All languages are equally fit
Trond Engen
In another group, when someone recently opined that some languages are
more elevated and thus more fit for abstraction, I answered that
linguists agree that all languages are equally fit to express any
thought, with the qualification that some may have a pre-existing
lexicon in a field but all are equally fit to create new terms for
anything from internal resources. I was challenged to provide support
for that, and now I realize that if I've ever seen it in print in a
linguistics book I can't remember where. Any suggestions?
--
Trond Engen
Peter T. Daniels
Surely it's in every intro book on linguistics for the gen pub?
Anything by David Crystal or Jean Aitchison?
And there's Sapir's famous "When it comes to linguistic form, Plato
walks with the Macedonian swineherd, Confucius with the head-hunting
savage of Assam." (Language, 1921: 219)
DKleinecke
The world is full of language chauvinists. Some times they are quite
amusing. For example, numerous Arab-speaking writers, completely
ignorant of any other language, have rhapsodized over the superior
eloquence of Arabic. There is, of course, no metric for measuring
eloquence - nor is there one for abstraction - but that doesn't stop
them.
I think you scarcely need the help of linguists to shoot down the idea
that set this off.
Marti...@forumbreak.com
Or are they equally fit?
---------------------------------
Posted at: http://www.ForumBreak.com
---------------------------------
John Atkinson
> The subject looks like trolling but isn't.
>
> In another group, when someone recently opined that some languages are
> more elevated and thus more fit for abstraction, I answered that
> linguists agree that all languages are equally fit to express any
> thought, with the qualification that some may have a pre-existing
> lexicon in a field but all are equally fit to create new terms for
> anything from internal resources.
This isn't really an answer to the question of "fitness" (whatever that
means), just a possibly related comment: Some languages are good at
creating new words by combining existing ones (German, Mandarin), some
don't allow this at all but have abundant affixes that can be used to
create new words (Eskimo languages), while others have few facilities
(possibly none?) for creating new words from "internal resources", but
readily borrow them from other languages.
I was challenged to provide support
> for that, and now I realize that if I've ever seen it in print in a
> linguistics book I can't remember where. Any suggestions?
I've recently been reading Sampson, Gill and Trudgill (eds) "Language
Complexity...", a series of articles providing evidence against the
axiom that all languages are similar in complexity (I know, not really
the point at issue). Sampson mentions Charles Hockett, "A Course in
Modern Linguistics" (1958) pp 180-81 as a particularly clear early
advocate of this idea: "[T]he total grammatical complexity of any
language, counting both morphology and syntax, is about the same as that
of any other."
It occurred to me that the existence or otherwise of a variety of
techniques for creating new words from internal resources might well be
a significant component of a complexity metric. However, a quick look
just now seems to show that none of the authors in this book have
investigated this. So, sorry, not much help.
John.
Trond Engen
> On Sep 27, 5:49 pm, Trond Engen <trond...@engen.priv.no> wrote:
>
>> The subject looks like trolling but isn't.
>>
>> In another group, when someone recently opined that some languages
>> are more elevated and thus more fit for abstraction, I answered that
>> linguists agree that all languages are equally fit to express any
>> thought, with the qualification that some may have a pre-existing
>> lexicon in a field but all are equally fit to create new terms for
>> anything from internal resources. I was challenged to provide
>> support for that, and now I realize that if I've ever seen it in
>> print in a linguistics book I can't remember where. Any suggestions?
>
> Surely it's in every intro book on linguistics for the gen pub?
> Anything by David Crystal or Jean Aitchison?
My problem seems to be that I haven't read any of the intro books except
Deutscher's The Unfolding of Language, if that counts as one, but he
doesn't touch that. And I can't find it in Trask/McColl Millar's
Historical Linguistics.
> And there's Sapir's famous "When it comes to linguistic form, Plato
> walks with the Macedonian swineherd, Confucius with the head-hunting
> savage of Assam." (Language, 1921: 219)
Thanks, I should have remembered that.
--
Trond Engen
Trond Engen
> On Sep 27, 2:49 pm, Trond Engen <trond...@engen.priv.no> wrote:
>
>> The subject looks like trolling but isn't.
>>
>> In another group, when someone recently opined that some languages
>> are more elevated and thus more fit for abstraction, I answered that
>> linguists agree that all languages are equally fit to express any
>> thought, with the qualification that some may have a pre-existing
>> lexicon in a field but all are equally fit to create new terms for
>> anything from internal resources. I was challenged to provide
>> support for that, and now I realize that if I've ever seen it in
>> print in a linguistics book I can't remember where. Any suggestions?
>
> The world is full of language chauvinists. Some times they are quite
> amusing. For example, numerous Arab-speaking writers, completely
> ignorant of any other language, have rhapsodized over the superior
> eloquence of Arabic. There is, of course, no metric for measuring
> eloquence - nor is there one for abstraction - but that doesn't stop
> them.
In this case it's more like inverse chauvinism: The inability of the
poster's own language to deal with complex issues in the same way as its
more elevated neighbours.
> I think you scarcely need the help of linguists to shoot down the
> idea that set this off.
True, but I tried to make it simple for myself by a sweep in the general
direction of linguistics, and when some people just won't accept appeals
to anonymous authority I have a problem.
--
Trond Engen
Trond Engen
John Atkinson skreiv:
> Trond Engen wrote:
>
>> The subject looks like trolling but isn't.
>>
>> In another group, when someone recently opined that some languages are
>> more elevated and thus more fit for abstraction, I answered that
>> linguists agree that all languages are equally fit to express any
>> thought, with the qualification that some may have a pre-existing
>> lexicon in a field but all are equally fit to create new terms for
>> anything from internal resources.
>
> This isn't really an answer to the question of "fitness" (whatever
> that means), just a possibly related comment: Some languages are
> good at creating new words by combining existing ones (German,
> Mandarin), some don't allow this at all but have abundant affixes
> that can be used to create new words (Eskimo languages), while others
> have few facilities (possibly none?) for creating new words from
> "internal resources", but readily borrow them from other languages.
I wonder about that. Is it so, or do all languages have systems for word
formation (for some value of word), but more or less attractive for
cultural reasons? What will happen to a borrowing language if it's cut
off from its sources? Say English becoming the one world language?
Wouldn't it just activate its more or less sleeping resources?
>> I was challenged to provide support for that, and now I realize
>> that if I've ever seen it in print in a linguistics book I can't
>> remember where. Any suggestions?
>
> I've recently been reading Sampson, Gill and Trudgill (eds) "Language
> Complexity...", a series of articles providing evidence against the
> axiom that all languages are similar in complexity (I know, not
> really the point at issue). Sampson mentions Charles Hockett, "A
> Course in Modern Linguistics" (1958) pp 180-81 as a particularly
> clear early advocate of this idea: "[T]he total grammatical
> complexity of any language, counting both morphology and syntax, is
> about the same as that of any other."
>
> It occurred to me that the existence or otherwise of a variety of
> techniques for creating new words from internal resources might well
> be a significant component of a complexity metric. However, a quick
> look just now seems to show that none of the authors in this book
> have investigated this. So, sorry, not much help.
But interesting nevertheless.
--
Trond Engen
Panu
> Question should be: Which language is fitter (Czech or Slovak)?
> Or are they equally fit?
Why should either of them be fitter? Even if we - for the sake of
argument - accept the concept of linguistic "fitness" based on the
extent of literary and scientific cultivation, surely both Czech and
Slovak are modern and well-defined literary languages which readily
cope with new concepts.
António Marques
Well, I don't rwally think it works like that. The first question is
'more fit for what?'. In the case you mention, it's 'abstraction'. I
happen to feel that's quite vague, but whatever. Whenever you are
challenged to support something, you need to know what kind of arguments
the other people are expecting. So, the second question would be 'what
is it there in language A that makes you think it is more fit?'. Your
argument then will have to consist of showing that those elements are
either present as well in any language B they choose, or that they would
develop naturally if needed.
I have little doubt that some languages *are* more fit than others for
very specific purposes - and I don't even mean those that they're
usually used for* -, and possibly some will always have an edge over the
rest, but it's very specific concrete purposes - nothing so vague and
vast as 'abstraction'.
(*) Somehow I'm reminded of the way that logarithms were one day found
to make trigonometrical calculations much easier and reliable, when at
first sight there's no relation between one field and the other.
Ruud Harmsen
sci.lang:
>Well, I don't rwally think it works like that. The first question is
>'more fit for what?'. In the case you mention, it's 'abstraction'. I
>happen to feel that's quite vague, but whatever. Whenever you are
>challenged to support something, you need to know what kind of arguments
>the other people are expecting. So, the second question would be 'what
>is it there in language A that makes you think it is more fit?'. Your
>argument then will have to consist of showing that those elements are
>either present as well in any language B they choose, or that they would
>develop naturally if needed.
In the daily translation practice, I constantly find English and
German ways of saying things that aren't easy to do in Dutch. So they
must be richer languages than Dutch. But when I write a webpage in
Dutch and then translate it to English, I find that Dutch too has many
unique ways of expressing things. So Dutch is richer AND poorer.
That's I know often start webpages in English and then translate them
to Dutch. That's much easier.
>I have little doubt that some languages *are* more fit than others for
>very specific purposes -
See the Wikipedia pages in languages like Latin, Old-English and
Gothic.
http://got.wikipedia.org/wiki
(Unfortunately, my Vista doesn't support the script!)
--
Ruud Harmsen, http://rudhar.eu
Ruud Harmsen
sci.lang:
>That's I know often
That's what I now ...
>start webpages in English and then translate them
>to Dutch. That's much easier.
--
Ruud Harmsen, http://rudhar.eu
Ruud Harmsen
sci.lang:
>Mon, 28 Sep 2009 13:43:58 +0200: Ruud Harmsen <r...@rudhar.eu>: in
>sci.lang:
>
>>That's I know often
>
>That's what I now ...
That's why I now.
>>start webpages in English and then translate them
>>to Dutch. That's much easier.
Yes, I hate myself for this too.
Trond Engen
> In the daily translation practice, I constantly find English and
> German ways of saying things that aren't easy to do in Dutch. So they
> must be richer languages than Dutch. But when I write a webpage in
> Dutch and then translate it to English, I find that Dutch too has
> many unique ways of expressing things. So Dutch is richer AND poorer.
>
> That's I know often start webpages in English and then translate them
> to Dutch. That's much easier.
My wife ended up doing the same thing writing her bilingual blog. I
think the reason is that her vocabulary limits what she can express in
English. Thus, she can translate her English text into Norwegian but not
always the opposite. Of course, there's a danger that the Norwegian text
becomes a too literal translation, carrying the limitations of the
second language back into the mother tongue.
--
Trond Engen
Hans Aberg
>> In another group, when someone recently opined that some languages
>> are more elevated and thus more fit for abstraction, I answered that
>> linguists agree that all languages are equally fit to express any
>> thought, with the qualification that some may have a pre-existing
>> lexicon in a field but all are equally fit to create new terms for
>> anything from internal resources. I was challenged to provide support
>> for that, and now I realize that if I've ever seen it in print in a
>> linguistics book I can't remember where. Any suggestions?
>
> Well, I don't rwally think it works like that. The first question is
> 'more fit for what?'.
Perhaps one can get input from the situation of computer languages,
which are mostly designed to fit they needs of humans (computers just
use machine code, and seem perfectly happy with that). If they just have
some basic logical constructs, they become Turing complete, meaning that
formally, given arbitrary amounts of time and space, they can compute
exactly the same set of algorithms.
But they differ vastly in what structures they are fit to describe, and
features like time and space, and interfacing, are of utmost importance
in practical applications. These languages are usually quite fit for
what they were designed for; heated discussions about their cons and
pros may be around features that were not part of the original design.
From natural languages, I recall an example of an English song
expressing sadness through weather by "Ever since you left me, it has
rained all day" or something, which was translated into some Indian(?)
language, spoken in a region where did not have much rain, and when it
came, it was one the most happy events. So the song was changed into
"Ever since you left me, the sun has been shining all day along".
So here, both languages have the equal capacity of expressing basic
human emotions such as sadness, but the differ (perhaps) in their
capabilities of doing it via weather. This is not really due to
limitations of the human language, but that human experience did not
find need to add it. It would be easy to add it, as illustrated by the
following case:
The codetalkers of World War II:
http://en.wikipedia.org/wiki/Codetalkers
The native languages did not have capability to express such military
operations, but a code was added as to provide it (for example "turtle"
might mean "tank"). At the time this was much faster than encode and
decode English by some encryption method. The Japanese caught a Navajo,
but though he recognized the language in the intercepted military
communication, he could not understand the dialog.
Hans
Panu
>
> The world is full of language chauvinists. Some times they are quite
> amusing. For example, numerous Arab-speaking writers, completely
> ignorant of any other language, have rhapsodized over the superior
> eloquence of Arabic.
That might partly be due to the fact that literary Arabic needs to be
mastered by the Arabs themselves, being so different. Whatever you
produce in standard Arabic, sounds eloquent and profound.
Peter T. Daniels
> See the Wikipedia pages in languages like Latin, Old-English and
> Gothic.http://got.wikipedia.org/wiki
> (Unfortunately, my Vista doesn't support the script!)
No, it just means you don't happen to have a font with Unicode-encoded
Gothic in it. The three I have are AlphabetumUnicode (which Juan-Jose
Marcos will sell you cheap; he's a professor of linguistics in
northern Spain), Code2001 (free download), and MPH 28 Damase (which is
the only downloadable font that supports a few of the more obscure
ranges, but it has mistakes in both the ones I've tried to use --
Buginese and Hanunoo).
Peter T. Daniels
Czech is more fitted for native speakers of Czech, and Slovak is more
fitted for native speakers of Slovak.
Hans Aberg
John Atkinson
believe they were, what he believed was that there was no correlation
between language complexity and level of civilization. Peter's source
has left out the previous sentence from that quote. It goes:
"Both simple and complex types of language ... may be found spoken at
any desired level of cultural advance. When it comes to linguistic
form, Plato walks ..."
As for Confucius and the headhunter, Classical Chinese (at least as
written) is unusually simple syntactically and morphologically,
although, obviously, this doesn't mean that it can't be used to discuss
extremely subtle philosophical ideas. Some of the Sino-Tibetan
languages of Assam, OTOH, look much more "complex". Here's a one-word
sentence in Garo (a "very ordinary" one, according to Robbins Burling):
Aganchaktaijawakon
(Unusually for ST, Garo isn't tonal.) As is common in ST languages, all
morphemes are monosyllabic. Translating syllable by syllable:
Speak-answer-again-not-future-probably
That is:
[He] will probably not answer again
John.
Ruud Harmsen
<gram...@verizon.net>: in sci.lang:
>On Sep 28, 7:43�am, Ruud Harmsen <r...@rudhar.eu> wrote:
>> See the Wikipedia pages in languages like Latin, Old-English and
>> Gothic.http://got.wikipedia.org/wiki
>> (Unfortunately, my Vista doesn't support the script!)
>
>No, it just means you don't happen to have a font with Unicode-encoded
>Gothic in it.
Yes, that's the same thing in different words.
Panu
There is the interesting problem that both (literary) Czech and
(literary) Slovak are artificial constructions and what we have is a
dialect continuum.
garabik-ne...@kassiopeia.juls.savba.sk
And even more interestingly, standard Slovak differs slightly from
literary Slovak (in pronunciation, but in such a way that a person
speaking literary Slovak sound "rural" and uneducated), and colloquial
Czech (what roughly corresponds to standard Slovak in sociolinguistic
terms) differs quite significantly from literary Czech in both
pronunciation and morphology, if not the vocabulary. And of course, the
eastern Slovak dialects, which differ from everything else and therefore
have almost negligible impact on both standard and literary Slovak.
Summarizing, (surprise!) literary Slovak is unfit for making profound,
official and eloquent public speeches, whereas literary Czech is perfect
for that. Literary Slovak is really good only for written documents
(where it is identical with standard Slovak). That is, of course,
when putting sociolinguistics into the picture.
--
-----------------------------------------------------------
| Radovan Garabík http://kassiopeia.juls.savba.sk/~garabik/ |
| __..--^^^--..__ garabik @ kassiopeia.juls.savba.sk |
-----------------------------------------------------------
Antivirus alert: file .signature infected by signature virus.
Hi! I'm a signature virus! Copy me into your signature file to help me spread!
Joachim Pense
We had this discussion here sometimes. My impression in those exchanges was
that "all languages are equally expressive" is just a sexed up version
of "there is no metric on expressiveness, so you can never tell (and it
makes no sense to state) that language A might be more expressive than
language B".
Joachim
Christian Weisgerber
> In the daily translation practice, I constantly find English and
> German ways of saying things that aren't easy to do in Dutch.
Probably, but this may also be an artifact of your limited command
of Dutch. Certainly my passive command of English (foreign/second
language) has long surpassed my active command of German (native).
> That's I know often start webpages in English and then translate them
> to Dutch. That's much easier.
Good for you. I have found that I can't even translate my own
English into German. What works for me is writing both versions
in parallel or shortly after each other, when the topic is still
fresh in my mind.
--
Christian "naddy" Weisgerber na...@mips.inka.de
Nathan Sanders
Joachim Pense <sn...@pense-mainz.eu> wrote:
That's probably the best way to look at it.
Let's suppose we did have a metric for measuring expressibility, say,
the number of phones needed to express a concept, essentially, a
relatively concrete measure of word length (we could even be more
concrete, and just use number of msec needed to say the word).
It's not clear that having either a smaller number or a larger number
in this metric corresponds to the "fitness" of a language to express
that concept.
Shorter words are of course typically easier to say for the speaker,
but by being shorter, they are more likely to be homophones or
near-homophones with unrelated words, increasing the probability of
confusion on the part of the listener.
So in addition to word length, you'd also need to figure out the
appropriate weighted combination of ease of articulation and
understandability. (This is incredibly difficult to do!)
Furthermore, that's only for a single, individual word. The original
question centers around "abstraction", but let's say they meant
something more general, like words that are all in the same semantic
domain (technology, mathematics, nature, whatever). Then you have to
first define these groups (which groups do we measure? what do we do
about overlap among the groups?), and figure out how to average the
weighted combination of articulation and understandability over the
entire group (how do we deal with type versus token frequency? what
do we do about words that are homophones with words outside the
semantic group?).
It seems to me to be a vastly complex problem well beyond anything we
are capable of coming close to solving with our current knowledge of
language.
Nathan
--
Nathan Sanders
Linguistics Program
Williams College
http://wso.williams.edu/~nsanders/
Harold Johanssen
> On Sep 27, 5:49 pm, Trond Engen <trond...@engen.priv.no> wrote:
>> The subject looks like trolling but isn't.
>>
>> In another group, when someone recently opined that some languages are
>> more elevated and thus more fit for abstraction, I answered that
>> linguists agree that all languages are equally fit to express any
>> thought, with the qualification that some may have a pre-existing
>> lexicon in a field but all are equally fit to create new terms for
>> anything from internal resources. I was challenged to provide support
>> for that, and now I realize that if I've ever seen it in print in a
>> linguistics book I can't remember where. Any suggestions?
>
> Surely it's in every intro book on linguistics for the gen pub? Anything
> by David Crystal or Jean Aitchison?
I would be interested to understand the status of this further. I
mean, is this something that can be proven, or are we talking an axiom
accepted without proof by the academic community?
Christian Weisgerber
> Perhaps one can get input from the situation of computer languages,
I find any comparisons between natural languages and computer
languages strained at best. Really, they are very different things
that just happen to share the word "language".
> which are mostly designed to fit they needs of humans (computers just
> use machine code, and seem perfectly happy with that).
Machine code is just another computer language. Instruction sets
are designed just like high-level languages.
> From natural languages, I recall an example of an English song
> expressing sadness through weather by "Ever since you left me, it has
> rained all day" or something, which was translated into some Indian(?)
> language, spoken in a region where did not have much rain, and when it
> came, it was one the most happy events. So the song was changed into
> "Ever since you left me, the sun has been shining all day along".
This strikes me as a _cultural_ difference rather than one relating
to language, although it might be difficult to separate the concepts,
I admit.
There is no good German translation for "high school", but this
doesn't tell us much beyond the fact that the American and German
education systems are organized differently and that Germans have
little need to talk about American education.
Christian Weisgerber
> The subject looks like trolling but isn't.
Most trolls say that. ;-)
> In another group, when someone recently opined that some languages are
> more elevated and thus more fit for abstraction, I answered that
> linguists agree that all languages are equally fit to express any
> thought, with the qualification that some may have a pre-existing
> lexicon in a field but all are equally fit to create new terms for
> anything from internal resources. I was challenged to provide support
> for that, and now I realize that if I've ever seen it in print in a
> linguistics book I can't remember where. Any suggestions?
If I wasn't so lazy, I would get up and check for it in David
Crystal's _Cambridge Encyclopedia of Language_.
More interesting than some linguist's bald-faced assertion (argument
by authority) would be what scientific support they can actually
offer for this position.
Mind you, I think it is a reasonable default assumption, especially when
confronted with the ravings of
(1) the chauvinists who proclaim the self-evident superiority of
their language;
(2) the "inverse chauvinists", as you called them, who insist that their
language is inferior to some prestige language;
(3) the cultural imperialists who know that the language of a
Western, industrialized, etc. society must be superior to that
of the primitive natives;
(4) the romantics, who are certain that the language of the noble
savage has not suffered the degradations imposed by Western,
industrialized, etc. society.
Actually, I don't remember encountering (4), but I expect they are
out there somewhere.
António Marques
It's not an axiom nor dogma. It's just an observation: every known
language has devices that enable it to reach 'fitness' in any given
domain. If one hadn't, it probably woulsn't be considered a complete
language. So you might say it's on the axiomatic level, but not as 'all
languages are equally fit' and rather as 'languages possess these and
these enabling capabilities [which incidentally allow them to attain
'fitness']'; something not exhibiting those capabilities is not
recognised as a language.
António Marques
> (4) the romantics, who are certain that the language of the noble
> savage has not suffered the degradations imposed by Western,
> industrialized, etc. society.
>
> Actually, I don't remember encountering (4), but I expect they are
> out there somewhere.
Where do you think all the lovers of 'celtic' are coming from?
Herman Rubin
>In another group, when someone recently opined that some languages are
>more elevated and thus more fit for abstraction, I answered that
>linguists agree that all languages are equally fit to express any
>thought, with the qualification that some may have a pre-existing
>lexicon in a field but all are equally fit to create new terms for
>anything from internal resources. I was challenged to provide support
>for that, and now I realize that if I've ever seen it in print in a
>linguistics book I can't remember where. Any suggestions?
First of all, it is not "abstraction" which is the
problems, but the expression of abstract terms and
concepts. An abstract concept may (or may not) have
started out as an abstracrion of something more
concrete, but my decades of working with them in
mathematics, in which these are the basics, has shown
me that they have an existence by themselves. Teaching
has shown me that they are simpler than the concrete
ideas leading up to them, and they should be presented
first.
I do not believe that any "natural" language is at all
suited to a good expression of them. The so-called
"mathematical language", which has nothing to do with
mathematics, but is needed for it, seems to be a
satisfactory language for that purpose; it is essentially
pure grammar with vocabulary added as needed. The grammar
needs to be strict.
As for non-scientific abstract ideas, it is not adequate,
and I do not believe the existing languages are, either.
--
This address is for information only. I do not claim that these views
are those of the Statistics Department or of Purdue University.
Herman Rubin, Department of Statistics, Purdue University
hru...@stat.purdue.edu Phone: (765)494-6054 FAX: (765)494-0558
Peter T. Daniels
> Mon, 28 Sep 2009 06:56:40 -0700 (PDT): "Peter T. Daniels"
>
> >On Sep 28, 7:43 am, Ruud Harmsen <r...@rudhar.eu> wrote:
> >> See the Wikipedia pages in languages like Latin, Old-English and
> >> Gothic.http://got.wikipedia.org/wiki
> >> (Unfortunately, my Vista doesn't support the script!)
>
> >No, it just means you don't happen to have a font with Unicode-encoded
> >Gothic in it.
>
> Yes, that's the same thing in different words.
No, it most certainly is not the same thing in different words.
EVERYONE's Vista supports Gothic script. Gothic was added to Unicode
at some point between the release of Windows XP and the release of
Windows Vista. Vista supports everything included in Unicode 5.0 (but
unfortunately nothing added since then).
Peter T. Daniels
> Hans Aberg <haberg_20080...@math.su.se> wrote:
> > Perhaps one can get input from the situation of computer languages,
>
> I find any comparisons between natural languages and computer
> languages strained at best. Really, they are very different things
> that just happen to share the word "language".
>
> > which are mostly designed to fit they needs of humans (computers just
> > use machine code, and seem perfectly happy with that).
>
> Machine code is just another computer language. Instruction sets
> are designed just like high-level languages.
>
> > From natural languages, I recall an example of an English song
> > expressing sadness through weather by "Ever since you left me, it has
> > rained all day" or something, which was translated into some Indian(?)
> > language, spoken in a region where did not have much rain, and when it
> > came, it was one the most happy events. So the song was changed into
> > "Ever since you left me, the sun has been shining all day along".
>
> This strikes me as a _cultural_ difference rather than one relating
> to language, although it might be difficult to separate the concepts,
> I admit.
Manuals for Bible translators _teem_ with examples like that. What do
you do in a society that values goats and disvalues sheep? What do you
do in a society where the pig is the core of the economy?
Peter T. Daniels
And no human society has ever been found (except, supposedly, the
Piraha) that does not possess a language with those capabilities.
Hans Aberg
>> Perhaps one can get input from the situation of computer languages,
>
> I find any comparisons between natural languages and computer
> languages strained at best.
You mean, though the situation happens to be the same, it is by accident.
> Really, they are very different things
> that just happen to share the word "language".
Designed by and for humans.
>> which are mostly designed to fit they needs of humans (computers just
>> use machine code, and seem perfectly happy with that).
>
> Machine code is just another computer language.
Yes, but mainly as an interface for the CPU.
> Instruction sets
> are designed just like high-level languages.
Nowadays, but not originally.
>> From natural languages, I recall an example of an English song
>> expressing sadness through weather by "Ever since you left me, it has
>> rained all day" or something, which was translated into some Indian(?)
>> language, spoken in a region where did not have much rain, and when it
>> came, it was one the most happy events. So the song was changed into
>> "Ever since you left me, the sun has been shining all day along".
>
> This strikes me as a _cultural_ difference rather than one relating
> to language, although it might be difficult to separate the concepts,
> I admit.
Right: it is hard to create a natural language without a cultural context.
> There is no good German translation for "high school", but this
> doesn't tell us much beyond the fact that the American and German
> education systems are organized differently and that Germans have
> little need to talk about American education.
So one would probably not consider a limitation of the German language.
Hans
Panu
wrote:
> Panu <craoibhi...@gmail.com> wrote:
> > On Sep 28, 4:58 pm, "Peter T. Daniels" <gramma...@verizon.net> wrote:
> >> Czech is more fitted for native speakers of Czech, and Slovak is more
> >> fitted for native speakers of Slovak.
>
> > There is the interesting problem that both (literary) Czech and
> > (literary) Slovak are artificial constructions and what we have is a
> > dialect continuum.
>
> And even more interestingly, standard Slovak differs slightly from
> literary Slovak (in pronunciation, but in such a way that a person
> speaking literary Slovak sound "rural" and uneducated),
Would you please elaborate? Back in the nineties I did some serious
study of both Czech and Slovak, but I have forgotten most of them.
Peter T. Daniels
Doesn't the claim that _literary_ Slovak has a pronunciation strike
you as odd? _Literary_ Slovak is by definition a written language. It
should be pronounceable in any existing dialect. (Or worse -- cf. the
Chinese classics, which supposedly are not understandable when each
character is simply pronounced in series.)
Hans Aberg
>> In another group, when someone recently opined that some languages are
>> more elevated and thus more fit for abstraction, I answered that
>> linguists agree that all languages are equally fit to express any
>> thought, with the qualification that some may have a pre-existing
>> lexicon in a field but all are equally fit to create new terms for
>> anything from internal resources.
> We had this discussion here sometimes. My impression in those exchanges was
> that "all languages are equally expressive" is just a sexed up version
> of "there is no metric on expressiveness, so you can never tell (and it
> makes no sense to state) that language A might be more expressive than
> language B".
All languages are rather differently expressive, and it is no objective
way to tell that one is better than another, since they may be good at
different things.
But the question is: is there some kind of human thought that is
expressible in one language and inexpressible in another, if one is
allowed to make suitable word additions? It goes back this this, I think:
http://en.wikipedia.org/wiki/Linguistic_relativity
Clearly, the influence of the environment in which one grows up affects
the cognition, and becomes hard wired. But is it possible to attribute
that difference in cognition to the language?
Hans
Ruud Harmsen
<gram...@verizon.net>: in sci.lang:
>On Sep 28, 11:04�am, Ruud Harmsen <r...@rudhar.eu> wrote:
>> Mon, 28 Sep 2009 06:56:40 -0700 (PDT): "Peter T. Daniels"
>> <gramma...@verizon.net>: in sci.lang:
>>
>> >On Sep 28, 7:43�am, Ruud Harmsen <r...@rudhar.eu> wrote:
>> >> See the Wikipedia pages in languages like Latin, Old-English and
>> >> Gothic.http://got.wikipedia.org/wiki
>> >> (Unfortunately, my Vista doesn't support the script!)
>>
>> >No, it just means you don't happen to have a font with Unicode-encoded
>> >Gothic in it.
>>
>> Yes, that's the same thing in different words.
>
>No, it most certainly is not the same thing in different words.
>
>EVERYONE's Vista supports Gothic script.
Mine doesn't. If I'm alone in this, so be it.
>Gothic was added to Unicode
>at some point between the release of Windows XP and the release of
>Windows Vista. Vista supports everything included in Unicode 5.0 (but
>unfortunately nothing added since then).
Please explain how an OS needs to have "support" for specific ranges
of Unicode, other than providing glyphs for them in standard fonts (at
least one)?
Perhaps you have better insight in the technical details than I have.
In a certain period, I made serious attempts to get a Windows program
to work with Polish and Czech text, under Windows 95/98/Me, and I
never got it to work really reliably. So obviously my knowledge of
this field is insufficient.
Hans Aberg
>>>> See the Wikipedia pages in languages like Latin, Old-English and
>>>> Gothic.http://got.wikipedia.org/wiki
>>>> (Unfortunately, my Vista doesn't support the script!)
>>> No, it just means you don't happen to have a font with Unicode-encoded
>>> Gothic in it.
>> Yes, that's the same thing in different words.
>
> No, it most certainly is not the same thing in different words.
>
> EVERYONE's Vista supports Gothic script. Gothic was added to Unicode
> at some point between the release of Windows XP and the release of
> Windows Vista. Vista supports everything included in Unicode 5.0 (but
> unfortunately nothing added since then).
In computer lingo, "support" refers to features that are actively
maintained by the developers. So if the Gothic font is not part of the
OS distribution it may or may not be supported, depending on what
attitude they have towards it. It may be supported as a part of the OS
distribution, or as a third part font, in which case one would expect
there to be some joint efforts between the two groups of developer to
make it work.
Most likely, the OS just supports the Unicode 5.0 standard, or on this
particular platform, some "improved" version of it, plus certain types
of font interfaces, and then a third party develops the font on its own.
In this case, the OS distribution does not support the font directly.
This has some practical significance, when contacting people over bugs:
they will normally only deal with what they support. Implemented, but
deprecated or unsupported features are less likely to be dealt with,
though that may happen, if for example it makes legacy code breaking or
the the unsupported features are about to become supported.
So, if there is an bug in the font, would the bug report go to the OS or
font developers?
Hans
Ruud Harmsen
<gram...@verizon.net>: in sci.lang:
>EVERYONE's Vista supports Gothic script.
http://rudhar.com/cgi-bin/shunicod.cgi
Code range started at hex 10330.
>Vista supports everything included in Unicode 5.0 (but
>unfortunately nothing added since then).
Nothing seems to work above 10000. Perhaps Vista works under the false
assumption the all of Unicode should fit in 16 bits. Music symbols are
above that too (1D100), I've seen them working anywhere either.
Ruud Harmsen
Weisgerber): in sci.lang:
>Ruud Harmsen <r...@rudhar.eu> wrote:
>
>> In the daily translation practice, I constantly find English and
>> German ways of saying things that aren't easy to do in Dutch.
>
>Probably, but this may also be an artifact of your limited command
>of Dutch.
Hahaha.
But seriously, my command of the language is good enough so I always
manage to arrive at an acceptable solution. That's my job. I can even
do it quickly enough so I nearly always meet my deadline.
But then things are often worded differently than in the original
language. That was my point.
>Certainly my passive command of English (foreign/second
>language) has long surpassed my active command of German (native).
Strange. My active _and_ passive of my native tongue is still better
than of any other language I know. Until now, I assumed that was
normal.
>> That's I know often start webpages in English and then translate them
>> to Dutch. That's much easier.
>
>Good for you. I have found that I can't even translate my own
>English into German. What works for me is writing both versions
>in parallel or shortly after each other, when the topic is still
>fresh in my mind.
Yes, I did that too.
Christian Weisgerber
> There is the interesting problem that both (literary) Czech and
> (literary) Slovak are artificial constructions and what we have is a
> dialect continuum.
Isn't there a continuum with Polish and Ukrainian, too?
Peter T. Daniels
> Mon, 28 Sep 2009 10:55:24 -0700 (PDT): "Peter T. Daniels"
>
>
>
>
>
> >On Sep 28, 11:04 am, Ruud Harmsen <r...@rudhar.eu> wrote:
> >> Mon, 28 Sep 2009 06:56:40 -0700 (PDT): "Peter T. Daniels"
> >> <gramma...@verizon.net>: in sci.lang:
>
> >> >On Sep 28, 7:43 am, Ruud Harmsen <r...@rudhar.eu> wrote:
> >> >> See the Wikipedia pages in languages like Latin, Old-English and
> >> >> Gothic.http://got.wikipedia.org/wiki
> >> >> (Unfortunately, my Vista doesn't support the script!)
>
> >> >No, it just means you don't happen to have a font with Unicode-encoded
> >> >Gothic in it.
>
> >> Yes, that's the same thing in different words.
>
> >No, it most certainly is not the same thing in different words.
>
> >EVERYONE's Vista supports Gothic script.
>
> Mine doesn't. If I'm alone in this, so be it.
>
> >Gothic was added to Unicode
> >at some point between the release of Windows XP and the release of
> >Windows Vista. Vista supports everything included in Unicode 5.0 (but
> >unfortunately nothing added since then).
>
> Please explain how an OS needs to have "support" for specific ranges
> of Unicode, other than providing glyphs for them in standard fonts (at
> least one)?
Operating systems do not provide glyphs. Fonts provide glyphs. Windows
Vista comes with several fonts that cover large amounts of Unicode
(such as Tahoma and Arial Unicode), but it doesn't come with fonts
that cover _every_ Unicode range, rightly recognizing that most users
of personal computers have no use for Gothic (or Hanunoo).
> Perhaps you have better insight in the technical details than I have.
I know no technical details. I only know that Windows XP could handle
everything included in Unicode 2.0 (if I had the fonts for it) and
nothing more; and Windows Vista can handle everything included in
Unicode 5.0 (if I have the fonts for it). Presumably Windows 7, to go
on sale next month, can handle everything in Unicode 5.2 (if the user
has the fonts for it).
> In a certain period, I made serious attempts to get a Windows program
> to work with Polish and Czech text, under Windows 95/98/Me, and I
> never got it to work really reliably. So obviously my knowledge of
> this field is insufficient.
I don't know what "work with" might mean. Electronic keyboards other
than the standard US keyboard have been available from the beginning.
Code pages for Eastern Europe were available long before Unicode.
don't know when Windows began to incorporate Unicode capability, but
that was irrelevant to Eastern Europe. I also don't know when MS Word
introduced the "Insert Symbol" panel, or when it became possible to
assign your own keyboard shortcuts for any character you'd like. I
also don't know when MS products began to be available with various
operating languages, so that if you wanted your computer to show you
all its menus and labels in Polish or Czech you could do that. You
certainly can now.
Peter T. Daniels
> Peter T. Daniels wrote:
> >>>> See the Wikipedia pages in languages like Latin, Old-English and
> >>>> Gothic.http://got.wikipedia.org/wiki
> >>>> (Unfortunately, my Vista doesn't support the script!)
> >>> No, it just means you don't happen to have a font with Unicode-encoded
> >>> Gothic in it.
> >> Yes, that's the same thing in different words.
>
> > No, it most certainly is not the same thing in different words.
>
> > EVERYONE's Vista supports Gothic script. Gothic was added to Unicode
> > at some point between the release of Windows XP and the release of
> > Windows Vista. Vista supports everything included in Unicode 5.0 (but
> > unfortunately nothing added since then).
>
> In computer lingo, "support" refers to features that are actively
> maintained by the developers. So if the Gothic font is not part of the
> OS distribution it may or may not be supported, depending on what
> attitude they have towards it. It may be supported as a part of the OS
> distribution, or as a third part font, in which case one would expect
> there to be some joint efforts between the two groups of developer to
> make it work.
>
> Most likely, the OS just supports the Unicode 5.0 standard, or on this
> particular platform, some "improved" version of it, plus certain types
> of font interfaces, and then a third party develops the font on its own.
> In this case, the OS distribution does not support the font directly.
It doesn't _provide_ the font. It _supports_ the font if you provide
it.
> This has some practical significance, when contacting people over bugs:
> they will normally only deal with what they support. Implemented, but
> deprecated or unsupported features are less likely to be dealt with,
> though that may happen, if for example it makes legacy code breaking or
> the the unsupported features are about to become supported.
>
> So, if there is an bug in the font, would the bug report go to the OS or
> font developers?
What would a bug in a font be? Gothic is a straightforward, right-to-
left alphabet with nothing at all special about it. If someone has
drawn an incorrect glyph, obviously that's not the fault of the OS.
Peter T. Daniels
> Mon, 28 Sep 2009 10:55:24 -0700 (PDT): "Peter T. Daniels"
>
> >EVERYONE's Vista supports Gothic script.
>
> http://rudhar.com/cgi-bin/shunicod.cgi
> Code range started at hex 10330.
>
> >Vista supports everything included in Unicode 5.0 (but
> >unfortunately nothing added since then).
>
> Nothing seems to work above 10000. Perhaps Vista works under the false
> assumption the all of Unicode should fit in 16 bits. Music symbols are
> above that too (1D100), I've seen them working anywhere either.
10000 happens to be the beginning of the Linear B range. I have no
trouble using either AlphabetumUnicode or Aegean in that range. (Oh,
and again Code2001 and MPH 28 Damase. I checked via BabelMap.)
Do you have Word2003? It also did not support anything later than
Unicode 2.0. Word2007 supports everything through Unicode 5.0, and
presumably Word2010 will support Unicode 5.2 (it wouldn't make great
sense for it to be able to do more than the OS, would it?)
Trond Engen
> Trond Engen wrote:
>
>> Peter T. Daniels skreiv:
>>
>>> On Sep 27, 5:49 pm, Trond Engen <trond...@engen.priv.no> wrote:
>>>
>>>> The subject looks like trolling but isn't.
>>>>
>>>> In another group, when someone recently opined that some languages
>>>> are more elevated and thus more fit for abstraction, I answered
>>>> that linguists agree that all languages are equally fit to express
>>>> any thought, with the qualification that some may have a
>>>> pre-existing lexicon in a field but all are equally fit to create
>>>> new terms for anything from internal resources. I was challenged
>>>> to provide support for that, and now I realize that if I've ever
>>>> seen it in print in a linguistics book I can't remember where. Any
>>>> suggestions?
>>>
>>> Surely it's in every intro book on linguistics for the gen pub?
>>> Anything by David Crystal or Jean Aitchison?
>>
>> My problem seems to be that I haven't read any of the intro books
>> except Deutscher's The Unfolding of Language, if that counts as one,
>> but he doesn't touch that. And I can't find it in Trask/McColl
>> Millar's Historical Linguistics.
>>
>>> And there's Sapir's famous "When it comes to linguistic form, Plato
>>> walks with the Macedonian swineherd, Confucius with the
>>> head-hunting savage of Assam." (Language, 1921: 219)
>>
>> Thanks, I should have remembered that.
>
> Sapir didn't say that all languages were equally complex. He didn't
> believe they were, what he believed was that there was no correlation
> between language complexity and level of civilization. Peter's
> source has left out the previous sentence from that quote. It goes:
>
> "Both simple and complex types of language ... may be found spoken at
> any desired level of cultural advance. When it comes to linguistic
> form, Plato walks ..."
Thanks, but it's OK. Since I wasn't really after complexity anyway I
didn't use it. Instead I followed Kleinecke's advise and argued from,
er, my own resources.
> As for Confucius and the headhunter, Classical Chinese (at least as
> written) is unusually simple syntactically and morphologically,
> although, obviously, this doesn't mean that it can't be used to
> discuss extremely subtle philosophical ideas. Some of the
> Sino-Tibetan languages of Assam, OTOH, look much more "complex".
> Here's a one-word sentence in Garo (a "very ordinary" one, according
> to Robbins Burling):
>
> Aganchaktaijawakon
>
> (Unusually for ST, Garo isn't tonal.) As is common in ST languages,
> all morphemes are monosyllabic. Translating syllable by syllable:
>
> Speak-answer-again-not-future-probably
But I get seven syllables. Is -jawa- a diphtong?
> That is:
>
> [He] will probably not answer again
The definition of "word" as opposed to "morpheme" may of course be quite
arbitrary in such a language. I think it was you who had some examples
on that from Bantu languages?
--
Trond Engen
Hans Aberg
>> In computer lingo, "support" refers to features that are actively
>> maintained by the developers. So if the Gothic font is not part of the
>> OS distribution it may or may not be supported, depending on what
>> attitude they have towards it. It may be supported as a part of the OS
>> distribution, or as a third part font, in which case one would expect
>> there to be some joint efforts between the two groups of developer to
>> make it work.
>>
>> Most likely, the OS just supports the Unicode 5.0 standard, or on this
>> particular platform, some "improved" version of it, plus certain types
>> of font interfaces, and then a third party develops the font on its own.
>> In this case, the OS distribution does not support the font directly.
>
> It doesn't _provide_ the font. It _supports_ the font if you provide
> it.
The Gothic font, no. But certainly the OS comes with a bunch of fonts.
Mac OS X comes with Times (Roman) and Times New Roman, etc.
>> This has some practical significance, when contacting people over bugs:
>> they will normally only deal with what they support. Implemented, but
>> deprecated or unsupported features are less likely to be dealt with,
>> though that may happen, if for example it makes legacy code breaking or
>> the the unsupported features are about to become supported.
>>
>> So, if there is an bug in the font, would the bug report go to the OS or
>> font developers?
>
> What would a bug in a font be? Gothic is a straightforward, right-to-
> left alphabet with nothing at all special about it. If someone has
> drawn an incorrect glyph, obviously that's not the fault of the OS.
It can be anything from a design error, to an error in the interface
that does not make it work properly. For Code2001 I think it was, Mac OS
X warned against using it, but it seemed working anyhow; hard to tell if
was due to the OS or the font. Verdana has some bugs in it:
http://en.wikipedia.org/wiki/Verdana
Hans
garabik-ne...@kassiopeia.juls.savba.sk
> On Sep 28, 2:13 pm, Panu <craoibhi...@gmail.com> wrote:
>> On Sep 28, 6:49 pm, garabik-news-2005...@kassiopeia.juls.savba.sk
>
>> > And even more interestingly, standard Slovak differs slightly from
>> > literary Slovak (in pronunciation, but in such a way that a person
>> > speaking literary Slovak sound "rural" and uneducated),
>>
>> Would you please elaborate? Back in the nineties I did some serious
>> study of both Czech and Slovak, but I have forgotten most of them.
>
le, li are never palatalized in standard Slovak, whereas in the literary
Slovak the palatalization is compulsory (or rather, these are true
palatals...), standard Slovak is slowly developing "semi-palatalized" l,
but never in front of e, i.
ä is never pronounced as /æ/ in standard Slovak, but this
is accepted in literary language.
Now, saying /ʎɛ/ /ʎi/ and /æ/ identifies you firmly to be a central
Slovak dialect user - and talking in a dialect is a big no-no in educated
circles.
Or you are hypercorrecting, and that's even worse - that identifies you
to be a jerk :-)
> Doesn't the claim that _literary_ Slovak has a pronunciation strike
> you as odd? _Literary_ Slovak is by definition a written language.
Yes, I agree it's odd (a misnomer, in fact, "literary language" is NOT a
good translation of "spisovný jazyk", but it is the established one),
but that is how the situation with Czech and Slovak developed. The
official, prescribed language(s) have their pronunciation rules
specified rather precisely - and in certain situation, we are supposed
to talk in that way (if it were not for the skew between standard and
literary Slovak...).
In Czech, it is even more prominent - the literary language is what you
use to write, and to read aloud (it has different inflectional morphology,
remember), and to talk when you want to sound VERY formal
(schoolchildren and their teachers, official speeches and ceremonies,
TV announcers...)
> It should be pronounceable in any existing dialect.
No, it does not work like in English or Chinese (but you know that...)
The orthography is almost (but not quite) phonemic, so if you follow the
orthography when talking/reading aloud, you get something very close to
middle Slovak dialects, which literary Slovak has been based on.
--
-----------------------------------------------------------
| Radovan Garabík http://kassiopeia.juls.savba.sk/~garabik/ |
| __..--^^^--..__ garabik @ kassiopeia.juls.savba.sk |
-----------------------------------------------------------
Antivirus alert: file .signature infected by signature virus.
Hi! I'm a signature virus! Copy me into your signature file to help me spread!
Trond Engen
> DKleinecke skreiv:
>
>> On Sep 27, 2:49 pm, Trond Engen <trond...@engen.priv.no> wrote:
>>
>>> The subject looks like trolling but isn't.
>>>
>>> In another group, when someone recently opined that some languages
>>> are more elevated and thus more fit for abstraction, I answered
>>> that linguists agree that all languages are equally fit to express
>>> any thought, with the qualification that some may have a
>>> pre-existing lexicon in a field but all are equally fit to create
>>> new terms for anything from internal resources. I was challenged to
>>> provide support for that, and now I realize that if I've ever seen
>>> it in print in a linguistics book I can't remember where. Any
>>> suggestions?
>>
>> The world is full of language chauvinists. Some times they are quite
>> amusing. For example, numerous Arab-speaking writers, completely
>> ignorant of any other language, have rhapsodized over the superior
>> eloquence of Arabic. There is, of course, no metric for measuring
>> eloquence - nor is there one for abstraction - but that doesn't stop
>> them.
>
> In this case it's more like inverse chauvinism: The inability of the
> poster's own language to deal with complex issues in the same way as
> its more elevated neighbours.
... and it was really a rather innocent claim, more about
sociolinguistics than the properties of language.
>> I think you scarcely need the help of linguists to shoot down the
>> idea that set this off.
I ended up doing just that, arguing the case in my own words, with a
long digression into sociolinguistics. I didn't wait for the discussion
today, though, so I see I may have missed a point or two.
> True, but I tried to make it simple for myself by a sweep in the
> general direction of linguistics, and when some people just won't
> accept appeals to anonymous authority I have a problem.
I should add that the one who challenged my assertion wasn't the same
who made the original claim. I believe he asked from a sincere interest
in literature on the subject, and my argumentation didn't really help
with that.
--
Trond Engen
Harold Johanssen
> Harold Johanssen wrote:
>> On Sun, 27 Sep 2009 17:38:35 -0700, Peter T. Daniels wrote:
>>
>>> On Sep 27, 5:49 pm, Trond Engen<trond...@engen.priv.no> wrote:
>>>> The subject looks like trolling but isn't.
>>>>
>>>> In another group, when someone recently opined that some languages
>>>> are more elevated and thus more fit for abstraction, I answered that
>>>> linguists agree that all languages are equally fit to express any
>>>> thought, with the qualification that some may have a pre-existing
>>>> lexicon in a field but all are equally fit to create new terms for
>>>> anything from internal resources. I was challenged to provide support
>>>> for that, and now I realize that if I've ever seen it in print in a
>>>> linguistics book I can't remember where. Any suggestions?
>>>
>>> Surely it's in every intro book on linguistics for the gen pub?
>>> Anything by David Crystal or Jean Aitchison?
>>
>> I would be interested to understand the status of this further. I
>> mean, is this something that can be proven, or are we talking an axiom
>> accepted without proof by the academic community?
>
> It's not an axiom nor dogma. It's just an observation: every known
> language has devices that enable it to reach 'fitness' in any given
> domain. If one hadn't, it probably woulsn't be considered a complete
> language.
Hmm... Let me play devil's advocate here. That comes across as
circular reasoning: All human languages are equally fit, because if there
is one that is not equally fit then it is not a human languages, for all
languages are equally fit.
Anyway, I find it acceptable that all languages spoken by people,
and studied so far, have been found to be equally fit to express any
thought potentially. It would seem, however, that what we have is
empirical evidence alone, correct?
António Marques
Besides cheerfully ignoring what I said (hint: I didn't say all
languages were 'equally fit'), that makes about much sense as
complaining that "All woodpeckers have vertebrae, because if there is
one that hasn't, then it is not a woodpecker" is circular. Don't mix
cause with consequent.
- A language must have certain traits. If it doesn't, it won't be
recognised as a complete language.
- Those traits, among other things, will allow it to become fit for
whatever purpose.
- Ergo, all languages can achieve fitness for the purpose at hand.
> Anyway, I find it acceptable that all languages spoken by people,
> and studied so far, have been found to be equally fit to express any
> thought potentially. It would seem, however, that what we have is
> empirical evidence alone, correct?
What we have is empirical evidence that every known people has a
complete language. It's conceiveable that some people might be found
somewhere who didn't have one (cf. the Piraha thing, if it were true),
but the fact that no such people has been found casts doubt on it - it
would seem that either languages are complete, or don't exist at all
(arguably, there was a time when there was no language; if language
could be less than complete, then one would expect at least some
people to have reverted to a less-than-complete one; rather it's as if
what characterises language - and makes all of them 'fittable' - is
tied to biology, which is pretty much all that all people on Earth
have in common - the same could, of course, be said about other
phenomena than language).
Harold Johanssen
> On 28 Set, 22:29, Harold Johanssen <noem...@please.net> wrote:
> - A language must have certain traits. If it doesn't, it won't be
> recognised as a complete language.
>
> - Those traits, among other things, will allow it to become fit for
> whatever purpose.
>
> - Ergo, all languages can achieve fitness for the purpose at hand.
I see. So it is true by definition - therefore, it is an axiom.
John Atkinson
> John Atkinson:
>
[...]
>
>> As for Confucius and the headhunter, Classical Chinese (at least as
>> written) is unusually simple syntactically and morphologically,
>> although, obviously, this doesn't mean that it can't be used to
>> discuss extremely subtle philosophical ideas.
>>
>
>> Some of the
>> Sino-Tibetan languages of Assam, OTOH, look much more "complex".
>> Here's a one-word sentence in Garo (a "very ordinary" one, according
>> to Robbins Burling):
>>
>> Aganchaktaijawakon
>>
>> (Unusually for ST, Garo isn't tonal.) As is common in ST languages,
>> all morphemes are monosyllabic. Translating syllable by syllable:
>>
>> Speak-answer-again-not-future-probably
>
> But I get seven syllables. Is -jawa- a diphtong?
No. My mistake. Looking at the reference again (Routledge's "The
Sino-Tibetan Languages" p 390) I see that "speak" is a-gan, a
two-syllable morpheme. Lots of verbs seem to start with "a-". I suppose
it derives from a prefix, but isn't recognised as such synchronically.
>
>> That is:
>>
>> [He] will probably not answer again
>
> The definition of "word" as opposed to "morpheme" may of course be quite
> arbitrary in such a language. I think it was you who had some examples
> on that from Bantu languages?
Yes, that's right, that seems to be true especially for ST languages.
However, in this case all the morphemes except the first are affixes
that have to be suffixed on a "base", so it's clear they're not "real"
words -- although they could perhaps be clitics.
I don't know anything about the stress pattern of Garo, so I couldn't
say if suffixes like these are stressed differently from syllables that
are proper grammatical words -- that is, whether or not they're
phonological words. (The Swahili prefixes and suffixes definitely
aren't phonological words by the stress criterion, although, as you say,
they were once written as separate words.)
John.
António Marques
No. As I said before, "[s]o you might say it's on the axiomatic level,
but not as 'all languages are equally fit' and rather as 'languages
possess these and these enabling capabilities [which incidentally
allow them to attain 'fitness']'".
It's not an axiom, it's derived from them (as true propositions
usually are). The important part here is that no one is defining
language so that all may be equally able to be fit for some purpose;
rather, language is understood as having a set of traits, and it just
so happens that those allow it to do such and such.
This argument is as if I'd say there was a law against robbery, that
Jesse James was guilty of robbery and hence that Jesse James was to be
taken to jail, and you'd complain that the essence of the law was
sending Jesse James to jail.
DKleinecke
> Trond Engen <trond...@engen.priv.no> wrote:
> > The subject looks like trolling but isn't.
>
>
> > In another group, when someone recently opined that some languages are
> > more elevated and thus more fit for abstraction, I answered that
> > linguists agree that all languages are equally fit to express any
> > thought, with the qualification that some may have a pre-existing
> > lexicon in a field but all are equally fit to create new terms for
> > anything from internal resources. I was challenged to provide support
> > for that, and now I realize that if I've ever seen it in print in a
> > linguistics book I can't remember where. Any suggestions?
>
> Crystal's _Cambridge Encyclopedia of Language_.
>
> More interesting than some linguist's bald-faced assertion (argument
> by authority) would be what scientific support they can actually
> offer for this position.
>
> Mind you, I think it is a reasonable default assumption, especially when
> confronted with the ravings of
> (1) the chauvinists who proclaim the self-evident superiority of
> their language;
> (2) the "inverse chauvinists", as you called them, who insist that their
> language is inferior to some prestige language;
> (3) the cultural imperialists who know that the language of a
> Western, industrialized, etc. society must be superior to that
> of the primitive natives;
> (4) the romantics, who are certain that the language of the noble
> savage has not suffered the degradations imposed by Western,
> industrialized, etc. society.
>
> Actually, I don't remember encountering (4), but I expect they are
> out there somewhere.
Wouldn't Arabic serve as an example of (4)?
Panu
> On Sep 28, 2:13 pm, Panu <craoibhi...@gmail.com> wrote:
>
>
>
> > On Sep 28, 6:49 pm, garabik-news-2005...@kassiopeia.juls.savba.sk
> > wrote:
>
> > > Panu <craoibhi...@gmail.com> wrote:
> > > > On Sep 28, 4:58 pm, "Peter T. Daniels" <gramma...@verizon.net> wrote:
> > > >> Czech is more fitted for native speakers of Czech, and Slovak is more
> > > >> fitted for native speakers of Slovak.
>
> > > > There is the interesting problem that both (literary) Czech and
> > > > (literary) Slovak are artificial constructions and what we have is a
> > > > dialect continuum.
>
> > > And even more interestingly, standard Slovak differs slightly from
> > > literary Slovak (in pronunciation, but in such a way that a person
> > > speaking literary Slovak sound "rural" and uneducated),
>
> > Would you please elaborate? Back in the nineties I did some serious
> > study of both Czech and Slovak, but I have forgotten most of them.
>
> Doesn't the claim that _literary_ Slovak has a pronunciation strike
> you as odd?
Well, the way he expresses it does sound odd, but I think what he is
getting at is the spelling pronunciation of standard written Slovak.
PaulJK
> Panu <craoi...@gmail.com> wrote:
>
>> There is the interesting problem that both (literary) Czech and
>> (literary) Slovak are artificial constructions and what we have is a
>> dialect continuum.
>
> Isn't there a continuum with Polish and Ukrainian, too?
Of course they are continua with those as well as other Slavic languages.
IIRC, Radovan was responding to a question which was specifically
referring to relationship between Czech and Slovak.
However, I'd say, these continua are not necessarily smooth. There
are sudden sharp changes, isoglossies, and not all of them are in the
expected geographical direction.
pjk
Panu
> Panu <craoibhi...@gmail.com> wrote:
> > There is the interesting problem that both (literary) Czech and
> > (literary) Slovak are artificial constructions and what we have is a
> > dialect continuum.
>
> Isn't there a continuum with Polish and Ukrainian, too?
Not such a smooth one as with Czech and Slovak, I'd say. Especially
not after the population surgery at the end of the Second World War.
But remember that Ukrainian is an East Slavic language, which suggests
that its nearness to Polish is more due to relatively recent
influences than to original similarity.
garabik-ne...@kassiopeia.juls.savba.sk
> Well, the way he expresses it does sound odd, but I think what he is
> getting at is the spelling pronunciation of standard written Slovak.
yes, more or less
however, 'literary Slovak' is an established (established as in
"Slovak authors are using it when they write in English") English
language term for the language that is both spoken and written,
whereas 'standard Slovak' is not a well established concept
(pronunciation of what I call 'standard Slovak' is, almost by a
definition, called 'substandard' !)
PaulJK
> Peter T. Daniels <gram...@verizon.net> wrote:
>> On Sep 28, 2:13 pm, Panu <craoibhi...@gmail.com> wrote:
>>> On Sep 28, 6:49 pm, garabik-news-2005...@kassiopeia.juls.savba.sk
>>>> literary Slovak (in pronunciation, but in such a way that a person
>>>> speaking literary Slovak sound "rural" and uneducated),
>>>
>>> Would you please elaborate? Back in the nineties I did some serious
>>> study of both Czech and Slovak, but I have forgotten most of them.
>
> le, li are never palatalized in standard Slovak, whereas in the literary
> Slovak the palatalization is compulsory (or rather, these are true
> palatals...), standard Slovak is slowly developing "semi-palatalized" l,
> but never in front of e, i.
> ä is never pronounced as /æ/ in standard Slovak, but this
> is accepted in literary language.
>
> Now, saying /ʎɛ/ /ʎi/ and /æ/ identifies you firmly to be a central
> Slovak dialect user - and talking in a dialect is a big no-no in educated
> circles.
> Or you are hypercorrecting, and that's even worse - that identifies you
> to be a jerk :-)
>
>> Doesn't the claim that _literary_ Slovak has a pronunciation strike
>> you as odd? _Literary_ Slovak is by definition a written language.
>
> Yes, I agree it's odd (a misnomer, in fact, "literary language" is NOT a
> good translation of "spisovný jazyk", but it is the established one),
The adjective "spisovný" is a cognate of "písmo" (writing) or
"písmeno" (a letter). The prefix "s-" means roughly "with-".
The English term "literary language" means a written language.
The Cz/Sk term "spisovný jazyk", however, doesn't mean just
a written language, it means a formal language used in writing
as well as in educated or formal speech. In fact, when you
consider the prefix "s-", "spisovný jazyk" = "with-writing language".
> but that is how the situation with Czech and Slovak developed. The
> official, prescribed language(s) have their pronunciation rules
> specified rather precisely - and in certain situation, we are supposed
> to talk in that way (if it were not for the skew between standard and
> literary Slovak...).
> In Czech, it is even more prominent - the literary language is what you
> use to write, and to read aloud (it has different inflectional morphology,
> remember), and to talk when you want to sound VERY formal
> (schoolchildren and their teachers, official speeches and ceremonies,
> TV announcers...)
As Radovan mentioned both Czech and Slovak are almost phonemic
languages. An exhaustive list of exceptions (often to do with voicing
and devoicing) is defined and well known. Regular spelling reforms
make sure that the formal written form reflects the live spoken
educated formal language.
Take, for example, word <vtip> (a joke) (<ť> is palatalized).
Czechs find it difficult to pronounce <vť>, they can either say
<vď> or <fť>. The voicing of /v/ to /f/ in <vtip> is permissible,
IIRC, it may be required.
Another example is "na shledanou" (see you later). It's particularly
interesting because it shows existence of isoglosy between
spoken Bohemian and Moravian Czech.
Czechs devoice <h> and pronounce it as if it were written
<Na schledanou>. (Remember <ch> in Cech is a single letter,
not a diagraph, pronounced as /x/ in Scotish <loch>.)
The Moravians, however, voice <s> and thus pronounce it as if
it were written <Na zhledanou>.
The Czech so called literary language in English is a language
heard and spoken on all formal ocassions.
pjk
>> It should be pronounceable in any existing dialect.
>
> No, it does not work like in English or Chinese (but you know that...)
> The orthography is almost (but not quite) phonemic, so if you follow the
> orthography when talking/reading aloud, you get something very close to
> middle Slovak dialects, which literary Slovak has been based on.
>
>> Radovan Garabík
PaulJK
That's true, AFAIK, there is quite a wide gap between the easternmost
Slovak dialect and Ukrainian. What's more, there's another language,
Lepko, sitting between them.
pjk
Cluster User
Some people would say that of Latin as well.
Yusuf B Gursey
> garabik-news-2005...@kassiopeia.juls.savba.sk wrote:
sounds a little like the situation with standard arabic
PaulJK
> Panu <craoi...@gmail.com> wrote:
>
>> Well, the way he expresses it does sound odd, but I think what he is
>> getting at is the spelling pronunciation of standard written Slovak.
>
> yes, more or less
> however, 'literary Slovak' is an established (established as in
> "Slovak authors are using it when they write in English") English
> language term for the language that is both spoken and written,
> whereas 'standard Slovak' is not a well established concept
> (pronunciation of what I call 'standard Slovak' is, almost by a
> definition, called 'substandard' !)
And there are more terms which when literarily translated to
English imply something else than the original: :-)
"Obecná čeština" (common Czech) = substandard (though not vulgar) Czech
"Hovorová čeština" (spoken Czech) =not formal spoken Czech
plus the already mentioned
"Spisovná čeština" = formal (or correct) written *and* spoken Czech
pjk
PaulJK
If they call it "standard" and not "literary arabic" they avoid all these
misunderstandings in English language environment.
Yusuf B Gursey
: Yusuf B Gursey wrote:
:>>
:>> The Czech so called literary language in English is a language
:>> heard and spoken on all formal ocassions.
:>
:> sounds a little like the situation with standard arabic
: If they call it "standard" and not "literary arabic" they avoid all these
: misunderstandings in English language environment.
: pjk
well, in arabic they use a word that originally meant "pure", but some
western writers do in fact use the term "written arabic", ignoring the
fact that it is indeed spoken in formal occasions, such as on radio, TV,
speeches, religious settings and even soemtimes when two arabs don't quite
understand each other's colloquial (though if informal, it usually is a
low register version of it).
:>>>> It should be pronounceable in any existing dialect.
Ruud Harmsen
<gram...@verizon.net>: in sci.lang:
>Operating systems do not provide glyphs. Fonts provide glyphs.
Yes. In this case, I assumes an OS to include the fonts that are there
without post-installing them.
Or with a narrower sense of "OS": Vista to me is the OS proper plus
its standard fonts.
>Windows
>Vista comes with several fonts that cover large amounts of Unicode
>(such as Tahoma and Arial Unicode), but it doesn't come with fonts
>that cover _every_ Unicode range, /
Right. We agree on that. To me, that is identical with saying "Vista
supports large parts of Unicode, but not _every_ range.
>rightly recognizing that most users
>of personal computers have no use for Gothic (or Hanunoo).
Right. So in my terms, Vista doesn't support Gothic, and cannot
properly display the Gothic Wikipedia.
>> Perhaps you have better insight in the technical details than I have.
>
>I know no technical details.
You use a different meaning of OS than I do. Yours is more like
"kernel". I sometimes use that meaning too. But not this time. That
causes our misunderstanding each other.
>I only know that Windows XP could handle
>everything included in Unicode 2.0 (if I had the fonts for it) and
>nothing more; and Windows Vista can handle everything included in
>Unicode 5.0 (if I have the fonts for it). Presumably Windows 7, to go
>on sale next month, can handle everything in Unicode 5.2 (if the user
>has the fonts for it).
Again, I don't understand why anything should be changed in the OS
proper (meaning kernel) to support specific definitions of Unicode. So
at least half of the above (anything that's not about fonts) is
meaningless to me.
>> In a certain period, I made serious attempts to get a Windows program
>> to work with Polish and Czech text, under Windows 95/98/Me, and I
>> never got it to work really reliably. So obviously my knowledge of
>> this field is insufficient.
>
>I don't know what "work with" might mean.
That the text on the screen (and on paper, if needed) looks OK.
>Electronic keyboards other
>than the standard US keyboard have been available from the beginning.
Nothing to do with keyboards. The program's text files had been
translated by competent translators and were available in readible
encodings.
>Code pages for Eastern Europe were available long before Unicode.
Yes. But they proved difficult to handle in programs.
>don't know when Windows began to incorporate Unicode capability, but
>that was irrelevant to Eastern Europe. I also don't know when MS Word
>introduced the "Insert Symbol" panel, or when it became possible to
>assign your own keyboard shortcuts for any character you'd like. I
>also don't know when MS products began to be available with various
>operating languages, so that if you wanted your computer to show you
>all its menus and labels in Polish or Czech you could do that.
It proved very difficult. It seemed Windows Explorer and Word used
tricks not available in standard API libraries.
--
Ruud Harmsen, http://rudhar.eu
Ruud Harmsen
<gram...@verizon.net>: in sci.lang:
>> Most likely, the OS just supports the Unicode 5.0 standard, or on this
>> particular platform, some "improved" version of it, plus certain types
>> of font interfaces, and then a third party develops the font on its own.
>> In this case, the OS distribution does not support the font directly.
>
>It doesn't _provide_ the font. It _supports_ the font if you provide
>it.
Every Windows I've seen so far (3.10, 3.11, 95, 98, ME, XP, Vista)
came with a number of fonts. Word (6, 97, 2007) came with several
more. So yes, they provided me with fonts.
Even MSDOS (3.3, 4.1. 5.1. 6.0) provided fonts.
>> So, if there is an bug in the font, would the bug report go to the OS or
>> font developers?
>
>What would a bug in a font be?
A Unicode code point for a letter looking like a is displayed as b.
>Gothic is a straightforward, right-to-
>left alphabet with nothing at all special about it. If someone has
>drawn an incorrect glyph, obviously that's not the fault of the OS.
It is if it happens in a font that is provided with the OS as
standard. It's not if it happens in a third-party add-on font.
Ruud Harmsen
<gram...@verizon.net>: in sci.lang:
>> Nothing seems to work above 10000. Perhaps Vista works under the false
>> assumption the all of Unicode should fit in 16 bits. Music symbols are
>> above that too (1D100), I've seen them working anywhere either.
>
>10000 happens to be the beginning of the Linear B range. I have no
>trouble using either AlphabetumUnicode or Aegean in that range. (Oh,
>and again Code2001 and MPH 28 Damase. I checked via BabelMap.)
OK, so Vista doesn't have the 16 bits limit I feared it might have.
>Do you have Word2003?
No, 2007.
>It also did not support anything later than
>Unicode 2.0. Word2007 supports everything through Unicode 5.0, and
>presumably Word2010 will support Unicode 5.2 (it wouldn't make great
>sense for it to be able to do more than the OS, would it?)
Meaningless question, because the OS doesn't "do" anything, not about
ranges. It does something about Unicode proper (any range) and
probably about right-to-left and things like that.
Ruud Harmsen
<gram...@verizon.net>: in sci.lang:
>Do you have Word2003? It also did not support anything later than
>Unicode 2.0. Word2007 supports everything through Unicode 5.0, and
>presumably Word2010 will support Unicode 5.2
In this case (Word), you seem to use "support" in the sense "provided
proper fonts for", but in the case of Windows, you don't.
Why this difference?
>(it wouldn't make great
>sense for it to be able to do more than the OS, would it?)
If "do" and "support" means "provide fonts for", then this happens all
the time: Windows usually provides quite a lot of fonts, then Word
provides many more. So Word provides a lot that the OS didn't
do/have/support, yes.
Ruud Harmsen
garabik-ne...@kassiopeia.juls.savba.sk: in sci.lang:
>le, li are never palatalized in standard Slovak, whereas in the literary
>Slovak the palatalization is compulsory (or rather, these are true
>palatals...), standard Slovak is slowly developing "semi-palatalized" l,
>but never in front of e, i.
>� is never pronounced as /�/ in standard Slovak, but this
>is accepted in literary language.
How can you tell, if literature is by definition written language?
This means if literature is read aloud in standard Slowak, it loses
its true identity and becomes the wrong language?
Panu
Well, IMHO it's not that bad, but the comparison is justified.
Adam Funk
> The subject looks like trolling but isn't.
>
> In another group, when someone recently opined that some languages are
> more elevated and thus more fit for abstraction, I answered that
> linguists agree that all languages are equally fit to express any
> thought, with the qualification that some may have a pre-existing
> lexicon in a field but all are equally fit to create new terms for
> anything from internal resources. I was challenged to provide support
> for that, and now I realize that if I've ever seen it in print in a
> linguistics book I can't remember where. Any suggestions?
I think most introductory and popular texts on linguistics say this,
but I've never seen a "proof" or even an outline or proposal for a
method of demonstrating it.
I suspect that if someone devised a method of analysing and measuring
this and got the (unwanted) result --- that all languages are not
equally versatile --- linguists would go nuts and attack it (or him),
because this is required dogma. (I'm not claiming it's wrong, just
that it's something linguistics students are told they have to
believe.)
I'm curious as to how "fit to create new terms for anything from
internal resources", can be reconciled (redefined?) with the frequency
of borrowing words from other languages.
--
Taken on the whole however this is a fine disc and a good example of
the current pop scene attempting to break out of its vulgarisms and
sometimes downright obscene derivative hogwash.
(Julian Stone-Mason B.A., 1972)
Peter T. Daniels
wrote:
> Panu <craoibhi...@gmail.com> wrote:
> > Well, the way he expresses it does sound odd, but I think what he is
> > getting at is the spelling pronunciation of standard written Slovak.
>
> yes, more or less
> however, 'literary Slovak' is an established (established as in
> "Slovak authors are using it when they write in English") English
> language term for the language that is both spoken and written,
> whereas 'standard Slovak' is not a well established concept
> (pronunciation of what I call 'standard Slovak' is, almost by a
> definition, called 'substandard' !)
That simply means that they have chosen to use an English expression
designed to confuse.
Peter T. Daniels
> Mon, 28 Sep 2009 13:35:49 -0700 (PDT): "Peter T. Daniels"
>
> >Operating systems do not provide glyphs. Fonts provide glyphs.
>
> Yes. In this case, I assumes an OS to include the fonts that are there
> without post-installing them.
(a) Why would you assume such a thing?
(b) Why would you drive all the type-manufacturers out of business,
let alone the hobbyists who create fonts for fun?
(c) What about scripts the OS-makers happened not to have thought of
yet?
> Or with a narrower sense of "OS": Vista to me is the OS proper plus
> its standard fonts.
And Vista's standard fonts happen not to include ones that cover the
Gothic range. Do you not see why it might be more (commercially)
desirable to include Hindi and Thai fonts but not Gothic or Hanunoo?
> >Windows
> >Vista comes with several fonts that cover large amounts of Unicode
> >(such as Tahoma and Arial Unicode), but it doesn't come with fonts
> >that cover _every_ Unicode range, /
>
> Right. We agree on that. To me, that is identical with saying "Vista
> supports large parts of Unicode, but not _every_ range.
You are simply wrong. There are Unicode ranges that Vista _does not_
support, so that if you install a font covering that range, it doesn't
know what to do with it. An example is Phags pa, which is a recent
addition to Unicode.
> >rightly recognizing that most users
> >of personal computers have no use for Gothic (or Hanunoo).
>
> Right. So in my terms, Vista doesn't support Gothic, and cannot
> properly display the Gothic Wikipedia.
Your terms are nonsensical.
If the Gothic wikipedia people don't provide a link to a free Gothic
font download right there on every page, they are being very stupid
indeed.
(The very concept of a Gothic wikipedia is very strange. Who is
expected to read it?)
> >> Perhaps you have better insight in the technical details than I have.
>
> >I know no technical details.
>
> You use a different meaning of OS than I do. Yours is more like
> "kernel". I sometimes use that meaning too. But not this time. That
> causes our misunderstanding each other.
A kernel is something that appears on an ear of corn. A kernel used to
be the simplest syntactic structure generated by the simplest rewrite
rules in pre-Aspects transformational grammar.
> >I only know that Windows XP could handle
> >everything included in Unicode 2.0 (if I had the fonts for it) and
> >nothing more; and Windows Vista can handle everything included in
> >Unicode 5.0 (if I have the fonts for it). Presumably Windows 7, to go
> >on sale next month, can handle everything in Unicode 5.2 (if the user
> >has the fonts for it).
>
> Again, I don't understand why anything should be changed in the OS
> proper (meaning kernel) to support specific definitions of Unicode. So
> at least half of the above (anything that's not about fonts) is
> meaningless to me.
Nor do I know any reason "why" Microsoft should expend energy on
accommodating users of obscure fonts of interest to a dozen or so
people around the world just because people like Michael Everson
decide they ought to be included, but they do. It's convenient for me,
but we made *The World's Writing Systems* without Unicode.
> >> In a certain period, I made serious attempts to get a Windows program
> >> to work with Polish and Czech text, under Windows 95/98/Me, and I
> >> never got it to work really reliably. So obviously my knowledge of
> >> this field is insufficient.
>
> >I don't know what "work with" might mean.
>
> That the text on the screen (and on paper, if needed) looks OK.
>
> >Electronic keyboards other
> >than the standard US keyboard have been available from the beginning.
>
> Nothing to do with keyboards. The program's text files had been
> translated by competent translators and were available in readible
> encodings.
>
> >Code pages for Eastern Europe were available long before Unicode.
>
> Yes. But they proved difficult to handle in programs.
>
> >don't know when Windows began to incorporate Unicode capability, but
> >that was irrelevant to Eastern Europe. I also don't know when MS Word
> >introduced the "Insert Symbol" panel, or when it became possible to
> >assign your own keyboard shortcuts for any character you'd like. I
> >also don't know when MS products began to be available with various
> >operating languages, so that if you wanted your computer to show you
> >all its menus and labels in Polish or Czech you could do that.
>
> It proved very difficult. It seemed Windows Explorer and Word used
> tricks not available in standard API libraries.
You have suddenly descended into the realm of technical details (from
20 years ago, no less). All I know is that I had no trouble typing
East European languages in those days, before Unicode. I simply needed
to have fonts with accented letters in them. Anyone who had the same
fonts could see the texts just as I saw them.
Peter T. Daniels
> Mon, 28 Sep 2009 13:37:50 -0700 (PDT): "Peter T. Daniels"
>
> >> Most likely, the OS just supports the Unicode 5.0 standard, or on this
> >> particular platform, some "improved" version of it, plus certain types
> >> of font interfaces, and then a third party develops the font on its own.
> >> In this case, the OS distribution does not support the font directly.
>
> >It doesn't _provide_ the font. It _supports_ the font if you provide
> >it.
>
> Every Windows I've seen so far (3.10, 3.11, 95, 98, ME, XP, Vista)
> came with a number of fonts. Word (6, 97, 2007) came with several
> more. So yes, they provided me with fonts.
>
> Even MSDOS (3.3, 4.1. 5.1. 6.0) provided fonts.
>
> >> So, if there is an bug in the font, would the bug report go to the OS or
> >> font developers?
>
> >What would a bug in a font be?
>
> A Unicode code point for a letter looking like a is displayed as b.
Has anything like that EVER happened?
Do you even have an example involving the thousands of CJK characters?
> >Gothic is a straightforward, right-to-
> >left alphabet with nothing at all special about it. If someone has
> >drawn an incorrect glyph, obviously that's not the fault of the OS.
>
> It is if it happens in a font that is provided with the OS as
> standard. It's not if it happens in a third-party add-on font.
Has anything like that EVER happened? Do you think MS would do
business with that supplier again if it ever did?
Peter T. Daniels
> Mon, 28 Sep 2009 13:42:35 -0700 (PDT): "Peter T. Daniels"
You simply don't know what you're talking about. Download the free
Phags pa font from the BabelMap site and see if you can use it (with
the contextual forms properly created).
Peter T. Daniels
> Mon, 28 Sep 2009 13:42:35 -0700 (PDT): "Peter T. Daniels"
>
> >Do you have Word2003? It also did not support anything later than
> >Unicode 2.0. Word2007 supports everything through Unicode 5.0, and
> >presumably Word2010 will support Unicode 5.2
>
> In this case (Word), you seem to use "support" in the sense "provided
> proper fonts for", but in the case of Windows, you don't.
Don't make up lies.
I don't even know which fonts came with Windows and which fonts came
with Office. What I know is that Microsoft does not bother to provide
fonts that almost no one in the world has any use for. What I know is
that if I acquire a font for a non-included range that was in Unicode
2.0, then XP and Office2003 can handle it, but they cannot handle
fonts from ranges that were added in later versions of Unicode.
> Why this difference?
>
> >(it wouldn't make great
> >sense for it to be able to do more than the OS, would it?)
>
> If "do" and "support" means "provide fonts for", then this happens all
> the time: Windows usually provides quite a lot of fonts, then Word
> provides many more. So Word provides a lot that the OS didn't
> do/have/support, yes.
Name one Unicode range covered by a font which Office2003 includes
that is not part of Unicode 2.0. Name one Unicode range covered by a
font which Office2007 includes that is not part of Unicode 5.0.
Whereas I can tell you that fonts for neither Burmese nor Khmer are
included with Vista or Office2007, even though both ranges are
included in Unicode 5.0 (and probably all the way back in Unicode 1.0
-- my Unicode 1.0 book vol. 1 is unaccountably not on the shelf beside
vol. 2 where it belongs). Microsoft made the decision that there was
no commercial value in including them in Tahoma or Arial Unicode,
because they didn't expect to sell (many) computers in Burma or
Cambodia. It turns out the Burmese government made (and distributes
for free) a Unicode font in 1996 or so, but it is not fully compliant
with (later?) Unicode standards, so I uninstalled that one and use one
distributed by SIL instead, which works perfectly.
António Marques
> On 2009-09-27, Trond Engen wrote:
>
>> The subject looks like trolling but isn't.
>>
>> In another group, when someone recently opined that some languages are
>> more elevated and thus more fit for abstraction, I answered that
>> linguists agree that all languages are equally fit to express any
>> thought, with the qualification that some may have a pre-existing
>> lexicon in a field but all are equally fit to create new terms for
>> anything from internal resources. I was challenged to provide support
>> for that, and now I realize that if I've ever seen it in print in a
>> linguistics book I can't remember where. Any suggestions?
>
> I think most introductory and popular texts on linguistics say this,
> but I've never seen a "proof" or even an outline or proposal for a
> method of demonstrating it.
>
> I suspect that if someone devised a method of analysing and measuring
> this and got the (unwanted) result --- that all languages are not
> equally versatile --- linguists would go nuts and attack it (or him),
> because this is required dogma. (I'm not claiming it's wrong, just
> that it's something linguistics students are told they have to
> believe.)
That's a completely unfair accusation: saying you know how other people
would behave in case a condition you have no idea how it could be met
would be met. That's about the most condescending and intellectually
dishonest slimy thing that can be done. You, sir, have crossed the line
that separates man from ---.
Besides, it's not even likely. You have no one in this thread arguing
all languages have equal versatility for all purposes. Quite the contrary.
> I'm curious as to how "fit to create new terms for anything from
> internal resources", can be reconciled (redefined?) with the frequency
> of borrowing words from other languages.
What have apples got to do with oranges? English and german are quite
similar in their word-creation abilities, yet the former went the route
of borrowing long ago where the latter has preferred using its internal
mechanisms. Though both have experienced periods of reversal of those
trends.
Harold Johanssen
> On Sep 29, 12:47 am, Harold Johanssen <noem...@please.net> wrote:
>> On Mon, 28 Sep 2009 14:56:18 -0700, António Marques wrote:
>> > On 28 Set, 22:29, Harold Johanssen <noem...@please.net> wrote: - A
>> > language must have certain traits. If it doesn't, it won't be
>> > recognised as a complete language.
>>
>> > - Those traits, among other things, will allow it to become fit for
>> > whatever purpose.
>>
>> > - Ergo, all languages can achieve fitness for the purpose at hand.
>>
>> I see. So it is true by definition - therefore, it is an
>> axiom.
>
> No. As I said before, "[s]o you might say it's on the axiomatic level,
> but not as 'all languages are equally fit' and rather as 'languages
> possess these and these enabling capabilities [which incidentally allow
> them to attain 'fitness']'".
>
> It's not an axiom, it's derived from them (as true propositions usually
> are).
Which ones are the traits, and how do they enable languages in
the way you describe? When you say "Those traits, among other things,
will allow it to become fit for whatever purpose" that "will" implies
that there is a well-defined process to show how the traits enable a
language in this sense, correct?
Christopher Culver
> Doesn't the claim that _literary_ Slovak has a pronunciation strike
> should be pronounceable in any existing dialect.
It's not at all unusual for literary languages to have different
pronunciation than normal speech. In the Finnish spoken language, the
affricate /ts/ and the cluster /hd/ have been (universally?)
simplified to /tt/ and /h/, but people reading a text off a page are
likely to pronounce them as written.
Panu
<crcul...@christopherculver.com> wrote:
Note that both spelling pronunciations are essentially, uhm, artifacts
and don't reflect any vernacular or dialectal pronunciation, anywhere.
Ruud Harmsen
<gram...@verizon.net>: in sci.lang:
>On Sep 29, 3:55�am, Ruud Harmsen <r...@rudhar.eu> wrote:
>> Mon, 28 Sep 2009 13:35:49 -0700 (PDT): "Peter T. Daniels"
>> <gramma...@verizon.net>: in sci.lang:
>>
>> >Operating systems do not provide glyphs. Fonts provide glyphs.
>>
>> Yes. In this case, I assume an OS to include the fonts that are there
>> without post-installing them.
>
>(a) Why would you assume such a thing?
Because since about 1991 that's how I have observed the real world to
be behaving. Perhaps you got your OS'es delivered without any
pre-installed fonts? Live and learn.
>(b) Why would you drive all the type-manufacturers out of business,
>let alone the hobbyists who create fonts for fun?
Do I? It's allright ifthey earn money making fonts, but I rarely
needed them. If I do, maybe I'll be willing to pay for them.
>(c) What about scripts the OS-makers happened not to have
>thought of yet?
Extra post-installed fonts will then come in handy. But that doesn't
contradict my statement "Vista doesn't support Gothic script, which I
know because in the Gothic Wikipedia, I see only rectangles".
>> Or with a narrower sense of "OS": Vista to me is the OS proper plus
>> its standard fonts.
>
>And Vista's standard fonts happen not to include ones that cover the
>Gothic range.
Right. Hence my statement (see above).
>Do you not see why it might be more (commercially)
>desirable to include Hindi and Thai fonts but not
>Gothic or Hanunoo?
Of course I do. And my observation was that Win98 and WinMe didn't
support Hindi (Devanagari), Vista does, but it doesn't support Gothic.
>> >Windows
>> >Vista comes with several fonts that cover large amounts of Unicode
>> >(such as Tahoma and Arial Unicode), but it doesn't come with fonts
>> >that cover _every_ Unicode range, /
>>
>> Right. We agree on that. To me, that is identical with saying "Vista
>> supports large parts of Unicode, but not _every_ range.
>
>You are simply wrong. There are Unicode ranges that Vista _does not_
>support, so that if you install a font covering that range, it doesn't
>know what to do with it. An example is Phags pa, which is a recent
>addition to Unicode.
OK, I believe you when you say so. Only I wonder how that could work
technically. It requires that when a browser program asks the OS to
display certain Unicode scalars it finds in a web page, the OS looks
which range they belongs to, and based on that range decides "what to
do with it". I don't see any technical justification for that. Why not
simply pass on the Unicode scalar, ANY Unicode scalar, to fonts
routines, which then decide if corresponding glyphs are available
given installed and activated fonts?
In other words, what other support should an OS provide for "Unicode
ranges" (which is not the same as support for "Unicode" proper) than
having pre-installed fonts, and allowing additional fonts to be
post-installed, that cover those ranges?
>> >rightly recognizing that most users
>> >of personal computers have no use for Gothic (or Hanunoo).
>>
>> Right. So in my terms, Vista doesn't support Gothic, and cannot
>> properly display the Gothic Wikipedia.
>
>Your terms are nonsensical.
Why? I still haven't seen any explanation from you yet.
Your Vista is what is installed by the standard installated (or was
pre-installed by the vendor of the computer you bought in a shop),
MINUS any pre-installed fonts? Why?
>If the Gothic wikipedia people don't provide a link to a free Gothic
>font download right there on every page, they are being very stupid
>indeed.
No they are not, because it is up to user, browser, OS-included of
third-party fonts, and OS to support Gothic or not, not up to
Wiki-writers.
(BTW. http://got.wikipedia.org/wiki/???????????????????????? does
provides links to Runic script and Latin script versions. The Runic
version also just showed rectangle, the Latin version of the page was
readible, but links were still in Gothic, so contained only
rectangles).
>(The very concept of a Gothic wikipedia is very strange. Who is
>expected to read it?)
I agree, but that's a different subject.
>> >> Perhaps you have better insight in the technical details than I have.
>>
>> >I know no technical details.
>>
>> You use a different meaning of OS than I do. Yours is more like
>> "kernel". I sometimes use that meaning too. But not this time. That
>> causes our misunderstanding each other.
>
>A kernel is something that appears on an ear of corn. A kernel used to
>be the simplest syntactic structure generated by the simplest rewrite
>rules in pre-Aspects transformational grammar.
This shows that your understanding of OS'es and software in general is
simply insufficient for a meaningful discussion of these matters.
>> Right. So in my terms, Vista doesn't support Gothic, and cannot
>> properly display the Gothic Wikipedia.
>
>Your terms are nonsensical.
And because of you fundamental lack of understanding of software, your
above statement is unjustified.
>> Again, I don't understand why anything should be changed in the OS
>> proper (meaning kernel) to support specific definitions of Unicode. So
>> at least half of the above (anything that's not about fonts) is
>> meaningless to me.
>
>Nor do I know any reason "why" Microsoft should expend energy on
>accommodating users of obscure fonts of interest to a dozen or so
>people around the world ...
But they do.
Why does Microsoft incourage people to send e-mails using Office 10,
which in its messages (even by people who only know English, or
perhaps English and Dutch) always includes info about the fonts that
MIGHT be used if the message would ever contain Thai, Tibetan or
Ethopian?
I'm not joking, they really actually do that.
>just because people like Michael Everson
>decide they ought to be included, but they do. It's convenient for me,
>but we made *The World's Writing Systems* without Unicode.
It would be much easier now, with Unicode.
Ruud Harmsen
<gram...@verizon.net>: in sci.lang:
>> >> So, if there is an bug in the font, would the bug report go to the OS or
>> >> font developers?
>>
>> >What would a bug in a font be?
>>
>> A Unicode code point for a letter looking like a is displayed as b.
>
>Has anything like that EVER happened?
I think there have been errors in Rumanian letters with cedilles
versus comma below, yes.
>Has anything like that EVER happened? Do you think MS would do
>business with that supplier again if it ever did?
Third-party font supplier do not need to do business with MS in order
for their fonts being user-installable on top of an MS OS.
Ruud Harmsen
<gram...@verizon.net>: in sci.lang:
>> >It also did not support anything later than
>> >Unicode 2.0. Word2007 supports everything through Unicode 5.0, and
>> >presumably Word2010 will support Unicode 5.2 (it wouldn't make great
>> >sense for it to be able to do more than the OS, would it?)
>>
>> Meaningless question, because the OS doesn't "do" anything, not about
>> ranges. It does something about Unicode proper (any range) and
>> probably about right-to-left and things like that.
>
>You simply don't know what you're talking about.
Do you?
>Download the free
>Phags pa font from the BabelMap site and see if you can use it (with
>the contextual forms properly created).
That could only mean MS OS software is doing unnecessary checks behind
the scenes. Wouldn't surprise me, isn't the first time. One of the
reasons why computers are still slow although the hardware is now over
a 1000 times faster than it was in 1990.
Ruud Harmsen
<gram...@verizon.net>: in sci.lang:
>On Sep 29, 4:07�am, Ruud Harmsen <r...@rudhar.eu> wrote:
>> Mon, 28 Sep 2009 13:42:35 -0700 (PDT): "Peter T. Daniels"
>> <gramma...@verizon.net>: in sci.lang:
>>
>> >Do you have Word2003? It also did not support anything later than
>> >Unicode 2.0. Word2007 supports everything through Unicode 5.0, and
>> >presumably Word2010 will support Unicode 5.2
>>
>> In this case (Word), you seem to use "support" in the sense "provided
>> proper fonts for", but in the case of Windows, you don't.
>
>Don't make up lies.
Where is the lie and why is it a lie?
How can a statement containing the verb "to seem" be a lie? It
describes an impression I got, how can it be up to you to deny I got
that impression (whether false or accurate)?
If I say it seem to be so TO ME, that's how it is.
>I don't even know which fonts came with Windows and which fonts came
>with Office. What I know is that Microsoft does not bother to provide
>fonts that almost no one in the world has any use for. What I know is
>that if I acquire a font for a non-included range that was in Unicode
>2.0, then XP and Office2003 can handle it, but they cannot handle
>fonts from ranges that were added in later versions of Unicode.
>
>> Why this difference?
>>
>> >(it wouldn't make great
>> >sense for it to be able to do more than the OS, would it?)
>>
>> If "do" and "support" means "provide fonts for", then this happens all
>> the time: Windows usually provides quite a lot of fonts, then Word
>> provides many more. So Word provides a lot that the OS didn't
>> do/have/support, yes.
>
>Name one Unicode range covered by a font which Office2003 includes
>that is not part of Unicode 2.0. Name one Unicode range covered by a
>font which Office2007 includes that is not part of Unicode 5.0.
>
>Whereas I can tell you that fonts for neither Burmese nor Khmer are
>included with Vista or Office2007,
So Vista and Office2007 do not support Burmese of Khmer script. I said
the same about Gothic, and you called that nonsense. Why? What the
difference between what I said (about Gothic) and what you said (about
Burmese and Khmer)?
>even though both ranges are
>included in Unicode 5.0 (and probably all the way back in Unicode 1.0
>-- my Unicode 1.0 book vol. 1 is unaccountably not on the shelf beside
>vol. 2 where it belongs). Microsoft made the decision that there was
>no commercial value in including them in Tahoma or Arial Unicode,
>because they didn't expect to sell (many) computers in Burma or
>Cambodia. It turns out the Burmese government made (and distributes
>for free) a Unicode font in 1996 or so, but it is not fully compliant
>with (later?) Unicode standards, so I uninstalled that one and use one
>distributed by SIL instead, which works perfectly.
How does any of this contradict any of my statements?
António Marques
> On Mon, 28 Sep 2009 17:27:06 -0700, António Marques wrote:
>
>> On Sep 29, 12:47 am, Harold Johanssen<noem...@please.net> wrote:
>>> On Mon, 28 Sep 2009 14:56:18 -0700, António Marques wrote:
>>>> On 28 Set, 22:29, Harold Johanssen<noem...@please.net> wrote: - A
>>>> language must have certain traits. If it doesn't, it won't be
>>>> recognised as a complete language.
>>>
>>>> - Those traits, among other things, will allow it to become fit for
>>>> whatever purpose.
>>>
>>>> - Ergo, all languages can achieve fitness for the purpose at hand.
>>>
>>> I see. So it is true by definition - therefore, it is an
>>> axiom.
>>
>> No. As I said before, "[s]o you might say it's on the axiomatic level,
>> but not as 'all languages are equally fit' and rather as 'languages
>> possess these and these enabling capabilities [which incidentally allow
>> them to attain 'fitness']'".
>>
>> It's not an axiom, it's derived from them (as true propositions usually
>> are).
>
> Which ones are the traits, and how do they enable languages in
> the way you describe?
Do you really want me to enumerate the porperties of human language (as
opposed, say, to the 'language' of traffic signs)?
> When you say "Those traits, among other things,
> will allow it to become fit for whatever purpose" that "will" implies
> that there is a well-defined process to show how the traits enable a
> language in this sense, correct?
No, you keep centering it on 'this sense', which isn't at all how it's
done. To keep on with analogies, cars have the ability to move over
smooth surfaces. That means they can eventually get from Sweden to
Thailand. It doesn't mean that someone actually keeps in their mind a
well-defined picture of how it can be done.
If you want to play the game, give an example of 'fitness' and ask how
some hypothetical language can attain it.
Peter T. Daniels
> Harold Johanssen wrote:
> > On Mon, 28 Sep 2009 17:27:06 -0700, António Marques wrote:
>
> >> On Sep 29, 12:47 am, Harold Johanssen<noem...@please.net> wrote:
> >>> On Mon, 28 Sep 2009 14:56:18 -0700, António Marques wrote:
> >>>> On 28 Set, 22:29, Harold Johanssen<noem...@please.net> wrote: - A
> >>>> language must have certain traits. If it doesn't, it won't be
> >>>> recognised as a complete language.
>
> >>>> - Those traits, among other things, will allow it to become fit for
> >>>> whatever purpose.
>
> >>>> - Ergo, all languages can achieve fitness for the purpose at hand.
>
> >>> I see. So it is true by definition - therefore, it is an
> >>> axiom.
>
> >> No. As I said before, "[s]o you might say it's on the axiomatic level,
> >> but not as 'all languages are equally fit' and rather as 'languages
> >> possess these and these enabling capabilities [which incidentally allow
> >> them to attain 'fitness']'".
>
> >> It's not an axiom, it's derived from them (as true propositions usually
> >> are).
>
> > Which ones are the traits, and how do they enable languages in
> > the way you describe?
>
> Do you really want me to enumerate the porperties of human language (as
> opposed, say, to the 'language' of traffic signs)?
You don't need to; C. F. Hockett did so, beginning around 1958. I
think his final version is in his 1977 collection *The View from
Language*, which contains annotated versions of a number of important
articles. He always got really annoyed when people cited the September
1960 Scientific American version, because that was a non-technical
treatment for the general public.
You can presumably find full discussions of his proposals by googling
"design features of language" and (if necessary) his name.
> > When you say "Those traits, among other things,
> > will allow it to become fit for whatever purpose" that "will" implies
> > that there is a well-defined process to show how the traits enable a
> > language in this sense, correct?
>
> No, you keep centering it on 'this sense', which isn't at all how it's
> done. To keep on with analogies, cars have the ability to move over
> smooth surfaces. That means they can eventually get from Sweden to
> Thailand. It doesn't mean that someone actually keeps in their mind a
> well-defined picture of how it can be done.
>
> If you want to play the game, give an example of 'fitness' and ask how
> some hypothetical language can attain it.-
Peter T. Daniels
> Tue, 29 Sep 2009 05:14:08 -0700 (PDT): "Peter T. Daniels"
>
> >On Sep 29, 3:55 am, Ruud Harmsen <r...@rudhar.eu> wrote:
> >> Mon, 28 Sep 2009 13:35:49 -0700 (PDT): "Peter T. Daniels"
> >> <gramma...@verizon.net>: in sci.lang:
>
> >> >Operating systems do not provide glyphs. Fonts provide glyphs.
>
> >> Yes. In this case, I assume an OS to include the fonts that are there
> >> without post-installing them.
>
> >(a) Why would you assume such a thing?
>
> Because since about 1991 that's how I have observed the real world to
> be behaving. Perhaps you got your OS'es delivered without any
> pre-installed fonts? Live and learn.
Early computers had the basic ASCII range (0-127) built in. The
typography was determined strictly by the printer hardware. Miguel
Civil (the Sumerologist) hooked up a Selectric typewriter to his TRS
box.
My first computer (Kaypro 4/84) had a daisy-wheel printer.
My second computer had a dot-matrix (8-dot, IIRC) printer.
I don't remember whether I used the same printer with my third
computer, which was an early "portable."
My fourth computer was a 1992 Macintosh, provided by OUP so I could
produce WWS, because DOS could still not deal with typography. The
system included a laser printer.
I now use a PC, thanks to my employer, who put one in my house, though
no longer that same one. Microsoft adopted Apple's approach to
graphics and typography, eventually.
MS products come with a fairly large range of fonts -- the vast
majority of which are "display fonts" that are of no use whatsoever to
me. MS's range of fonts is far more than adequate for the vast
majority of their customers, even those who might need to write to
South Asia or East Asia or Thailand (though not the rest of Southeast
Asia).
> >(b) Why would you drive all the type-manufacturers out of business,
> >let alone the hobbyists who create fonts for fun?
>
> Do I? It's allright ifthey earn money making fonts, but I rarely
> needed them. If I do, maybe I'll be willing to pay for them.
You think fonts should be provided by OS-makers.
> >(c) What about scripts the OS-makers happened not to have
> >thought of yet?
>
> Extra post-installed fonts will then come in handy. But that doesn't
> contradict my statement "Vista doesn't support Gothic script, which I
> know because in the Gothic Wikipedia, I see only rectangles".
Then, as I have been saying, you are using the word "support" in a
_very_ peculiar way.
Doubtless my computer supports all sorts of games. But I don't have
any beyond the few solitaires that came with it.
> >> Or with a narrower sense of "OS": Vista to me is the OS proper plus
> >> its standard fonts.
>
> >And Vista's standard fonts happen not to include ones that cover the
> >Gothic range.
>
> Right. Hence my statement (see above).
>
> >Do you not see why it might be more (commercially)
> >desirable to include Hindi and Thai fonts but not
> >Gothic or Hanunoo?
>
> Of course I do. And my observation was that Win98 and WinMe didn't
> support Hindi (Devanagari), Vista does, but it doesn't support Gothic.
Of course it does. If you had a Gothic-encoded font, you could read
the Gothic wikipedia. But if you had a Phags pa-encoded font, you
couldn't read the Phags pa wikipedia, because Vista does not support
Phags pa.
> >> >Windows
> >> >Vista comes with several fonts that cover large amounts of Unicode
> >> >(such as Tahoma and Arial Unicode), but it doesn't come with fonts
> >> >that cover _every_ Unicode range, /
>
> >> Right. We agree on that. To me, that is identical with saying "Vista
> >> supports large parts of Unicode, but not _every_ range.
>
> >You are simply wrong. There are Unicode ranges that Vista _does not_
> >support, so that if you install a font covering that range, it doesn't
> >know what to do with it. An example is Phags pa, which is a recent
> >addition to Unicode.
>
> OK, I believe you when you say so. Only I wonder how that could work
> technically. It requires that when a browser program asks the OS to
> display certain Unicode scalars it finds in a web page, the OS looks
> which range they belongs to, and based on that range decides "what to
> do with it". I don't see any technical justification for that. Why not
> simply pass on the Unicode scalar, ANY Unicode scalar, to fonts
> routines, which then decide if corresponding glyphs are available
> given installed and activated fonts?
Now it sounds like you don't understand how Unicode works. The first
500+ pages of the Unicode 5.0 book set forth the details of how _each
individual range_ of characters needs to be implemented in the OS.
> In other words, what other support should an OS provide for "Unicode
> ranges" (which is not the same as support for "Unicode" proper) than
> having pre-installed fonts, and allowing additional fonts to be
> post-installed, that cover those ranges?
Once you get beyond the simple alphabets, the handling of scripts
becomes more and more complex. Adjacent characters combine, characters
appear in non-linear orders, forms vary contextually. Every single
feature of every single script has to be encoded in the OS for you to
be able to type in a language without thinking about such things.
> >> >rightly recognizing that most users
> >> >of personal computers have no use for Gothic (or Hanunoo).
>
> >> Right. So in my terms, Vista doesn't support Gothic, and cannot
> >> properly display the Gothic Wikipedia.
>
> >Your terms are nonsensical.
>
> Why? I still haven't seen any explanation from you yet.
> Your Vista is what is installed by the standard installated (or was
> pre-installed by the vendor of the computer you bought in a shop),
> MINUS any pre-installed fonts? Why?
Now you are making no sense whatsoever.
> >If the Gothic wikipedia people don't provide a link to a free Gothic
> >font download right there on every page, they are being very stupid
> >indeed.
>
> No they are not, because it is up to user, browser, OS-included of
> third-party fonts, and OS to support Gothic or not, not up to
> Wiki-writers.
If they want their stuff to be read, and they use an arcane script,
they need to provide access to that script.
Virtually every ethnic-solidarity-group's web page offers free font
downloads. That's how I found a nice Pahawh Hmong font, for instance.
> (BTW.http://got.wikipedia.org/wiki/????????????????????????does
> provides links to Runic script and Latin script versions. The Runic
> version also just showed rectangle, the Latin version of the page was
> readible, but links were still in Gothic, so contained only
> rectangles).
>
> >(The very concept of a Gothic wikipedia is very strange. Who is
> >expected to read it?)
>
> I agree, but that's a different subject.
>
> >> >> Perhaps you have better insight in the technical details than I have.
>
> >> >I know no technical details.
>
> >> You use a different meaning of OS than I do. Yours is more like
> >> "kernel". I sometimes use that meaning too. But not this time. That
> >> causes our misunderstanding each other.
>
> >A kernel is something that appears on an ear of corn. A kernel used to
> >be the simplest syntactic structure generated by the simplest rewrite
> >rules in pre-Aspects transformational grammar.
>
> This shows that your understanding of OS'es and software in general is
> simply insufficient for a meaningful discussion of these matters.
I know how the fonts work. I don't need to know anything else about
OSs, any more than I need to know how the thermostat in my window air
conditioner works..
> >> Right. So in my terms, Vista doesn't support Gothic, and cannot
> >> properly display the Gothic Wikipedia.
>
> >Your terms are nonsensical.
>
> And because of you fundamental lack of understanding of software, your
> above statement is unjustified.
>
> >> Again, I don't understand why anything should be changed in the OS
> >> proper (meaning kernel) to support specific definitions of Unicode. So
> >> at least half of the above (anything that's not about fonts) is
> >> meaningless to me.
>
> >Nor do I know any reason "why" Microsoft should expend energy on
> >accommodating users of obscure fonts of interest to a dozen or so
> >people around the world ...
>
> But they do.
>
> Why does Microsoft incourage people to send e-mails using Office 10,
> which in its messages (even by people who only know English, or
> perhaps English and Dutch) always includes info about the fonts that
> MIGHT be used if the message would ever contain Thai, Tibetan or
> Ethopian?
"Office 10"? Is that even supported any more? Office2007 is Office 12.
> I'm not joking, they really actually do that.
>
> >just because people like Michael Everson
> >decide they ought to be included, but they do. It's convenient for me,
> >but we made *The World's Writing Systems* without Unicode.
>
> It would be much easier now, with Unicode.
To an extent. But Unicode is much less flexible than being able to
fire up Fontographer and add to a font some contextual variety of some
character that we happened not to think of the first time around.
Peter T. Daniels
> Tue, 29 Sep 2009 05:15:58 -0700 (PDT): "Peter T. Daniels"
>
> >> >> So, if there is an bug in the font, would the bug report go to the OS or
> >> >> font developers?
>
> >> >What would a bug in a font be?
>
> >> A Unicode code point for a letter looking like a is displayed as b.
>
> >Has anything like that EVER happened?
>
> I think there have been errors in Rumanian letters with cedilles
> versus comma below, yes.
When I was dealing with Latvian and Lithuanian, I went the the (no
longer extant) Baltic and Slavic Division of the New York Public
Library to discover exactly what the diacritics -- especially on the g
-- should _really_ look like. There is no consistency among the many
reference works on the open shelves -- and the differences do not
correlate to the pre-Soviet, Soviet, and post-Soviet periods of Baltic
history.
In fact, most East European printing -- during _all_ those periods --
was of a quality such that cedilla, comma, or ogonek could often not
even be identified. The question may thus be undecidable for Rumanian,
as well. (There was some fine printing in Imperial Russia, but that's
not relevant to any of these cases.)
> >Has anything like that EVER happened? Do you think MS would do
> >business with that supplier again if it ever did?
>
> Third-party font supplier do not need to do business with MS in order
> for their fonts being user-installable on top of an MS OS.
But you decreed that all fonts must be supplied along with the OS.
Peter T. Daniels
> Tue, 29 Sep 2009 05:17:28 -0700 (PDT): "Peter T. Daniels"
No, it means that Windows Vista is not set up to accommodate the
special contextual rules needed for Phags pa.
Peter T. Daniels
> Tue, 29 Sep 2009 05:27:07 -0700 (PDT): "Peter T. Daniels"
> >On Sep 29, 4:07 am, Ruud Harmsen <r...@rudhar.eu> wrote:
> >> Mon, 28 Sep 2009 13:42:35 -0700 (PDT): "Peter T. Daniels"
> >> <gramma...@verizon.net>: in sci.lang:
>
> >> >Do you have Word2003? It also did not support anything later than
> >> >Unicode 2.0. Word2007 supports everything through Unicode 5.0, and
> >> >presumably Word2010 will support Unicode 5.2
>
> >> In this case (Word), you seem to use "support" in the sense "provided
> >> proper fonts for", but in the case of Windows, you don't.
>
> >Don't make up lies.
>
> Where is the lie and why is it a lie?
That I use "support" in different ways.
Of course they support Burmese and Khmer -- I can show you the texts I
typed in them. I just had to locate fonts for them first.
> >even though both ranges are
> >included in Unicode 5.0 (and probably all the way back in Unicode 1.0
> >-- my Unicode 1.0 book vol. 1 is unaccountably not on the shelf beside
> >vol. 2 where it belongs). Microsoft made the decision that there was
> >no commercial value in including them in Tahoma or Arial Unicode,
> >because they didn't expect to sell (many) computers in Burma or
> >Cambodia. It turns out the Burmese government made (and distributes
> >for free) a Unicode font in 1996 or so, but it is not fully compliant
> >with (later?) Unicode standards, so I uninstalled that one and use one
> >distributed by SIL instead, which works perfectly.
>
> How does any of this contradict any of my statements?
You claim (falsely) that Vista does not support Burmese and Khmer.
Ruud Harmsen
<gram...@verizon.net>: in sci.lang:
>On Sep 28, 7:43�am, Ruud Harmsen <r...@rudhar.eu> wrote:
>
>> See the Wikipedia pages in languages like Latin, Old-English and
>> Gothic.http://got.wikipedia.org/wiki
>> (Unfortunately, my Vista doesn't support the script!)
>
>No, it just means you don't happen to have a font with Unicode-encoded
>Gothic in it. The three I have are AlphabetumUnicode (which Juan-Jose
>Marcos will sell you cheap; he's a professor of linguistics in
>northern Spain), Code2001 (free download), and MPH 28 Damase /
After downloading and installing Code2001 from
http://www.code2000.net/code2001.htm, I tested
http://got.wikipedia.org . This works well in all four browsers I
tried: Explorer 8.0.6001.18813, Firefox 3.5.3, Google Chrome
3.0.195.21 and Opera 9.64.
When testing with my own http://rudhar.com/cgi-bin/shunicod.cgi
however, starting at 10000 or 10300, with the max. range of 2048, it
worked in Firefox and Chrome, but NOT in Explorer and Opera. This may
be because the first two are smart enough to continue to other fonts
if the current font contains nothing for the requested Unicode scalar.
How Wikipedia manages to get it work anyway, I don't know.
António Marques
> On Sep 29, 10:27 am, Ruud Harmsen<r...@rudhar.eu> wrote:
>> Tue, 29 Sep 2009 05:15:58 -0700 (PDT): "Peter T. Daniels"
>> <gramma...@verizon.net>: in sci.lang:
>>
>>>>>> So, if there is an bug in the font, would the bug report go to the OS or
>>>>>> font developers?
>>
>>>>> What would a bug in a font be?
>>
>>>> A Unicode code point for a letter looking like a is displayed as b.
>>
>>> Has anything like that EVER happened?
>>
>> I think there have been errors in Rumanian letters with cedilles
>> versus comma below, yes.
>
> When I was dealing with Latvian and Lithuanian, I went the the (no
> longer extant) Baltic and Slavic Division of the New York Public
> Library to discover exactly what the diacritics -- especially on the g
> -- should _really_ look like. There is no consistency among the many
> reference works on the open shelves -- and the differences do not
> correlate to the pre-Soviet, Soviet, and post-Soviet periods of Baltic
> history.
>
> In fact, most East European printing -- during _all_ those periods --
> was of a quality such that cedilla, comma, or ogonek could often not
> even be identified. The question may thus be undecidable for Rumanian,
> as well. (There was some fine printing in Imperial Russia, but that's
> not relevant to any of these cases.)
Romanians insist that their letters are t-comma and s-comma, but the
Unicde 'experts' originally decided that those should be merely a
typeface variant of s-cedilla (which was already provided for turkish)
and t-cedilla (which didn't exist before but was created for this
purpose). In later revisions, s-comma and t-comma were added to the
standard, but in the meanwhile romanians have had to use the -cedilla
forms in a lot of material, furthering the confusion. T-cedilla, which
could be thrown out now, remains.
Ruud Harmsen
<gram...@verizon.net>: in sci.lang:
>MS products come with a fairly large range of fonts -- the vast
>majority of which are "display fonts" that are of no use whatsoever to
>me.
Truetype fonts can be used for displays and printers equally well.
That's because they're vector fonts, not raster fonts.
>> >(b) Why would you drive all the type-manufacturers out of business,
>> >let alone the hobbyists who create fonts for fun?
>>
>> Do I? It's allright ifthey earn money making fonts, but I rarely
>> needed them. If I do, maybe I'll be willing to pay for them.
>
>You think fonts should be provided by OS-makers.
No I don't, I just observe that many are. I said nothing about what
should happen, only about what I see happens.
I personally think an OS should contain only three fonts, similar to
Arial, Times New Romand and Courier New, supporting as many Unicode
scalars as possible.
Anything else should belong to text processors and independent font
designers.
But the simple fact is that Windows versions have always provided many
more than just three fonts. So be it, I don't really mind.
>> >(c) What about scripts the OS-makers happened not to have
>> >thought of yet?
>>
>> Extra post-installed fonts will then come in handy. But that doesn't
>> contradict my statement "Vista doesn't support Gothic script, which I
>> know because in the Gothic Wikipedia, I see only rectangles".
>
>Then, as I have been saying, you are using the word "support" in a
>_very_ peculiar way.
I'm sure nearly everybody I know uses it that way, except you.
>Doubtless my computer supports all sorts of games. But I don't have
>any beyond the few solitaires that came with it.
That is exactly the same way _I_ use the terms. It means Vista
supports solitaire but not chess. You can however install several free
and commercial chess programs if you wish.
So the situation is exactly the same with games as with fonts. Thanks
for the clarifying analogy.
(Maybe I should say: "Vista features solitaire but not chess." Or
"comes with".)
>> Of course I do. And my observation was that Win98 and WinMe didn't
>> support Hindi (Devanagari), Vista does, but it doesn't support Gothic.
>
>Of course it does. If you had a Gothic-encoded font, you could read
>the Gothic wikipedia.
Meanwhile I have. So now my enhanced Vista includes, supports,
features, has and does Gothic. Great!
>But if you had a Phags pa-encoded font, you couldn't read the
>Phags pa wikipedia, because Vista does not support Phags pa.
I'll be testing that soon.
>> OK, I believe you when you say so. Only I wonder how that could work
>> technically. It requires that when a browser program asks the OS to
>> display certain Unicode scalars it finds in a web page, the OS looks
>> which range they belongs to, and based on that range decides "what to
>> do with it". I don't see any technical justification for that. Why not
>> simply pass on the Unicode scalar, ANY Unicode scalar, to fonts
>> routines, which then decide if corresponding glyphs are available
>> given installed and activated fonts?
>
>Now it sounds like you don't understand how Unicode works. The first
>500+ pages of the Unicode 5.0 book set forth the details of how _each
>individual range_ of characters needs to be implemented in the OS.
Details about left-to-right, ligatures, kerning etc.?
Any online references where I can read that? I'm not going to buy
paper Unicode books.
>Once you get beyond the simple alphabets, the handling of scripts
>becomes more and more complex. Adjacent characters combine,
>characters appear in non-linear orders, forms vary contextually.
OK, I agree that such things may need support in the OS.
>Every single feature of every single script has to be encoded
>in the OS for you to be able to type in a language without
>thinking about such things.
Typing is a different subject. We're still on displaying.
>> Why? I still haven't seen any explanation from you yet.
>> Your Vista is what is installed by the standard installated (or was
>> pre-installed by the vendor of the computer you bought in a shop),
>> MINUS any pre-installed fonts? Why?
>
>Now you are making no sense whatsoever.
You've finally explained what you mean by 'support'.
>> >If the Gothic wikipedia people don't provide a link to a free Gothic
>> >font download right there on every page, they are being very stupid
>> >indeed.
>>
>> No they are not, because it is up to user, browser, OS-included of
>> third-party fonts, and OS to support Gothic or not, not up to
>> Wiki-writers.
In fact, they do link to fonts:
===
Gothic is written in an alphabet of its own. You shall find fonts
there
[http://got.wikipedia.org/wiki/Wikipedia:Dala%C3%BEla%C3%BEan_Gutiska_Meljanboeka],
keyboard help here
[http://got.wikipedia.org/wiki/Wikipedia:Gutiska_Soetelkartus] and
transliteration rules here
[http://got.wikipedia.org/wiki/Wikipedia:Alja-Kuns_Waurds].
/===
>> Why does Microsoft incourage people to send e-mails using Office 10,
>> which in its messages (even by people who only know English, or
>> perhaps English and Dutch) always includes info about the fonts that
>> MIGHT be used if the message would ever contain Thai, Tibetan or
>> Ethopian?
>
>"Office 10"? Is that even supported any more? Office2007 is Office 12.
OK. I mean any Outlook (not to be confused with Outlook Express) that
is part of any Office version. They produce hopelessly superfluous and
bloated e-mail messages.
>To an extent. But Unicode is much less flexible than being able to
>fire up Fontographer and add to a font some contextual variety of some
>character that we happened not to think of the first time around.
If Fontographer can create TrueType or OpenType fonts, couldn't you
still do that?
Ruud Harmsen
<gram...@verizon.net>: in sci.lang:
>No, it means that Windows Vista is not set up to accommodate the
>special contextual rules needed for Phags pa.
If you'd said that 15 messages ago, we could have each other lots of
meaningless non-discussion messages.
I know nothing about Tibetan, you do.
Joachim Pense
> On Mon, 28 Sep 2009 18:27:05 +0100, António Marques wrote:
>
>> Harold Johanssen wrote:
>>> On Sun, 27 Sep 2009 17:38:35 -0700, Peter T. Daniels wrote:
>>>
>>>> On Sep 27, 5:49 pm, Trond Engen<trond...@engen.priv.no> wrote:
>>>>> The subject looks like trolling but isn't.
>>>>>
>>>>> In another group, when someone recently opined that some languages
>>>>> are more elevated and thus more fit for abstraction, I answered that
>>>>> linguists agree that all languages are equally fit to express any
>>>>> thought, with the qualification that some may have a pre-existing
>>>>> lexicon in a field but all are equally fit to create new terms for
>>>>> anything from internal resources. I was challenged to provide support
>>>>> for that, and now I realize that if I've ever seen it in print in a
>>>>> linguistics book I can't remember where. Any suggestions?
>>>>
>>>> Surely it's in every intro book on linguistics for the gen pub?
>>>> Anything by David Crystal or Jean Aitchison?
>>>
>>> I would be interested to understand the status of this further. I
>>> mean, is this something that can be proven, or are we talking an axiom
>>> accepted without proof by the academic community?
>>
>> It's not an axiom nor dogma. It's just an observation: every known
>> language has devices that enable it to reach 'fitness' in any given
>> domain. If one hadn't, it probably woulsn't be considered a complete
>> language.
>
> Hmm... Let me play devil's advocate here. That comes across as
> circular reasoning: All human languages are equally fit, because if there
> is one that is not equally fit then it is not a human languages, for all
> languages are equally fit.
>
> Anyway, I find it acceptable that all languages spoken by people,
> and studied so far, have been found to be equally fit to express any
> thought potentially. It would seem, however, that what we have is
> empirical evidence alone, correct?
Do we actually have empirical evidence, or do we have only results of
philosophical reasoning?
Joachim
Ruud Harmsen
<gram...@verizon.net>: in sci.lang:
>That I use "support" in different ways.
You use it in a very special way (namely: to implement the
peculiarities of lots of exotic scripts) without explaining that and
without realising that I didn't understand that. That caused us to
talk past each other for a long time.
>> So Vista and Office2007 do not support Burmese of Khmer script. I said
>> the same about Gothic, and you called that nonsense. Why? What the
>> difference between what I said (about Gothic) and what you said (about
>> Burmese and Khmer)?
>
>Of course they support Burmese and Khmer -- I can show you the texts I
>typed in them. I just had to locate fonts for them first.
Your very specialised meaning of "support" again. Now I understand
what you mean: Khmer and Burmese are pretty straightforward, so Vista
can handle them as standard.
>> How does any of this contradict any of my statements?
>
>You claim (falsely) that Vista does not support Burmese and Khmer.
Only with your very specialised meaning of "support".