What makes an apriori language easy?

What is easy?
There are few levels to what is easy. Is it easy to create the language, is it easy for the creator to learn the language, is it easy for some particular audience to learn the language, with an emphasis on “what particular audience.”

What is a priori? What are the degrees of Apriori-ness? Why a priori?
A constructed language, in my opinion, should be different. Already many people have said harsh things about relexes–which charitably could be called English con-dialects. A language can be the same or different on a variety of levels. It can have much the same vocabulary as other European languages but different grammar, like Esperanto. It can have different phonetics, but the same morpho-syntactic alignment. Of all the parts of a language, what is it that is important to be different? An pure exercise in con-phonetics may as well be a relex if the rest of the language isn’t breaking any new ground..

I’ve in mind a language that will induce different metaphysical obsessions, so phonetics don’t have to be a priori (and how could they be? We can only speak what our mouths can speak!) But the morphosyntax and semantics should be a priori.

But Pidgins, Creoles and Hyper-loan-word languages are easy.
Yes, but why bother? If you optimize this strategy for one English speaking person, then the optimally easy constructed language is English with minimal changes. For much the same reason, I’m not interested in hybrid languages. A hybrid language just converges on an easy version of the language family’s proto-form.

Your Culture is Easy. An Alien Culture is Hard. Neutral Cultures aren’t that easy.
[I’m excluding metaphysical obsessions—see below--from culture for the moment. I have in mind things like, do you have a word for pizza, Sir, and the like]

By easy, I mean, it is easy to come to peace with how the language copes with those points where culture and language intersect. Dritok is harder to speak because of the mixture of a conworld with the conlang. When there is a conworld, sentences have to be evaluated against syntax that you are still learning and appropriateness in a conculture, which you may not know very well. Languages without a conculture are easier, because a speaker is able to bring along as much cultural baggage as they like.

There are linkages between language and culture. If the culture is a priori, then you will be embarking on a world building adventure, which makes perfect sense for a book or a movie, or if you just enjoy writing ethnological studies of fake cultures. A lot of work can be saved by omitting the wholesale world building.

Neutral cultures are difficult because culture is in the metaphors and choice of root words. It is easier to make a language if you needn’t sweat bullets every time you use a metaphor or pick a root word.

I suppose to make an a priori language easy for a target audience, one might imagine a hybrid culture, say the language that might be used in a household of a Japanese and American couple—something with more mechanisms for politeness and manners than English but fewer than Japanese.

I plan to make a language with my own contemporary culture in mind. A hypothetical fan of my language wouldn’t pay much attention to the implied culture in my conlang. They will have to pay attention to metaphysical obsessions, like if the language insists on marking time, completion, biological gender, etc.

Phonetics.
If I can hear a language, I can learn it’s phonetics. The languages that must be spoken entirely from a written description are hard. With the internet it is easier to get mp3 snippets of exotic sounds, but for a lot of exotic sounds, a short mp3 snippet isn’t really enough. It would be better to have hours and hours of MP3 news, say on Voice of America or what have you.

Constructing a phonotactics system that sounds good is hard.

So I plan to use the phonetics and phonotactics of an existing language, but the lexicon, syntax, semantics will all be a priori.

Size
The language description should be small. The vocabulary should be small, but if it is too small, then people will resort to memorized phrase, which is hard.

The number of linguistic mechanisms total should be relatively small as well. If there are inflections, there should be few. If there are bound morphemes, there should be a modest number. This biases the easy languages towards being analytic, which maybe provably easy, at least many people seem to think lexical and analytic processes are easier to learn and use than morphological (inflections, prefixes, suffixes, etc).
At the moment, there are a lot of linguists noticing that languages spoken over large areas, languages spoken by populations that have recently merged from two original groups tend to be more analytic and isolating, which lends some weight to the idea that analytic/isolating languages are easier.

Regularity
Irregularity may appear in the language as you teach it to yourself. The irregularities will be easier for you to leave in than to remove, but will be easier for other if you remove the irregularities.

Easy to Make
Machine generate the lexicon. This doesn’t reduce the burden of defining the words, but it makes it easier to list them.
http://whee.dk/ Word Builder

Try a self-elicitation approach to conlanging. Babble, and then see if you can find any patterns in what you’ve written. Alternatively, take simple model sentences, babble and then match up the babbling with some possible glosses. If the model sentences are simple, it follows that the syntax is likely to be simple as well. Choose your model sentences well—if your sentences are in English, then they will all have time markers in them, which is a peculiarity of English and not a necessity for a conlang.

Self-elicitation is likely to be very fast and easy, but there is no particular reason why the result would especially a priori, especially if you accidentally start babbling in some mixture of all the languages you’ve ever studied before.

Avoid churning design decisions, especially on phonetics and phonotactics-that is don’t make lot of rapid and fundamental changes. Avoid churning on other features to make it easier for you to learn. Churning design decisions may or may not make it easier for others to learn, it depends on if it was a good decision. After a language is published and has a fan, design decision churn is just about impossible, except maybe in the hyper-technical details. If you do publish, get a fan and then attempt to make radical changes, then it’s going to be extremely difficult, on the scale of learning a new language. It’s been estimated that a mere 10% of words being different can make two language mutually incomprehensible.

This entry was posted in conlang design. Bookmark the permalink.

Comments are closed.