An amateur linguist loses control of the language he invented (2012)

In fact it doesn't just sound like an experiment worth doing. It sounds like something that somebody somewhere might have already done.

After reading it, I would say it is really written badly, but the author's point seems to be that most of Esperanto's structure was basically designed at random, with no particular goals in mind, other than looking superficially similar to other European languages. His substantive objections are as follows:

* the language has an excessive number of phonemes

me: On this point I disagree strongly. A language with more phonemes can form shorter words and convey information faster. that's just basic math (43^9 >> 16^9)

* the vocabulary is too large to be practical and has dubious links to other languages. specifically, Basic English requires far fewer (>10x fewer!) words for competence and is more recognizeable due to worldwide borrowing.

me: Here I agree. The amount of memorization required is quite large and borrowing words doesn't make up for it. Interlingua took a better path, but it targeted the only continent that doesn't need it.

* synthesis mechanisms are irregular and insufficiently general. in particular, esperanto's semantic structure and word synthesis fail to allow the speaker to compensate for missing vocabulary by using compound words or overly general words with attached descriptors

[I don't understand this sort of thing, but it sounds serious]

* the alphabet is needlessly complex. it uses symbols people do not recognize and cannot type in favor of ASCII digraphs with wide recognition.

me: Also, it does not accord significant importance to the way that an orange bikeshed would clearly clash with the trees in the background.

Okay, fine, I agree. This is just putting the nail in the coffin at this point.

* the noun-to-verb-to-adjective declension is not compatible with the structure of meaning, and so cannot be extended in a consistent, predictable way. that is to say, the variety of possible verbs is such that it is basically undecidable what noun should correspond to a given verb in the general case, and vice versa.

me: I don't know this but as far as I can tell he's right. For example, what is the "verb" associated with "electron"?

* the language is inherently sexist.

me: This is not a trivial criticism.

So there are some substantial criticisms in the article, methinks. And you can't discount the impact of difficulty on network effects: if one person has too much trouble learning Esperanto, they're not going to pass it on to their friends, who will not pass it on to their friends.

I wonder if a language designed to be usable with as small a core as possible -- something the author suggests -- could have a substantially better chance of catching on? Perhaps if it were also good at borrowing words from other languages ("extensible"), and grew in the right community for a little while...

Spanish is a beautiful language, I speak some myself, it's probably better than Esperanto or the more European-focused Interlingua, but... consider the number of forms of a regular verb:

(lavar (present lavo lavas lava lavamos lavan) (past lavé lavaste lavó lavámos lavarón) (imperfect lavaba lavabas lavabamos lavaban) (future lavaré lavarás lavará lavaramos lavarán) (conditional lavaría lavarías lavaríamos lavarían) (present-subjunctive lave laves lavemos laven) (gerund lavando) (participle lavado))

And that's ignoring the twenty or so irregular verbs and roughly ten "irregular" patterns (e.g. querer). It's a lot simpler than French or Italian, to be sure, or Latin, for that matter, but it could be a lot easier.

To be glib, it has been done. We call it language.

Seriously, though, take a look at the link I posted in https://news.ycombinator.com/item?id=8180924

One of the techniques used is to computationally create a space of possible ways to partition semantic domains on a plane whose dimensions are simplicity and informativeness, in order to look at where in the possible space it is that real languages lie. While it's not been done (to my knowledge) for a whole language, it's potential direction to go.

58 different phonemes means the language is just substituting one form of complexity for another.

And apq’uxasiu for 'gawk' doesn't strike me as a particularly huge win for conciseness.