Comments on Solresol

This is probably petty, but before getting to the more serious matters, I would like to mention a pet peeve about some conlangs: only a twit would give a project a generic name. Solresol means language; Babm means international language; and Lojban means logical language. (Loglan, meanwhile, is not generic within that system: Loglan is a name formed regularly from logla [is a part/instance of Loglan].) Loglan is in fact one of the only major systems to have an actual name rather than a sobriquet. The list of offenders is long: Glosa, Interglossa, Interlingua (either one), etc. Even Occidental, Frater, and (for that matter) Esperanto use normal words for names. (Though the choice was not Zamenhof's, Lingvo Internacia is hardly an improvement.) If you ever try creating a language, PLEASE use some imagination in naming it.

End of soapbox; on to the serious commentary.

It is easy to find problems and flaws in Solresol. Couturat and Leau were mystified by its early popularity, and presented a very dark picture of its workings--a little too dark, perhaps. Some of the flaws they cite are exaggerations, and even most of the real flaws have if not an equal, at least an opposite, positive. Consider the following cases:

Overdependence on French. Various features of the grammar and lexicon are based slavishly on French: the lack of the numbers 70 and 90 (soixante-dix [sixty-ten] and quatre-vingt dix [four-score ten]), the tense system, and the apparent range of meanings for some words all derive from French.
On the other hand, many features are far removed from French thought: though the tense system indeed uses the convoluted French system of tenses and moods as a reference, it is much simpler. Just look at how many different forms are subsumed under each of the particles. For that matter, the use of particles for tense and mood is quite revolutionary for a time when agglutination and inflection ruled: try comparing Solresol with Volapük sometime!
The syntax is also a vast simplification over French: adjectives (as opposed to determiners) invariably follow their head, the partitive is gone, the expletive ne likewise.

Difficulty in resolution. Gajewski himself notes several times that speakers must pause between words to keep them from running together. This is an awkward feature at best: who wants to keep pausing between words? Yet without some method of separating words, resolution is nearly impossible. It would perhaps be sufficient to isolate the one- and two-syllable words, assuming that the three- and four-syllable ones would then sort themselves out. This might not be all that hard to do, either, as the shorter words tend to be loners anyway (except for the determiners): they are mostly pronouns, adverbs, and the like, easily set off by a slight pause.

Arbitrariness. Couturat and Leau note that the derivational system is not always followed, giving a case where fsolso (basic meaning: boat) supposedly takes on different shades of meaning as a noun (referring to different types of boats) rather than changing part of speech (38). I find it easier to believe that they misread the entry in Sudre's dictionary than to think that he would have utterly reversed his usual approach to semantics and the lexicon. Then they observe that “la classification des idées correspondant aux combinaisions successives de notes n'est pas plus régulière, et est faite sans aucun principe logique: elles sont rangées dans un ordre à peu près arbitraire, et en tout cas absolument empirique” (the classification of ideas corresponding to successive combinations of notes is no more regular [than the derivational system] and does not follow any logical principle: the ideas are arranged in an order that is almost arbitrary and in any case absolutely unpredictable p. 38).
Now it is true that some of the classifications are a bit hard to follow, and they certainly have little in common with Bishop Wilkin's Real Character. This is absolutely crucial to understanding the nature of Solresol, and one reason it should be considered a classic conlang. Couturat and Leau utterly miss the point. Note what they say shortly before attacking Solresol's arbitrariness: “En somme, le Solrésol présente, à un degré suprème, tous les défauts pratiques des langues philosophiques, sans en avoir les avantages théoriques. En effet, la logique est la moindre qualité de ce système.” (To sum up, Solresol presents in the highest degree all the practical flaws of philosophical languages, without having any of their theoretical advantages. Indeed, logic is the least of this system's qualities. p. 37) But that is precisely the point: unlike other a priori schemes, Solresol is not truly a philosophical language. The other systems attempted to produce a thorough classification of ideas, a philosophical and semantic taxonomy. Solresol merely uses classification as a mnemonic device. The other systems usually boasted a large vocabulary; Solresol has a comparatively small one--2,660 roots. In fact, Solresol has less in common with schemes such as Ro and em sigh ay than it does with suma, another a priori maverick. It may have been the first non-rationalistic a priori project, and that, together with its advanced concept of analytical conjugation and declension, makes it a classic.
But Solresol's status among conlangs and auxlangs aside, how serious is the charge of arbitrariness? The lexicon is not well designed (see below), but is logic a help or a hindrance to learnability? I would say that rigorous classification (at least of the classical type found in the Real Character) decreases a language's learnability.
Consider: while it is true that certain sounds can have associations (some limited to a specific language, a very few perhaps universal), no natural lexicon appears based on such factors. Indeed, it would probably be easier to argue for dissimilation in most cases. What is the result? When a language has an easily detected lexical taxonomy, the very regularity of its forms will tend to reinforce the speaker's awareness that he is dealing with something artificial. I know of people who belittle Esperanto as being just an overly-predictable toy--and Esperanto is hardly a model of regularity or predictability, though it far exceeds natural languages. Will not a system such as Ro therefore destroy its own credibility? Solresol, by contrast, has just enough order for a mnemonic, but perhaps not quite enough to seem a counterfeit. If this is so, it is a point well worth learning.

The principle of antonym inversion, though clever, is impractical and inconsistently used. In a way, this is a subcategory of the arbitrariness argument. Again, the inversion exists merely as a mnemonic, not as a philosophical statement. In the cases where it is used, I suppose that the learner would simply learn the two forms together: dmrs/srmd hear/be deaf. (This should again remind those familiar with suma of its unique lexical strategy.)

The lexicon is badly designed. This really is a problem. One of the great discoveries found in both Volapük and Esperanto is that a well-designed system of affixes can greatly reduce the number of morphemes one has to learn for basic, ordinary communication. (Of course, for advanced usage--especially for literature--a large vocabulary is still necessary; see Fernando de Diego, Pri Esperanta Tradukarto [Saarbrücken: Artur E. Iltis, 1979], especially “Esperanto en rilato al la naciaj lingvoj” and “Arkaismoj kaj neologismoj kiel helpiloj de tradukado.”) Although trying to reduce the number of basic words, Sudre did not appreciate the power of compounding; only the fs/sf pair makes a token effort in this direction.
Perhaps the problem here is a matter of linguistic background. I once (privately) corrected a professor with a degree in English Linguistics for his analysis of Shelley's Ode to the West Wind, because he referred to the various thous as subjects. They are in fact vocatives (“O thou West Wind...”). Now this professor was no dope; his knowledge of Germanic languages far exceeded my own, and he had been doing linguistics for quite a while. Why did I catch the vocative? I suggested (and still believe) that it was a matter of linguistic background: he knew only Germanic languages, which lack a true vocative case, while I knew various languages (Koine Greek, Latin, and Russian) that have at least vestiges of such a case.
Applying the same idea, compare Sudre's French with Schleyer's German and Zamenhof's Polish (Russian, German, etc.). Which of these people will be least likely to think of compounding as a major theme of their lexicon? French has been nearly dead derivationally for some time, especially as compared to the Germanic and Slavic languages. So it should come as no surprise that this concept was pioneered by people with such a background. (Of course, as a professor, Sudre must have known at least a little Latin, and perhaps even some Greek--languages that should have alerted him to compounding's usefulness. Yet his interest in reducing the need for language study--one of the goals of Solresol--points to less of a love for language than one finds in Schleyer and Zamenhof, so he may have missed the point these non-native languages could have taught him.)

So what should we remember about Solresol? Not its complicated derivational system, not its dependence on French, not even its clever or fanciful variant forms, but that

  1. it pioneered the field of practical artificial languages--not erudite philosophical constructs, but systems using ordinary vocabulary, and reserving the shortest forms for the most common ideas;
  2. it paved the way for the analytical approach to grammar;
  3. it attempted to be the most phonologically accessible system--its syllables could be mangled considerably and still be perfectly recognizable, and its sounds were generally accessible to anyone from a proper phonemic standpoint (for example, those who aren't comfortable with the l/r distinction could still differentiate between re and la, and sol could be pronounced so without much confusion); and
  4. it was the project most concerned with the needs of the handicapped, the first (and still about the only) system to make a selling point of mainstreaming the blind and deaf.

The Future of Solresol

Does Solresol have a future? In itself, no. It does have a past, and most of that could be resurrected by anyone who could unearth Sudre's Double dictionnaire. Failing that, one could sort the vocabulary I offer here into its proper groups, then fill in the empty spaces using either the classification system or inversion. The result would not be “canonical” Solresol, but it would be a fair approximation. For anyone wishing to try this, I would suggest giving inversion priority--see whether the blanks correspond to the inversion of a known form whose antonym seems like a practical, common-use word, and if so, use it. Otherwise, follow the classification. I would estimate that the vocabulary given here (about 500 forms) could be expanded to about seven or eight hundred forms with reasonable confidence.

As to finding a future for Solresol, the system has some interesting features, but as it stands, it is a relic of a bygone age. However, this does not mean that something of the sort could not be designed. The following are the major possibilities:

  1. A musical language using the solfeggio (scale names). This could work much like Solresol, with the following exceptions. First, the vocabulary would be better designed, using affixoids to reduce the need for extra words. Second, it would not mess around with suprafixes (stress, length, etc. used for derivational/syntactic purposes). Third, it would likely be built to auto-resolve. I am very skeptical of this feature in most modern conlangs (Loglan, Lojban, etc.), though I have used it myself on occasion; yet here it would be a virtual necessity. A means of handling other languages (including borrowings) would have to be found--this might not be an outright improvement on Solresol, as Sudre had the wit to construct some such system, but I don't know how his system worked, so another would be needed.
  2. Another syllabic system. One comparison I perhaps should have made earlier is with Babm (boabomu), another syllabic a priori system. Babm is a typical philosophical scheme, but the syllabic idea has its points. Many people have expressed concern over the pronounceability of Esperanto, Interlingua, etc. A system based on open syllables could be written compactly and feature enough redundancy to make phonological sloppiness trivial, while permitting a fairly large effective phoneme pool. As noted above, even people who have trouble distinguishing l and r should be able to produce re and la distinctly. Although such a scheme would lend itself to an a priori approach, an a posteriori system is equally possible, with distortions of source words about like that found in Japanese borrowings.
Back to the Conlangs page.

Back to the Solresol page.

Copyright © 1997, Stephen L. Rice

Last update: Nov. 19, 1997

URL: http://www2.polarnet.com/~srice/solresol/comments.htm