old Norse verse forms
[[from [ju'i lobypli|ju'i lobypli]] number 10: November-December 1989]
Skaldic Poetry and Lojban, by Athelstan Berfoetskald
English speakers may be unaware of the wealth of possible forms of poetry that exist in the world's literature. Most English poetry is dominated by a single form, 'end-rhyme', in which the final word of each line rhymes with one or more other lines; the exact lines in a stanza which are paired or grouped in rhyme differ according to the specific form, giving us such end-rhyme forms as doggerel, limericks, and sonnets. End-rhyme is often elaborated by requiring particular matching stress patterns in each rhyming pair or group of lines. End-rhyme has proven less interesting for modern poets, who abandon rhyme, stress pattern, or both to achieve a density or abstractness of expression unconstrained by form.
There are many other possible forms, including forms that ignore rhyme and which count only stressed syllables (Germanic long-line, the form of "Beowulf", is an example). In languages that do not have stress as a distinguishing characteristic of tone, the syllable count in each line is important. Japanese 'haiku' is probably the most well-known of these forms; it requires no rhyme or meter; instead it has exactly three lines with 17 syllables in a 5/7/5 pattern, with the added requirement that it demonstrate a particular 'balance' in harmony/disharmony of subject, sound, and meter.
Of poetry's many forms, English end-rhyme does not seem especially suited for Lojban. This is but a guess, based on the limited number of gismu that rhyme and the lujvo architecture that limits the variety of ideas that will end in matching sounds. Many other poetic forms may be more apropos to Lojban's sound and word structures. In this article, I shall concentrate on the Norse alliterative poetry of the skalds, and in particular the 'drottkvaet', or 'court metre'.
First some definitions:
The common unit of skaldic poetry is the strophe, or stanza, which consists of four 'lines', or eight 'half-lines'. Unlike much of English poetry, these lines are usually not grammatically intact sentences or phrases--sentence structure is independent of line structure. In fact, such poetry often has several independent 'sentence' thoughts going at once, with the harmony of sound helping the listener piece the structures together.
Each line consists of two half-lines. There is an odd half-line with exactly six syllables, three of which are accented. Each odd half-line is followed by an even half-line, also with exactly six syllables, three accented. Unlike English end-rhyme forms, all alliteration and rhyme takes place in the accented syllables, or beats. I will for this discussion use the most common stress pattern, 'trochaic'. That is, each beat is followed by an unstressed syllable.
Two syllables 'alliterate' if their initial soulnds are alike: the words "fight", "far" and "phantasm" alliterate. Any two syllables that begin with vowels also alliterate. In each line of skaldic poetry, exactly two beats in the odd half-line and the first beat in the even half-line alliterate.
A 'full rhyme' is one that shares the same vowel and final consonant sounds in one syllable: "rat"/"cat", "rut"/"cut" and "rough"/"cuff" are full rhymes. A 'half rhyme' differs from a full rhyme in that the vowel sound must be different: "rat"/"cut", "rut"/"cat" and "rough"/"cough" are half rhymes. In each line of skaldic poetry, a half rhyme appears in exactly two of the three beats in the odd half-line, including the third beat. In the even half-line, a full rhyme appears in exactly two beats, also including the third, but the pattern need not be the same as in the odd half-line.
This is an unfamiliar (and complicated) form to many, so I include here an English language example of a skaldic strophe. (It's not really SUPPOSED to have great meaning in it - this is only an example.) Please note that the last syllables of each half-line are unstressed and so neither rhyme nor alliterate. The alliterations appear in UPPER case. Vowel alliterations are marked with a period.
Sing with me a Song of
Soaring birds and words that
tell of Hawks; the Hall of
Heaven flows with prose of
.owls and .airborne furless
.awful bats with hats on.
Night and day are Not the
Nicest times for rhymes' sound.
Strophes were informally composed and recited singly or in pairs, but more formal occasions demanded the 'long lay', or "drapa" form. This consists of twenty or more strophes, partitioned into three or more parts by a chorus of one, two or four lines which ended a strophe or stood alone.
Skaldic poetry often used a figurative metaphor known as a "kenning", whose canonical English form is 'x of the y'. For example, "steed of the waves" is a kenning for 'ship', and "whales' road" stands for 'ocean'. In some drapas, kennings are all-pervasive and no thing is directly named.
I think that Lojban is well suited phonologically and grammatically to the use of skaldic form. Lojban gismu are trochair in meter, and Lojban's penultimate word-stress system is conducive to the regular patterns required. The large number of independently elidable cmavo words allows one to alter the rhythm of the utterance to suit the form. Moreover, the morphology allows us to vary the length and form of a lujvo, allowing the poet to choose the word form that best suits the meter without changing the meaning.
We have the ability in Lojban to mark a word as figurative in its use, and so we may use kennings at will, but we also have the option of redefining the grammar for the poem to treat all two word metaphors as kennings, and may thus accommodate the all-kenning form as well.
I shall attempt such poetry in the near future as my schedule permits, but I put it forth to all aspiring Lojban poets to explore this and other alliterative poetry forms. In addition to formal skaldic poetry, there may be other patterns, more or less structured, that will project the full color of Lojban's expressive nature. A new language calls for new ideas, and for the re-examination of old ones.
Michael Helsem on Lojban Poetry Forms
... Lojban poetry. I am eager to see Athelstan's proposal, being a
skaldic aficionado myself (albeit non-Norskophone), but I have grave
doubts already. (I do admire the grandeur of its absurdity --akin to
one of the projects to write uantitative verse in English!) Lojban is
a non-inflected language with numerous non-elidable function words (at
least in its grammatically correct form), whereas such rigorous modes of
versification as Dro"ttkvaett (or haiku for that matter) require either a
language that is inherently terse (i.e. inflected), or else a polyvalent
grammar like the anarchical tradition of English poetry gives access to
(--my English skaldic verse must seem totally twisted to anyone who
doesn't know those conventions). (Or a tradition of broken utterance,
which is haiku's, as any literal rendering will show.) Athelstan is absolutely correct in perceiving Lojban's consonantal richness as the
salient poetic character. But that applies only to its core vocabulary.
Include all the function words as they naturally occur, and Lojban has a
greater resemblance to a pizzicato language like Hawaiian or Japanese,
than to the surflike pounding of Germanic lines in Alliterative-
Accentual. Another thing: it takes many near-synonyms to be able to
say what you want and have it alliterate. I don't foresee this ever
being true of Lojban unless it swamps its carefully-distinguishable
word-hoard with a ton of redundant mports...
I imagine Lojban poetry will eventually create brand new forms out of
its own vast uniqueness which is hardly perceptible yet to those who
think in English (or other natural language) first. I've got a few
noodly intimations I'm reluctant to pontificate upon before I can
substantiate them with practical excellence, but I can say right now
that the "abstract ideas of modern poetry" find a greatly expanded range
in this language with few concrete noun-distinctions (names of birds,
flowers, and feelings), but capable of phrases such as "mu'e laxyzva" --
"The achievement of balance-presence" (which may mean just "poise", or
imply a whole world of Taoist philosophy). --so naturally and yet so
I guess a lot of people still identify "poetry" with rhyme+meter,
although almost no poets even attempt them anymore outside of popular
music. For Lojban to be shoehorned into existing English poetic forms
is not quite the contortionist feat that skaldic verse would be, but my
ear-lier point still applies. In addition, Lojban has few natural
rhymes, and hardly any 'good' ones at all. 'Love' (prami) rhymes only
with 'computer' (skami) and 'acid' (slami), while 'desire' (djica)
rhymes with 'differ' (frica) and --too closely --'deceive' (tcica), for
example. Compare fire/desire or Schmerz/Herz et al... This is not as
bad as Esperanto, which rhymes Vulva and Gunpowder; still, artificial
languages seem uniformly unfortunate in this respect. More
sophisticated devices, like slant-rhyme, assonance, and particularly the
combination of an initial and a medial consonance (muzga zgana; lujvo
jvinu), I suggest, are well worth trying in an irregular way. But, to
use Pound's nomenclature, Lojban excels in aptness for meaning-subtlety
(Logopoeia) rather than phonic luxuriance (Melapoeia) or pictorial
Lojban is an athletic language; it stretches your mind with challenging
concept-divergences and novel connections. Making lujvo will probably
be its first and most popular word game always; just as Crosswords
exploit the hodgepodge of English megavocabulary, lujvo involve the very
quiddity of Lojban. I predict that it will eventually become a
much-used source of return-borrowings for many natural languages,
somewhat as classical Greek has continued to provide new loans for
One last thing. I think there might be a definite limerick potential.
Anapests seem to come easier than iambs; and what rhymes exist, have a
quality that reminds me of a triple rhyme in English --as can be found
in the dustier corners of a rhyming dictionary: mentor-centaur, gurgle-
burgle, muscular-crepuscular. This is a Special Effect, like fireworks,
not the constant recurrence of a mild harmony (like rhyme in Italian or
Athelstan responds with two notes:
1. Germanic languages were not as terse as they are now, precisely
because they were more highly inflected. Latin and Greek were very
highly inflected, and they are anything but terse.
2. There are usually several directions from which to approach a
concept: function, appearance, resemblance, effects, etc. Many gismu
are also expressible as conversions of other gismu. Combined with the
further tool of kenning description, Lojban promises as rich and diverse
a phonetic realm as is available in any language, past or present.
Bob adds some other points:
Lojban has more rhyme capability than Michael indicates. It is true
that there isn't much rhyming capability among gismu, but among lujvo
there should be considerably more rhymes. After all, all Lojban brivla
end in a vowel or diphthong, and there are only so many possible
It is true, though, that many of the rhyming syllables would be based on
sharing the same rafsi, leading to a lack of variety of pure rhyming
forms. But moving beyond pure rhyme, into Athelstan's half-rhyme and
alliterative schemes should allow Lojban to show great richness. After
all, Lojban has a much more restricted set of permitted word endings,
and a smaller set of phonemes, than does English. Since there are, at
least theoretically, more possible Lojban words than English words, this
suggests a higher density of rhymes and near-rhymes than English has.
Time will tell if this is a real feature of the language, and not a
Lojban has many non-elidable structure words, but it has many elidable
ones as well. It can even be said that Lojban's optional tense system
is an elidable form of inflection. "pu klama" can be said to be an
inflected form of "klama"; however, Lojban inflections are both optional
and completely regular. It is unclear why this would be a disadvantage
in poetry as compared with inflected languages, as Michael implies.
With regard to sound and rhythm, I cannot say whether Lojban is more
like Japanese than like German. However, since Lojban is a language
with stress-oriented pronunciation, one would suspect that rhythm, (as
opposed to rhyme) would follow the strong-stressed patterns of Germanic
lan-guages, as opposed to the syllabic rhythms that I believe are
typical of most oriental languages (as well as French). The consonant
clustering, is of course reminiscent of Slavic languages; perhaps we
should examine Slavic poetry forms to determine some additional
possibilities for Lo-jban.
I am not well-versed in poetry, but I am going to be getting a bit of
experience in devising Lojban forms: the Arabian Nights tales that I am
working on are scattered with embedded verse, as well as a frequently
alliterative and simply-rhyming prose. I'll be trying to capture the
Arabic richness of pattern in Lojban, just as Burton tried to capture it
into English. As an example, which helps belie Michael's perception of
a lack of synonyms, Burton's English translation of a vocative (near the
beginning of the first story) reads "O King of the time and Caliph of
the tide ..."; the "tide" here obviously means "season", as in
Lojban is fully capable of expressing this parallel metaphor of nearly
identical meanings with two totally unrelated expressions, just as
English does: "doi nolrai co turni be le temci be'o je catni be le
cabna" (O superlative-noble type-of [governor of the time-interval and
authority over the present]), and I even captured the alliteration
(which is highly valued in Arabic poetry and a feature that Burton tries
to em-ulate) at least as well as the English does. (I won't promise
such successful Arabic poeticity in the rest of my translation.)
In the final analysis, both Michael and Athelstan may be right. We have
to try new ideas and see what works for Lojban. If we can give rebirth
to old forms that have been lost to English, great. If people choose to
develop a new cultural form of poetry, this may even be evi-dence of a
new Sapir-Whorf effect of a type never consider by Jim Brown and others.