the solresol effect is when you look at how bad solresol was and think "i could write a better language" and then do
I won't. But I could.
The idea of a "musical language" that you could play on an instrument is appealing. It's extreme conlangerbait in the best possible way, something a lot of people independently converge on. The thing that makes it especially conlangerbait is that the original musical language solresol is not great.
Problem 1: It's hard to understand. To quote Prof. Gajewski, the author of the grammar:
Important Note: When speaking Solresol, one should take great care to pause after every word; this slight pause is necessary to separate the words, so that the listener does not become confused but understands easily.
This is partly because the vocabulary is not super spaced-out. There are only eight syllables, and every possible one- and two-syllable word is defined. I'm not saying it's literally impossible that fluent solresol speakers could get to a point where they're singing the equivalent of "Djeetyet?" for "Did you eat yet?", but it's hard to imagine.
Problem 2: It's hard to sing. This is coming from someone who got into choir all-state based solely off a perfect sight-reading. The tonic gets kind of lost, and I can't imagine a non-musician singing it quickly in a way that doesn't confuse "sol" and "la", for example-- especially because many of the passages aren't particularly musical, and a non-musician might try to "snap" them to something more musical.
Problem 3: Solresol is extremely a conlang from the 1800s. Gajewski again:
To avoid giving any national language an advantage, François SUDRE created a
language that does not resemble any other, and, as a result, is absolutely neutral.
Oh, good, absolutely neut--
The class of RE is for clothing, the house, furniture, housekeeping and the family.
The words of four syllables beginning with DO are dedicated to [religion].
EXAMPLES:
Ddsod, gospel, bible;
Dmss, superstition;
Dffd, Easter
The feminine alone is indicated, in speech, by accenting the last vowel. (The masculine gender is the unmarked default.)
To put that in perspective, the Swadesh list at the back doesn't list a word for big. But God forbid we can't say Satan and Easter.
In fact, the thing repeatedly touted as giving the language its neutrality (its lack of resemblance to natural language) ironically makes it unable to take new loanwords. The already lacking vocabulary might be fine if it were possible to loan in words like "ramen", "LGBT", or "Africa", but it isn't. I'm not a Toki Pona enjoyer by any means, but at least you can import people's names. It's hard to imagine an auxlang being able to truly connect people across the globe when you can't name your local food in it.
(Editing to clarify: Of course it's possible for a language that doesn't use IPA symbols to loan words. ASL, for example, has a vocabulary for nationalities, and the ability to fingerspell. The difference is that Solresol is already struggling for inventory space, and can't make room for hundreds of nationalities and/or dozens of letters. I'm also willing to bet that Sudre and Gajewski were too auxlangpilled to accept a syncretic relationship with a natural language, like the one ASL has with English.)
Solresol's main weakness is a weakness of western music theory-- it confuses notes for music. That means that, other than "auxiliary" aspects like accenting feminine syllables, it's a language with an extremely small, not-even-that-distinct syllabry. To quote legendary bassist Victor Wooten:
The reason many musicians get frustrated when start to play... is that they rely on notes to express themselves. There are only twelve notes. Imagine trying to speak a whole language using only twelve words.
(To be clear, he's not talking about Solresol. The absurdity of a musical language spoken in only notes is just such a delicious metaphor that it comes up unrelatedly.)
Music is made in rhythm! If the boundaries of words are getting lost, just give them a predictable beat to latch onto! Imagine a system of "triconcosonantal" (tri-tonal?) roots, with different forms being made by partitioning a measure of 3/4 time differently-- maybe do-re-mi sung in even quarters is "music", but do-re-mi sung as two eighths, a quarter note, and a quarter rest is "to play music". And so on, and so forth. That's more words with more visible grammar (that is, if you don't quite catch a word but hear it's even quarters, you can guess it's a noun)-- we have our cake and eat it too.
You could even reserve an easy-to-catch rhythm like double dotted quarters for particles that begin a sentence Lojban-style, perhaps encoding information about the subject or the mood.
In fact, with rhythm adding more words into the language, we could get away with "taking away" some of the harder-to-distinguish or harder-to-sing ones. For example, I think non-musicians have difficulty with ascending DO-LA, particularly distinguishing it from DO-SOL, and that's fine now-- we can space things out a little better. That's progress toward the problem that solresol is hard to sing.
The "conlang from the 1800s" thing is easy enough to solve (just don't add a contrived femininity marker as an afterthought!) but the loaning problem is not. Perhaps we accept our fate as "mostly worse toki pona but at least the word for news is the NBC jingle", and join the ranks of meme languages alongside "conlang where every word is a variation of uwu" and "conlang where every word is a false cognate of English". Perhaps we need to leave behind the dream of a musical auxlang, and with it the humanist dream that mutual unintelligibility is the root of the world's problems, that class enemies and powerful bigots could be debated down from their lofty positions if only we had the words.
...although there are some natural languages where speakers can speak at long distances with drums and/or whistling, encoding rhythm and pitch. Inshallah Enlightenment humanism doesn't have a chance, but loanwords do. I leave it as an exercise to the reader.