Quick Links

Constructed Languages

Interlinear text example


Skerre: An Invented Language

By Doug Ball

Given at Rochester URC Conference April 7, 2000

I have invented a language, called Skerre. There are many terms for invented languages: imaginary languages, glossopoeisis, international auxiliary languages, but the preferred term among language inventors is "constructed language," which is most often abbreviated to "conlang." Skerre is a type of conlang called an art language or artlang. This means it is not a language game like Pig Latin, nor is it an attempt to create a language in order to promote world peace, like Esperanto. It is not an attempt to create a computer language, or a logical language. Rather an artlang is an attempt to simulate a real natural language for the purpose of exploring the workings of language. Complete simulation is indeed an impossible dream, but that is not the point — does a model railroader try to simulate the entire railroad in his layout? Rather the point is to further explore languages through the hands-on way provided by a conlang. And since no language exists in a vacuum, I have also invented a race of people, also called the Skerre, who speak my language, and provide it with a cultural context.

I'm not the only one

Am I the only one who endeavors in this pursuit? Certainly not. There is an international conlang community, in touch with each other through a listserv called CONLANG, who discuss their invented languages through the listserv. And doubtless there are numerous other people not online. There are also several famous conlangers. J.R.R. Tolkien, author of The Lord of the Rings Trilogy, invented both Quenya and Sindarin for his elves, and Marc Okrand invented the now popular Klingon, which, contrary to popular belief and The Onion, does not have more speakers than Navajo.

Foremost on anyone's mind when they hear that I've invented a language are the questions: Why would you invent a language? How do you do it? The why question is difficult to respond to, since the answer isn't consciously apparent to me. The short answer is because I'm interested in language, but that doesn't tell you much.

It started with...

So I will give a little bit of background of how I got started. I started inventing the Skerre language on July 30,1994 — I was 13. The Skerre people proceeded the language in invention by a few months. I have long had a fascination with the invention of make-believe worlds and with languages. I have created numerous countries (complete with maps), TV shows, stories, alien races, and imaginary sports leagues. My creativity also extends to music, since I also very much enjoy composing.

My fascination with language started early. I took German in elementary school — I didn't learn much, but I was interested nevertheless. I took French late in seventh grade (Spring 1994) as six-week introductory class, and was so enamored with it that I signed up to take all of eighth grade. I even tried to create some conlangs in fifth and sixth grade without much success since I didn't know what I was doing. But in the summer of 1994, two things happened. First, I created the Skerre people, and I began developing their culture. Second, I went to a summer school, and learned about a semester's worth of Latin. This was the catalyst I needed because I began to understand syntax. I now had a window into a system for language, and I jumped at the chance. I had this people that needed a language, so at the end of three weeks of Latin, I decided to give them one.

Tools of the trade

However, I soon realized that I didn't really have the proper tools to invent my language. So, I turned to two sources. One was the French that I began to take in the fall. French provided a welcome alternative to English, especially in instances where I hadn't learned the Latin construction. French, Latin, and, more recently, Ancient Greek, have provided inspiration for many grammatical features of Skerre, and it seems that study of natural languages invigorates my interest in my own language. Besides taking from the languages I was learning, I broadened my understanding of languages by reading grammars. I collected grammars of many diverse languages: Norwegian, Russian, German, Turkish, and later Basque. I would read these books voraciously, and in the process of learning these languages, I would find things to add to Skerre. There was a slight influence from other conlangs. I borrowed some words from the appendices of Tolkien's Return of the King and The Silmarillion. I also picked books about languages in general, anything that I could get my hands on, both to learn about how the languages worked, and to see if I could borrow from them. It was through this that I was introduced to the discipline of Linguistics, the key to answering the question, how do you invent a conlang.

Linguistics and conlanging

Linguistics is the scientific study of language, in general, that is the study of the system of language — not learning to speak, write, read, or understand numerous languages. The simple answer to the question, how you invent languages, is merely to apply the discoveries of linguistics backwards. This, however, isn't how most conlangers explicitly think of language inventing. On a more practical level, conlangers must invent a grammar — a set of rules for their sounds, for their words, and for their sentences, and then also a lexicon, all the words and affixes of meaning. For most of my conlanging career, the focus has been on the grammar part of the language - I have only invented about 880 words in about six years of conlanging.

How does the discipline of linguistics interact with conlanging?. Actually, in the early days of my conlanging career, there wasn't much interaction. I would borrow and choose features at random from a wide variety of languages, and as a result Skerre was not terribly coherent. Also, when I was faced with a problem I didn't really understand, I would run back to "Mother English" and borrow from her. However, as I further encountered languages away from the Indo-European family that most western European languages, including English, belong to, I was able to move Skerre away from being a relex of English. One of the most important things I discovered that would impact Skerre grammar was the idea of word-order typology universals.

This idea derives from Joseph Greenberg's 1966 work, "Some Universals of Grammar with Particular Reference to the Order of Meaningful Elements," but I first learned of the concepts in Winfred Lehmann's book "The Theoretical Bases of Indo-European Linguistics."


Typology is the study of how languages resemble each other in their structural parts. These languages don't have to be related, but Greenberg and others discovered that if the verb comes after the object instead of before, that is, "I him saw" instead of "I saw him," one could draw certain generalizations about the order of the other elements in the sentence. This chart shows the six possible word orders, divided into Verb-Object and Object-Verb, since the subject need not be explicit in many languages, and it shows the characteristics of each type. For Verb-Object languages, such as Spanish, adjectives and other noun modifiers usually follow their nouns, adverbs are put after their verbs, prepositions are present, relative pronouns are used in forming relative clauses, and prefixing is common.

The sentence "I saw an old man in the park" becomes "I saw a man old in the park." For Object-Verb languages, such as Basque, the opposite order occurs: adjectival modifiers come before the noun, adverbs precede the verb, there are postpositional phrases, relative pronouns aren't used for relative clauses, and suffixing is preferred. Thus our sentence becomes "I old man park-in saw." Since I decided that Skerre should be of the Verb-Object, I promptly changed the grammar to fall in line with the Verb-Object pattern. Thus the sentence in Skerre is "E har ok kelios vio ja tanes'"

Examining English, one can see that it does not neatly fall into one of the patterns. The original theory does not account for exceptions like English, and therefore has been quite controversial. Why then did I decide to be so straightforward anyway? One reason is simply personal preference — I like the order that this brings to the system. But also more importantly I found that a number of unrelated VO natural languages do pattern like Skerre: French, Vietnamese, and Maori, for example, and such blanket patterning is not as unnatural as one might think.


Not all my ideas are borrowed directly from languages or linguistics. There are times when I innovate things myself, drawing on my own linguistic intuitions, and then I must research to see if such ideas are reasonable within linguistic possibility. For example, earlier this semester I changed the plural formation to a reduplicative prefix. This means the plural prefix is a copy of the first consonant and vowel of the word.

For instance: s'al is mountain; s'as'al is mountains. This rule is very similar to reduplication processes found in many languages, and indeed the overall rule was largely inspired by the reduplication in the present form of Greek -mi verbs. However, within the reduplicative prefix there is a simplification of initial consonant clusters. Skerre permits a large number of possible consonant clusters which I figured might be a lot to say twice, so I thought when words begin with a cluster, the cluster should simplify.

So instead of the plural *skiskias for skias, "warrior" the correct plural is kiskias. Instead of *krkr for kr, truth, the correct plural is kakr.

I now sought to find out whether this pattern occurred in natural languages. And looking in Prosodic Morphology (Prince and McCarthy 1986), I found that such consonant simplification does take place in Greek and Sanskrit, and it follows the same hierarchy as the Skerre, thus validating the Skerre rule.

Overall, this need to research languages for such a personal reason leads to a type of ownership of the knowledge that I feel that I could not get by just studying the languages. And using all the various different grammatical features that I have in Skerre allows me to better understand these processes, having worked with them on such an intimate level.

What to do with a conlang

What do you do with a conlang, especially after six years of work? One such task is to further modify and expand your grammar; another task is to further expand your vocabulary. And, occasionally, there are chances to share your language.

The first one for me was in eighth grade. A teacher there wanted to have a class where we would invent a language and learn how language and culture interact by writing a play in that language. So I offered Skerre. With the help of my classmates, we wrote and performed a play in Skerre, with subtitles provided on overheads, for my middle school. I had another wonderful experience in twelfth grade when I explained a bit of the grammar of Skerre to my French class in French. And finally, this semester I have had the unique experience of combining linguistic and compositional exercises in an independent study with Professors Jeff Runner and Sarah Higley.

With Professor Runner, I have been working to write a grammar of Skerre. With Professor Higley, I have engaged in correspondence with a "speaker" of her invented language and a "speaker" of Skerre, where the dialogue are in our respective languages, with English glosses. This exercise has been a wonderful way to further develop Skerre culture, and to give my language a badly needed road test.

It seems that when I usually tell people about Skerre it involves a story. But unlike most stories, this one isn't over yet. So I cannot tie together the loose ends and conclude. Instead, all I can say is that the story of Skerre will continue to transpire. Who knows what the future will bring — I just hope that it will be as rewarding as what has already happened.


Further Conlang References

  • Conlang Listserv Homepage. http://listserv.brown.edu/archives/conlang.html
  • Higley, Sarah L. "Audience, Uglossia, and CONLANG: Inventing Languages on the Internet." M/C: A Journal of Media and Culture 3.1 (2000). [March 17, 2000] http://www.uq.edu.au/mc/0003/languages.html (no longer available).
  • Kennaway, Richard. Some Internet Resources Relating to Constructed Languages. . (The most comprehensive list (with links) of invented languages on the Internet.)


YearMy Grade in SchoolEvent
1991-924th and 5thTook German
1992-935th and 6thEarly conlang attempts
1994(April/May)7thTook 6-week introductory French
1994(May 24)7thInvention of Skerre people
1994(July)Inter 7th 8thTook Latin
1994(late July)Inter 7th 8thBegan inventing Skerre
19948thBegan French instruction (thru 12th grade)
199913thBegan Ancient Greek

An Example of Skerre

U tita enska aka ca vik i totar i tari
U kan�a tella hok su ta cerren i kaska ra
ov u krein korska ja t�,
a ta u reta tovo sujar

U    tita enska    aka  ca   vik i   totar i   tari.
IMPF be   language path with way GEN river GEN time.

U     kan�a   tella hok su cerren i  kaska ra
IMPF leads ERG this  us  to home  GEN ones  ancient

ov   u   krein  korska ja t�
but IMPF arrive no.one at that,

a   ta   u    reta  tovo  sujar
ERG REL IMPF fear   water deep

Language is the path through the river of time.
It leads us to the dwelling of the ancestors,
but no one arrives there
who is afraid of deep water.

-Vladislav Illich-Svitych

REL=relative pronoun


Word Order Types

VO                    OV
SVO----I saw him      SOV----I him saw
VSO----Saw I him      OSV----Him I saw
VOS---Saw him I       OVS----Him saw I

VO Characteristics

  • Noun-Modifier (adjective or genitive)
  • adverbs after the verb
  • prepositional phrases
  • relative clauses with relative pronouns
  • prefix-preference

OV Characteristics

  • Modifier-Noun,
  • adverbs before the verb
  • postpositional phrases
  • relative clauses expressed without relative pronouns
  • a suffix preference

Sentence Comparison:

  • English: I saw the old man in the park.
  • VO-prototype: I saw the man old in the park.
  • OV-prototype: I old man park-in saw.
Skerre: E    har ok  kelios vio ja tanes'
        PERF I   see man    old in park"


This process is used to form the plural in Skerre. It is a prefixed formed from the first consonant and vowel of the word, as the following examples illustrate:

As in the following examples


Examples with cluster simplification

skias'warrior'kiskias(not skiskias)
kr'truth'kakr(not krkr)

Example from "Prosodic Morphology" of Sanskrit reduplication featuring the same concept:

base formreduplicated form

About the author

Doug Ball was born in Denver, Colorado, USA on August 30, 1980. He graduated May 1999 from Arapahoe High School in Littleton, Colorado, where he lived with his mother, stepfather, and younger sister. His father, maternal grandparents, and several other relatives also live in the Denver metro area.

Doug is majoring in Linguistics at the University of Rochester in Rochester, New York. When not "conlanging," he enjoys composing music (which he is studying adjunctly at the Eastman School of Music) and listening to his eclectic collection of classical and jazz.

Doug is very interested in sports. He closely follows the action of the Denver Broncos, Colorado Avalanche and Colorado Rockies. Other interests include reading (especially science fiction), writing and enjoying the humor of Saturday Night Live, The Simpsons, and The Tonight Show with Jay Leno.


  • 25-04-2000 - Creation

text © 2000 Doug Ball