Some interesting features

In the following, we shall use the following notations:

  • C and V stand for Lojban consonants and vowels; the usual CV, CV'V, VV (and so on…) notations keep the same meaning;
  • l / t / d and o stand for full-letter consonants and vowel tengwar , resp., as defined previously, where t stands for any voiceless consonant, d for any voiced consonant, and l for any consonant;
  • L / T / D and O stand for consonant and vowel tengwar, resp., over which a vowel tehta is applied, with the exception of CY for which [tdl]y will be used instead.

Additionnally, please remember that most of what follows is about the last mode defined in the previous part, in which tehtar and full-letter vowels are mixed. The two others modes are far less interesting, and I didn’t (yet!) take much time to study them.

Lojban morphology rules applied to our Tengwar mode

In this chapter I shall be somewhat following the order of Chapter 4 of the Lojban Grammar book.

The groups of Lojban letters

Lojban construct V VV V'V C CV C/C CC CVV CV'V
Tengwar construct o O o'o l L tt, dd tt, dd [1] lO L'o
[1]the set of permissible initial consonant pairs is described fully in Chapter 4 of the Lojban Grammar.

The first point here is that the Tengwar system makes it very easy to separate voiced from unvoiced consonants, because the voicing of a consonant is strictly indicated by a second bow in the tengwa. Therefore, the rule stating that one cannot use two adjacent consonants of different voicing can be easily rephrased as “two adjacent consonant tengwar cannot have the same number of bows”.

Lojban’s cmavo

The Tengwar forms of Lojban cmavo are .o, .O, L and L'o.

Lojban’s brivla

All Lojban brivla must end with a vowel, which gives, when using Tengwar, a mandatory L or L'o ending for every brivla.

They also contain a consonant pair in the first 5 Latin letters, where “y” and the apostrophe are not counted as letters. In other words, this gives the only following possibilities as beginnings for brivla: CVVCC, CV'VCC, CVCCV (and CVC/CCV), CCVCC, CCVCV (CCVV not being possible). In the Tengwar notation, this gives lOll (or lOlL), L'oll (or L'olL), LlL (and LllL), lLll (or lLlL), lLL. Hence the easy rule: “there must be an unaccented above (the underdot doesn’t count) tengwa within the first three tengwar which make up the brivla”.

Finally, all brivla must have more than one syllabe, therefore must contain at least three consonant tengwar and two vowels (for the above rule to be also respected)

Lojban’s gismu

Gismu must be of the form CVC/CV or CCVCV and cannot contain the apostrophe (thus cannot contain VV nor V'V). In Tengwar terms, this means that gismu have exactly three consonant tengwar and two vowel tehtar, with one tehta set over the final consonant.

Lojban’s rafsi and lujvo

Short rafsi are of the form CVC, CCV or CVV. Their Tengwar forms are therefore either Ll, lL, lO or L'o. For four-latin-letter rafsi, it gives Lll or lLl. (Mind the restrictions on ll and lL, about the voicedness of consonants.)

When grouping rafsi to form lujvo, one must take care of including a consonant pair in the first three consonant tengwar, and must put a dot below the first tengwa of a forbidden consonant pair (in Latin terms, insert a “y” between the two consonants).

What is interesting about this last point is that the discreetness of the tehta for “y” makes it very easy to break up lujvo into rafsi; it suffices to follow the following rule, after stripping all underdot (“y”) tehtar.

  • Break after each pair of tengwar (either vowel or consonant) which contains only one tehta.
  • Take care not to leave standalone or pairs of tengwar with tehtar over them.
  • If two ajacent tengwar both have a covering tehta, try to isolate a three-tengwar rafsi form instead.
  • In case of a standalone “r” or “n” (it would then be an hyphenation letter), leave it away.

The justification to this rule is not straightforward; however, if you think about every combination of two rafsi, using the Tengwar notations introduced above, it just reflects the official Lojban rule.

To illustrate this rule, let’s take some random examples (taken randomly from the lujvo list):

LUJVO TENGWAR NOTATION BREAKS UP AS
zvastejbu lLlLlL lL+lL+lL (zva+ste+jbu)
xalbarja LlLlL Ll+LlL (lL+Ll+L not possible) (xal+barja)
nancyprali LlllLL (LllyllL) Lll+lLL (nanc+prali)
gacytisna LlLlL (LlyLlL) Ll+LlL (gac+tisna)
cavysezlu’i LlLlL'o Ll+Ll+L'o (cav+sez+lu’i)
daurda’a lOlL'o lOl+L'o, then lO+r+L'o (dau+da’a)
gairgau lOllO lOl+lO, then lO+r+lO (gai+gau)
fi’ortutra L'olLlL L'ol+LlL, then L'o+r+LlL (fi’o+tutra) (BAD LUJVO)
cacusku LLlL L+L+lL? L+Ll+L? can’t break up with the above rule
fuzpida LlLL Ll+L+L? L+lLL? LlL+L? can’t break up with the above rule

I strongly believe that no badly formed lujvo can be broken up at all when using the above rule, although I didn’t demonstrate it yet.

Some notes on typography

What is interesting with our system of vowels is that it eases a lot the process of separation of syllabes. So much that even to a certain extent, it makes the comma almost useless.

Indeed, where is the comma actually used? It is mandatory in names to indicate a non-standard syllabic break (to turn an otherwise diphtong into two distinct vowels, for example), or can serve as a verbose indication in other words. However, it quite generally remains only used for breaks between vowels. Therefore, as our Tengwar mode states that only diphtongs should be written as a full-letter vowel tengwa with a tehta over it, and that other forms of vowels should be either written as a tehta (over a preceding consonant tengwa) or a single vowel tengwa (in case the vowel is standalone), breaking up a diphtong into two distinct vowels is just done by writing the two parts as full-letter tengwar, which is necessary to have the two sounds separated, but also sufficient the sense that the even presence of two adjacent vowels written in full indicates syllabic separation. No extra comma sign is thus needed in this case.

Of course, in the case where the two vowels to be separated follow a consonant, when the separation occurs the first one should be transformed into a tehta over that consonant.

Conclusion

Our Tengwar mode for Lojban is now defined, and ready for use. However, several features of the Tengwar system have not been used, as I didn’t figure out how they could be useful. Therefore, while the preceding baseline is a good start for a sensible use of the Tengwar system for Lojban, further additions must be expected, as a regular use of Fëanorian letters is likely to induce some more practical considerations.

As it was initially announced, you should have noticed through this page how Lojban’s morphology is superceded in the very structure of the Tengwar system. It is not a coincidence, though, because most of Lojban’s morphology follow the same work direction than the one followed by Tolkien. However, it is important to notice how “logical” these rules are, and how sensible it is to include them within the writing system itself.

See also: