I would like to write software that can convert a list of actions and
dialog into a story.
The system would have a knowledge base of the objects, scenes, and actors
and would then be fed a list of actions and dialogue involving the actors
and objects. The result would be a readable short story...
I would appreciate any comments, feedback, or pointers to any other work
that is being done that is similar in nature.
thanks
Bryan
This is something I have always wanted to do, too...Long ago, I wrote
something similar, but was rather fixed...basically, it was one
script with variations...
>The system would have a knowledge base of the objects, scenes, and actors
>and would then be fed a list of actions and dialogue involving the actors
>and objects. The result would be a readable short story...
>
>I would appreciate any comments, feedback, or pointers to any other work
>that is being done that is similar in nature.
Um, check into work by Roger Schank (sp?)...Tailspin, etc. It's a
little dated, but it might get you oriented...I think Tailspin uses
complex scripts and interactions between scripts?
-t
A good place to learn about Tailspin is in Inside Computer Understanding:
Five Programs Plus Minatures (Roger C. Schank and Christopher K.
Riesbeck), 1981, Lawrence Erlbaum.
[The End]
--
Will Fitzgerald 708.491.3500 (voice), 708.491.5258 (fax)
Institute for the Learning Sciences, Northwestern University
1890 Maple Ave. Evanston IL, 60201
Yes, it is relatively easy to learn another Turkic language once you have
learned one or two. There is a high degree of mutual intelligibility, and
the degree varies by geographic distance and by historical development.
There is a marked division between the languages of Islamicized Turkic
peoples and those that are Buddhist, Shamanist or Christian, since Arabic
and Iranian lexical influences tend to be strong among the former. It is
rather easy two switch for example between Turkish and Azeri, between
Tatar and Bashkir, between Kazakh and Kyrghyz, and between Uzbek and
Uyghur.
Learning both Uzbek and Kazakh is a wonderful start. With some practice
and a dictionary or two and perhaps a grammar sketch you'll soon be able
to read texts in other Turkic languages.
As for "how easy," that's relative, of course. I'd say you won't find
that learning Turkic languages is harder than learning Slavic languages.
There are far fewer grammatical forms to be learned. However, you'll have
to acquaint yourself with a different type of morphological and syntactic
structure. Like the structure of other Altaic languages, Turkic structure
may be generally explained as agglutinative: suffixes and enclitics are
attached to stems. A "word" thus derived tends to be the equivalent of a
phrase or even of a sentence in Indo-European languages. The basic
sentence structure is SOV, e.g., Uyghur ...
Ata+m+ning kitab+I ana+ng+GA ba"r+d+m
FATHER+1sg+gen BOOK+acc MOTHER+dat GIVE+past+1sg
Atamning kitabi ananggha ba"rdim.
I gave my father's book to your mother.
Good luck and lots of fun!
Ron Hahn
If you're intersted, E-Mail me and I'll give you a list of text and tape
materials.
Paul Tholfsen
Bellingham, WA
congratulations.
>While the agglutinative structure is difficult for those accustomed to
>Indo-European languages, once you learn the rules, things are completely
>predictable. The pronunciation is very easy - the only non-English sounds
>are the rounded o and u (as in French buerre, and sur), and the
>root vocabulary seems to consist mostly of one and two syllable words - a
>great relief after Russian!
I am surprized that you didn't mention the 'i' without the dot. What
sound in English is similar (I am not a native English speaker)
Also, although not often used, there is the soft g that is different.
And although dropped in the last 20 years, there is an accent that goes
on 'a' and 'i' which makes them sound different. A word can even have
different meanings without it, for example profit <-> snow , kAr <-> kar
>
>If you're intersted, E-Mail me and I'll give you a list of text and tape
>materials.
The Turkish typing tutor should also help. It used to be available on
wuarchive.
>
>Paul Tholfsen
>Bellingham, WA
>
--
-------------------------------------------------------------------------------
Mustafa Soysal MS57 mso...@mistik.express.net
-------------------------------------------------------------------------------
Tarkan
>Gorusmek uzere......
>
> Tarkan
>
>tarkan....@individual.puug.pt
Rumors are, that you are not Turkish, and not a native speaker.
Rumors are, that you are a provocator pretending to be Turkish.
So am I: this is the Turkish sound that gives me the most trouble. It
does show up in English, but only in some dialects.
>Also, although not often used, there is the soft g that is different.
Is this ever pronounced? I thought all it did was lengthen the pre-
ceeding vowel.
>And although dropped in the last 20 years, there is an accent that goes
>on 'a' and 'i' which makes them sound different. A word can even have
>different meanings without it, for example profit <-> snow , kAr <-> kar
I believe the "^a" is roughly equivalent to the [%] in English "cat" or
"bad." Or is it just an [a] that palatalises the preceeding consonant?
--
Daniel "Da" von Brighoff /\ Dilettanten
(de...@midway.uchicago.edu) /__\ erhebt Euch
/____\ gegen die Kunst!
: So am I: this is the Turkish sound that gives me the most trouble. It
: does show up in English, but only in some dialects.
Native English speaker here....
It has been approximated as the sound midway between "i" in "bit" and "u"
in "but". A high central unrounded lax vowel...
I've heard a sound like it in Black English pronounciation of "e" in
"America"
: >Also, although not often used, there is the soft g that is different.
: Is this ever pronounced? I thought all it did was lengthen the pre-
: ceeding vowel.
The "rule" is that it:
1) lengthens a back or front vowel,
dag^ = daa
2) acts as a glottal stop between two back vowels,
og^ul = o'ul
3) acts as a glide between two front vowels.
deg^il = deyil
: >And although dropped in the last 20 years, there is an accent that goes
: >on 'a' and 'i' which makes them sound different. A word can even have
: >different meanings without it, for example profit <-> snow , kAr <-> kar
: I believe the "^a" is roughly equivalent to the [%] in English "cat" or
: "bad." Or is it just an [a] that palatalises the preceeding consonant?
I thought it has a pronounciation of:
1) [aw] when long (non-palatalizing)
2) [a] (regular Turkish) in front of palatalized consonant (with that
symbol the consonants being "k", "g", and "l"
just my thoughts on this...
: --
: Daniel "Da" von Brighoff /\ Dilettanten
: (de...@midway.uchicago.edu) /__\ erhebt Euch
: /____\ gegen die Kunst!
--
--Reid
Corey Alton Reid pea...@wam.umd.edu
6 Cornerwood Court
Gaithersburg, MD 20878-1904
(301) 670-6995
ok, that would not be the same sound. I have not come across any
similar sounds in English in daily life here (Ann Arbor, Michigan)
>
>: >Also, although not often used, there is the soft g that is different.
>
>: Is this ever pronounced? I thought all it did was lengthen the pre-
>: ceeding vowel.
>
>The "rule" is that it:
>
>1) lengthens a back or front vowel,
> dag^ = daa
>2) acts as a glottal stop between two back vowels,
> og^ul = o'ul
>3) acts as a glide between two front vowels.
> deg^il = deyil
Actually, it may sound like that for someone who is not familiar with
the language, but the movements in the mouth and throat as well as the
sound produces are not the same. Therefore, soft g is pronounced, but
'soft'ly as the name indicates. This becames even more significant if
you add suffixes to the word.
>
>
>: >And although dropped in the last 20 years, there is an accent that goes
>: >on 'a' and 'i' which makes them sound different. A word can even have
>: >different meanings without it, for example profit <-> snow , kAr <-> kar
>
>: I believe the "^a" is roughly equivalent to the [%] in English "cat" or
>: "bad." Or is it just an [a] that palatalises the preceeding consonant?
>
>I thought it has a pronounciation of:
>1) [aw] when long (non-palatalizing)
>2) [a] (regular Turkish) in front of palatalized consonant (with that
>symbol the consonants being "k", "g", and "l"
>
>just my thoughts on this...
>
>: --
>: Daniel "Da" von Brighoff /\ Dilettanten
>: (de...@midway.uchicago.edu) /__\ erhebt Euch
>: /____\ gegen die Kunst!
>
>--
>--Reid
>
>Corey Alton Reid pea...@wam.umd.edu
>6 Cornerwood Court
>Gaithersburg, MD 20878-1904
>(301) 670-6995
--
[about undotted-i, I]
The closest thing I can come to (in my norhtern Californian
native English) is "Mrs."... while some pronounce it with an /s/ final,
some pronounce it with a /z/. Lending [missIz] or [missIs].. which seems
to me approximating (if not on the nose) the Republican Turkish
undotted-i.
>>: >Also, although not often used, there is the soft g that is different.
[rules snipped]
>Actually, it may sound like that for someone who is not familiar with
>the language, but the movements in the mouth and throat as well as the
>sound produces are not the same. Therefore, soft g is pronounced, but
>'soft'ly as the name indicates. This becames even more significant if
>you add suffixes to the word.
Hmm.. in my experience with Republican Turkish and other Turkic
languages, it seems that the rules were actually pretty on target.
While Republican Turkish retains the "soft-g" in orthography,
lending to the funky phonological contortions one must go through ;), the
soft-g is in NO way an actual "soft-g" as can be found in the eastern
Turkic languages, like Uzbek, Qazaq, Qyrghyz, Tyva, etc., which are
actual voiced velar (or pharyngeal) fricatives.
>Mustafa Soysal MS57
-Patrick (pahriddin)
--
_____________________________________________________________________
"je voudrais voir mourir d'amour ou bien d'haine.." T. Klingsor, "Bolero"
A language spoken by termites? (tahta = wood)
Maybe you mean Tajik/Tadzhik/Tadjik/..., which is spoken (where else?)
in Tajikistan, a former Soviet republic in Central Asia. As far as I
know, it's closely related to Persian, which makes it Indo-European
instead of Turkic. Lots of Turkic-speaking neighbors though...
Stephen
>I would appreciate information, preferably from anyone with personal
>experience, on the extent to which Turcic languages are mutually
>intelligible to their respective native speakers.
From what I've seen of many Turkic speakers coming over and
co-mingling here in the US (viz. Univ of Washington, and environs), the
mutual intelligibility levels vary from speaker to speaker. Those who
have had more contact with other Turkic language speakers (and have made
an effort to try and understand) ofttimes have it the easiest in
understanding other Turkic forms.
>I have read that, for
>example, a native speaker of Turkish has no problem speaking with natives
>speakers of, for example, Azeri, Kazah, Uzbek, etc.. Is this an
>overstatement?
Yes. Turkey Turks ofttimes have a hard time understanding their
eastern Turkic "cousins." The easiest I've seen a Republican Turk
communicate is with an Azeri... then again, those two languages are much
closer in development, etc., while the Turk (who had no other training in
many other eastern Turkic languages) had difficulty while learning Uzbek,
had a _very_ hard time understanding Qazaq (even trying to filter through
with the aid of Uzbek) and couldn't understand a thing a Tyvan or Sakha
(Yakut) said.
In a previous post, someone had posted some groupings which
seemed _very_ good and follow what I've personally seen: Turkish and
Azeri, Uzbek and Uyghur, Qazaq and Qyrghyz, with Tyva coming slightly
closer to the Qazaq and Qyrghyz, and Sakha being quite removed.. though
not as far as Chuvash.
My two instructors were Uyghurs from $arki turkestan ("Eastern
Turkestan", aka: Xinjiang) who were able to converse and communicate
without problems in most eastern Turkish forms.. of course they lived in
an area where many various Turkic peoples intermingled. They had some
problems with Republican Turkish, but fudged enough to be understood
well.
I couldn't think of any real applicable analogies or parallels
with the Turkic languages and English, sorry..
-Patrick
Indeed, that one is right on.
>
>>>: >Also, although not often used, there is the soft g that is different.
> [rules snipped]
>>Actually, it may sound like that for someone who is not familiar with
>>the language, but the movements in the mouth and throat as well as the
>>sound produces are not the same. Therefore, soft g is pronounced, but
>>'soft'ly as the name indicates. This becames even more significant if
>>you add suffixes to the word.
>
> Hmm.. in my experience with Republican Turkish and other Turkic
>languages, it seems that the rules were actually pretty on target.
Depends on who is speaking. It keeps getting worse every day. You do
know that removed the ^ hat from the 'a' etc which can cause trouble in
the context.
> While Republican Turkish retains the "soft-g" in orthography,
>lending to the funky phonological contortions one must go through ;), the
>soft-g is in NO way an actual "soft-g" as can be found in the eastern
>Turkic languages, like Uzbek, Qazaq, Qyrghyz, Tyva, etc., which are
>actual voiced velar (or pharyngeal) fricatives.
That might be the case, but as a native Turkish speaker, I cannot agree
that soft-g is not pronounced. Remember, we are not comparing.
>
>>Mustafa Soysal MS57
>
>-Patrick (pahriddin)
>
>
>--
>_____________________________________________________________________
>"je voudrais voir mourir d'amour ou bien d'haine.." T. Klingsor, "Bolero"
--
again, depends on who on both sides is engaged in the conversation
attempt. There are regions in TR which I would have a hard time
understanding what they say, much worse than compared with Azeri.
Talking speed makes a big difference too, as well as vocabulary used.
>eastern Turkic "cousins." The easiest I've seen a Republican Turk
>communicate is with an Azeri... then again, those two languages are much
>closer in development, etc., while the Turk (who had no other training in
>many other eastern Turkic languages) had difficulty while learning Uzbek,
>had a _very_ hard time understanding Qazaq (even trying to filter through
>with the aid of Uzbek) and couldn't understand a thing a Tyvan or Sakha
>(Yakut) said.
> In a previous post, someone had posted some groupings which
>seemed _very_ good and follow what I've personally seen: Turkish and
>Azeri, Uzbek and Uyghur, Qazaq and Qyrghyz, with Tyva coming slightly
>closer to the Qazaq and Qyrghyz, and Sakha being quite removed.. though
>not as far as Chuvash.
> My two instructors were Uyghurs from $arki turkestan ("Eastern
>Turkestan", aka: Xinjiang) who were able to converse and communicate
>without problems in most eastern Turkish forms.. of course they lived in
>an area where many various Turkic peoples intermingled. They had some
>problems with Republican Turkish, but fudged enough to be understood
>well.
>
> I couldn't think of any real applicable analogies or parallels
>with the Turkic languages and English, sorry..
>
>-Patrick