Anything but Romance... why auxlangs fail

IMPORTANT UPDATE: 4 September 2011: I no longer agree with the opinions that I expressed below. In fact, I have chosen to return to Interlingua as it has become my language of first choice in which to read literature! The opinions I expressed below do make sense if you expect (a kind of magical thinking) to be fluent in an auxlang within a few months or a year. But if you accept that to be an unrealistic goal and instead you view learning an auxlang to fluency as a five-year process, then auxlangs based on Romance languages are fine and in fact have many advantages. My frustration and disappointment expressed below were the direct result of my unrealistic expectations of how quickly a language could be learned. Those unrealistic expectations were the result of unwise propaganda from organisations and persons promoting various international auxiliary languages, including speakers of Esperanto and Interlingua, which suggested that rapid fluency was likely. I now think that auxlangs should be promoted as "Five-Year Languages", thus preventing unrealistic expectations and preventing gross disappoinment and disillusionment such as I experienced below. Had I fully understood right from the start that five years are, on average, required to master any of the major auxlangs to fluency, I never would have become disillusioned but rather would have happily continued to make slow progress towards the very worthwhile goal of fluency. In fact, auxlangs based on Romance languages are very advantageous because studying them gives you the advantage of subsequently being able to understand a considerable amount of Spanish, Portuguese and Italian. Moreover they allow you to easily learn and understand scientific terminology, including Latin terms (for example, used to name features on the Moon and on Mars). It simply takes a few years of study to achieve, and that is worthwhile. END OF UPDATE

One of the great challenges of writing this blog is to demand one thing of myself: complete honesty. I have always been of the belief that if I am not honest then this blog is of no value. I think whatever value this blog has is due to its honesty.

Accordingly, I have been willing to make a complete fool of myself in public. Again and again readers have seen me wildly enthusiastic about one language after another only to go on to... get very poor results. You have seen me honestly become enthralled by various languages and then honestly become disappointed by them; you have seen me go through extreme changes of opinion, swinging from one view to another, as I have desperately scrambled to find a language that works for me. Regular readers will know I am a novelist and I am seeking a language suitable for writing international literature in... and I mean novels, not just short stories. Therefore auxlangs which many others find quite acceptable (for general conversation and for writing short articles in magazines) have been unacceptable to me because they are not up to the requirements of writing a good novel (this requires a very large vocabulary, unambiguous grammar, and most importantly the ability to dismantle the idiomatic expressions of natural languages either into unambiguous plain language or into well-established idioms in the auxlang, requirements that very few constructed languages meet... in fact virtually none and certainly not, for example, Interlingua).

Late last year I came to the conclusion that I had completely wasted my time by studying auxlangs and that I wanted my nine months back!

Then, at the eleventh hour, I managed to salvage a literary result by producing a translation into Occidental, eventually of more than 5700 words in length, of the beginning of the great French novel, La Chartreuse de Parme. This translation might even have the benefit of being somewhat comprehensible to some Continental Europeans (but not to most British readers) with perhaps only a few weeks of prior study and access to a dictionary; some especially well educated readers who already speak several European languages might even be able to get the gist of it without any prior study. This is unlike Esperanto, the world's most popular constructed language, which is nearly completely incomprehensible without extensive prior study.

But... so what?

Here's what happened after I published my translation, which was for me a great personal achievement: nothing. About three people congratulated me, on an Occidental-language Yahoo group, where my translation will sit unnoticed for the next ten years gathering dust. This is because there is no active community (other than perhaps a couple of dozen people who sporadically produce very little content in very inconsistent styles and in minor variants of the language, which frankly are not absolutely mutually comprehensible) for the Occidental language. Now, this lack of any truly significant community is a clue. Let's investigate.

Here is a list of the constructed languages with a truly significant community, say more than 1000 users of the language, other than Esperanto:

• none

Okay, that didn't go too well. Let's try a different approach. Here, to the best of my knowledge, is a list of the constructed languages other than Esperanto in which published novels are available (either in print or just as e-books):

• Interlingua

Hmmm, that's weird. If we have a language with huge dictionaries, as Interlingua has, in which novels have been published, how come it has no significant community of users? Aha. Here is the answer. To save time I will cut straight to the chase and will reveal the secret finding of my one year of study:

Because it does not work. It is gibberish. For the novelist, most auxlangs based on Romance languages are a complete waste of time and, ultimately, are either fundamentally gibberish or too difficult to understand without extensive knowledge of relevant natural languages. Auxlangs based on Romance languages are nearly always not "standalone" languages which one can study on their own without reference to any natural language. This is why nobody, other than a tiny handful of language enthusiasts, uses them (even if you disagree with my conclusions here, the fact remains that no Romance-language auxlang has a truly significant community of users, something which is hard to explain unless my hypothesis here is correct... Romance-language auxlangs do not work).

Here is the best way to communicate internationally using a language which is based on a Romance language or which resembles a Romance language:

Don't. Use a real Romance language instead.

I totally understand now why the auxlang movement has failed, except Esperanto which has had a limited degree of success and actually has a small but significant community of users, maybe a few thousand people around the world. It is because nearly every significant auxlang is based on Romance languages and auxlangs based on Romance languages do not work. If they worked, people would be using them. They don't work.

Why don't they work? Because dumbing down Romance-language grammar produces auxlangs which are inherently ambiguous gibberish despite looking very impressive at first sight and despite in some cases have massive vocabularies. Proper Romance languages (not counting creoles here, to keep the discussion simple) have difficult and complex grammars: that is how they work, and this ultimately inherits from Latin. Furthermore, Romance languages inherently work by virtue of thousands of highly idiomatic expressions, which are unique to each Romance language, and upon which intelligible communication depends. Auxlangs which use Romance-language vocabulary attempt to simplify this and it simply does not work. This is why Interlingua drifts in the direction of the native Romance language of its speaker or writer, becoming at best ambiguous and at worst gibberish to others.

Even Occidental, by far the best of the group for novelists, is too inconsistent among its tiny group of users and quite frankly is too difficult to learn unless you already know two European languages other than English, consisting of one Romance language and one other language, preferably a Germanic one. By the way, Occidental is the best of the bunch because its grammar has Germanic influences which make it possible to use it more easily without having to have such extensive knowledge of natural Romance languages.

The other approach, that of Lingua Franca Nova, to use Romance vocabulary with a creole grammar, in some ways has more merit because it is further from natural Romance languages and thus produces less drift, but unfortunately it does not work either as unless one is an absolute and seasoned expert in the language it is impossible to write without a massive degree of ambiguity, which is a drawback of its creole grammar. I tried again recently and failed. That is, Romance-based auxlangs have a choice between: (a) using difficult grammars and unreliable idioms which require extensive prior knowledge of various natural languages (often resulting in gibberish); or (b) dumbing down the grammar to that of a creole (not necessarily a bad strategy) which usually fails and usually produces a language which ultimately is equally or more ambiguous.

I'm tired of writing this post now. I give up. I leave you with this advice:

(1) Do not use any auxlang based primarily on Romance languages. My intention is to never do so again. This leaves only Esperanto, Lingwa de Planeta, Sambahsa and Frenkisch to choose from out of the most significant auxlangs.

(2) Do not use any auxlang based on an Interlingua-like process of averaging the vocabulary of related natural languages. This rules out Frenkisch*. This is because the same kind of Interlingua-like drift problems will occur and you will not be able to learn and use the language as a "standalone" language without needing to simultaneously study several natural languages.

(3) For the novelist, you need extensive vocabulary: this currently rules out Lingwa de Planeta, although it will get there eventually one day in the future. Lingwa de Planeta shows a lot of promise but is currently in an early a stage of its evolution, suitable only for short stories at this stage, not novels.

(4) For the novelist, you need extensive multi-lingual documentation and reference materials: this rules out Sambahsa, although one day hopefully it will have these things. However, possibly Sambahsa may still require too much knowledge of various natural languages to be used "standalone" without simultaneous study of natural languages; the jury is still out on that.

* Update 30 September 2011: I sincerely regret saying the above about Frenkisch. Now I understand that just as Interlingua is a useful language, so too could Frenkisch be a useful language. Frenkisch is a fine project of high quality and I am embarrassed now that I said such negative things about it here. Again, this was a consequence of my "magical thinking" than somehow a language could be learned to fluency in a few months. Now that I realise that learning any language to fluency takes years, Frenkisch appears to me to be a valuable linguistic resource and one which could grow into a viable international language with a little more development of its grammar in the years to come. 
END OF UPDATE

This leaves the novelist with only one reasonable current choice: Esperanto. Please bear in mind I am not recommending Esperanto, because I do not speak Esperanto yet. However I am going to learn it. At this stage the most likely outcome is probably that I will also be disappointed by Esperanto and probably will choose not to write novels in it, but at least it is in with a chance. The most likely outcome is I will choose to use Afrikaans or Indonesian instead.

Well, that's it. My study plan for 2011 is:

Mondays, Thursdays: Afrikaans

Tuesdays, Fridays: Esperanto

Wednesdays, Saturdays: Indonesian

Any day (unrestricted): French, English

A few parting points:

(a) Occidental is a nice language fully capable of having novels written in it, but basically depends on the reader already speaking at least two European languages to be able to read it with minimal prior study; this is pointless since the language is not a "standalone" language but depends on knowledge of other languages. Were it massively popular I would write novels in it, and those novels would not be excessively ambiguous because idioms would have been established by popular use. However it is not massively popular and never will become so, because it requires too much knowledge of other languages at the current point in time at which its multi-lingual documentation is virtually non-existent.

(b) Occidental would be a better language if it had more non-Romance elements. Which is essentially what Sambahsa is, a very admirable language.

(c) An auxlang based very closely on a simplification of just one Romance language, rather than averaging several Romance languages, could work, since all the idioms could be unambiguously inherited from that single language. Thus, Basic Spanish something like Basic English. But I would not bother because languages like Afrikaans and Indonesian are fully capable and are easier.

(d) Esperanto will have to prove itself to me to be able to rival Afrikaans and Indonesian both in terms of power and ease. I am highly doubtful it will.

(e) At-sight languages are dead. Forget them. (An at-sight language is one designed to be comprehensible without prior study or with little prior study; while this sounds like a nice idea I now realise this is usually a fatal flaw since it utterly depends upon one having knowledge of certain natural languages, thus making the auxlang not a "standalone" language which one can study or use without having to simultaneously study other languages. Interlingua is the classic example of this and it suffers from the worst of the drawbacks of this, making it essentially inconsistent gibberish. The absurdity of this was recently driven home to me when I realised I can now read French more easily than reading Interlingua!)

Onward...