Balancing a Conlang

       “Fast, Cheap, Good : pick any two”.

       I am often reminded of this when considering constructed languages. Unfortunately the wishlist is not that simple!

       One thing that might be desired is for the new language to be easy to use and quick to learn. This suggests that the language should have a logical and consistent construction. Verbs should all be regular and there should be very few “exceptions to the rule”.

       Easy to use implies that the reader/ speaker is not required to have a particularly in-depth knowledge of grammatical theory. Esperanto and several other constructed languages have distinct endings for adjectives, adverbs and/or other word types. It is quite possible to be fluent in a natural language without being particularly conscious of such distinctions. I managed several decades of speaking English without even knowing the distinction between transitive and intransitive verbs.

       Familiarity might seem an attractive feature in making a conlang user friendly. Many conlangs have drawn on existing natural languages for their vocabulary. Typically the romance languages are used. Examples include Novial and Lingua Franca Nova. An English based example is Inlis. Some of these conlangs have the advantage that it is often relatively easy to work out the gist of the meaning. Easy, that is, if you have some grounding in the languages on which the conlang is based. Knowledge of the romance languages is not as widespread as one might assume from the number of speakers. I once overhead a long drawn out conversation between a French speaker with limited English and an English speaking Chinese girl. The stumbling block was the question “Quel age as-tu?” Another stumbling block is that the same or similar words have different meanings in some natural languages. English has so many homophones and irregular verbs that attempting to make it into something more consistent and logical rapidly transforms it to something that is not that recognizable.

       Using existing words and words that are logical or easily learnt meshes somewhat uneasily. The word for plumber in many European languages is based on a word for lead, but it is not necessarily based on the word “plumbum”. Plumbers seldom use lead for modern plumbing so the name is not obvious for those that do not know its historical origins. “Water pipe worker” seems more obvious but that could also describe a urologist! We could replace the word for “worker” with “doctor” or “healer”, but a urologist may be an academic rather than a medical man. “Water pipe studier” could mean someone who designs drainage systems!

       Brevity is another desirable characteristic. Obviously it would be desirable if the most commonly used words are the shortest and/or easiest to say. This may clash with using existing or familiar words. It may also clash with more logical construction of words. Even if we assign only a syllable to “water pipe worker” we have a word of three syllables, four if “worker” is formed as an agent noun of “work”.

 

Lego Words

  One of the current problems with the Diinlang project has been the lack of vocabulary. Most of my work has been with respect to verb structures, pronouns, noun genders and similar subjects. Essentially I have built some bones for Diinlang but we also need some “meat” to better illustrate how the bones are working.

A number of approaches can be made towards creating a vocabulary. The one that I favour at present is the use of “lego” syllables. Each syllable has a specific meaning so the meaning of a word can be deduced by consideration of the “bricks” it was built with. An often repeated piece of information on language webpages is that the Mandarin word for “plumber” is “water pipe technician”. The three characters used can be read in a number of ways and an alternative might be “pipe working expert”. This does, however illustrate the lego syllable idea. If we have syllables that mean “water”, “pipe/tube” and “work/ worker” we can build a word for plumber and have a reasonable chance that a Diinlang speaker who has never before encountered that word would have a good idea of its meaning. For example, if water was “kwa”, pipe was “piy” and work “gung” our plumber would be a “kwapiygungzo”, or perhaps just a “piygungzo”.

This approach has been tried with some other artificial languages. Ithkuil is an example of a language with information dense words. Possibly the best know example is Searight’s Sona language. Sona uses syllables as “radicals” and the meanings and associations with related syllables can sometimes be more intricate that you might first assume. Like many conlangs it is geared more towards written rather than spoken use and some of the distinct radicals are phonically similar. See this article for an essay on the use of radicals and a convenient list here.

One of the initial concepts of Diinlang was that syllables should have a “CVn” format, where “C” is a consonant, “V” is a vowel and “n” is a nasal such as “m”, “n” or “ng”. It should be understood that C and V represent phonemes rather than single letters. In practice Diinlang has expanded to use phonically clear mora (“CV” and “VC”) and some “CVC” constructions, particularly when the final C has a hard sound.

There are therefore plenty of syllables to choose from. The real work is selecting what the building blocks to assign them to should be!

Ving, Vang, Vong

Before I decided to place my thoughts on Diinlang on this blog I was using the word sequence “ving”, “vang”, “vong” for the words “here”, “there” and “yonder”. The acoustics of ving, vang, vong made them easy to remember that they were related. I expected to use the same system for other sequences of related words.

The start of the blog coincided with an interest in Dutton Speedwords so I considered a shorter series of words and proposed “sa”, “si” and “so”. Despite the brevity I have not been particularly satisfied with this change. “Si” and “sa” also had the meaning of “this” or “that” and by implication “so” could mean “that which is very distant”.

If the use of ving, vang and vong is reinstituted it may be more logical to place them in alphabetical order so the sequence becomes “vang”, “ving”, “vong”. The concept of “near” could also be added to the sequence, giving us “vang”, “veng”, “ving” and “vong” for “here”, “near”, “there/ far” and “yonder/ very far”. If used as demonstrative pronouns (this, that, these, those) the definitive article or another determiner is used to indicated noun/ pronoun status. Compare to "det här" in Swedish. Hence “de vang, de ving, de vangz/ dez vang, de vingz/ dez ving”. These may mutate to “dang” and “ding” so keep these words free for this potential use.

This also makes it more practical to resume use of “se” as a reflexive pronoun, as it is used in Portuguese and some Scandinavian languages. Pronunciation would most likely be “she” in Diinlang.

Whether to use “ving” also as a relative pronoun needs to be considered. Interglossa uses “su” from “subject” as a relative pronoun, which I quite like. It is possible Diinlang could use both “ving” and “su” but have them fully interchangeable.

Agent Nouns in Diinlang

In English agent nouns are created by adding “-er” to a verb. English being English this is sometimes done with “-ir” or “-or”. “-er” is also used for comparatives. Some agent nouns, such as “artist” use “-ist”, although “-ist” is more commonly used for someone who holds a belief or follows a philosophy rather than someone who performs an action.

For Diinlang it is obviously desirable that there is only one way to create an agent noun, and that this be of a form that is distinct from other word types.

An option I considered was to use the suffix “-or”. This could be gendered as “-oro” for males and “-ora” for females. Inanimate objects that perform an action would be designated by “-it”. Thus if we used the work “kuk” for the action of cooking a cooker or stove would be a “kukit” and the person using it a “kukor”, “kukoro” or “kukora”. (This is just for illustrative purposes. The final world for “cook” may be quite different)

It occurs to me that things can be made simpler for the learner. The word “du” is used for the verb “to do”. We also have the pronouns “zo”, “za” and “ze” to indicate male, female and neuter/unknown. This might give us “kukdu/ kukduze”, “kukduzo” and “kukduza”. Unfortunately if the verb we are modifying is “du” this gives us “dudu” or “duduze”. The syllable “du” is probably redundant which suggest that agent nouns be created by the simple expedient of using “-ze”, “-z0” and “-za” as suffixes.