Loading ...
Sorry, an error occurred while loading the content.
 

An algorithm for adding new words

Expand Messages
  • lingwadeplaneta
    Salam, lidepla-jenta! Recently I had an idea of adopting a certain algorithm for importing new words. To have such an algorithm would be extremely useful,
    Message 1 of 25 , Sep 25, 2011

      Salam, lidepla-jenta!


      Recently I had an idea of adopting a certain algorithm for importing new words. To have such an algorithm would be extremely useful, because it would allow everybody to immediately make justified choices, and the language would begin to develop as if on its own. However, generally importing roots is handwork, no algorithm can take into account all nuances including sound symbolism. So I think algorithmic suggestions should be approved by a certain committee or something.Currently the committee may include everybody who wants to participate.

       

      Let's imagine we already have an algorithm. How approvement can be organized? A page on Textalok wiki may be created where everyone can submit their suggestions based on the algorithm.

      Let's say so: an algorithmic choice is regarded to be approved if there are no objections to it within one month after that the suggestion has been made. An algorithmic choice is regarded to be declined if the majority of committee is against it,for whatever reason. A reason may be, for example, that someone suggests another form, say more beautiful or reflecting the sound symbolism to a more  degree. It would be good if decisions on every suggestion take place within one month, so that to always go forward, not to be at a standstill. If someone takes to write something in LdP and makes some suggestions, he/she won't be really frustrated if his/her choice is declined quickly and another form is adopted instead.

       

      Now about the very algorithm. Omitting various reservations and details, I can see it like this. The main principles are: take the widespreadest, the shortest, and the easiest to pronounce.

       

      There are 2pools of source languages. The first pool is 10 languages: English, Chinese,Hindi-Urdu, Spanish, Portuguese, Italian, German, French, Russian, Arabic.Italian isn't a really big language but it deserves being included into the first pool as Latin's descendant. In this pool we can mark out a subpool of the6 West-European languages and a subpool of the 4 others. The second pool is: Japanese,Indonesian, Pashtoo, Swahili, Persian, Turkish... Enough? Including too many will increase difficulty.

       

      Step 1. Look at every language of pool 1 to see how the needed concept is expressed there.If in 5 or more languages there is no special word for this concept but compounds are used instead, then make a compound basing on the available dictionary. Follow the bulk in making the compound. For example, "railway"is expressed roughly as "iron+road" in 7 cases out of 10, so this should be the choice.

      In case more than half of the pool 1 languages have a specific word for the needed concept, follow step 2.

       

      Step 2. If all the 6 languages of the WE subpool + at least 1 of the 4 others agree, i.e.use similar forms for the needed concept, take it. (That must be a very widespread Latinate root like "kultura" or a widespread English word like "blog".) If it is not the case, follow step 3.

       

      Step 3. Out of 10 variants, look at those having one or two syllables (in already imported form) and choose among them the shortest and having least consonant clusters. If the shortest form has 3 consonants in a row, take a form which is one syllable longer but easier. Now check if the choice fits into the existing system. Don't forget that words may not differ only through J - CH.

      If no one-syllable or two-syllable form can be integrated, go to step 4.

       

      Step 4. Include into consideration the forms of the second pool. If there neither any one-syllable or two-syllable form fits the system, consider three-syllabled forms of both pools and choose the one with least consonant clusters. If nothing fits, go to step 5.

       

      Step 5.Look for something that would fit in other languages.

       

      There are pitfalls here, of course. Thus, it isn't desirable to take for a noun something that ends in –Ci or –Ce, where C is a consonant.

      Maybe it is not yet time to begin such activity and it's better to pick a couple thousands words "manually" first. But why not try. The procedure may be improved. Any thoughts?

       

      Dimitri

    • Attilio Liotto
      Idea es gro hao, bat komo realisi it? Swasti Ati 2011/9/25 lingwadeplaneta ... Idea es gro hao, bat komo realisi it? Swasti Ati
      Message 2 of 25 , Sep 25, 2011
        Idea es gro hao, bat komo realisi it?

        Swasti
        Ati

        2011/9/25 lingwadeplaneta <lingwadeplaneta@...>
         

        Salam, lidepla-jenta!


        Recently I had an idea of adopting a certain algorithm for importing new words. To have such an algorithm would be extremely useful, because it would allow everybody to immediately make justified choices, and the language would begin to develop as if on its own. However, generally importing roots is handwork, no algorithm can take into account all nuances including sound symbolism. So I think algorithmic suggestions should be approved by a certain committee or something.Currently the committee may include everybody who wants to participate.

         

        Let's imagine we already have an algorithm. How approvement can be organized? A page on Textalok wiki may be created where everyone can submit their suggestions based on the algorithm.

        Let's say so: an algorithmic choice is regarded to be approved if there are no objections to it within one month after that the suggestion has been made. An algorithmic choice is regarded to be declined if the majority of committee is against it,for whatever reason. A reason may be, for example, that someone suggests another form, say more beautiful or reflecting the sound symbolism to a more  degree. It would be good if decisions on every suggestion take place within one month, so that to always go forward, not to be at a standstill. If someone takes to write something in LdP and makes some suggestions, he/she won't be really frustrated if his/her choice is declined quickly and another form is adopted instead.

         

        Now about the very algorithm. Omitting various reservations and details, I can see it like this. The main principles are: take the widespreadest, the shortest, and the easiest to pronounce.

         

        There are 2pools of source languages. The first pool is 10 languages: English, Chinese,Hindi-Urdu, Spanish, Portuguese, Italian, German, French, Russian, Arabic.Italian isn't a really big language but it deserves being included into the first pool as Latin's descendant. In this pool we can mark out a subpool of the6 West-European languages and a subpool of the 4 others. The second pool is: Japanese,Indonesian, Pashtoo, Swahili, Persian, Turkish... Enough? Including too many will increase difficulty.

         

        Step 1. Look at every language of pool 1 to see how the needed concept is expressed there.If in 5 or more languages there is no special word for this concept but compounds are used instead, then make a compound basing on the available dictionary. Follow the bulk in making the compound. For example, "railway"is expressed roughly as "iron+road" in 7 cases out of 10, so this should be the choice.

        In case more than half of the pool 1 languages have a specific word for the needed concept, follow step 2.

         

        Step 2. If all the 6 languages of the WE subpool + at least 1 of the 4 others agree, i.e.use similar forms for the needed concept, take it. (That must be a very widespread Latinate root like "kultura" or a widespread English word like "blog".) If it is not the case, follow step 3.

         

        Step 3. Out of 10 variants, look at those having one or two syllables (in already imported form) and choose among them the shortest and having least consonant clusters. If the shortest form has 3 consonants in a row, take a form which is one syllable longer but easier. Now check if the choice fits into the existing system. Don't forget that words may not differ only through J - CH.

        If no one-syllable or two-syllable form can be integrated, go to step 4.

         

        Step 4. Include into consideration the forms of the second pool. If there neither any one-syllable or two-syllable form fits the system, consider three-syllabled forms of both pools and choose the one with least consonant clusters. If nothing fits, go to step 5.

         

        Step 5.Look for something that would fit in other languages.

         

        There are pitfalls here, of course. Thus, it isn't desirable to take for a noun something that ends in –Ci or –Ce, where C is a consonant.

        Maybe it is not yet time to begin such activity and it's better to pick a couple thousands words "manually" first. But why not try. The procedure may be improved. Any thoughts?

         

        Dimitri


      • Stephen Rice
        Having a time limit is important. One of the differences between Ido and Esperanto is that Eists just coin words as needed, while Idists feel obliged to stick
        Message 3 of 25 , Sep 25, 2011
          Having a time limit is important. One of the differences between Ido
          and Esperanto is that Eists just coin words as needed, while Idists
          feel obliged to stick asterisks on neologisms. Then they have to wait
          a few decades to see whether it's legalized.

          Other than that, the algorithm seems reasonable--but is ch/j the only
          proscribed minimal pair?

          Steve

          On 9/25/11, Attilio Liotto <attilio.liotto@...> wrote:
          > Idea es gro hao, bat komo realisi it?
          >
          > Swasti
          > Ati
          >
          > 2011/9/25 lingwadeplaneta <lingwadeplaneta@...>
          >
          >> **
          >>
          >>
          >> ********
          >>
          >> Salam, lidepla-jenta!
          >>
          >>
          >> Recently I had an idea of adopting a certain algorithm for importing new
          >> words. To have such an algorithm would be extremely useful, because it
          >> would
          >> allow everybody to immediately make justified choices, and the language
          >> would begin to develop as if on its own. However, generally importing
          >> roots
          >> is handwork, no algorithm can take into account all nuances including
          >> sound
          >> symbolism. So I think algorithmic suggestions should be approved by a
          >> certain committee or something.Currently the committee may include
          >> everybody
          >> who wants to participate. ****
          >>
          >> ** **
          >>
          >> Let's imagine we already have an algorithm. How approvement can be
          >> organized? A page on Textalok wiki may be created where everyone can
          >> submit
          >> their suggestions based on the algorithm. ****
          >>
          >> Let's say so: an algorithmic choice is regarded to be approved if there
          >> are
          >> no objections to it within one month after that the suggestion has been
          >> made. An algorithmic choice is regarded to be declined if the majority of
          >> committee is against it,for whatever reason. A reason may be, for example,
          >> that someone suggests another form, say more beautiful or reflecting the
          >> sound symbolism to a more degree. It would be good if decisions on every
          >> suggestion take place within one month, so that to always go forward, not
          >> to
          >> be at a standstill. If someone takes to write something in LdP and makes
          >> some suggestions, he/she won't be really frustrated if his/her choice is
          >> declined quickly and another form is adopted instead. ****
          >>
          >> ** **
          >>
          >> Now about the very algorithm. Omitting various reservations and details, I
          >> can see it like this. The main principles are: take the widespreadest, the
          >> shortest, and the easiest to pronounce. ****
          >>
          >> ** **
          >>
          >> There are 2pools of source languages. The first pool is 10 languages:
          >> English, Chinese,Hindi-Urdu, Spanish, Portuguese, Italian, German, French,
          >> Russian, Arabic.Italian isn't a really big language but it deserves being
          >> included into the first pool as Latin's descendant. In this pool we can
          >> mark
          >> out a subpool of the6 West-European languages and a subpool of the 4
          >> others.
          >> The second pool is: Japanese,Indonesian, Pashtoo, Swahili, Persian,
          >> Turkish... Enough? Including too many will increase difficulty. ****
          >>
          >> ** **
          >>
          >> Step 1. Look at every language of pool 1 to see how the needed concept is
          >> expressed there.If in 5 or more languages there is no special word for
          >> this
          >> concept but compounds are used instead, then make a compound basing on the
          >> available dictionary. Follow the bulk in making the compound. For example,
          >> "railway"is expressed roughly as "iron+road" in 7 cases out of 10, so this
          >> should be the choice.****
          >>
          >> In case more than half of the pool 1 languages have a specific word for
          >> the
          >> needed concept, follow step 2.****
          >>
          >> ** **
          >>
          >> Step 2. If all the 6 languages of the WE subpool + at least 1 of the 4
          >> others agree, i.e.use similar forms for the needed concept, take it. (That
          >> must be a very widespread Latinate root like "kultura" or a widespread
          >> English word like "blog".) If it is not the case, follow step 3.****
          >>
          >> ** **
          >>
          >> Step 3. Out of 10 variants, look at those having one or two syllables (in
          >> already imported form) and choose among them the shortest and having least
          >> consonant clusters. If the shortest form has 3 consonants in a row, take a
          >> form which is one syllable longer but easier. Now check if the choice fits
          >> into the existing system. Don't forget that words may not differ only
          >> through J - ****CH.********
          >>
          >> If no one-syllable or two-syllable form can be integrated, go to step 4.**
          >> **
          >>
          >> ** **
          >>
          >> Step 4. Include into consideration the forms of the second pool. If there
          >> neither any one-syllable or two-syllable form fits the system, consider
          >> three-syllabled forms of both pools and choose the one with least
          >> consonant
          >> clusters. If nothing fits, go to step 5.****
          >>
          >> ** **
          >>
          >> Step 5.Look for something that would fit in other languages.****
          >>
          >> ** **
          >>
          >> There are pitfalls here, of course. Thus, it isn't desirable to take for a
          >> noun something that ends in –Ci or –Ce, where C is a consonant. ****
          >>
          >> Maybe it is not yet time to begin such activity and it's better to pick a
          >> couple thousands words "manually" first. But why not try. The procedure
          >> may
          >> be improved. Any thoughts?****
          >>
          >> ** **
          >>
          >> Dimitri****
          >>
          >>
          >>
          >
        • Attilio Liotto
          Aŭdu ba! Presentation de Interlingu: http://www.youtube.com/watch?v=sp5O0J6-4bk&feature=related Swasti Ati 2011/9/25 Stephen Rice ...
          Message 4 of 25 , Sep 26, 2011
            Aŭdu ba!

            Presentation de Interlingu:

            http://www.youtube.com/watch?v=sp5O0J6-4bk&feature=related

            Swasti

            Ati


            2011/9/25 Stephen Rice <ansrith@...>
            Having a time limit is important. One of the differences between Ido
            and Esperanto is that Eists just coin words as needed, while Idists
            feel obliged to stick asterisks on neologisms. Then they have to wait
            a few decades to see whether it's legalized.

            Other than that, the algorithm seems reasonable--but is ch/j the only
            proscribed minimal pair?

            Steve

            On 9/25/11, Attilio Liotto <attilio.liotto@...> wrote:
            > Idea es gro hao, bat komo realisi it?
            >
            > Swasti
            > Ati
            >
            > 2011/9/25 lingwadeplaneta <lingwadeplaneta@...>
            >
            >> **
            >>
            >>
            >> ********
            >>
            >> Salam, lidepla-jenta!
            >>
            >>
            >> Recently I had an idea of adopting a certain algorithm for importing new
            >> words. To have such an algorithm would be extremely useful, because it
            >> would
            >> allow everybody to immediately make justified choices, and the language
            >> would begin to develop as if on its own. However, generally importing
            >> roots
            >> is handwork, no algorithm can take into account all nuances including
            >> sound
            >> symbolism. So I think algorithmic suggestions should be approved by a
            >> certain committee or something.Currently the committee may include
            >> everybody
            >> who wants to participate. ****
            >>
            >> ** **
            >>
            >> Let's imagine we already have an algorithm. How approvement can be
            >> organized? A page on Textalok wiki may be created where everyone can
            >> submit
            >> their suggestions based on the algorithm. ****
            >>
            >> Let's say so: an algorithmic choice is regarded to be approved if there
            >> are
            >> no objections to it within one month after that the suggestion has been
            >> made. An algorithmic choice is regarded to be declined if the majority of
            >> committee is against it,for whatever reason. A reason may be, for example,
            >> that someone suggests another form, say more beautiful or reflecting the
            >> sound symbolism to a more  degree. It would be good if decisions on every
            >> suggestion take place within one month, so that to always go forward, not
            >> to
            >> be at a standstill. If someone takes to write something in LdP and makes
            >> some suggestions, he/she won't be really frustrated if his/her choice is
            >> declined quickly and another form is adopted instead. ****
            >>
            >> ** **
            >>
            >> Now about the very algorithm. Omitting various reservations and details, I
            >> can see it like this. The main principles are: take the widespreadest, the
            >> shortest, and the easiest to pronounce. ****
            >>
            >> ** **
            >>
            >> There are 2pools of source languages. The first pool is 10 languages:
            >> English, Chinese,Hindi-Urdu, Spanish, Portuguese, Italian, German, French,
            >> Russian, Arabic.Italian isn't a really big language but it deserves being
            >> included into the first pool as Latin's descendant. In this pool we can
            >> mark
            >> out a subpool of the6 West-European languages and a subpool of the 4
            >> others.
            >> The second pool is: Japanese,Indonesian, Pashtoo, Swahili, Persian,
            >> Turkish... Enough? Including too many will increase difficulty. ****
            >>
            >> ** **
            >>
            >> Step 1. Look at every language of pool 1 to see how the needed concept is
            >> expressed there.If in 5 or more languages there is no special word for
            >> this
            >> concept but compounds are used instead, then make a compound basing on the
            >> available dictionary. Follow the bulk in making the compound. For example,
            >> "railway"is expressed roughly as "iron+road" in 7 cases out of 10, so this
            >> should be the choice.****
            >>
            >> In case more than half of the pool 1 languages have a specific word for
            >> the
            >> needed concept, follow step 2.****
            >>
            >> ** **
            >>
            >> Step 2. If all the 6 languages of the WE subpool + at least 1 of the 4
            >> others agree, i.e.use similar forms for the needed concept, take it. (That
            >> must be a very widespread Latinate root like "kultura" or a widespread
            >> English word like "blog".) If it is not the case, follow step 3.****
            >>
            >> ** **
            >>
            >> Step 3. Out of 10 variants, look at those having one or two syllables (in
            >> already imported form) and choose among them the shortest and having least
            >> consonant clusters. If the shortest form has 3 consonants in a row, take a
            >> form which is one syllable longer but easier. Now check if the choice fits
            >> into the existing system. Don't forget that words may not differ only
            >> through J - ****CH.********
            >>
            >> If no one-syllable or two-syllable form can be integrated, go to step 4.**
            >> **
            >>
            >> ** **
            >>
            >> Step 4. Include into consideration the forms of the second pool. If there
            >> neither any one-syllable or two-syllable form fits the system, consider
            >> three-syllabled forms of both pools and choose the one with least
            >> consonant
            >> clusters. If nothing fits, go to step 5.****
            >>
            >> ** **
            >>
            >> Step 5.Look for something that would fit in other languages.****
            >>
            >> ** **
            >>
            >> There are pitfalls here, of course. Thus, it isn't desirable to take for a
            >> noun something that ends in –Ci or –Ce, where C is a consonant. ****
            >>
            >> Maybe it is not yet time to begin such activity and it's better to pick a
            >> couple thousands words "manually" first. But why not try. The procedure
            >> may
            >> be improved. Any thoughts?****
            >>
            >> ** **
            >>
            >> Dimitri****
            >>
            >>
            >>
            >


            ------------------------------------

            Yahoo! Groups Links

            <*> To visit your group on the web, go to:
               http://groups.yahoo.com/group/lingwadeplaneta/

            <*> Your email settings:
               Individual Email | Traditional

            <*> To change settings online go to:
               http://groups.yahoo.com/group/lingwadeplaneta/join
               (Yahoo! ID required)

            <*> To change settings via email:
               lingwadeplaneta-digest@yahoogroups.com
               lingwadeplaneta-fullfeatured@yahoogroups.com

            <*> To unsubscribe from this group, send an email to:
               lingwadeplaneta-unsubscribe@yahoogroups.com

            <*> Your use of Yahoo! Groups is subject to:
               http://docs.yahoo.com/info/terms/


          • lingwadeplaneta
            ... Treba experimenti. Sey algoritma es fo toy jen kel yo jan lidepla, jan suy fonetika e osobitaa. Yoshi sey jen mus bi kapable de findi ol treba-ney info in
            Message 5 of 25 , Sep 26, 2011
              --- In lingwadeplaneta@yahoogroups.com, Attilio Liotto <attilio.liotto@...> wrote:
              >
              > Idea es gro-hao, bat komo realisi it?
              >
              > Swasti
              > Ati


              Treba experimenti.
              Sey algoritma es fo toy jen kel yo jan lidepla, jan suy fonetika e osobitaa. Yoshi sey jen mus bi kapable de findi ol treba-ney info in lexikon de mucho lingwa. Bu es fasile, yedoh ye kelke moden media kel mog helpi: pa exampla, Gugl-tradukter lekti lautem den skriba de arabi, hindi, putunhua, nihon...
            • lingwadeplaneta
              ... Yes. But there are many other details to take into account, which require knowledge not only of LdP but also of at least basics of Mandarin, Arabic, etc.
              Message 6 of 25 , Sep 26, 2011
                --- In lingwadeplaneta@yahoogroups.com, Stephen Rice <ansrith@...> wrote:
                >
                > Having a time limit is important. One of the differences between Ido
                > and Esperanto is that Eists just coin words as needed, while Idists
                > feel obliged to stick asterisks on neologisms. Then they have to wait
                > a few decades to see whether it's legalized.
                >
                > Other than that, the algorithm seems reasonable--but is ch/j the only
                > proscribed minimal pair?
                >

                Yes. But there are many other details to take into account, which require knowledge not only of LdP but also of at least basics of Mandarin, Arabic, etc. For example, when importing a Mandarin word which contains JI, should it transform into DI or into ZI? A lot, a lot of such things. But even the presented general outlines may be helpful for someone who would be guessing how to find a word.

                Dmitry
              • lingwadeplaneta
                Experiment #1 Let s find a lidepla word for raccoon . In the first pool we have: English: raccoon Arabic: rakiun Hindi: raikun Chinese: huanxiong Russian:
                Message 7 of 25 , Sep 27, 2011
                  Experiment #1

                  Let's find a lidepla word for "raccoon".
                  In the first pool we have:

                  English: raccoon
                  Arabic: rakiun
                  Hindi: raikun
                  Chinese: huanxiong
                  Russian: yenot
                  German: Waschbaer
                  Spanish: mapache
                  Portuguese: guaxinim
                  Italian: procione
                  French: raton laveur

                  Step 1. In the majority of cases it's a specific word, so a specific word is needed.

                  Step 2. No universal agreement, so move on.

                  Step 3. The shortest forms are rakun (English) and yenot (Russian). Both good, without consonant clusters, but rakun has the advantage that Arabic and Hindi, the languages of different language families, have similar forms. (Surely they just imported this word, but it doesn't matter).

                  The result: rakun.

                  (A glimpse into pool 2 makes us completely certain about rakun:
                  Indonesian: rakun
                  Turkish: rakun
                  Japanese: araiguma...
                  I begin to think that raccoons live mostly in English-speaking countries, as almost everybody imports this word.)


                  Experiment #2

                  Let's find a lidepla word for "linden".
                  The first pool:

                  English: linden
                  Arabic: zeizafun
                  Hindi: linden
                  Chinese: duanshu
                  Russian: lipa
                  German: Linde
                  Spanish: tilo
                  Portuguese: tilia
                  Italian: tiglio
                  French: tilleul

                  Steps 1 and 2 make us proceed to step 3.

                  Step 3. The shortest forms are lipa, Linde and tilo. "nd" in Linde isn't a consonant cluster and I don't think that Linde is really longer than lipa or more difficult. So any of the three could go.

                  Let's see how they fit into the existing dictionary. Lipa doesn't fit because there is already lipi. Linde is OK and tilo too (hardly tilo may clash with the preposition "til"). If import Linde or tilo, they should rather become linda and tila. Both are good. So, which of them?

                  Tila is supported by similar Romance forms:
                  Portuguese: tilia
                  Italian: tiglio
                  French: tilleul

                  Linda is supported by:
                  English: linden
                  Hindi: linden

                  I think there should be a kind of rule that similarity within a language family weighs less than similarity between different language families. So similarity between Romance languages is sort of balanced by the similarity between Germanic ones (English and German), and there is also Hindi to support the latters.
                  Maybe in such cases both choices should be presented to the committee.

                  As a member of the committee, I would prefer linda, because 8 of 10 forms of pool 1 contain the sound L and 4 of them begin with L. So the importance of L is high, and I'd prefer to see the word begin with L.







                  --- In lingwadeplaneta@yahoogroups.com, "lingwadeplaneta" <lingwadeplaneta@...> wrote:
                  >
                  >
                  > Salam, lidepla-jenta!
                  >
                  >
                  >
                  >
                  >
                  > Recently I had an idea of adopting a certain algorithm for importing new
                  > words. To have such an algorithm would be extremely useful, because it
                  > would allow everybody to immediately make justified choices, and the
                  > language would begin to develop as if on its own. However, generally
                  > importing roots is handwork, no algorithm can take into account all
                  > nuances including sound symbolism. So I think algorithmic suggestions
                  > should be approved by a certain committee or something.Currently the
                  > committee may include everybody who wants to participate.
                  >
                  >
                  >
                  > Let's imagine we already have an algorithm. How approvement can be
                  > organized? A page on Textalok wiki may be created where everyone can
                  > submit their suggestions based on the algorithm.
                  >
                  > Let's say so: an algorithmic choice is regarded to be approved if there
                  > are no objections to it within one month after that the suggestion has
                  > been made. An algorithmic choice is regarded to be declined if the
                  > majority of committee is against it,for whatever reason. A reason may
                  > be, for example, that someone suggests another form, say more beautiful
                  > or reflecting the sound symbolism to a more degree. It would be good if
                  > decisions on every suggestion take place within one month, so that to
                  > always go forward, not to be at a standstill. If someone takes to write
                  > something in LdP and makes some suggestions, he/she won't be really
                  > frustrated if his/her choice is declined quickly and another form is
                  > adopted instead.
                  >
                  >
                  >
                  > Now about the very algorithm. Omitting various reservations and details,
                  > I can see it like this. The main principles are: take the widespreadest,
                  > the shortest, and the easiest to pronounce.
                  >
                  >
                  >
                  > There are 2pools of source languages. The first pool is 10 languages:
                  > English, Chinese,Hindi-Urdu, Spanish, Portuguese, Italian, German,
                  > French, Russian, Arabic.Italian isn't a really big language but it
                  > deserves being included into the first pool as Latin's descendant. In
                  > this pool we can mark out a subpool of the6 West-European languages and
                  > a subpool of the 4 others. The second pool is: Japanese,Indonesian,
                  > Pashtoo, Swahili, Persian, Turkish... Enough? Including too many will
                  > increase difficulty.
                  >
                  >
                  >
                  > Step 1. Look at every language of pool 1 to see how the needed concept
                  > is expressed there.If in 5 or more languages there is no special word
                  > for this concept but compounds are used instead, then make a compound
                  > basing on the available dictionary. Follow the bulk in making the
                  > compound. For example, "railway"is expressed roughly as "iron+road" in 7
                  > cases out of 10, so this should be the choice.
                  >
                  > In case more than half of the pool 1 languages have a specific word for
                  > the needed concept, follow step 2.
                  >
                  >
                  >
                  > Step 2. If all the 6 languages of the WE subpool + at least 1 of the 4
                  > others agree, i.e.use similar forms for the needed concept, take it.
                  > (That must be a very widespread Latinate root like "kultura" or a
                  > widespread English word like "blog".) If it is not the case, follow step
                  > 3.
                  >
                  >
                  >
                  > Step 3. Out of 10 variants, look at those having one or two syllables
                  > (in already imported form) and choose among them the shortest and having
                  > least consonant clusters. If the shortest form has 3 consonants in a
                  > row, take a form which is one syllable longer but easier. Now check if
                  > the choice fits into the existing system. Don't forget that words may
                  > not differ only through J - CH.
                  >
                  > If no one-syllable or two-syllable form can be integrated, go to step 4.
                  >
                  >
                  >
                  > Step 4. Include into consideration the forms of the second pool. If
                  > there neither any one-syllable or two-syllable form fits the system,
                  > consider three-syllabled forms of both pools and choose the one with
                  > least consonant clusters. If nothing fits, go to step 5.
                  >
                  >
                  >
                  > Step 5.Look for something that would fit in other languages.
                  >
                  >
                  >
                  > There are pitfalls here, of course. Thus, it isn't desirable to take for
                  > a noun something that ends in –Ci or –Ce, where C is a
                  > consonant.
                  >
                  > Maybe it is not yet time to begin such activity and it's better to pick
                  > a couple thousands words "manually" first. But why not try. The
                  > procedure may be improved. Any thoughts?
                  >
                  >
                  >
                  > Dimitri
                  >
                • Stephen Rice
                  That looks workable, especially when you weight the results to favor forms widely borrowed beyond their own family. Steve
                  Message 8 of 25 , Sep 27, 2011
                    That looks workable, especially when you weight the results to favor
                    forms widely borrowed beyond their own family.

                    Steve
                  • cafaristeir
                    Hao dey, kare Dmitry ! Working with pools of language is good, but I think that an automatic algorithm can t work. A human must always be there to ensure
                    Message 9 of 25 , Sep 27, 2011
                      Hao dey, kare Dmitry !

                      Working with "pools" of language is good, but I think that an "automatic" algorithm can't work. A human must always be there to ensure that this system doesn't generate ridiculous, unpronounceable words, or homophones.

                      Beware of Google translator, which tends to repeat the English word when it finds no translation in its data base !

                      For "Linden", Sambahsa has "liep", from Balto-Slavic (and a vague Welsh correspondance). "Linden" comes from PIE *lenteH2, but it would have created a confusion with "lent" = "slow" in Sambahsa. Indeed, they may share the same root *len- = "flexible, lazy", found in "lenient" or in Russian "lenj".

                      The "racoon" comes indeed from North America, and the English word from an Indian dialect. Though the French were the first Europeans to trade furs in Northern America, she seems to have no particular name, as Quebeckers say "chat sauvage" and Cajuns "chaoui" (from an Indian dialect).
                      If we look closer at the various translations, we often name like "washing-bear", "washing-rat", because this animal washes her food before eating. When new people settle into an unknown area, they tend to create names of animals with a compound containing the name of an already known animal.
                      For example, Sambahsa has "kurmiaend" for "platypus", i.e. "mole-duck". The Russian name "yenot" is cognate with "genet", a small mammal whose name comes from Arabic, and which has been taken back by Sambahsa : "jarnayt". Hence, I suggest "lavjarnayt" = "racoon" in Sambahsa ("lav" = to wash).

                      Swasti-gro !

                      Olivier



                      --- In lingwadeplaneta@yahoogroups.com, "lingwadeplaneta" <lingwadeplaneta@...> wrote:
                      >
                      >
                      >
                      > Experiment #1
                      >
                      > Let's find a lidepla word for "raccoon".
                      > In the first pool we have:
                      >
                      > English: raccoon
                      > Arabic: rakiun
                      > Hindi: raikun
                      > Chinese: huanxiong
                      > Russian: yenot
                      > German: Waschbaer
                      > Spanish: mapache
                      > Portuguese: guaxinim
                      > Italian: procione
                      > French: raton laveur
                      >
                      > Step 1. In the majority of cases it's a specific word, so a specific word is needed.
                      >
                      > Step 2. No universal agreement, so move on.
                      >
                      > Step 3. The shortest forms are rakun (English) and yenot (Russian). Both good, without consonant clusters, but rakun has the advantage that Arabic and Hindi, the languages of different language families, have similar forms. (Surely they just imported this word, but it doesn't matter).
                      >
                      > The result: rakun.
                      >
                      > (A glimpse into pool 2 makes us completely certain about rakun:
                      > Indonesian: rakun
                      > Turkish: rakun
                      > Japanese: araiguma...
                      > I begin to think that raccoons live mostly in English-speaking countries, as almost everybody imports this word.)
                      >
                      >
                      > Experiment #2
                      >
                      > Let's find a lidepla word for "linden".
                      > The first pool:
                      >
                      > English: linden
                      > Arabic: zeizafun
                      > Hindi: linden
                      > Chinese: duanshu
                      > Russian: lipa
                      > German: Linde
                      > Spanish: tilo
                      > Portuguese: tilia
                      > Italian: tiglio
                      > French: tilleul
                      >
                      > Steps 1 and 2 make us proceed to step 3.
                      >
                      > Step 3. The shortest forms are lipa, Linde and tilo. "nd" in Linde isn't a consonant cluster and I don't think that Linde is really longer than lipa or more difficult. So any of the three could go.
                      >
                      > Let's see how they fit into the existing dictionary. Lipa doesn't fit because there is already lipi. Linde is OK and tilo too (hardly tilo may clash with the preposition "til"). If import Linde or tilo, they should rather become linda and tila. Both are good. So, which of them?
                      >
                      > Tila is supported by similar Romance forms:
                      > Portuguese: tilia
                      > Italian: tiglio
                      > French: tilleul
                      >
                      > Linda is supported by:
                      > English: linden
                      > Hindi: linden
                      >
                      > I think there should be a kind of rule that similarity within a language family weighs less than similarity between different language families. So similarity between Romance languages is sort of balanced by the similarity between Germanic ones (English and German), and there is also Hindi to support the latters.
                      > Maybe in such cases both choices should be presented to the committee.
                      >
                      > As a member of the committee, I would prefer linda, because 8 of 10 forms of pool 1 contain the sound L and 4 of them begin with L. So the importance of L is high, and I'd prefer to see the word begin with L.
                      >
                      >
                      >
                      >
                      >
                      >
                      >
                      > --- In lingwadeplaneta@yahoogroups.com, "lingwadeplaneta" <lingwadeplaneta@> wrote:
                      > >
                      > >
                      > > Salam, lidepla-jenta!
                      > >
                      > >
                      > >
                      > >
                      > >
                      > > Recently I had an idea of adopting a certain algorithm for importing new
                      > > words. To have such an algorithm would be extremely useful, because it
                      > > would allow everybody to immediately make justified choices, and the
                      > > language would begin to develop as if on its own. However, generally
                      > > importing roots is handwork, no algorithm can take into account all
                      > > nuances including sound symbolism. So I think algorithmic suggestions
                      > > should be approved by a certain committee or something.Currently the
                      > > committee may include everybody who wants to participate.
                      > >
                      > >
                      > >
                      > > Let's imagine we already have an algorithm. How approvement can be
                      > > organized? A page on Textalok wiki may be created where everyone can
                      > > submit their suggestions based on the algorithm.
                      > >
                      > > Let's say so: an algorithmic choice is regarded to be approved if there
                      > > are no objections to it within one month after that the suggestion has
                      > > been made. An algorithmic choice is regarded to be declined if the
                      > > majority of committee is against it,for whatever reason. A reason may
                      > > be, for example, that someone suggests another form, say more beautiful
                      > > or reflecting the sound symbolism to a more degree. It would be good if
                      > > decisions on every suggestion take place within one month, so that to
                      > > always go forward, not to be at a standstill. If someone takes to write
                      > > something in LdP and makes some suggestions, he/she won't be really
                      > > frustrated if his/her choice is declined quickly and another form is
                      > > adopted instead.
                      > >
                      > >
                      > >
                      > > Now about the very algorithm. Omitting various reservations and details,
                      > > I can see it like this. The main principles are: take the widespreadest,
                      > > the shortest, and the easiest to pronounce.
                      > >
                      > >
                      > >
                      > > There are 2pools of source languages. The first pool is 10 languages:
                      > > English, Chinese,Hindi-Urdu, Spanish, Portuguese, Italian, German,
                      > > French, Russian, Arabic.Italian isn't a really big language but it
                      > > deserves being included into the first pool as Latin's descendant. In
                      > > this pool we can mark out a subpool of the6 West-European languages and
                      > > a subpool of the 4 others. The second pool is: Japanese,Indonesian,
                      > > Pashtoo, Swahili, Persian, Turkish... Enough? Including too many will
                      > > increase difficulty.
                      > >
                      > >
                      > >
                      > > Step 1. Look at every language of pool 1 to see how the needed concept
                      > > is expressed there.If in 5 or more languages there is no special word
                      > > for this concept but compounds are used instead, then make a compound
                      > > basing on the available dictionary. Follow the bulk in making the
                      > > compound. For example, "railway"is expressed roughly as "iron+road" in 7
                      > > cases out of 10, so this should be the choice.
                      > >
                      > > In case more than half of the pool 1 languages have a specific word for
                      > > the needed concept, follow step 2.
                      > >
                      > >
                      > >
                      > > Step 2. If all the 6 languages of the WE subpool + at least 1 of the 4
                      > > others agree, i.e.use similar forms for the needed concept, take it.
                      > > (That must be a very widespread Latinate root like "kultura" or a
                      > > widespread English word like "blog".) If it is not the case, follow step
                      > > 3.
                      > >
                      > >
                      > >
                      > > Step 3. Out of 10 variants, look at those having one or two syllables
                      > > (in already imported form) and choose among them the shortest and having
                      > > least consonant clusters. If the shortest form has 3 consonants in a
                      > > row, take a form which is one syllable longer but easier. Now check if
                      > > the choice fits into the existing system. Don't forget that words may
                      > > not differ only through J - CH.
                      > >
                      > > If no one-syllable or two-syllable form can be integrated, go to step 4.
                      > >
                      > >
                      > >
                      > > Step 4. Include into consideration the forms of the second pool. If
                      > > there neither any one-syllable or two-syllable form fits the system,
                      > > consider three-syllabled forms of both pools and choose the one with
                      > > least consonant clusters. If nothing fits, go to step 5.
                      > >
                      > >
                      > >
                      > > Step 5.Look for something that would fit in other languages.
                      > >
                      > >
                      > >
                      > > There are pitfalls here, of course. Thus, it isn't desirable to take for
                      > > a noun something that ends in –Ci or –Ce, where C is a
                      > > consonant.
                      > >
                      > > Maybe it is not yet time to begin such activity and it's better to pick
                      > > a couple thousands words "manually" first. But why not try. The
                      > > procedure may be improved. Any thoughts?
                      > >
                      > >
                      > >
                      > > Dimitri
                      > >
                      >
                    • Stephen Rice
                      ... Howdy, kare Olivier! ... 1. It s not automated, so bad forms shouldn t arise, though homophones could, so 2. there s supposed to be a group of actual human
                      Message 10 of 25 , Sep 27, 2011
                        On 9/27/11, cafaristeir <cafaristeir@...> wrote:
                        > Hao dey, kare Dmitry !

                        Howdy, kare Olivier!

                        > Working with "pools" of language is good, but I think that an "automatic"
                        > algorithm can't work. A human must always be there to ensure that this
                        > system doesn't generate ridiculous, unpronounceable words, or homophones.

                        1. It's not automated, so bad forms shouldn't arise, though homophones could, so

                        2. there's supposed to be a group of actual human beings to examine
                        proposed neologisms.

                        I don't see a problem. Of course, eventually, when LdP has thousands
                        of enthusiastic users, they'll have to allow a more natural approach
                        to lexical growth. But this should work for now.

                        > Beware of Google translator, which tends to repeat the English word when it
                        > finds no translation in its data base !

                        Lazy devils! They should pick a word at random just to keep it interesting.

                        > The "racoon" comes indeed from North America, and the English word from an
                        > Indian dialect. Though the French were the first Europeans to trade furs in
                        > Northern America, she seems to have no particular name, as Quebeckers say
                        > "chat sauvage" and Cajuns "chaoui" (from an Indian dialect).
                        > If we look closer at the various translations, we often name like
                        > "washing-bear", "washing-rat", because this animal washes her food before
                        > eating. When new people settle into an unknown area, they tend to create
                        > names of animals with a compound containing the name of an already known
                        > animal.
                        > For example, Sambahsa has "kurmiaend" for "platypus", i.e. "mole-duck". The
                        > Russian name "yenot" is cognate with "genet", a small mammal whose name
                        > comes from Arabic, and which has been taken back by Sambahsa : "jarnayt".
                        > Hence, I suggest "lavjarnayt" = "racoon" in Sambahsa ("lav" = to wash).

                        This demonstrates the difference (one of them, anyway) between
                        Sambahsa and LdP: Sambahsa has etymological concerns that I don't
                        think matter to LdP, while LdP has concerns with current forms and
                        usage that are largely foreign to Sambahsa. Sambahsa asks, "Where did
                        it come from?" LdP asks, "What's out there right now that I can use?"

                        Steve
                      • lingwadeplaneta
                        ... Dey hao, kares! Imitating assalam aleikum - aleikum assalam! : - Hao dey! - Dey hao! ... And I ve noticed that it translates any plant or animal into
                        Message 11 of 25 , Sep 28, 2011
                          --- In lingwadeplaneta@yahoogroups.com, Stephen Rice <ansrith@...> wrote:
                          >
                          > On 9/27/11, cafaristeir <cafaristeir@...> wrote:
                          > > Hao dey, kare Dmitry !
                          >
                          > Howdy, kare Olivier!

                          Dey hao, kares!

                          Imitating "assalam aleikum - aleikum assalam!":
                          - Hao dey!
                          - Dey hao!

                          >
                          > > Working with "pools" of language is good, but I think that an "automatic"
                          > > algorithm can't work. A human must always be there to ensure that this
                          > > system doesn't generate ridiculous, unpronounceable words, or homophones.
                          >
                          > 1. It's not automated, so bad forms shouldn't arise, though homophones could, so
                          >
                          > 2. there's supposed to be a group of actual human beings to examine
                          > proposed neologisms.
                          >
                          > I don't see a problem. Of course, eventually, when LdP has thousands
                          > of enthusiastic users, they'll have to allow a more natural approach
                          > to lexical growth. But this should work for now.
                          >
                          > > Beware of Google translator, which tends to repeat the English word when it
                          > > finds no translation in its data base !
                          >
                          > Lazy devils! They should pick a word at random just to keep it interesting.
                          >

                          And I've noticed that it translates any plant or animal into Hindi as "a type of plant" or "a kind of animal".
                          Still it may be useful for reading aloud Arabic script. Look for Arabic word in a good English-Arabic dictionary, then paste it into Google and listen. Surely, one must also check the exact meaning of the word in Arabic-English dictionary.

                          Dmitry
                        • cafaristeir
                          Hao dey kare Steve ! ... I don t bother about the lack of etymological concern, but about the now . Now , US-American is extremely influential and provides a
                          Message 12 of 25 , Sep 28, 2011
                            Hao dey kare Steve !

                            > This demonstrates the difference (one of them, anyway) between
                            > Sambahsa and LdP: Sambahsa has etymological concerns that I don't
                            > think matter to LdP, while LdP has concerns with current forms and
                            > usage that are largely foreign to Sambahsa. Sambahsa asks, "Where did
                            > it come from?" LdP asks, "What's out there right now that I can use?"
                            >

                            I don't bother about the lack of etymological concern, but about the "now". "Now", US-American is extremely influential and provides a lot of neologisms to other languages. But "what if" is this situation changes ? For words which are linked to fashion, culture, a certain kind of civilisation (ex: hit-parade, hamburger, vuvuzela...), this is not a problem. But this is not so for biological realities or names of species.
                            This concern for Sambahsa does indeed not come from etymology, but from the fact I have often translated gamebooks whose action takes place at other times, on other worlds. That's why I try to refrain from using too recent americanisms for Sambahsa.

                            Olivier
                          • Stephen Rice
                            ... Then go elsewhere. But I don t foresee a swift change, and already established words, especially borrowings, will probably endure easily enough. For words
                            Message 13 of 25 , Sep 28, 2011
                              On 9/28/11, cafaristeir <cafaristeir@...> wrote:
                              >
                              >> This demonstrates the difference (one of them, anyway) between
                              >> Sambahsa and LdP: Sambahsa has etymological concerns that I don't
                              >> think matter to LdP, while LdP has concerns with current forms and
                              >> usage that are largely foreign to Sambahsa. Sambahsa asks, "Where did
                              >> it come from?" LdP asks, "What's out there right now that I can use?"
                              >>
                              >
                              > I don't bother about the lack of etymological concern, but about the "now".
                              > "Now", US-American is extremely influential and provides a lot of neologisms
                              > to other languages. But "what if" is this situation changes ?

                              Then go elsewhere. But I don't foresee a swift change, and already
                              established words, especially borrowings, will probably endure easily
                              enough.

                              For words
                              > which are linked to fashion, culture, a certain kind of civilisation (ex:
                              > hit-parade, hamburger, vuvuzela...), this is not a problem. But this is not
                              > so for biological realities or names of species.

                              Isn't it? In this case, we are dealing with a North American animal,
                              and the English word has been borrowed elsewhere (Hindi, Indonesian,
                              Macedonian, Slovene, Serbo-croatian, Turkish, more or less in
                              Japanese, and (dispositively) in Volapük). This isn't likely to be
                              undone anytime soon. No other form is so widespread. Germanic
                              languages generally have "wash-bear," though there are apparently
                              related forms in a few Romance languages. But if I considered an
                              alternative, it would likely be the Slavic yanot/yenot.

                              > This concern for Sambahsa does indeed not come from etymology, but from the
                              > fact I have often translated gamebooks whose action takes place at other
                              > times, on other worlds. That's why I try to refrain from using too recent
                              > americanisms for Sambahsa.

                              I doubt that in this case at least the borrowing is recent.

                              While my linguistic analyses are typically diachronic, in this sort of
                              work I tend in a more synchronic direction, and dmitry's method is a
                              good one. The alternative (for plants and animals) would be to adapt
                              Linnaean names. For example, I'm currently translating
                              "Rikki-Tikki-Tavi" into Occidental, and I can't find a word for
                              "muskrat." Do I use *musco-ratte or *ondatra? (Perhaps actually
                              *ondatre.) I suspect I'll use the first, especially as it lends itself
                              to the idea that Chuchundra's cousin is a regular rat. (It's also the
                              solution Dmitry (or whoever) used in the LdP version.) But since I
                              intend to do an Interlingua version as well, I'll run into the problem
                              again--and worse, because I can't find an Interlingua word for
                              "mongoose."

                              Steve
                            • cafaristeir
                              . The alternative (for plants and animals) would be to adapt ... For Interlingua, I suggest manguste , which is the most widespread Romance (and even Slavic)
                              Message 14 of 25 , Sep 29, 2011
                                . The alternative (for plants and animals) would be to adapt
                                > Linnaean names. For example, I'm currently translating
                                > "Rikki-Tikki-Tavi" into Occidental, and I can't find a word for
                                > "muskrat." Do I use *musco-ratte or *ondatra? (Perhaps actually
                                > *ondatre.) I suspect I'll use the first, especially as it lends itself
                                > to the idea that Chuchundra's cousin is a regular rat. (It's also the
                                > solution Dmitry (or whoever) used in the LdP version.) But since I
                                > intend to do an Interlingua version as well, I'll run into the problem
                                > again--and worse, because I can't find an Interlingua word for
                                > "mongoose."
                                >

                                For Interlingua, I suggest "manguste", which is the most widespread Romance (and even Slavic) term.
                                "Musco-ratte" looks good, since she owes her name to her musk-glands.

                                Olivier
                              • lingwadeplaneta
                                Konstantin has uploaded a file with translations of the word lily of the valley / May lily into various languages. Let s follow the algorithm and suggest a
                                Message 15 of 25 , Oct 7, 2011
                                  Konstantin has uploaded a file with translations of the word "lily of the valley / May lily" into various languages. Let's follow the algorithm and suggest a word for it.
                                  My investigations gave different translations for Hindi and Arabic.
                                  What we have in the first pool:

                                  English: lily of the valley / May lily
                                  Arabic: konfalria ayyar (konfalria + May)
                                  Hindi: repeats English names / saman (from Persian)
                                  Chinese: ling lan (bell orchid)
                                  Russian: landysh
                                  German: Maigloeckchen (little May bell)
                                  Spanish: muguete / lirio de los valles (lily of the valley)
                                  Portuguese: lirio do vale
                                  Italian: mughetto
                                  French: muguet

                                  Step 1. "If in 5 or more languages there is no special word for this concept but compounds are used instead, then make a compound".
                                  6 languages use compounds, although only 4 of them (English, Chinese, German, and Portuguese) have no other specific word.
                                  Arabic simply copies the Latin name (are these flowers around there at all?). I am not sure if it can be called compounding.
                                  In Hindi the English names seem to be much more used than "saman".
                                  It's pretty much 50-50 situation. I'd say, let's make a compound.

                                  We have 4 lilies of the valley and 4 mentioning of May, but patterns with May differ while the pattern "lily of the valley" is stable, so it wins: valley-lily. Oh, we don't yet have a word for lily, but it's easy: lilia (universal agreement at the step 2). The result: dol-lilia.
                                • Attilio Liotto
                                  Kare amigas, plis pardoni ba me por sey tarde jawaba. Ya me samaji pyu hao esperanto e interlingua kem inglish. Om lingwa yusibil hir, me opini ke es hao adi
                                  Message 16 of 25 , Oct 7, 2011

                                     

                                    Kare amigas, plis pardoni ba me por sey tarde jawaba.

                                    Ya me samaji pyu hao esperanto e interlingua kem inglish.

                                     

                                    Om lingwa yusibil hir, me opini ke es hao adi toshi interlingwa por ke it es samajibil sin lerning, bay 1.000.000.ooo (un-bilion) de jen (oli kel janmi-lingwa es fon-latina.

                                     

                                    Lingwistas kel afte 25 jar de gunsa,  in yar 1951 publikisi interlingwa zwo-te gro hao leksikon basen in sey nichen regula:

                                     

                                    1)   Un parola es adoptate in interlingua, si illo es commun a  alminus tres del linguas anglese, francese, espagnol/portugese o italian.

                                     

                                    Un worda es adopten pa “interligua”, si it es komun a amini tri aus lingwas inglen, fransen, espanen, portugalen, o italen.

                                     

                                    2)   Si le parola es commun in duo de iste linguas, su existentia in germano e/o russo es decisive.

                                     

                                    Si worda es komun in dwa de sey lingwa, suy yenesa in doichen e/o rusen es desidive.

                                     

                                    („-ive“ es hao afixe „produkti > produktive, isoli > isolive, frukti > frutive, akti > aktive, it es aktive forma de “-ibil(e)“.

                                     

                                    3)   Le forma standardisate de un parola debe essere maximalmente conforme al linguas mentionate, isto es apparer como le denominator commun del vocabulos e elementos in le linguas fonte.

                                     

                                    4)   Forma “standardisen” de worda mus bi “maxi-nem” konforme a sursa lingwa e apari-ney komo un worda de sey lingwa.

                                     

                                     

                                    Me opini ke wud bi hao adopti sey regulas toshi en lidepla, bat limiti sey wodas tip pa maxi 33 %. Se fo demokrasi-nem lasi lok a otre lingwas.

                                     

                                     

                                    Kan ba: “Presentation de interlingua”

                                    http://www.youtube.com/watch?v=sp5O0J6-4bk&feature=related

                                     

                                    *

                                    Om “bu” e “no”, segun me, regula es: wen “bu” sufi, bu yusi „no-„:

                                     

                                    bu dumi, bu es fogeti

                                    bu ami, bu es heni

                                    bu heni, bu es ami

                                     

                                    Me bu he dumi a yu bat me bu fogeti yu (“nodumi” es sinsignifa)

                                    Me bu pyu ami yu, bat me haishi lyan yu

                                    Me bu heni bat nopri ela

                                     

                                    Ela bu jivi hao, bu signifi ela no-jivi hao (sey laste expresa es sinsignifa)

                                    Lu bu lai manya, bu signifi lu no-lai manya (sey laste expresa es sinsignifa)

                                    Me mus bu pi, bu signifi me mus no-pi (signifa de sey laste expresa es ga otre)

                                     

                                    *

                                    dol-lilia es muy hao

                                     

                                    *

                                     

                                    Swasti!

                                     

                                    Attilio

                                     

                                    P. S.

                                     Kwo yu pyu pri inu lidepla?
                                    www.facebook.com/groups/lidepla
                                    .
                                    Kion vi pli ŝatas en "lidepla" ?
                                    Kwo yu pyu pri inu lidepla?
                                    .
                                    Si yu haishi bu jawabi a sey kwesta kliki ba uupar linka.

                                    Afte shiki ba may kwesti-tabula e kliki ba kwadra-ki a lefta.

                                     

                                    Shukran!

                                    Attilio


                                    2011/10/7 lingwadeplaneta <lingwadeplaneta@...>
                                     



                                    Konstantin has uploaded a file with translations of the word "lily of the valley / May lily" into various languages. Let's follow the algorithm and suggest a word for it.
                                    My investigations gave different translations for Hindi and Arabic.
                                    What we have in the first pool:

                                    English: lily of the valley / May lily
                                    Arabic: konfalria ayyar (konfalria + May)
                                    Hindi: repeats English names / saman (from Persian)
                                    Chinese: ling lan (bell orchid)
                                    Russian: landysh
                                    German: Maigloeckchen (little May bell)
                                    Spanish: muguete / lirio de los valles (lily of the valley)
                                    Portuguese: lirio do vale
                                    Italian: mughetto
                                    French: muguet

                                    Step 1. "If in 5 or more languages there is no special word for this concept but compounds are used instead, then make a compound".
                                    6 languages use compounds, although only 4 of them (English, Chinese, German, and Portuguese) have no other specific word.
                                    Arabic simply copies the Latin name (are these flowers around there at all?). I am not sure if it can be called compounding.
                                    In Hindi the English names seem to be much more used than "saman".
                                    It's pretty much 50-50 situation. I'd say, let's make a compound.

                                    We have 4 lilies of the valley and 4 mentioning of May, but patterns with May differ while the pattern "lily of the valley" is stable, so it wins: valley-lily. Oh, we don't yet have a word for lily, but it's easy: lilia (universal agreement at the step 2). The result: dol-lilia.


                                  • Konstantin Stolyarov
                                    Privet, Dimi! Shukran por jawaba. Me simplem dukti-te exampla. E me joy ke Ati se pri. A me sembli ke dwashi lingwa es sufi-shem fo algoritma. Apropoo
                                    Message 17 of 25 , Oct 7, 2011
                                      Privet, Dimi!
                                      Shukran por jawaba.
                                      Me simplem dukti-te exampla. E me joy ke Ati se pri. A me sembli ke dwashi lingwa es sufi-shem fo algoritma.

                                      Apropoo "landysh'.
                                      In may garden "landysh" flori in mes-pet, e garden bu loki in dol, e "landysh" es bu lily. Es syao blan glok-ki!
                                    • lingwadeplaneta
                                      ... Ver! Muy hao ke yu skribi se. Dol-lilia es sol proposa. Aproba de proposa es sekwi-she fasa de worda-adopting. Si yu fai objeta kontra dol-lilia, dan
                                      Message 18 of 25 , Oct 8, 2011
                                        --- In lingwadeplaneta@yahoogroups.com, "Konstantin Stolyarov" <gostomysl790@...> wrote:
                                        >
                                        >
                                        >
                                        >
                                        > Privet, Dimi!
                                        > Shukran por jawaba.
                                        > Me simplem dukti-te exampla. E me joy ke Ati se pri. A me sembli ke dwashi lingwa es sufi-shem fo algoritma.
                                        >
                                        > Apropoo "landysh'.
                                        > In may garden "landysh" flori in mes-pet, e garden bu loki in dol, e "landysh" es bu lily. Es syao blan glok-ki!
                                        >

                                        Ver! Muy hao ke yu skribi se. "Dol-lilia" es sol proposa. Aproba de proposa es sekwi-she fasa de worda-adopting. Si yu fai objeta kontra dol-lilia, dan treba proposi koy otre varianta. Es ver ke "mey" e "glok" es muhim fo deskribi sey flor, also oni mog proposi "mey-glok".
                                        Me yus he kreati un nove kwesta in Feisbuk-grupa: "Ob yu wud preferi dol-lilia oda mey-glok?" Yoshi dar ye posiblitaa de adi otre varianta. Partisipi ba!

                                        Dimi
                                      • Stephen Rice
                                        ... Considering that lily of the valley comes from the Old Testament and thus from Hebrew, I would guess that there is in fact an Arabic word for it
                                        Message 19 of 25 , Oct 8, 2011
                                          On 10/7/11, lingwadeplaneta <lingwadeplaneta@...> wrote:

                                          > Konstantin has uploaded a file with translations of the word "lily of the
                                          > valley / May lily" into various languages. Let's follow the algorithm and
                                          > suggest a word for it.

                                          > Arabic simply copies the Latin name (are these flowers around there at
                                          > all?).

                                          Considering that "lily of the valley" comes from the Old Testament and
                                          thus from Hebrew, I would guess that there is in fact an Arabic word
                                          for it somewhere. Let's see... Got it: it's

                                          زنبق الوادي

                                          which is "lily of the valley." (Duh!) I should have known Arabic would
                                          use the same pattern as Hebrew. (But don't tell anybody...)

                                          Steve
                                        • cafaristeir
                                          Sellamat quantims ! In France, our custom is to give lilies-of-the valley on the 1st of May. As I live in front of forests, which I know very well, I wanted to
                                          Message 20 of 25 , Oct 9, 2011
                                            Sellamat quantims !

                                            In France, our custom is to give lilies-of-the valley on the 1st of May.
                                            As I live in front of forests, which I know very well, I wanted to find some for my mother.
                                            To tell the truth, it would be impossible to find such lilies in valleys ! No, the only places where I could find some are shadowy places in the highest places of forests, where very few humans tread... This has really nothing to do with valleys !
                                            Furthermore, because of global warming, spring came earlier and the few lilies I found already began to rot...

                                            Thus, I neither chose "lily of the valley" nor "bell of may" for Sambahsa, I took "convall" from Latin, a word which is found in some European languages (and of course, the Latin scientific name is international). There is a connection of "convall" with "valley/vallis", but this is not obvious nowadays.

                                            Swasti a oli !

                                            Olivier
                                          • Konstantin Stolyarov
                                            Me dumi, ke fo oli titla planta e animal treba yusi latina termin.
                                            Message 21 of 25 , Oct 10, 2011
                                              Me dumi, ke fo oli titla planta e animal treba yusi latina termin.
                                            • Stephen Rice
                                              ... In Israel and thereabouts, the valley plain is where agriculture is usually easiest, the hills being eroded. Hebrew ‘emeq doesn t refer so much to a
                                              Message 22 of 25 , Oct 10, 2011
                                                On 10/9/11, cafaristeir <cafaristeir@...> wrote:

                                                > To tell the truth, it would be impossible to find such lilies in valleys !
                                                > No, the only places where I could find some are shadowy places in the
                                                > highest places of forests, where very few humans tread... This has really
                                                > nothing to do with valleys !

                                                In Israel and thereabouts, the valley plain is where agriculture is
                                                usually easiest, the hills being eroded. Hebrew ‘emeq doesn't refer so
                                                much to a ravine but to a plain from which mountains rise--though the
                                                "mountains" there are scarcely worthy of the name.

                                                > Thus, I neither chose "lily of the valley" nor "bell of may" for Sambahsa, I
                                                > took "convall" from Latin, a word which is found in some European languages
                                                > (and of course, the Latin scientific name is international). There is a
                                                > connection of "convall" with "valley/vallis", but this is not obvious
                                                > nowadays.

                                                In cases where there is no clear international form, this may be the
                                                best way, as Konstantin has also said. The problem in some cases is
                                                that the scientific name is clear only if the complete binomial form
                                                is used, which is awkward. (And sometimes the name changes, though
                                                regular English retains brontosaur rather than switching to
                                                apatosaur.)

                                                Steve
                                              • lingwadeplaneta
                                                ... In biologia, mogbi. Bat in pinchan jiva, ob inplas doga yu ve shwo Canis lupus familiaris ? To wud bi drole.
                                                Message 23 of 25 , Oct 10, 2011
                                                  --- In lingwadeplaneta@yahoogroups.com, "Konstantin Stolyarov" <gostomysl790@...> wrote:
                                                  >
                                                  > Me dumi, ke fo oli titla planta e animal treba yusi latina termin.
                                                  >

                                                  In biologia, mogbi. Bat in pinchan jiva, ob inplas "doga" yu ve shwo "Canis lupus familiaris"? To wud bi drole.
                                                • Konstantin Stolyarov
                                                  ... Es ya tro! Troisa. Ïåðåáîð!
                                                  Message 24 of 25 , Oct 14, 2011
                                                    >"Canis lupus familiaris"?

                                                    Es ya tro! Troisa. Ïåðåáîð!
                                                  • lingwadeplaneta
                                                    ... I am thinking about the notion of an open system , whether it s correct to use it to demonstrate LdP s difference from all the other projects. In physics,
                                                    Message 25 of 25 , Oct 27, 2011
                                                      --- In lingwadeplaneta@yahoogroups.com, Stephen Rice <ansrith@...> wrote:
                                                      >
                                                      > This demonstrates the difference (one of them, anyway) between
                                                      > Sambahsa and LdP: Sambahsa has etymological concerns that I don't
                                                      > think matter to LdP, while LdP has concerns with current forms and
                                                      > usage that are largely foreign to Sambahsa. Sambahsa asks, "Where did
                                                      > it come from?" LdP asks, "What's out there right now that I can use?"
                                                      >
                                                      > Steve
                                                      >

                                                      I am thinking about the notion of "an open system", whether it's correct to use it to demonstrate LdP's difference from all the other projects. In physics, biology, etc. an open system is a system where matter or energy can flow into and/or out of the system.
                                                      The very idea of LdP is to suggest a variant of integration of the most spoken languages. So it proceeds from the current reality in its fullness, it's open to it.

                                                      Another thing that shows openness is that LdP doesn't change imported words too much, permitting variations for parts of speech, ex. both "gao" and "normale" can be adjectives. This feature has drawbacks: 1) an impression of "salad" might occur; 2) derivation becomes polytypic. These, however, are not necessarily drawbacks.

                                                      If it's correct to declare that LdP is an open system exchanging with the environment, then we can go further and set out some principles of its further development...

                                                      Dmitry
                                                    Your message has been successfully submitted and would be delivered to recipients shortly.