very close (so that it's
recognizable when the person goes to another language), even if it's
not really a synonym. It becomes a locus that you can hang related
meaning and nuance upon. And then you just explain the translation
as such a varation.
For example, the Spanish word "simp�tico" doesn't really mean the
usual sense of "sympathetic" in English. It would normally translate
as "nice" or "likeable" if maintaining a correspondence didn't matter.
But if this were a Lisp word, I might prefer to translate it as
"sympathetic" for the sake of the mnemonic, and just teach people that
there is a meaning of sympathetic that can stretch to fit... e.g.,
the use as in "The protagonist in that story was a sympathetic character."
In other words, It's enough that there's an explanation of the term's
use; the use does not have to be something that is self-apparent.
Tons of computer-science terms have to be explained (things like
queue, stack, buffer, process, character, etc. and verbs like
abort) to avoid their ordinary English meaning being too vague.
Dunno if any of this helps...
Bulent Murtezaoglu wrote:
> The latest
> instance of this is the IMHO misguided direct translation of 'lexical'
> into the Turkish 'sozcuksel' which which is a plain derivation from
> 'word' (sozcuk) that's readily apparent to any native-speaker. It
> would translate back to English roughly as "word-wise" or, umm,
> wordic(?) I think this obscures the technical meaning and causes the
> novice learner of the concepts (as in lexical scope or lexical
> closures) to mistakenly concentrate on individual words. I think we
> can safely translate lexical as 'metinsel' (textual) w/o being unduly
> inaccurate and capture the right connotation (as in the programmer
> and/or the compiler being able to tell what is going on by looking at
> the text of the code rather than the dynamic execution contours). I
> also think we can do his and still leave 'sozcuksel' as the
> translation of lexical in the context of lexical analysis for
> compilers etc. so our usage is not in conflict with the existing
> dictionaries.
I do not know Turkish, so my advice is perhaps of limited utility.
But I think that the way programmers use "lexical" would be best
captured by a word referring to the *structure* of a document
rather than to the document itself, or its basic components(words).
So if Turkish has words for phrase, sentence, or paragraph structure,
I'd be looking at them as the most logical words to attach jargon
meanings to or to derive logical new words from.
That said, the extraordinarily precise jargon uses of words like
'lexical' in English largely result from the fact that a few
technical developers or researchers decided to use them that way
and wrote papers whose very precise jargon definitions were adopted
by the community. The way computer scientists use the word 'lexical'
is a single spot in a much more general usage of the term by
linguists, which again is only a tiny region of the vague and
imprecise usage of the word by ordinary people. If you asked
my relatives who are innocent of programming or linguistic studies
whether our particular usage of 'lexical' captured its meaning
correctly, they'd shrug, and maybe look confused.
> I am in a position to bully the translators (ill-deserved authority)
> into adopting this translation scheme. But would I be wrong to do
> this? We're confusing enough people as it is.
My advice is to pick (or construct) a good word that doesn't clash
with the rest of the language or sound unpleasant, which Turkish-
speaking people will recognize, write down a very precise jargon
definition for the word, and use it absolutely consistently,
across every context you encounter; to the extent that people
adopt it, you will *make* the word mean exactly that.
Bear