Proper Lojban Orthography: Difference between revisions

From Lojban
Jump to navigation Jump to search
mNo edit summary
 
mNo edit summary
Line 1: Line 1:


You expect to be able to use ''pu'' in place of ''cu'' in front of a selbri to close the front sumti - right?  Wrong - if it's an abstraction!
Lojban is a logical language and, therefore, may be presumed to be reasonably scientific. Why then the horribly unscientific alphabet in which it is written.  The symbols used to record the language -- the spoken language that is -- ought to be clearly related to the sounds used, not the product of irrelevant historical accidents (ultimately what sound started various Egyptian words and how the objects referred to by those words were drawn -- the Just-So version makes more sense).


''lenu mi klama pu vajni'' is a sumti with parse:
The sounds of Lojban divide fairly accurately into vowels and consonants (two vowels also have cononantal uses, four consonants vocalic).  For vowels, the only significant distinguishing features are height of tongue at the narrowest point and whether that narrowest point is front, back, or central.  Six of the nine positions are filled: high and mid front and back, low (also central) and central non-low.  There may also be heard a different central vowel, but it is not significant and so needs no symbol. The two high vowels also appear as on- and off-glides in vowel clusters.


''({le < [[jbocre: nu (mi {klama < [jbocre: pu KU|pu KU]] VAU>}) KEI] vajni> KU} VAU)''
The consonants divide on three basic factors: fortis v. lenis (whether voiceless-voiced, or cough-sigh, or a combination, or something else altogether, stop v. continuant, and point of articulation: lips, teeth, hard palate, back of throat.  Not all combinations are recognized, and there are also the added factors: nasality, lambdacism and rhotacism, that make unique contributions outside some or all these categories. What we have are, in fact, fortis and lenis stops and continuants at lips, teeth and somewhere back of the teeth/alveolum -- call it the hard palate, against future reference.  There are then nasals (which are also technically stops, though the air flows is not interrupted, merely diverted, and which are fortis)at lips and teeth (well, mouth roof, since the "teeth" one has palatal and further back allophones).  And a rho and a lambda, both also voiced and both more or less teeth (though the rho may go back into the palate some)and both continuants (though the rho may be a trill or even a flap).  These four non-systemic additions may also be syllabic.  Finally, there are a back fortis fricative, clearly behind the hard palate (conceptually), and what is technically a lenis vowel (its character depends on the vowels it comes between), which is treated as a continuant of indefinite -- but unique -- articulation. (There is also a glottal stop -- back fortis stop -- which is significant but always predictable within the full language -- though not merely phonologically).


I don't. Sounds like someone mislearned that tenses mark the start of selbri.
The task then is to find a symbolism that reflects the definition of each sound and yet keeps the various sounds visually distinct.  The mythic Hangul idea of portraying the articulation immediately suggests itself, but -- in its pure form, at least -- fails the distinctness test: in vowels the difference between high and mid (and possibly between front and back) is not always easy to spot. Further, vowels are probably too important to be givens such minimalist representations (on the other hand, something brief is desirable, since vowels occur much more frequently than consonants).


To be fair, the book says (10.1, p. 216) ''The placement of a tense construct within a Lojban bridi is easy: right before the selbri. It goes immediately after the "cu", '''and can in fact always replace the "cu"'''.'' Of course, it then goes on to say ''(although in very complex sentences the rules for eliding terminators may be changed as a result).''
Similarly, the front-middle-back of consonants portayed by a bulge at the appropriate place on a line, the closed v. slightly open marked by the height of a hump, the fortis-lenis marked by a tag at the bottom back, and other iconic factors (just what to do with rho and lambda is unclear)all tend to require careful writing and a keen sense of what to look for in the reader. Still, this is a place to start.
 
So, the vowel sign is a pair of parallel vertical lines, running from capital space (or at least cross-bar) to below the writing line.  The differentiation is then by a loop in the appropriate place: "i" (high front) looks like a paragraph sign with its loop whited in and its stem going below the line, "u" (high back) looks like capital "P" with its stem dobled and going below the line.  "e" and "o" (mid front and back) look like "d or "q" with the stem extended and "b" or "p" extended.  "a" (low, central) is a loop between the lines and below the writing line, "y" (central, non-low) is a loop between the lines at the writing level.  Only vowels go below the line This makes them easy to spot, too).
 
For consonants, rather than the spot along the lines, each position might get an iconic main component a rounded "e" (backward 3) for labials, something sharp (maybe an "x" for meeting teeth) for dentals and something opening the throat (a backward "c" say) for velars.  (All of these suggestions need modification at least for handwriting.)  The basic symbol is a stop, the continuant is differentiated by a horizontal line through the center of the base. Lenis is marked by a vertical stroke at the back (vocal cors in operation) and nasals by a tilde over the top -- something above is open. (These last two are really problematic.)  The rho, lambda, and chi clearly do not fit this iconic pattern, but need symbols all their own.  Lambda might be an inverted "u" to show the lateral tongue drop that is its characteristic feature.  Rho's characteristic feature is a buzz (which it shares, alas, with spirants at least and which is missing in at least one major variant); perhaps something like a "w," with the dental points and the iterations would be close enough.  The voiceless vowel is just a vowel with its descender cut off and without a loop, a squared off "u." And chi?  No bright ideas at all.
 
Have fun correcting this.
 
pc
 
*I actually looked at some of these last night.  The vowels aren't too bad, but the regular consonants are way too busy to be actually used.  I think a different set of icons is needed -- or maybe skipping iconicity altogether. pc
*Compare Alexander Melville Bell's "Visible Speech", which is an entire phonetic alphabet based on a similar notion of iconicity.  In fact, the vowels are quite similar to what you get here.  The consonants are all based off the same shape (a c-like loop) which is turned into different orientations to show lips, velum, tongue-tip, and tongue-body (palatal sounds, basically.  Phonetics in Bell's day wasn't what it is today).  Also Herman Miller's Lhoerr alphabet, which is also an attempt at an iconic phonetic alphabet.  Either of these actually could already be used with little or no modification (just in broad transcription).
 
**Both Bell and Miller give systems for "all" sounds, which are, thus, more complex than needed for a single language.  Bell's system pretty clearly fails the easy distinctness test -- the differentiations are often minute and the overall looks of many items are the same (as the comment notes: "based off the same shape").  Miller's is a bit more varied, but with harder distinctions to learn (the being less iconic -- to my eye). But there are true type fonts for both, and unicode.
*Oh, and it was Phoenecian, not Egyptian.
 
**Arguable.  Each older version found is from farther south and closer to some trasnitional hieroglyphic/demotic form. pc

Revision as of 17:09, 4 November 2013

Lojban is a logical language and, therefore, may be presumed to be reasonably scientific. Why then the horribly unscientific alphabet in which it is written. The symbols used to record the language -- the spoken language that is -- ought to be clearly related to the sounds used, not the product of irrelevant historical accidents (ultimately what sound started various Egyptian words and how the objects referred to by those words were drawn -- the Just-So version makes more sense).

The sounds of Lojban divide fairly accurately into vowels and consonants (two vowels also have cononantal uses, four consonants vocalic). For vowels, the only significant distinguishing features are height of tongue at the narrowest point and whether that narrowest point is front, back, or central. Six of the nine positions are filled: high and mid front and back, low (also central) and central non-low. There may also be heard a different central vowel, but it is not significant and so needs no symbol. The two high vowels also appear as on- and off-glides in vowel clusters.

The consonants divide on three basic factors: fortis v. lenis (whether voiceless-voiced, or cough-sigh, or a combination, or something else altogether, stop v. continuant, and point of articulation: lips, teeth, hard palate, back of throat. Not all combinations are recognized, and there are also the added factors: nasality, lambdacism and rhotacism, that make unique contributions outside some or all these categories. What we have are, in fact, fortis and lenis stops and continuants at lips, teeth and somewhere back of the teeth/alveolum -- call it the hard palate, against future reference. There are then nasals (which are also technically stops, though the air flows is not interrupted, merely diverted, and which are fortis)at lips and teeth (well, mouth roof, since the "teeth" one has palatal and further back allophones). And a rho and a lambda, both also voiced and both more or less teeth (though the rho may go back into the palate some)and both continuants (though the rho may be a trill or even a flap). These four non-systemic additions may also be syllabic. Finally, there are a back fortis fricative, clearly behind the hard palate (conceptually), and what is technically a lenis vowel (its character depends on the vowels it comes between), which is treated as a continuant of indefinite -- but unique -- articulation. (There is also a glottal stop -- back fortis stop -- which is significant but always predictable within the full language -- though not merely phonologically).

The task then is to find a symbolism that reflects the definition of each sound and yet keeps the various sounds visually distinct. The mythic Hangul idea of portraying the articulation immediately suggests itself, but -- in its pure form, at least -- fails the distinctness test: in vowels the difference between high and mid (and possibly between front and back) is not always easy to spot. Further, vowels are probably too important to be givens such minimalist representations (on the other hand, something brief is desirable, since vowels occur much more frequently than consonants).

Similarly, the front-middle-back of consonants portayed by a bulge at the appropriate place on a line, the closed v. slightly open marked by the height of a hump, the fortis-lenis marked by a tag at the bottom back, and other iconic factors (just what to do with rho and lambda is unclear)all tend to require careful writing and a keen sense of what to look for in the reader. Still, this is a place to start.

So, the vowel sign is a pair of parallel vertical lines, running from capital space (or at least cross-bar) to below the writing line. The differentiation is then by a loop in the appropriate place: "i" (high front) looks like a paragraph sign with its loop whited in and its stem going below the line, "u" (high back) looks like capital "P" with its stem dobled and going below the line. "e" and "o" (mid front and back) look like "d or "q" with the stem extended and "b" or "p" extended. "a" (low, central) is a loop between the lines and below the writing line, "y" (central, non-low) is a loop between the lines at the writing level. Only vowels go below the line This makes them easy to spot, too).

For consonants, rather than the spot along the lines, each position might get an iconic main component a rounded "e" (backward 3) for labials, something sharp (maybe an "x" for meeting teeth) for dentals and something opening the throat (a backward "c" say) for velars. (All of these suggestions need modification at least for handwriting.) The basic symbol is a stop, the continuant is differentiated by a horizontal line through the center of the base. Lenis is marked by a vertical stroke at the back (vocal cors in operation) and nasals by a tilde over the top -- something above is open. (These last two are really problematic.) The rho, lambda, and chi clearly do not fit this iconic pattern, but need symbols all their own. Lambda might be an inverted "u" to show the lateral tongue drop that is its characteristic feature. Rho's characteristic feature is a buzz (which it shares, alas, with spirants at least and which is missing in at least one major variant); perhaps something like a "w," with the dental points and the iterations would be close enough. The voiceless vowel is just a vowel with its descender cut off and without a loop, a squared off "u." And chi? No bright ideas at all.

Have fun correcting this.

pc

  • I actually looked at some of these last night. The vowels aren't too bad, but the regular consonants are way too busy to be actually used. I think a different set of icons is needed -- or maybe skipping iconicity altogether. pc
  • Compare Alexander Melville Bell's "Visible Speech", which is an entire phonetic alphabet based on a similar notion of iconicity. In fact, the vowels are quite similar to what you get here. The consonants are all based off the same shape (a c-like loop) which is turned into different orientations to show lips, velum, tongue-tip, and tongue-body (palatal sounds, basically. Phonetics in Bell's day wasn't what it is today). Also Herman Miller's Lhoerr alphabet, which is also an attempt at an iconic phonetic alphabet. Either of these actually could already be used with little or no modification (just in broad transcription).
    • Both Bell and Miller give systems for "all" sounds, which are, thus, more complex than needed for a single language. Bell's system pretty clearly fails the easy distinctness test -- the differentiations are often minute and the overall looks of many items are the same (as the comment notes: "based off the same shape"). Miller's is a bit more varied, but with harder distinctions to learn (the being less iconic -- to my eye). But there are true type fonts for both, and unicode.
  • Oh, and it was Phoenecian, not Egyptian.
    • Arguable. Each older version found is from farther south and closer to some trasnitional hieroglyphic/demotic form. pc