There are 23 messages in this issue.
Topics in this digest:
1a. Re: Information Density and Comprehension Speed
From: Matthew Turnbull
1b. Re: Information Density and Comprehension Speed
From: And Rosta
1c. Re: Information Density and Comprehension Speed
From: Gary Shannon
1d. Re: Information Density and Comprehension Speed
From: Jim Henry
1e. Re: Information Density and Comprehension Speed
From: Dale McCreery
1f. Re: Information Density and Comprehension Speed
From: Logan Kearsley
2a. How Does Everyone Translate?
From: Sam Stutter
2b. Re: How Does Everyone Translate?
From: Koppa Dasao
2c. Re: How Does Everyone Translate?
From: Charlie Brickner
2d. Re: How Does Everyone Translate?
From: Gary Shannon
2e. Re: How Does Everyone Translate?
From: Roger Mills
2f. Re: How Does Everyone Translate?
From: Matthew Turnbull
2g. Re: How Does Everyone Translate?
From: Logan Kearsley
3a. Re: Conlang Textbook Template
From: Gary Shannon
3b. Re: Conlang Textbook Template
From: Sam Stutter
3c. Re: Conlang Textbook Template
From: Gary Shannon
3d. Re: Conlang Textbook Template
From: Ben Scerri
3e. Re: Conlang Textbook Template
From: Logan Kearsley
3f. Re: Conlang Textbook Template
From: Ben Scerri
3g. Re: Conlang Textbook Template
From: Alex Fink
3h. Re: Conlang Textbook Template
From: Sam Stutter
4a. Re: Barsoomian Project
From: Fredrik Ekman
4b. Re: Barsoomian Project
From: Peter Cyrus
Messages
________________________________________________________________________
1a. Re: Information Density and Comprehension Speed
Posted by: "Matthew Turnbull" [email protected]
Date: Fri Dec 23, 2011 6:48 am ((PST))
Lol, never heard that one, I've heard people from the states say
canadians speak too fast, but in my experience we speak at about the
same speed as long as it's a plains dialect of english. And also I
think people tend to hear speach they have difficulty processing as
"faster". Of course then again every time I've heard someone speak
Cree or Ojibwe I've thought they were just drawling along, and it
feels like I have to slow my speach down just to pronouce anything
right. (something I don't have to do with Spanish, and with French I
think actually speed up).
On 12/23/11, Sam Stutter <[email protected]> wrote:
> On 23 Dec 2011, at 05:28, Dale McCreery <[email protected]> wrote:
>
>> I've listened to a lot of guys from eastern Canada who spoke at what
>> seemed to be at least twice as fast as I did at the time, and I was a fast
>> speaker.
>
> Isn't there a joke that Canadians. Speak. Very. Slowly? :)
>
>> Sort of based on that assumption (that some people speak all the
>> time much faster) I have a question and a supposition.
>>
>> First - do we speak anywhere nears as fast as we read?
>
> Largely, I'd say no. I'd also say I have three reading speeds: skimming,
> reading fully and "trying sounds out in my head". Skimming is probably twice
> as fast as speaking, reading fully about 50% faster and sounding exactly the
> same speed.
>
>> It seems to me
>> that while reading speed might max out based on purely cognitive limits,
>> spoken speed could very easily face other types of limitations...
>>
>> And secondly, I had the general impression that most of the guys I heard
>> speaking English ridiculously fast used a lot of set phrases, sort of
>> poetic turns of speech, etc., and was wondering if there have any been any
>> studies evaluating language speed as it related to the density of set
>> phrases and structures? I would suspect that the more poetic or formulaic
>> the speech (in a community where this style was heavily established) the
>> faster the general flow of speech would be...
>>
>
> Horse racing bookies are scarily fast.
>
>> -dale-
>
--
Sent from my mobile device
Messages in this topic (14)
________________________________________________________________________
1b. Re: Information Density and Comprehension Speed
Posted by: "And Rosta" [email protected]
Date: Fri Dec 23, 2011 7:14 am ((PST))
Logan Kearsley, On 23/12/2011 04:38:
> On 22 December 2011 18:34, And Rosta<[email protected]> wrote:
>> Logan Kearsley, On 22/12/2011 18:30:
>>> This has interesting implications for how fast humans generically are
>>> able to absorb information, which I'm sure will turn out to be useful
>>> to know for things like signage and human/computer interface design.
>>> From a conlanging perspective, though, this discovery has seriously
>>> updated (liberalized, I might say) some of my opinions on good
>>> language design. I'm a big fan of concision, making use of multiple
>>> channels (e.g., syllables+syntax+prosody), keeping words optimally
>>> short, etc. But it turns out that that all really makes no difference
>>> to have quickly you can communicate in speech or writing. So, making
>>> your words optimally short may be a fun project, but it turns out to
>>> be totally unnecessary for concision of speech
>>
>> How do you reach these conclusions? The web articles you point to claim that
>> information is transmitted at a constant rate. From that the conlanger may
>> conclude that there is no concision to be gained in speech from maximizing
>> the number of paradigmatic phonological contrasts; IOW the more bits of data
>> transmitted simultaneously, the slower the transmission rate. But you seem
>> to be leaping to the further unwarranted conclusion that there is no
>> concision to be gained from a redundancy-minimizing encoding that minimizes
>> the number of bits per word.
>
> I don't see how you gained that interpretation from what I wrote.
> Making words optimally short (maximizing bits per segment) and
> eliminating redundancy (minimizing the number of bits that are
> transmitted at all) are entirely different matters.
"Making words optimally short" is ambiguous. You were talking about comparing
phonologies, comparing how many bits per segment they encode, and then
proceeded to compare word lengths across phonologies in terms of number of
segments per word. I'd mistakenly thought you had proceeded to compare the
lengths of alternate phonological forms for a given word within a given
phonology, i.e. comparing the number of bits per candidate form; "making words
optimally short" to me meant "minimizing (or optimizing) the number of bits per
word".
Returning to the main point of the thread, I am reluctant to accept -- given
the currently available evidence -- that the rate of encoded-information
transmission is necessarily constant (-- your message was of course about the
rate of transmission of encoding-information, not encoded-information). I'm
reluctant in practice because the currently available evidence points to the
contrary conclusion: encodings can vary in their efficiency, and I know of no
evidence to indicate that human languages necessarily converge on encodings of
the same efficiency. And I'm reluctant in principle because to me the most
intellectually (rather than merely artistically & aesthetically) interesting
aspect of conlanging pertains to investigating the practicability of a loglang
(logical language), where a logical language is one that encodes logical form
fully and unambiguously while being as concise as natlangs -- in technical
terms, the full explicature should be unambiguous encoded; so this
necessarily means that the loglang has to achieve a greater rate of information
transmission than natlangs. (In practice I understand that to mean that the
loglang has to find a more efficient encoding, not that the bitrate has to be
increased; upping the bitrate would be essentially an extralinguistic solution,
and would not address the question of whether a human language can be a
loglang.) Now this research programme may eventually prove that it is
impossible, but it can't prove anything if the task is prematurely judged to be
impossible.
--And.
Messages in this topic (14)
________________________________________________________________________
1c. Re: Information Density and Comprehension Speed
Posted by: "Gary Shannon" [email protected]
Date: Fri Dec 23, 2011 8:30 am ((PST))
On Fri, Dec 23, 2011 at 6:47 AM, Matthew Turnbull <[email protected]> wrote:
> ... And also I
> think people tend to hear speach they have difficulty processing as
> "faster".
This will be old hat to polyglots, but it was a fascinating revelation
for me. I finally decided it was time for me to learn a second
language, so I took up the study of Spanish about 11 months ago. When
I listened to recordings of spoken Spanish it sounded impossibly fast
to me and I was sure that I would never be able to listen that fast,
let alone speak that fast.
So I put the MP3 files I'd downloaded aside and went on with my book
learning for 11 months.
Then, a few weeks ago, I stumbled across those MP3 files from 11
months ago and decided to listen to them to see if I could make out
any words yet. Those same recordings that were impossibly fast 11
months ago are almost leisurely now. What was just a blur of sounds is
now a sequence of clearly enunciated words.
There are only two possibilities. Either somebody snuck into my
computer and replaced the old MP3 recordings with slowed-down
versions, or back 11 months ago I was suffering from some transitory
hearing defect that prevented me from following speech at a perfectly
normal pace.
--gary
Messages in this topic (14)
________________________________________________________________________
1d. Re: Information Density and Comprehension Speed
Posted by: "Jim Henry" [email protected]
Date: Fri Dec 23, 2011 9:51 am ((PST))
On 12/22/11, Logan Kearsley <[email protected]> wrote:
> Even when
> using a more phonemic script, after all, a fluent reader recognizes
> entire words, not individual letters to be sounded out.
Do we know this to be the case with languages that make extensive use
of derivational morphology to form novel words? If, for instance,
I'm reading a text in Esperanto, there may be few or no morphemes I'm
unfamiliar with, but there are usually more than a few word-forms I've
never encountered before -- nonce compounds and derivations whose
component morphemes are all fairly common, but which as whole words
are very rare in the Esperanto corpus. Most of their meanings are
obvious from their component parts to a fluent speaker of the
language, but that doesn't mean it takes no more effort to parse them
than familiar word-forms. There are a number of natlangs with similar
properties. Have there been any studies of reading speeds and the way
fluent readers recognize words that included such languages?
--
Jim Henry
http://www.pobox.com/~jimhenry/
Messages in this topic (14)
________________________________________________________________________
1e. Re: Information Density and Comprehension Speed
Posted by: "Dale McCreery" [email protected]
Date: Fri Dec 23, 2011 1:40 pm ((PST))
>> And secondly, I had the general impression that most of the guys I heard
>> speaking English ridiculously fast used a lot of set phrases, sort of
>> poetic turns of speech, etc., and was wondering if there have any been
>> any
>> studies evaluating language speed as it related to the density of set
>> phrases and structures? I would suspect that the more poetic or
>> formulaic
>> the speech (in a community where this style was heavily established) the
>> faster the general flow of speech would be...
>
> That's a very interesting question. You'd want to have some way to
> distinguish whether particular formulaic phrases were conveying a
> particularly large density of information, or whether rate of speech
> is increased precisely because poeticisms expand the volume of speech
> without increasing its informative content. And the answer might
> actually turn out to be both, as some formulaic patterns are
> high-context means of suggesting larger blocks of information that the
> hearer is expected to already know and thus avoiding transmitting it.
>
> -l.
>
My expectation would be that a set phrase would require an overall
reduction of processing, even if it maintained an equivalent bits per
phoneme ratio, reflecting the fact that there would be a greater number of
exemplars for the phrase for the speakers. This would be correlated with
something Ive noticed for both my sister and myself - after we learnt
Russian, our English production slowed down, as we now had more optional
ways to express each concept, while the stereotypical old monolingual male
speaker can really put out a nice turn of phrase without near as much
thought.
Perhaps the very fact of learning a conlang renders us permanently slower
speakers, thereby defeating the concept of speedtalk for everyone but
hypothetical monolingual speakers :)
-dale-
Messages in this topic (14)
________________________________________________________________________
1f. Re: Information Density and Comprehension Speed
Posted by: "Logan Kearsley" [email protected]
Date: Fri Dec 23, 2011 5:00 pm ((PST))
On 23 December 2011 08:14, And Rosta <[email protected]> wrote:
> Logan Kearsley, On 23/12/2011 04:38:
[...]
>> I don't see how you gained that interpretation from what I wrote.
>> Making words optimally short (maximizing bits per segment) and
>> eliminating redundancy (minimizing the number of bits that are
>> transmitted at all) are entirely different matters.
>
> "Making words optimally short" is ambiguous. You were talking about
> comparing phonologies, comparing how many bits per segment they encode, and
> then proceeded to compare word lengths across phonologies in terms of number
> of segments per word. I'd mistakenly thought you had proceeded to compare
> the lengths of alternate phonological forms for a given word within a given
> phonology, i.e. comparing the number of bits per candidate form; "making
> words optimally short" to me meant "minimizing (or optimizing) the number of
> bits per word".
Ah, I understand the confusion.
> Returning to the main point of the thread, I am reluctant to accept -- given
> the currently available evidence -- that the rate of encoded-information
> transmission is necessarily constant (-- your message was of course about
> the rate of transmission of encoding-information, not encoded-information).
So, to summarize for clarity, there's the issue of 1) how big of a
word you need to convey a certain amount of information (encoding
information, what was measured and compared in the article), and then
there's the issue of 2) how much information you bother to put in any
particular word (encoded information). (2) probably has a significant
effect on the observation of (1), but was not controlled for, and thus
probably accounts for some 'random' variation in the data gathered in
the study on (1). I would not be at all surprised if
similarities/differences in what information is actually encoded
account for a great deal of the statistical grouping of English,
German, French, Spanish, and Italian, while Mandarin and Japanese are
kind of outliers.
> I'm reluctant in practice because the currently available evidence points to
> the contrary conclusion: encodings can vary in their efficiency, and I know
> of no evidence to indicate that human languages necessarily converge on
> encodings of the same efficiency. And I'm reluctant in principle because to
> me the most intellectually (rather than merely artistically & aesthetically)
> interesting aspect of conlanging pertains to investigating the
> practicability of a loglang (logical language), where a logical language is
> one that encodes logical form fully and unambiguously while being as concise
> as natlangs -- in technical terms, the full explicature should be
> unambiguous encoded; so this necessarily means that the loglang has to
> achieve a greater rate of information transmission than natlangs. (In
> practice I understand that to mean that the loglang has to find a more
> efficient encoding, not that the bitrate has to be increased; upping the
> bitrate would be essentially an extralinguistic solution, and would not
> address the question of whether a human language can be a loglang.) Now this
> research programme may eventually prove that it is impossible, but it can't
> prove anything if the task is prematurely judged to be impossible.
I for one am quite certain that natural languages do not converge on
the most efficient possible encoding, for several reasons- first,
because natural languages rely to greatly varying degrees on
pragmatics to avoid encoding things, second because languages are
always changing in stochastic ways that will find local optima but not
global (which is really just a cover for the big list of reasons for
why they change), and third because humans intentionally make use of a
lot of redundancy. (Perhaps they actually do converge on
close-to-optimal solutions to the problem that includes the constraint
of a certain amount of redundancy.) I suspect that they converge on a
certain range of efficiencies, though precisely how big that range is
I'm not sure. When you don't consider the average of the whole
language over a large corpus, and just consider small bits at a time,
it's clear that some languages are more efficient at certain domains
than others (which is why we develop jargons- to fix up our language
to be really effective at communicating about one particular little
thing).
On 23 December 2011 10:51, Jim Henry <[email protected]> wrote:
> On 12/22/11, Logan Kearsley <[email protected]> wrote:
>> Even when
>> using a more phonemic script, after all, a fluent reader recognizes
>> entire words, not individual letters to be sounded out.
>
> Do we know this to be the case with languages that make extensive use
> of derivational morphology to form novel words?
I don't; probably, that is yet another good research question. There's
some relevance here to the elsethread talk of the usage of whitespace
in writings systems- extrapolating from my self-observations of how I
deal with derivational neologisms in English, I would expect common
productive morphemes to be treated like individual symbols that just
happen to not be whitespace delimited. That only really works well
with concatenative morphology, though; how readers might deal with
fully productive templatic derivational morphology is entirely beyond
me.
> Have there been any studies of reading speeds and the way
> fluent readers recognize words that included such languages?
Haven't heard of them. Perhaps I shall have to remember to ask around
the Linguistics department when Christmas Break is over. My initial
hypothesis is that processing such words in writing would be roughly
the same effort as the normal processing load for a writing system
with no word boundary delimiters. But, I could easily turn out to be
entirely wrong about that.
On 23 December 2011 14:40, Dale McCreery <[email protected]> wrote:
[...]
> My expectation would be that a set phrase would require an overall
> reduction of processing, even if it maintained an equivalent bits per
> phoneme ratio, reflecting the fact that there would be a greater number of
> exemplars for the phrase for the speakers.
Possibly. I would expect it to be treated like interpreting any single
lexical item, thus involving fewer bits to distinguish it in the
mental lexicon than processing it as a sequence of individual words
would. But you raise another point that bears investigation, which is
how much effort does it take to interpret any single lexical item, and
what are the factors that influence that? Frequency of exposure /
number of exemplars I'm sure is one of them.
> This would be correlated with
> something Ive noticed for both my sister and myself - after we learnt
> Russian, our English production slowed down, as we now had more optional
> ways to express each concept, while the stereotypical old monolingual male
> speaker can really put out a nice turn of phrase without near as much
> thought.
I don't know references to any of the original research on this, but
I've been told that one of the cognitive effects of multilinguilism of
any sort is slower lexicon access. Apparently, this is entirely
independent of the size of your lexicon in any particular language
(e.g., an Oxford professor is not expected, on average, to have slower
access than Joe Average-Vocab), and solely a function of having
multiple languages. I'm not sure I fully believe that, without having
seen the research myself, but it does correspond to the fact that,
though my English competence has never decreased (unlike some people
in immersion language learning environments I know), I developed
greater and greater difficulty with recalling precisely the right word
(in circumstances that call for the usage of rare words) rapidly in
English as my command of Russian improved.
-l.
Messages in this topic (14)
________________________________________________________________________
________________________________________________________________________
2a. How Does Everyone Translate?
Posted by: "Sam Stutter" [email protected]
Date: Fri Dec 23, 2011 7:08 am ((PST))
I know it's a couple of weeks back, but when I wrote out my card exchange
letter I think I've managed to refine my translation method. I used to work
sentence by sentence / clause by clause.
I used to see the sentence "I went to the shops" and then translate
holistically: look up "go", find the preterite, declarative, 1st person
singular conjugation and then write that. Then "to" and then "shops" and
decline that for lative definite.
"Schàdha unla tyentàrré"
Now when I translate I break each English sentence down grammatically and
analyse how it might function in the target language. So "I went to the shops"
is transcribed "go"PRET.DECL.1PS.STILL *TOWARDS* "shop"NOM.DEF.PLU+LAT.
Eventually the whole document is converted into individual one line commands.
Even though it takes a lot more work, I find it quicker. It also means I
generate a list of words to look up in one go or to coin.
What I wondered was, how does everyone else go about translating from English
into their conlangs? Does anyone have a really efficient method?
Sam Stutter
[email protected]
"No e na il cu barri"
Messages in this topic (7)
________________________________________________________________________
2b. Re: How Does Everyone Translate?
Posted by: "Koppa Dasao" [email protected]
Date: Fri Dec 23, 2011 7:27 am ((PST))
It should depend on the conlang. Your conlang is so different from
English that a simple word for word translation doesn't fit, while
Delang is closer to English and for that reason can cope easier with a
simple word for word translation followed by a syntax check.
For Delang the analysis would be as follow
I went to the shops
1ps.sing go.past_tense to shop.simple_plural.definite
Az fremarzjy aunj deskopin
Syntax check
Az aunj deskopin fremarzjy
(ÐÑ Ò©Ð½Ñ Î´ÐµÑкопÑн ÆÅемаÅÒ«y)
Koppa Dasao
___
Norway isn't the solution, but the appendix that's cut out!
2011/12/23 Sam Stutter <[email protected]>:
> I know it's a couple of weeks back, but when I wrote out my card exchange
> letter I think I've managed to refine my translation method. I used to work
> sentence by sentence / clause by clause.
>
> I used to see the sentence "I went to the shops" and then translate
> holistically: look up "go", find the preterite, declarative, 1st person
> singular conjugation and then write that. Then "to" and then "shops" and
> decline that for lative definite.
>
> "Schàdha unla tyentàrré"
>
> Now when I translate I break each English sentence down grammatically and
> analyse how it might function in the target language. So "I went to the
> shops" is transcribed "go"PRET.DECL.1PS.STILL *TOWARDS*
> "shop"NOM.DEF.PLU+LAT. Eventually the whole document is converted into
> individual one line commands.
>
> Even though it takes a lot more work, I find it quicker. It also means I
> generate a list of words to look up in one go or to coin.
>
> What I wondered was, how does everyone else go about translating from English
> into their conlangs? Does anyone have a really efficient method?
>
> Sam Stutter
> [email protected]
> "No e na il cu barri"
Messages in this topic (7)
________________________________________________________________________
2c. Re: How Does Everyone Translate?
Posted by: "Charlie Brickner" [email protected]
Date: Fri Dec 23, 2011 8:06 am ((PST))
On Fri, 23 Dec 2011 15:07:57 +0000, Sam Stutter <[email protected]>
wrote:
>What I wondered was, how does everyone else go about translating from
English into their conlangs? Does anyone have a really efficient method?
The grammar of Senjecas is rather simple so I don't worry about that when
translating. And I'm becoming better at remembering the vocabulary. The
problem I have is with word order, especially when there is some kind of
subordinate clause.
I rewrite the sentence in Senjecan order, then translate. For this exercise I
had to create the word for "shop." I only had the word for outdoor places:
market, fair, and bazaar, which is "mèrkëdáros".
I went to the shops.
I shop to go.
1-NOM.s shop-LAT.s to PAST go-IND
m-us mèrkëdém-om do per át-a:
Charlie
Messages in this topic (7)
________________________________________________________________________
2d. Re: How Does Everyone Translate?
Posted by: "Gary Shannon" [email protected]
Date: Fri Dec 23, 2011 8:34 am ((PST))
I tend to take a grammar-first approach. I write an English gloss of
the sentence, and then relex the gloss. That's how I did it for that
30-day conlang I did back a year or so ago.
--gary
On Fri, Dec 23, 2011 at 7:07 AM, Sam Stutter <[email protected]> wrote:
> I know it's a couple of weeks back, but when I wrote out my card exchange
> letter I think I've managed to refine my translation method. I used to work
> sentence by sentence / clause by clause.
Messages in this topic (7)
________________________________________________________________________
2e. Re: How Does Everyone Translate?
Posted by: "Roger Mills" [email protected]
Date: Fri Dec 23, 2011 9:20 am ((PST))
From: Sam Stutter <[email protected]>
> I know it's a couple of weeks back, but when I wrote out my card exchange
> letter I think I've managed to refine my translation method. I used to work
> sentence by sentence / clause by clause.
> I used to see the sentence "I went to the shops" and then translate
> holistically: look up "go", find the preterite, declarative, 1st person
> singular conjugation and then write that. Then "to" and then "shops" and
> decline that for lative definite.
------------------------------------------------
I've been working with Kash for a sufficiently long time that I've internalized
a lot of the basic verbs, person markers, a lot of grammar and vocab., etc.
Your sentence would come out first as "macosasa ri SHOP-PL-DAT", then I'd have
to look up 'shop' and add the appropriate endings. Gwr is very analytical, it
would simply be a word for word translation (and at the moment I'd have to look
up everything). Prevli is VSO, and I have to keep charts of the phonological
changes to roots, and if it's a trans.verb I have to consult my chart of the
subj+obj pronoun forms. Some statements require an irrealis verb form; relative
clauses often involve passive voice.
--------------------------------------------------------
> Now when I translate I break each English sentence down grammatically and
> analyse how it might function in the target language. So "I went to the
> shops" is transcribed "go"PRET.DECL.1PS.STILL *TOWARDS*
> "shop"NOM.DEF.PLU+LAT. Eventually the whole document is converted into
> individual one line commands.
-------------------------------
That's how I usually translate e.g. a relay text
Messages in this topic (7)
________________________________________________________________________
2f. Re: How Does Everyone Translate?
Posted by: "Matthew Turnbull" [email protected]
Date: Fri Dec 23, 2011 9:58 am ((PST))
I consider the meaning of each sentance and cultural relevance of the
activity and then try and generate a novel phrase in Jorayn that may
or may not reflect the original translation. I also coin words as I
need to express the concept better. Slometimes it helps to consider
the text as a whole instead of one sentence at a time
When working with my other conlangs I generally have the grammar with
me, consider who is doing what, what the main verb/verbs are and then
translate the sentences one at a time.
On 12/23/11, Sam Stutter <[email protected]> wrote:
> I know it's a couple of weeks back, but when I wrote out my card exchange
> letter I think I've managed to refine my translation method. I used to work
> sentence by sentence / clause by clause.
>
> I used to see the sentence "I went to the shops" and then translate
> holistically: look up "go", find the preterite, declarative, 1st person
> singular conjugation and then write that. Then "to" and then "shops" and
> decline that for lative definite.
>
> "Schàdha unla tyentàrré"
>
> Now when I translate I break each English sentence down grammatically and
> analyse how it might function in the target language. So "I went to the
> shops" is transcribed "go"PRET.DECL.1PS.STILL *TOWARDS*
> "shop"NOM.DEF.PLU+LAT. Eventually the whole document is converted into
> individual one line commands.
>
> Even though it takes a lot more work, I find it quicker. It also means I
> generate a list of words to look up in one go or to coin.
>
> What I wondered was, how does everyone else go about translating from
> English into their conlangs? Does anyone have a really efficient method?
>
> Sam Stutter
> [email protected]
> "No e na il cu barri"
>
--
Sent from my mobile device
Messages in this topic (7)
________________________________________________________________________
2g. Re: How Does Everyone Translate?
Posted by: "Logan Kearsley" [email protected]
Date: Fri Dec 23, 2011 12:31 pm ((PST))
On 23 December 2011 10:58, Matthew Turnbull <[email protected]> wrote:
> I consider the meaning of each sentance and cultural relevance of the
> activity and then try and generate a novel phrase in Jorayn that may
> or may not reflect the original translation. I also coin words as I
> need to express the concept better. Slometimes it helps to consider
> the text as a whole instead of one sentence at a time
When I first started out, I'd look things up word-by-word and then do
the "syntax check"- rearrange the phrases so my SVO
conlang-gloss-of-English-sentence turned into a SOV conlang-sentence
(for example), move the adjectives to the other side of the noun, etc.
Then I learned to speak a non-romance language and got used to the
habit of interpreting meanings rather than translating sentences or
words (all of my foreign language teachers have considered
'translation' to be a dirty word, but the opinion didn't really 'take'
until I started to get fluent in Russian). Since then, I try to
internalize the grammar / pragmatics / social context of the conlang
and figure out how the idea of the English sentence would be
structured in that language, and then do a lot of dictionary look-ups
to try to pick the best words for rendering that concept, whether or
not they're good one-for-one matches with the English words or
phrases. With Russian, where I've got the lexicon internalized,
English isn't involved in the translation process at all; the English
provides an idea, and the idea then gets re-encoded directly as
Russian. With conlangs where I haven't memorized all the vocabulary
(partly because I keep changing it...) the spellout part of language
production still has to go through an English intermediary so I can
find what the right words are.
I suspect that making that realization / transition is part of the
process of going from the beginner stage of relexing your native
language to coming up with something genuinely new.
-l.
Messages in this topic (7)
________________________________________________________________________
________________________________________________________________________
3a. Re: Conlang Textbook Template
Posted by: "Gary Shannon" [email protected]
Date: Fri Dec 23, 2011 2:30 pm ((PST))
Lessons 1-14 are now available. http://fiziwig.com/conlang/template/
NOTE: I am re-working some of the vocabulary and exercises, and well
as re-arranging some of the grammar lessons a little. The original
organization was based on a 1917 Spanish textbook, which results in
two kinds of problems: 1) The chapter organization is, in places, too
specific to Spanish and needs to be made more generic; and 2) the
vocabulary and practice exercises are sometimes dated and obsolete,
making references to streetcars and coal-fired steam heat, and
treating electric lights as something quite revolutionary, or at
least, worthy of special mention.
You can, of course, use the lessons as they stand, but be aware that
as time goes on I will be revising and, hopefully, improving the
lessons to be less specific to any one language, and to make the
vocabulary and exercises more up to date. Of course if you are using
the lessons for a language used by bronze age hunter-gatherers then
the exercises and vocabulary will have to be adjusted accordingly
anyway.
In the most basic sense, syntax and grammar are about creating
sentences, and so the content of, for example, lesson 22 should be,
first and foremost, to answer the question: What do I need to know in
order to write the exercises of chapter 22 that I didn't need to know
to write the exercises of chapter 21. And that, in a nutshell, will be
the grammar lesson of chapter 22.
Comments, corrections, and suggestions are always welcome.
--gary
On Thu, Dec 22, 2011 at 9:11 PM, Gary Shannon <[email protected]> wrote:
> Lessons 1 through 11 are now available at http://fiziwig.com/conlang/template/
>
> I'm 1/5th of the way through. WOOT!
>
> --gary
Messages in this topic (12)
________________________________________________________________________
3b. Re: Conlang Textbook Template
Posted by: "Sam Stutter" [email protected]
Date: Fri Dec 23, 2011 2:39 pm ((PST))
A question about the first exercise: how did the original text handle "paper"?
As I interpret it, I would expect some sort of quantifier like "a sheet of
paper" or, in the English translations there's the assumption that "I have
paper" = "I have some paper". Did the original use the Spanish plural
indefinite article? If so, it's a very odd place to start...
Sam Stutter
[email protected]
"No e na il cu barri"
On 23 Dec 2011, at 22:30, Gary Shannon <[email protected]> wrote:
> Lessons 1-14 are now available. http://fiziwig.com/conlang/template/
>
> NOTE: I am re-working some of the vocabulary and exercises, and well
> as re-arranging some of the grammar lessons a little. The original
> organization was based on a 1917 Spanish textbook, which results in
> two kinds of problems: 1) The chapter organization is, in places, too
> specific to Spanish and needs to be made more generic; and 2) the
> vocabulary and practice exercises are sometimes dated and obsolete,
> making references to streetcars and coal-fired steam heat, and
> treating electric lights as something quite revolutionary, or at
> least, worthy of special mention.
>
> You can, of course, use the lessons as they stand, but be aware that
> as time goes on I will be revising and, hopefully, improving the
> lessons to be less specific to any one language, and to make the
> vocabulary and exercises more up to date. Of course if you are using
> the lessons for a language used by bronze age hunter-gatherers then
> the exercises and vocabulary will have to be adjusted accordingly
> anyway.
>
> In the most basic sense, syntax and grammar are about creating
> sentences, and so the content of, for example, lesson 22 should be,
> first and foremost, to answer the question: What do I need to know in
> order to write the exercises of chapter 22 that I didn't need to know
> to write the exercises of chapter 21. And that, in a nutshell, will be
> the grammar lesson of chapter 22.
>
> Comments, corrections, and suggestions are always welcome.
>
> --gary
>
>
> On Thu, Dec 22, 2011 at 9:11 PM, Gary Shannon <[email protected]> wrote:
>> Lessons 1 through 11 are now available at
>> http://fiziwig.com/conlang/template/
>>
>> I'm 1/5th of the way through. WOOT!
>>
>> --gary
Messages in this topic (12)
________________________________________________________________________
3c. Re: Conlang Textbook Template
Posted by: "Gary Shannon" [email protected]
Date: Fri Dec 23, 2011 3:22 pm ((PST))
Yeah, I thought it was a little odd too. I may go back and modify
that. The original used indefinite articles for OTHER nouns, but NO
article at all for paper: el lápiz - pencil; la pluma - pen; papel
(f.) - paper
And in a later chapter when plurals are covered it gave no plural for
"paper". It seems to be treating "paper" as a mass noun like "sugar"
or "water".
There are a number of things like that that I think need improvement,
so I appreciate your observations and suggestions.
--gary
On Fri, Dec 23, 2011 at 2:39 PM, Sam Stutter <[email protected]> wrote:
> A question about the first exercise: how did the original text handle
> "paper"? As I interpret it, I would expect some sort of quantifier like "a
> sheet of paper" or, in the English translations there's the assumption that
> "I have paper" = "I have some paper". Did the original use the Spanish plural
> indefinite article? If so, it's a very odd place to start...
>
> Sam Stutter
> [email protected]
> "No e na il cu barri"
>
> On 23 Dec 2011, at 22:30, Gary Shannon <[email protected]> wrote:
Messages in this topic (12)
________________________________________________________________________
3d. Re: Conlang Textbook Template
Posted by: "Ben Scerri" [email protected]
Date: Fri Dec 23, 2011 5:44 pm ((PST))
Interestingly enough, I am using your template to write the language of a
bronze age hunter-gatherer people... I've found a simple shift in focus
(from study to warfare) has resulted in some pretty clean transitions.
Sure, some of the sentences i need to rewrite, but only a few words here or
there. For example:
pencil > sword
pen > horse
book > amrour
paper > shield
write > ride
ink > bow
blackboard > battlefield
etc
I got ride of "an exercise" as I thought it didn't fit into the system I
had begun using. It seems to be going well, however, and I just finished
lesson 3 :)
Thank you HEAPS for this.
On 24 December 2011 10:21, Gary Shannon <[email protected]> wrote:
> Yeah, I thought it was a little odd too. I may go back and modify
> that. The original used indefinite articles for OTHER nouns, but NO
> article at all for paper: el lápiz - pencil; la pluma - pen; papel
> (f.) - paper
>
> And in a later chapter when plurals are covered it gave no plural for
> "paper". It seems to be treating "paper" as a mass noun like "sugar"
> or "water".
>
> There are a number of things like that that I think need improvement,
> so I appreciate your observations and suggestions.
>
> --gary
>
> On Fri, Dec 23, 2011 at 2:39 PM, Sam Stutter <[email protected]> wrote:
> > A question about the first exercise: how did the original text handle
> "paper"? As I interpret it, I would expect some sort of quantifier like "a
> sheet of paper" or, in the English translations there's the assumption that
> "I have paper" = "I have some paper". Did the original use the Spanish
> plural indefinite article? If so, it's a very odd place to start...
> >
> > Sam Stutter
> > [email protected]
> > "No e na il cu barri"
> >
> > On 23 Dec 2011, at 22:30, Gary Shannon <[email protected]> wrote:
>
Messages in this topic (12)
________________________________________________________________________
3e. Re: Conlang Textbook Template
Posted by: "Logan Kearsley" [email protected]
Date: Fri Dec 23, 2011 6:42 pm ((PST))
On 21 December 2011 20:57, Gary Shannon <[email protected]> wrote:
> I've taken an old out-of-copyright Spanish textbook and scanned it. It
> has 50 lessons which are a nice small convenient size for bite-sized
> language lessons. What I'm doing is changing it from a Spanish
> textbook to a generic language textbook template, as a set of
> interactive HTML pages. I'm translating all the exercises to English
> and replacing all the grammatical explanations with general
> descriptions of which grammatical principles need to be covered in a
> given paragraph.
Shall we be using this for the Conlang Textbook Relay, then?
I might end up using it for a collaborative language with my
soon-to-be-wife. I was going to start out with the Where Are Your Keys
dialogs, but no reason we couldn't use the textbook template, too.
> Of course it won't be suitable for all kinds of conlangs, but it
> should work for a great many of them, and certainly for IE languages
> in general. As always with stuff on my web page, it's there for anyone
> to use any way they like.
It is rather difficult to come up with a good pan-language textbook
template / guide. I'm not sure if that would make for a useful
long-term research project, or a totally pointless one. Perhaps we
could at least try putting together a set of templates that each cover
certain broad categories of languages.
-l.
Messages in this topic (12)
________________________________________________________________________
3f. Re: Conlang Textbook Template
Posted by: "Ben Scerri" [email protected]
Date: Fri Dec 23, 2011 6:50 pm ((PST))
Personally I don't think this would work for the relay. I thought the idea
of that was to let each person input whatever they wanted into eahc lesson
(maybe introducing 1-3 rules each one) and therefore making a quite unique
conlang. If we follow this template, it will lead to an IE language which
would basically allow the relayers to decide on sound, rather than
function...
On 24 December 2011 13:42, Logan Kearsley <[email protected]> wrote:
> On 21 December 2011 20:57, Gary Shannon <[email protected]> wrote:
> > I've taken an old out-of-copyright Spanish textbook and scanned it. It
> > has 50 lessons which are a nice small convenient size for bite-sized
> > language lessons. What I'm doing is changing it from a Spanish
> > textbook to a generic language textbook template, as a set of
> > interactive HTML pages. I'm translating all the exercises to English
> > and replacing all the grammatical explanations with general
> > descriptions of which grammatical principles need to be covered in a
> > given paragraph.
>
> Shall we be using this for the Conlang Textbook Relay, then?
>
> I might end up using it for a collaborative language with my
> soon-to-be-wife. I was going to start out with the Where Are Your Keys
> dialogs, but no reason we couldn't use the textbook template, too.
>
> > Of course it won't be suitable for all kinds of conlangs, but it
> > should work for a great many of them, and certainly for IE languages
> > in general. As always with stuff on my web page, it's there for anyone
> > to use any way they like.
>
> It is rather difficult to come up with a good pan-language textbook
> template / guide. I'm not sure if that would make for a useful
> long-term research project, or a totally pointless one. Perhaps we
> could at least try putting together a set of templates that each cover
> certain broad categories of languages.
>
> -l.
>
Messages in this topic (12)
________________________________________________________________________
3g. Re: Conlang Textbook Template
Posted by: "Alex Fink" [email protected]
Date: Fri Dec 23, 2011 7:35 pm ((PST))
On Sat, 24 Dec 2011 13:50:01 +1100, Ben Scerri <[email protected]> wrote:
>Personally I don't think this would work for the relay. I thought the idea
>of that was to let each person input whatever they wanted into eahc lesson
>(maybe introducing 1-3 rules each one) and therefore making a quite unique
>conlang. If we follow this template, it will lead to an IE language which
>would basically allow the relayers to decide on sound, rather than
>function...
Well, as Gary said earlier, we tried that one time and ended up with
http://www.frathwiki.com/Round_Robin_Conlang ; this is meant to be something
a little different.
At the same time, I'd also find it constraining and predetermining to follow
Ford-Hills to the letter. I'd advocate allowing liberties to be taken and
substitutions to be made as long as they don't greatly change the size of
the learning burden of each chapter -- e.g. at the point where a second
Spanish tense is first introduced, if you'd rather have aspect or
evidentiality or whatnot be the central verb contrast of the language,
introduce a second aspect or evidential or whatnot instead. Then again, I
don't know what havoc that would wreak on the stability of the game.
Alex
Messages in this topic (12)
________________________________________________________________________
3h. Re: Conlang Textbook Template
Posted by: "Sam Stutter" [email protected]
Date: Sat Dec 24, 2011 4:50 am ((PST))
I also intend to use this in the near future. Keep up the good work Gary :)
Sam Stutter
[email protected]
"No e na il cu barri"
On 24 Dec 2011, at 01:44, Ben Scerri wrote:
> Interestingly enough, I am using your template to write the language of a
> bronze age hunter-gatherer people... I've found a simple shift in focus
> (from study to warfare) has resulted in some pretty clean transitions.
> Sure, some of the sentences i need to rewrite, but only a few words here or
> there. For example:
>
> pencil > sword
> pen > horse
> book > amrour
> paper > shield
> write > ride
> ink > bow
> blackboard > battlefield
> etc
>
> I got ride of "an exercise" as I thought it didn't fit into the system I
> had begun using. It seems to be going well, however, and I just finished
> lesson 3 :)
>
> Thank you HEAPS for this.
>
> On 24 December 2011 10:21, Gary Shannon <[email protected]> wrote:
>
>> Yeah, I thought it was a little odd too. I may go back and modify
>> that. The original used indefinite articles for OTHER nouns, but NO
>> article at all for paper: el lápiz - pencil; la pluma - pen; papel
>> (f.) - paper
>>
>> And in a later chapter when plurals are covered it gave no plural for
>> "paper". It seems to be treating "paper" as a mass noun like "sugar"
>> or "water".
>>
>> There are a number of things like that that I think need improvement,
>> so I appreciate your observations and suggestions.
>>
>> --gary
>>
>> On Fri, Dec 23, 2011 at 2:39 PM, Sam Stutter <[email protected]> wrote:
>>> A question about the first exercise: how did the original text handle
>> "paper"? As I interpret it, I would expect some sort of quantifier like "a
>> sheet of paper" or, in the English translations there's the assumption that
>> "I have paper" = "I have some paper". Did the original use the Spanish
>> plural indefinite article? If so, it's a very odd place to start...
>>>
>>> Sam Stutter
>>> [email protected]
>>> "No e na il cu barri"
>>>
>>> On 23 Dec 2011, at 22:30, Gary Shannon <[email protected]> wrote:
>>
Messages in this topic (12)
________________________________________________________________________
________________________________________________________________________
4a. Re: Barsoomian Project
Posted by: "Fredrik Ekman" [email protected]
Date: Sat Dec 24, 2011 3:26 am ((PST))
Don,
I am going to comment on your blog posts one at a time, as time permits. I
start with the first one, titled "Ay, Ra, Co... Counting in Barsoomian!"
and available at http://library.conlang.org/blog/?p=583
Your scheme of connecting the planet name Barsoom with the numeral bar is
not novel. It has been suggested at least a couple of times in the past.
Unfortunately, it does not work very well, for at least two reasons.
First, we do not know the meaning of -soom. It is only attested in planet
names, so the "keep things simple" principle should caution us to perhaps
not use it with the sun and the moons. Now, if that was the only problem I
would be more than willing to let it go. After all, Burroughs' planet
names were clearly formulated by someone(s) well aquainted with astronomy;
who knew that Mars (Barsoom) and Earth (Jasoom) belong to the same class
of objects. However, by extension, your logic postulates that the moon
Phobos should be named Ovsoom, but Burroughs explicitly says Thuria. Of
course every language is full of exceptions, but the Burroughs I know is
not one to let hold of a good system that easily. If he had intended -soom
to mean heavenly body, then you can be sure that all the moons, th sun,
the meteors and asteroids would have been named accordingly.
You never explained further about Jupiter. Jupiter is Sasoom, so what
numeral is sa, after having counted all the moons?
In my world, the words Bar- and bar are just homonyms. Nothing more. And
why not? Bar- could mean almost anything, such as "home", "red", "jolly
nice", "slowly dying" or whatnot.
The second reason why your scheme works poorly is the term il. As you
suggest yourself, it becomes somewhat of an annoyance. But let us bring
some more logic into the equation.
Il-dur-en is the name of a hormad, just like Tor-dur-bar. We know that the
latter is four million eight, so il-dur-en is something million something.
Now, remember that all hormads are numbered according to the order in
which they came out of their culture vats, and they are all confined to
one small island. Fifty million of them? I think not. In fact, my
hypothesis is that Tor-dur-bar is one of the younger of the lot, and if
that is true (Burroughs says nothing specific on the subject) then il must
mean either two or three. At any rate, we know that there have to be AT
LEAST four million hormads. Since no higher numbers are confirmed,
statistical chances are better that il is two or three; certainly not
fifty.
But if you have to assign il (or any other word) to a numeral between ten
and hundred, I would suggest that twenty is a much better candidate than
fifty. If you want the full reasoning behind this, I would have to direct
you to my essay "Barsoomian 101, Part 4: Numbers" in ERB-APA #99.
(Back-issues may or may not be available. Let me know in private mail, and
I can put you in touch with the ERB-APA editor.)
Moving further down your list, I see that you have fallen into Henning's
"dar" trap. Henning, in his article, to which you link, assumes that since
the military units utan and umak contain the numerals tan and mak (the
former of which is attested in other contexts), then the military unit dar
must also be a numeral. This does not work. One reason is that the
military unit, by extension, should be *udar, which it is not. Another is
that one dar is not exactly 1,000 men, since Burroughs, in one book (I
forget which at the moment) writes about a dar with at least eleven utans.
The logical assumption must be that since *teeay-tan appears to mean
eleven hundred, *tee-tan means "ten hundred" or, in other words, one
thousand. My hypothesis is that Barsoomian does not have a unique word for
thousand. Again, for a more detailed discussion I refer you to the
abovementioned ERB-APA essay.
Hopefully, I will be back with another comment tomorrow or so. In the
meantime,
To each and all, a very Merry Christmas!
Fredrik
Messages in this topic (4)
________________________________________________________________________
4b. Re: Barsoomian Project
Posted by: "Peter Cyrus" [email protected]
Date: Sat Dec 24, 2011 3:44 am ((PST))
Bodies of the solar system (or more precisely, barycenters) should be
numbered by mass, not by orbital radius, with the Sun as zero.
Earth is Sol 5, and our Moon is Sol 5.1
On Sat, Dec 24, 2011 at 12:25 PM, Fredrik Ekman <[email protected]> wrote:
> Don,
>
> I am going to comment on your blog posts one at a time, as time permits. I
> start with the first one, titled "Ay, Ra, Co... Counting in Barsoomian!"
> and available at http://library.conlang.org/blog/?p=583
>
> Your scheme of connecting the planet name Barsoom with the numeral bar is
> not novel. It has been suggested at least a couple of times in the past.
> Unfortunately, it does not work very well, for at least two reasons.
>
> First, we do not know the meaning of -soom. It is only attested in planet
> names, so the "keep things simple" principle should caution us to perhaps
> not use it with the sun and the moons. Now, if that was the only problem I
> would be more than willing to let it go. After all, Burroughs' planet
> names were clearly formulated by someone(s) well aquainted with astronomy;
> who knew that Mars (Barsoom) and Earth (Jasoom) belong to the same class
> of objects. However, by extension, your logic postulates that the moon
> Phobos should be named Ovsoom, but Burroughs explicitly says Thuria. Of
> course every language is full of exceptions, but the Burroughs I know is
> not one to let hold of a good system that easily. If he had intended -soom
> to mean heavenly body, then you can be sure that all the moons, th sun,
> the meteors and asteroids would have been named accordingly.
>
> You never explained further about Jupiter. Jupiter is Sasoom, so what
> numeral is sa, after having counted all the moons?
>
> In my world, the words Bar- and bar are just homonyms. Nothing more. And
> why not? Bar- could mean almost anything, such as "home", "red", "jolly
> nice", "slowly dying" or whatnot.
>
> The second reason why your scheme works poorly is the term il. As you
> suggest yourself, it becomes somewhat of an annoyance. But let us bring
> some more logic into the equation.
>
> Il-dur-en is the name of a hormad, just like Tor-dur-bar. We know that the
> latter is four million eight, so il-dur-en is something million something.
> Now, remember that all hormads are numbered according to the order in
> which they came out of their culture vats, and they are all confined to
> one small island. Fifty million of them? I think not. In fact, my
> hypothesis is that Tor-dur-bar is one of the younger of the lot, and if
> that is true (Burroughs says nothing specific on the subject) then il must
> mean either two or three. At any rate, we know that there have to be AT
> LEAST four million hormads. Since no higher numbers are confirmed,
> statistical chances are better that il is two or three; certainly not
> fifty.
>
> But if you have to assign il (or any other word) to a numeral between ten
> and hundred, I would suggest that twenty is a much better candidate than
> fifty. If you want the full reasoning behind this, I would have to direct
> you to my essay "Barsoomian 101, Part 4: Numbers" in ERB-APA #99.
> (Back-issues may or may not be available. Let me know in private mail, and
> I can put you in touch with the ERB-APA editor.)
>
> Moving further down your list, I see that you have fallen into Henning's
> "dar" trap. Henning, in his article, to which you link, assumes that since
> the military units utan and umak contain the numerals tan and mak (the
> former of which is attested in other contexts), then the military unit dar
> must also be a numeral. This does not work. One reason is that the
> military unit, by extension, should be *udar, which it is not. Another is
> that one dar is not exactly 1,000 men, since Burroughs, in one book (I
> forget which at the moment) writes about a dar with at least eleven utans.
>
> The logical assumption must be that since *teeay-tan appears to mean
> eleven hundred, *tee-tan means "ten hundred" or, in other words, one
> thousand. My hypothesis is that Barsoomian does not have a unique word for
> thousand. Again, for a more detailed discussion I refer you to the
> abovementioned ERB-APA essay.
>
> Hopefully, I will be back with another comment tomorrow or so. In the
> meantime,
>
> To each and all, a very Merry Christmas!
>
> Fredrik
Messages in this topic (4)
------------------------------------------------------------------------
Yahoo! Groups Links
<*> To visit your group on the web, go to:
http://groups.yahoo.com/group/conlang/
<*> Your email settings:
Digest Email | Traditional
<*> To change settings online go to:
http://groups.yahoo.com/group/conlang/join
(Yahoo! ID required)
<*> To change settings via email:
[email protected]
[email protected]
<*> To unsubscribe from this group, send an email to:
[email protected]
<*> Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/
------------------------------------------------------------------------
