Auxlangers vs. Artlangers

NOTE: My viewpoint on this matter has changed. I no longer think that there are such clear-cut differences between auxlangers and artlangers beyond some tendencies, especially given the fact that both groups overlap to a certain degree.

Auxlangs and artlangs are usually quite different designs, simply due to their fundamentally different design goals. An auxlang is designed as a means of international communication, meant to be learned by many people (ideally, by everyone in the world) as a second language such that people can make themselves understood across language barriers. This means that an auxlang has to be easy to learn and use, and ought to be culturally neutral. (Which doesn't necessarily mean that these objectives are actually met - there are many baroque auxlang designs that fail to meet these criteria, often in spectacular ways.)

An artlang, in contrast, is designed with no such practical purpose in mind. It is simply a work of art that exists merely for itself, or as a part of a fictional world or culture. The design of the artlang simply follows the personal preferences of the designer, or whatever seems fit in the context of the fictional world it is designed for. The vast majority of artlangers prefer languages that look like real, natural languages, with all the complications and irregularities that are characteristic of the latter, and they like to 'spice' their languages with interesting linguistic features, be it unusual phonemes and phonotactic rules, exotic (from a western European standpoint) grammars (e.g. ergative constructions) or words with especially 'interesting' etymologies (in my own Albic languages, for example, the word for 'barbarian' literally means 'meat-eater': the Elves are vegetarians and consider the habit of eating meat disgusting and not worthy of a civilized person), or whatever. All these things are features that are generally expected to be absent from auxlangs: auxlangs are expected to be straightforward (whatever that may mean) and easy to learn.

Three observations

It seems, however, that auxlangers and artlangers, i.e. the people who design auxlangs and those who design artlangs, are also two very different sorts of people. Of course they are different, because they do different things. But there is more difference than one might expect at first glance. (After all, they both share the rather unusual disposition towards making up languages.) In this essay, I want to cast a light at these differences and try to find out in which way they relate to the different types of languages they construct.

First of all, here is a word of warning. Don't expect a perfectly balanced, objective evaluation of the facts. I am an artlanger myself, so I am biassed. But I think some things can be objectively observed, may it please anyone or not. Having said that, let's turn to the observations.

1. Artlangers tend to get along with each other much better than auxlangers

From looking at the virtual waterholes of both factions - the CONLANG and AUXLANG mailing lists, respectively - one can easily make out that artlangers get along with each other much better than auxlangers. OK, every now and then a flamewar erupts in the CONLANG mailing list, but that can be observed even in the best scholarly fora. (And when it does, it might have any reason but impolite criticism of a particular artlang, which is rare.) Most of the time, people on CONLANG devote their resources to more or less fruitful discussion, and the worst thing that happens to one is being ignored. More typical is that whenever someone discloses his latest creation to the public, a flurry of appreciatory comments is showered on him. Even if changes to the design are suggested, the tone remains friendly because both the original designer and the suggestor understand this to be well-meaning. Those who don't like his design simply don't bother commenting, and everyone is happy.

Now look at AUXLANG, and the picture is entirely different. The hottest topic on that list is which auxlang is best. While most of the time, the participants in this debate stay polite towards each other, they are firmly entrenched in their positions and won't budge, and every now and then it erupts into flamewars. The debate is entirely fruitless, because there is no consensus to be reached. (This constant warfare is the very reason why the AUXLANG list exists at all: all that used to molest the audience of CONLANG, until the CONLANG list admin set up a separate list - the AUXLANG list - for the auxlangers and explicitly banned auxlang advocacy from CONLANG.) It is sometimes astounding how much bile is shed on how minute details. Often the proposals at war are similar enough to be mutually intelligible (such as Esperanto and Ido), and the details in which they differ rather obscure and of little interest to the general audience and probably don't affect the usability of the languages in any appreciable way. But nevertheless, the auxlangers apparently cannot agree on them.

Why is it so? Why do artlangers appreciate each other's projects, while auxlangers are bickering about their proposals all the time? Well, it is very simple. It is because auxlangs suffer from what one could call (and at least one commentator - I have forgotten who - actually did call) the Highlander condition: There can be only one. After all, the goal of all that auxlanging is to set up a single language for use by everyone. The only thing that can be called success of an auxlang is its adoption as a globally valid language of international communication, and this can only be reached by one single proposal, if any - all others that have been made must be rejected. Thus, every new proposal infers an implicit rejection of every previous proposal, because if only one proposed auxlang was really fit for its purpose, there would be no need for any more proposals. Hence, someone who (seriously) proposes a new auxlang can (and will) be construed to claim that all previous proposals are inappropriate. And there have been lots of auxlang proposals.

Artlangs, in contrast, do not conflict each other like that in any way. There is always room for more. It is as with any art, be it music, paintings, architecture or literature. Just because so much good stuff has been composed, painted, built and written before, that's no reason to stop composing, painting, building and writing even more good stuff - to the contrary: the works that have been done inspire new artists to create new works. There are even projects on which several artlangers cooperate, each contributing their own language to a shared universe.

2. Artlangers tend to be better linguists than auxlangers

This, of course, cannot be generalized, there is everything from brilliant linguists to complete ignorants on both sides of the fence, but some general tendencies are observable. In both camps, the vast majority of contributors are amateurs. There are a few professional linguists among both artlangers and auxlangers, but the vast majority are not.

Now, many artlang sketches show that their authors have a more or less firm grasp of contemporary linguistic concepts. At least, they tend to understand the basic notions of phonology, morphology, syntax, historical linguistics, etc. Most artlangers understand what a phoneme is and that it is phonemes, not letters, that are the basic building blocks of langauges. Artlangers, for the most part, also try to look beyond the rim of their lingusitic bowl. They study descriptions of exotic languages in constant search of ideas for new, interesting features to build into their languages. They check out linguistics textbooks from libraries in order to learn about historical linguistics, language typology and other linguistic topics.

Auxlangers, however, seem to be less firmly seated in linguistics. One often sees things in auxlang proposals that make every (professional or good amateur) linguist's hair stand on end. To start with, most auxlangers describe their languages in terms of letters rather than phonemes, thinking that letters are the fundamental building blocks of language while they are really just symbols representing the actual building blocks, namely, phonemes. Instead of saying, 'My language has the following phonemes, which are represented thus in the orthography', they say, 'This language uses the following letters, which are pronounced thus'. (Some auxlang proposals do not even state how the letters are pronounced.) More generally, auxlangs are often poorly described. Auxlang grammar sketches tend to be full of expressions such as 'all the letters have their international values' or 'the words have natural stress'. What are 'international values' of letters, and which stress pattern is 'natural'? Or they say how the 'conditional' is formed, but not what this form means.

Another common fallacy of auxlangers, and far worse than confusing letters and phonemes, is their parochial perspective. They tend to mistake western European peculiarities for universals, and take grammatical categories specific to European languages for granted. The logic runs along the lines of 'English, German, French and Spanish are all like that, hence all languages of the world are probably like that'. Kawoom! This is obviously false, but most auxlangers fail to realize. This is the reason why 99% of all auxlang proposals are entirely eurocentric.

And then there are those mavericks who propose 'philosophical' languages and other schemes that claim to be 'perfect' or 'improved' languages which try to fix 'bugs' of human languages that are actually features. (More about these folks below.)

The question is of course, why is auxlanger linguistics so doggy? Perhaps this is because anyone who knows how languages work inevitably becomes aware of the countless problems that weigh down auxlangs, and loses interest in creating a new auxlang; or realizes that the minute differences between the mainstream IAL proposals (such as Esperanto or Novial) aren't all that important, and thus sees no point in all that bickering. So you are either a good linguist or an auxlanger, but never both, it seems. This question was also briefly discussed in the CONLANG mailing list in June 2003. Here are three answers:

I think that it is because auxlangers try to fix natural languages, while artlangers only try to imitate them (usually, at least). I think it's only a question of too much ego, thinking that they know better than Mother Nature ;-)
(Roberto Suarez Soto)

My guess is it's different intentions. Artlangers, generally, are after imitation or exploration. Auxlangers are after a language that everyone will use and will put them into history books. You see it so often: things designed to make its creator famous are often of a lower quality than things made for the fun of it (though of course not always).
(Tristan McLeay)

Or perhaps it is that people with a good feeling for language, skilled linguists or not, will realize that the little details don't really matter that much for learnability or usability. If they should want to further the cause of one common IAL, they will sit back and wait for the fanatics to agree on which one.
Or they speak Esperanto and are done with the discussion already.
(Lars Henrik Mathiesen)

This way or the other, it is apparently the different goals of auxlangers on one hand and artlangers on the other hand that are responsible for the observable difference, and that insight into the workings of language leads to a standpoint aloof of the pesky debates among auxlangers.

3. Auxlangers tend to be infected with all sorts of odd memes

Auxlangers tend to be infected with all sorts of odd memes that artlangers seem no more susceptible to than the general population. The chief of them all is the auxlang meme itself, of course: most social problems are caused by misunderstandings between speakers of different languages, and an artificial language could fix them. It is of course true that misunderstandings cause all sorts of problems, from unintelligible users' manuals to embarrassments on trips to foreign countries to controversies over the interpretation of international treaties; but auxlangers tend to overrate the magnitude of such problems as well as the ability of an IAL to fix political and social ills. It would not end wars, for example: there have been bitter wars where both sides spoke the same language. (Examples are, among countless others, the American War of Independence, the American Civil War and for that matter, most civil wars.)

The auxlang meme often combines with other memes from one or both of two classes. The first class are several radical political ideologies and philosophies - socialism, libertarianism, objectivism, whatever-ism. The second class can be summed up as various varieties of the natlangs-are-flawed meme: many auxlangers thus think that they can better the lot of mankind by 'fixing' certain 'flaws' of language, in many cases 'flaws' that are now recognized by linguists to be features necessary for language to work. Here are a few examples of such memes:

The philosophical language meme - the arbitrary nature of the relation between sound-shape and meaning is seen as a flaw, and 'fixed' by constructing lexical roots along a classificatory scheme of concepts. (The reason why such a beast is called a 'philosophical' language lies in the terminology of the 17th century when such schemes were popular; back then, 'philosophical' meant 'scientific'.) The fault in this lies in the fact that classification schemes are arbitrary by themselves, and also difficult to learn; it also leads to a lack of redundancy and an uncomfortable similarity of semantically related words. A milder version of this is evident in schematic derivational systems like that of Esperanto; while it doesn't make the language unworkable, it is far from running smoothly, and most later auxlang proposals have ditched it for good reasons.
The closed vocabulary meme, also called the Basic English meme after Ogden's Basic English scheme: this is closely related to the philosophical language meme. The assumption is that one doesn't need more than a few hundred words, as everything else can be semantically decomposed into simpler concepts - and it is wrong. Things aren't quite that simple, and if Basic English works, it only does so because it uses lots of idioms which must be learned just like the words they replace.
The Speedtalk meme - the idea that redundancy is evil, and information ought to be packed as tightly as possible in order to shorten texts and speed up discourse and thinking. This is often combined with the philosophical language and closed vocabulary memes. Bugger. Redundancy is not evil, but necessary to prevent information loss in noisy environments.
The 'to be'-is-evil meme - the proponents of this idea claim that the verb 'to be' causes a high degree of semantic sloppiness, resulting in all sorts of social problems. Thus, removing the verb 'to be' would lift all sorts of curses - wars, crime, racism, sexism, environmental pollution, etc. - from mankind. Blech. There are hundreds of languages which actually lack that verb, and the speakers of those aren't more saintly than Englishmen or Americans. (The terrorists who destroyed the World Trade Center spoke such a language, for instance.)

There are several others. The common denominator of all these is that allegedly, language, and by that means the condition of humanity, can be bettered by radically departing from the patterns of natural languages. In fact, this only leads to awkward, unworkable schemes. And that is the reason why international business is still not conducted in Esperanto, nor in Lojban, much less in John Wilkins's (or anyone else's) philosophical language - and there is little reason to assume that this will change in the foreseeable future.

Conclusion

What it seems to boil down to is a difference of personality structure between auxlangers and artlangers. The auxlanger is typically of the sort who thinks that (s)he has found a way to better the world; that the languages that are spoken in this world offer no solution and thus a new, artificial language is on the order of the day. One could say that the auxlanger is of a visionary leader personality. (L. L. Zamenhof, inventor of Esperanto and the world's best known auxlanger, is an excellent example of this personality type.) Such people tend to erect mental barricades against any kind of input that seems to threaten their world-view and thus endanger their mission. They have a grand plan in their head and live to fulfill it.

The artlanger, in contrast, has no such aspirations, or at least, they are not why (s)he creates languages. The artlanger wishes to create a language as a work of art; (s)he is of an artist personality as opposed to the visionary leader personality of the auxlanger. Like most artists, (s)he aims at mastership in the art (in this case, of making languages), and thus endeavours to obtain a good understanding of language.