Orphaned languages of Teppala

From FrathWiki
Jump to: navigation, search

These are languages that don't have a place in my current conworld because it's historically impossible for them to exist alongside the others. None of them has proper diachronics because I created these languages at a time when I didn't know better.


Early childhood languages

I did not create any organized a priori conlangs prior to 1994. For Camia I used distorted English when I was about 10 years old, of a type that changed from one story to the next and thus was never a proper conlang. I would say that all of the languages I created before #Moonshine belong together because I discontinued them all immediately when I started Moonshine and came to associate them with my childhood. By contrast, from Moonshine onward, I kept languages alive even through periods where I put little effort into them.


Beginning around age five I sometimes used distorted forms of English words and names. My parents didn't understand it and my kindergarten teacher said I was "playing with the letters". I mostly changed the vowels (including by omission) and left the consonants alone, though I turned an initial /k/ into /ŋ/ in one person's name. For common nouns, I mostly had names for numbers and a few other things. I learned by age five that English words are often spelled in manners not according to their pronunciations, so even when I made up new spellings for words, they didn't always align well with my pronunciations: chakun "chicken" rhymed with "man".

I never considered this a language since every word corresponded directly to its English word. I revived the idea in my early adult life to create a proper language, however: #He_pam.2C_e.CA.94oqaaniam.3F.

I went through a brief phase in which I coined new words in English, sometimes repurposing existing words. The only ones I can remember right now are listen, meaning "to sharpen a pencil", and conspiracy, meaning "role reversal". Soon, I began coining words that didn't resemble English words, but I still wrote them down as if they were. Some of these were essenblam and glomeran. I know there were more such words, but essenblam was always my favorite. I knew by this time that essen was German for "to eat", and that may have influenced me, but on a conscious level, the words I made up were entirely a priori.

These words perhaps show that my tastes have changed somewhat: I strongly favored voiced stops when I was young, and I can remember no words with /p/, and only one apiece with /k/ and /t/, both of which strongly resembled words I had seen elsewhere. Even so, the /bl/ cluster has been a favorite of mine for a long time and remains so today.

Another word I created was emblem "hallucination; vision", which I considered to be a new coinage even though I was aware it was already an English word. I pronounced both e's with full vowels. I remember seeing the word BLEM on a sign and thinking it was among the most beautiful words I had ever seen. This may have influenced me.


Camian was never really a single language since I changed it from day to day. One day Camian was Old English flawlessly preserved since 700 AD, the next its vocabulary was half Spanish loanwords but with so many diacritics stacked on that I could only spell it properly on paper. I felt that a language for a powerful nation with superheros would need to have a similarly overpowered phonology, a stereotype that has not fully left me, as I still feel the same way about Leaper.

I may have gotten the idea for the diacritic sea from seeing names spelled in a Bible where every vowel and many of the consonants had extra diacritics on; I believed at the time that they were accurately transcribing names from Hebrew and other languages, and thus that there were languages with dozens of vowels and even more consonants.

Early on, I came up with the name Jafa (/jafa/) for the language, although the Jafa people were actually a culture living in Camia rather than a linguistic group. I had learned of /j/ being spelled j in early childhood, and even mistakenly attributed it to Spanish at one point, perhaps because I confused it with IPA. I wasnt aware of the IPA's existence until I was about 12 years old, but I think I came across it in dictionaries, and this may have led me to confuse word pairs like bayou ~ bijou even in English. Similarly I still mentally pronounce the Japanese city Hakodate with a final silent -e, because even though I had all the vowels right I had somehow generalized the final silent -e of French and English to Japanese. (I may have also thought it was silent in German.)

I remember final-vowel alternations such as waro "fight" vs waru "deadly fight", an idea that I revived a few years later with consonants.


One of the few words I remember is clarasarapel. Clarasarta "light source" was my word for the sun, and clarasarapel meant "east-west; following the path of the sun". The lack of a t in the second word was simply my mistake, even though I was writing on paper and therefore it wasn't a typo.

I was fond of pseudo-intellectual Greco-Latin coinages like califloron "beautiful flower; flower" and astet "don't do that", although because I changed the language so frequently I sometimes used them and sometimes didn't. At one point I called my language Neo-Latin, even though this made little sense for a story set in the distant future. But I may have turned against Latin within mere months of this stage, as I remember rejecting the Latin word robur "strength" after reading that it could also mean "oak".

Relationship to English

Camian was never identified with English. Camians sometimes spoke English with an accent, as in "all right, all right, i'll giff you a penny" said by a Camian man who was robbing a Wamian store.

I wrote a story, o3dr, with words such as uberinvitet and discuberaton in what I considered to be a reform of English, but I came to identify this as "Camian" at some point even though it was illogical. This was a very late stage of Camian, and I think I may have already started work on Moonshine, which at one point had a series of words ending in -āteń, which may have pushed me to use -aton (no i) for standard English -ation.

There was also a stage with words like "fabicola" that I also considered English.

Dialects of Camian

I was fascinated with the concept of dialects early on, and from my earliest writing there was a distinction between the standard Camian language and the regional dialects, of which the superheros spoke East Camian. I carried this through even to the phase where I had decided to make Camian a very conservative language, and this is the origin of the split between Tarise and Amade in my later work.


The only common trait among the various iterations of the Camian language was that it was powerful. If I had finished it, I would have had Camian be the language that all of the other nations' languages borrowed their scientific terminology from, even nations that were politically hostile. Resisting change was also a sign of its power, such that it could stay the same for 1300 years; however, I wobbled between liking that idea and rejecting it, and some iterations of Camian were derived from English but had changed very quickly instead.


Also not a single language, "the Wamian language" is essentially a term for any language consisting of features I dont like, which goes to the nation of "Wamia", which itself is not a single nation but rather a term for any nation fulfilling a similar role.

When I was 10 years old, Wamian was essentially English spoken by a particular boy who had an apparent speech impediment. So severe was his speech impediment that he didnt replace all /l/ and /r/ sounds with /w/, as stereotypical toddlers did ,but instead replaced all /l/, /r/, and /w/ sounds with /b/. (As some toddlers do.) The boy's name was Gary, but he could only call himself "Gabby". Thus just as Wamians "couldnt even say their own country's name", he couldn't say his. I had planned to reveal in the storyline that Gary in fact did not have a speech impediment because all the Wamians spoke like that, but I never got around to it and I started making fun of Gary less often as I matured.

Later I seem to have decided that Wamians spoke a language that had changed very rapidly over time, whereas Camians spoke a language that was extremely conservative. If planet Teppala had been settled by the Proto-Germanic people, Wamians today would be speaking modern English, while Camians would be speaking something resembling Gothic or perhaps even a language still intelligible with proto-Germanic itself.


Wamia was identified as being settled from planet Earth hudnreds of years later than Camia was, and its people retained a close connection to those of planet Earth. At the time, I had said that there were only two languages on Earth: Camian and "a symbol language" which had no spoken form. This is a contradiction, though, since Camia and Wamia spoke mutually unintelligible languages, and Wamian had only recently come from planet Earth. If I had noticed this, I would likely have corrected myself by adding Wamian as a third language.

I never explicitly identified Wamian with English either. But it occupied the same role in the sense that it was the "corrupt" language I rebelled against, with Camian being the pure form, preserved against outside influences. When I needed to represent specifically Wamian speech in my comics, I left it untranslated whenever Camian was represented as English.



I know that I disliked French, and I even made a "French" section of planet Namma populated by people whose men only stood as tall as their women's hips,[1] which I later renamed Repilia. Repilia was a large area of land centered on the Equator with a hot climate consisting of both jungle and dried out savanna-like areas.

I associated French with feminine things like ballet, lingerie, and perfume, and therefore my calling Repilia "French" meant that I intended it to be an area of the planet where females were in control of everything and males woke up early every morning to do unpleasant chores while the women giggled and enjoyed themselves. It was a "Jack Spratt" society; women took all the food and, when they felt like it, gave a tiny amount to their men. (This nursery rhyme bothered me as a kid.)

The idea of making women taller than men had occurred to me when I was about five years old, but I did not build entire societies on alien planets; I just played with ideas for their own sake. I remember having a sport called "girling" where a boy tries to leap over the belly of a girl who's lying down, and in most cases can only just barely make it.

For the most part, Repilian women completely ignored their men, as if men were subhuman creatures and the best among men could only hope to be considered equal to house pets. But sometimes, the women had temper tantrums and argued with each other, and in Repilia a man trying to pick a side in an argument between two women would stand as much chance as a mouse in a fight between two housecats. Thus men had to scurry and hide whenever they heard an angry female voice, because even if that woman was not angry at him, any man would make an easy and unsympathizable victim.

Even though I was a feminist myself, and female characters in my stories nearly always appeared in positive roles, the people living on the Equator of Namma were a female dominated society "in all the wrong ways". Repilia could be seen as being the embodiment of my childhood "girls are icky" mentality which had otherwise faded away as I began to be interested in girls (I did not consider myself gay at the time). I had fallen for a story once where a girl told me that in French, word for unpleasant or embarrassing objects, such as toilets, are always masculine, and words for beautiful or impressive things are feminine. I remember being upset when I noticed that France had a city called Brest sticking out into the Atlantic Ocean, as if the entire city were somehow a celebration of the female body that I was not allowed to participate in because I was a boy. This is the kind of feminism that "French Namma" believed in.

By contrast, in the female dominated societies of the cold climates, such as #Moonshine, men obeyed women willingly and were only a little bit smaller rather than a lot. (Mappa at one point refers to her boyfriend as "a little Sakhi boy", even though he was tall for his age, because he was still shorter than most girls.)

The real French

The real French also appeared briefly in my stories; see below.

Alien languages

When spiders from planet Theta captured the superhero boys, they sometimes heard the spiders' language. I made no attempt to develop a serious language for the spiders, but I did specifically write that the spiders spoke only one language and that they had only one major city on their planet. This is considered a trope of bad science fiction, but I excuse myself because I used it to show how the spiders were different from humans not just physically but also culturally.

Nevertheless, the spiders' language was poorly developed. It was not even internally consistent; in one early story, a spider said "çëJBHHLKHJ. THB JHUIKIKJHKG HUG ç HO ì KJ YTTVJ¬ç EARTHNE," whereas in a later story, a spider said "blek farg offa" and one of the boys was modestly competent in their language. Also in the first story, two spiders spoke Jafa even to each other.


In 1993 I was working on a novel that I never finished, as a side project to my comics. In that story, two young boys visit planet Namma and meet up with a smaller boy who speaks a language that only one of the boys understands. (This was a plot hole that I didn't notice until much later, which is ironic because I spotted the same plot problem in The Time Machine movie.) The Namman boy's language, which I at the time called Danta,[2] was superficially similar to Pabappa.

I saw Danta as a stereotypically childlike language, as the villagers the boys interacted with were mostly children younger than themselves, who saw the boys as strong protectors. I was eleven years old at this time, and this was the first story I wrote where the main characters were children much younger than me rather than kids my age or my own self-insert character.

All the words were short, vowels were ample, and there were no guttural consonants at all except possibly word-initial /h/. The commonest consonants were /p/ and /b/; my tastes have not changed at all. The sentences I remember are

Ble Danta tapa.
Jo baba Treba? Har tapa ipo?
Mana laru home [placename].

In one draft of the story, the two boys figured they would need to give themselves new names because the villagers wouldn't be able to pronounce the consonants of their birth names. Thus, these kids were stereotyping other kids who just happened to be younger. Adults also made the same mistake: when the Namman kids arrived back on Teppala, the adults in charge of the missions saw them as so helpless and simple-minded that they would be more of a hassle than a help on the other kids' missions. They were wrong, as they soon realized that some of the kids had mental powers far exceeding those of any adults. I had planned to make the language more complicated, and I think by that I meant grammatically, meaning that I wanted to keep the baby talk phonology while changing the grammar to what I felt a "smart kid" society would want to use.

In-world history

I came up with the idea of humans living on other planets when I was just six years old, and it never occurred to me that I had no explanation for how they got there. By the time I was ten or eleven I decided that they had long ago come from Earth, and set my stories about 2,000-4,000 years in the future. Later I reduced the time interval to just 400 years, and it still didn't occur to me that a language could not have evolved from a known Earth language into Pabappa in just 400 years, but when I did realize the problems, I expanded the time interval back to about 2,000 years again and did not need to change any other important plot elements because that intervening time period had been primarily empty all along. Still later, I decided that the Pabaps' planet (Namma) had actually been settled tens of thousands of years earlier, long before civilization arose on Earth, and that although Earth was the original homeland of man, it was not the planet which developed the first advanced human civilizations. The mechanism for how humans got to other planets was still never set in stone, however.

Pabappa words with /k/

Though the snippets of the language I used in the book had no /k/, I did create quite a few proper names with /k/. Early on, the boys met up with an aggressive man named Nanuko who kicked them out of his village and promised to track them down if they caused trouble in any of the neighboring villages. One of these nearby villages was Kotona, whose men attacked the boys as they attempted to board their spaceship. There were also three young girls: Anitak, Kara, and Kamea, showing that the use of /k/ was not restricted to just the names of adults. I coined at least some of these names during a time when I was being influenced by Māori, but others may have existed even during the time when the Danta language seemed to have no velars.

Abandonment of Pabappa

After I gave up on the novel, I drew a few more comics and wrote plotlines for comics that I never got around to drawing. As I grew older, I moved the age of the self-insert character higher to match my own. Eventually I decided he should have a girlfriend and that it should be the girl he met on one of his early trips to Namma. I seemed to have felt that it was inappropriate for a teenager to be speaking a Pabappa-like language, and therefore moved her from the Southern Hemisphere of Namma, where most of the main characters had been young children, to the icecapped North, where men were tall and women were taller, and both spoke a language that I later came to call #Moonshine. Still, the stereotype of a deceptively simple language hiding secret complexity was in my head and I kept it there for the next 15 years or so while working on other languages.


My next a priori language was "Manni", spoken by a series of tribes of people in Camia who superficially resembled Native Americans, particularly the Algonquians who were native to the area where I lived. I actually pictured them with green skin, but I seem not to have taken this idea seriously, since when I drew Manni characters in my comics they looked the same as the white characters.[3]

Their language had voiceless nasals and I think four vowels. However, for some reason I decided to create an alphabet for this language and write the Manni words entirely in that alphabet, and when I lost the piece of paper that told me which letter was which, I couldn't even read my own notes and thus my entire memory of the language was lost. The only word I remember is man(n)i itself, the name of the language, which meant "beautiful", although it may have been a subconscious loan from Algonquian manito, or perhaps a blend of that word and mana. I mispronounced this word, giving it a Spanish-like pronunciation, and didn't realize my mistake for more than 20 years afterward.

My work on this language coincides with the brief time in which I introduced a character named Manitoman, who resembled Captain Planet in that he was an adult male superhero who was often summoned into battle by a team of child superheros, and also in superficial ways such as my picturing him with green skin (sic; not blue) and having a similar hairstyle. However, unlike Captain Planet, Manitoman had his own superpowers, and did not become stronger when summoned to battle by the kids. In fact, I soon reduced Manitoman to equal status with the kid superheros, saying that he traveled with them rather than leading or being led by them. By this time, the main "kid" characters were in their mid teens, because at that time I always made the main characters the same age as myself, so having a full-grown adult among them would not seem as strange to outsiders as it had a few years earlier when Nanuko and Bob were hitching rides aboard spaceships piloted by children 10 years of age and under.

I intended to pattern the Manni religion after that of the Algonquians, and (this was 1995 or so) I found a Native American man online who told me that their religion was similar to Christianity, which actually pleased me. I was still outwardly Christian but was feeling conflicting emotions regarding some of its basic ideas, and to find out that a completely unrelated religion that appealed to me shared much in common with Christianity made me think that I could follow both religions without any conflicts. However, I soon lost interest and returned to mainstream Christianity for a while, and the Manitans became a dead-end storyline in my work, which I eventually explained away as being the relic of an immigration from planet #Namma.

the "ice cream" language

Briefly I worked on a deformed English used by the teenagers in FILTER. it was modified English, just like most of my previous languages had been, but it was separate from those languages. It actually occupies the same place in the story as what later became #Moonshine. All I remember are a few words like boid "boy", boi "body", and gird "girl". I seem to have formed several derivations of each word, though, intending each word to be part of a word family rather than just existing by itself. The sound changes were entirely irregular.

This language was also a spelling-reform language, in that, despite being only slightly different from English, the spelling varied starkly from English. For example the word for "wound" was spelled oön despite being pronounced /wun/, essentially the same as English except for the loss of the final -d.

Since I had already started two a priori languages by this point, regressing to distorted English afterwards seems odd to me now. I think the explanation is that I intended this language to have evolved from English, but that I simply didn't understand how languages change over time and therefore produced something that was much lower-quality than what I had come up with when I was younger and worked on static a priori languages.

By this point, I had decided to purify the language of anything having to do with warm climates, and that meant getting rid of the Spanish loanwords. I hated French too, and even loanwords in English that came from those other languages. This is why I identify this language with ice cream and candy, and therefore with #Moonshine.


French reappeared at this point, this time as the real country rather than just an appellation. When the female-dominated army was able to take over the entire planet, they taught everyone their language, but all they managed to do in France was to get the French to change the word chaud "hot" to chau. I had also entertained the idea of making e (pronounced /ə/) the French word for anus, saying that French culture was so degenerate that the most basic word in the entire language was the word for anus. As a tribute to this idea, I've used ê (pronounced as IPA /(j)eˤ:˩/) as the word for anus in Leaper, though a fully-developed Leaper language would have at least dozen other words consisting of just a single vowel, and this could be arguably analyzed as being more than one phoneme.

Moonshine (1994)


This was a language I created in 1994, and was my first true a priori conlang. It had no ancestors or daughters, and I remember considering it to be eternal. It was very compact, and I remember writing "Power speaks a language that works in its written but not its spoken form". That is to say, Moonshine was so full of homophones that it was literally impossible to communicate in Moonshine out loud without being ambiguous. It could be compared to a hypothetical Japanese in which the kanji are used freely with their short pronunciations, meaning that, for example, the syllable /shō/ could refer to any of more than 70 kanji, which nevertheless would be perfectly understandable when written down.

Phonology of Moonshine

Early signs of my love for babylike speech were visible in Moonshine. On my very earliest sketch of the language, I gave it a phonology with no dorsal consonants at all: there were labials and coronals, and that's all. I had no conscious reason for this choice, but I know that I held to it because I remember at a stage when I still allowed loanwords that my word for clock was očo.

Early phonology

In the early days I created words by babbling as a baby does, usually imagining in my mind the thing I wanted a word for or in some cases (for verbs) acting it out with my hands. For some subconscious reason, most of the words I created were vowel-initial. Remembering as best I can, I had īsa "girl", ēbo "boy", èpèp "to drink", amñ "to eat", ūmu "water", and āme "love" among the very first words I created. I seem to have had about six distinct consonants, if onomatopoeic words such as amñ can be ignored (it was the only word with /ñ/). This is a very late post-analysis, though, as I used an extremely narrow phonetic transcription at the time and didn't think about phonemes. I even went so far as to write down fractional differences in how open my mouth was, so that for example I had two /ambo/'s, one slightly more open than the other. This was based on my mistaken assumption that there existed languages in the wild that did make these tiny distinctions phonemic.

The six consonants that I today analyze this stage of the language as having were /p b m w t n/. I remember making a few words with the cluster /nd/, but this could be analyzed as /nt/ since I never had a contrast between /mp/ and /mb/ either. Some words had [s], but only when it bordered /i/. /t/ > [s] before /i/ is apparently a common sound change in languages with simple phonologies, so perhaps I was pulling on something that made sense subconsciously that I didnt discover until much later.

I wrote with a five-vowel orthography, using accents for long vowels (and later for tones), but I think mentally I was pulling from a four-vowel setup of /a e i u/, where /u/ is lowered to [o] when not adjacent to another high vowel.

Had they not been on paper that got thrown out over the years, I could probably spend hours just poring over the complicated derivations I used to get from basic words like these to ideas for concepts like "lab rat", "brainchild", and so on, since I know that I never used babbling for complicated things like that.

Later phonology

I slowly drifted towards a more complicated phonology over time. I seemed to want to make all of the onomatopoeic words canonical, and therefore added fine distinctions such as labial vs labiodental and voiced vs voiceless nasal. Eventually I ended up with a setup of five POAs and six manners of articulation, for a total of 30 consonants.

There was a full series of bilabials and a full series of labiodentals. Likewise there was a full series of dentals, a full series of alveolars, and a full series of postalveolars, with no consonants further back than that. However, I slowly warmed to the idea of pushing the postalveolar series (/č ǯ š ž ñ ṇ̃/, the last being a voiceless nasal) into velars.

Another trait that Moonshine had in common with baby babbling was that there were no /r/ or /l/ sounds. However, this was because I wanted the language to have a perfectly symmetrical phonology, and could not figure out a way to work liquids into the system. Later on, I added them anyway, saying that /r/ was the opposite partner of /l/. All in all, the language had 32 consonants, which is far above average, but I'm not sure if I had any reason for making so many consonants other than that I felt I needed them for symmetry. I didn't explicitly identify the language as having a simple phonology or resembling baby talk so the large consonant inventory never bothered me.

There was no phonemic /j/ or /w/ in the language. I seem to have had diphthongs in a few words early on, but later on I flattened them out into long monophthongs. Later still, I decided to introduce /w hʷ/ by saying that the voiced and voiceless labiodental nasals had changed into these.

Moonshine was also tonal, but I literally didn't know what tones were, even if one might think it should have been obvious. I seemed not to realize that it was possible for people to change the pitch of their voice while speaking and thought that it must have been a metaphor for some concept I didn't fully understand. Later, I abandoned the tones and repurposed the high-tone vowels as diphthongs, meaning that there was exactly one diphthong for each pure vowel.


Almost every root, even those for complex things, was monosyllabic. I remember wanting a language that looked "modern" and compact, and I remember holding up Hungarian as a model to follow.

Consonant clusters were allowed, and in fact I decided early on to have no phonotactic rules at all, such that a word like vgt /vɣt/ "to injure" was allowed, but on the whole I seemed to prefer to have a vowel/consonant ratio close to 1.

Moonshine thus was very compact. Nearly every translation I did was shorter than its English original, sometimes by a lot.

Nevertheless, I spilled a little extra here and there, to give the language variety. I think the very first bisyllabic word with a monosyllabic English translation was ṃetit "to ask", and that was only because it was a compound something like ṃe-t-i-t "what [you] do when (t) [you] want to (i) finish (t) thinking (ṃe)". The morpheme order was commonly the precise opposite of English at the time. I remember writing in a comic book "ádas benuṗunūsītsén" (the spelling may be slightly off) and "ās vasapāmsītsén", thus Moonshine did sometimes have long words. But the English translations of these words are far longer than the Moonshine originals. The first, I think, means "one who wishes that those who are in power should remain in power" and the other means "supporter of heartbreakers" but I know it had some metaphorical meaning that I dont remember.

Moonshine culture

I always identified Moonshine with cold weather, imagining its population as somehow consisting entirely of teenagers and children eating ice cream outdoors in the middle of a blizzard. It was the language spoken "north and east of Russian". I remember distinctly being disappointed when I realized that Russian was totally unlike Moonshine since it seemed stand in the way of a trend of languages becoming distinctly more Moonshine-like as one moved east in Europe. However I soon learned that not just Russian, but all Slavic languages in general, had those characteristics I disliked.

Later, I realized that Moonshine was a much better fit for Greenland than for Russia, since the Moonshine speakers are most notable for living in climates with cold summers, but not extremely cold winters, since their habitat was mostly coastal.

I also associated humid weather with heavily inflected fusional languages, and said that Moonshine was spoken "in the ocean" because it so far exceeded even the most fusional of all languages on Earth, at least in its early form. Though I never really got around to making the fusional grammar, and I soon introduced grammatical ideas that made fusion impossible. The ocean imagery is a better fit overall for Ižda Mir, which occupied the same cultural position in my stories but was not the same language.

Tropical settlements

The Moonshine speakers lived only in cold climates, and most of them lived in climates that were too cold for ordinary humans, similar to the Camians at certain stages of my writing. On two separate occasions, I thought about extending Moonshine to a tropical rainforest habitat, but I wouldn't have changed the language to "feel warm". This was similar to how, a few years earlier, I had suddenly become obsessed with New Zealand but didn't bother incorporating Māori into my world because the NZ settlements were all run by kids from Namma who spoke Pabappa.


The Moonshine speakers were very feministic even in my earliest conception, but this had no reflection in the language. However, the word for female was a and the word for male was b, which meant that sometimes an extra vowel needed to be inserted into a word for a man or boy, which made me feel happy. (Note that my current Icecap Moonshine language, which I consider to be a continuation of the 1994 language, takes feminism to such an extreme that I suspect most people reading the grammar would interpret it as a parody.)


The name had nothing to do with alcohol, but rather with its speakers' apparent cultural habit of being nocturnal and avoiding the sun. I seem to have subconsciously associated it with icecream, which may come from a single dream I had one night where Old English had been flawlessly preserved on an island off the coast of Massachusetts and the population of that island seemed to consist of teenagers and children who were very fond of eating peppermint-stick ice cream cones outdoors even in cold weather.

Moonshine alphabet

I did my earliest Moonshine writings using a modified form of the Greek alphabet, even though the Greek alphabet was poorly suited to the language and I think I needed to borrow some letters from the Roman alphabet even then. (e.g. I have no idea what my /ñ/ letter would have been). Digamma was in the alphabet. There was no way to type this with the computer technology of the time, so I did my work on paper.

At one point I created an a priori script for Moonshine in which any two-syllable VCV word could be represented in just one character. I think I was inspired by Hangul since I remember the elements for /b/ and /p/ being straight lines (I dont remember what I used for /m/). I know that I developed this script into a fully functioning form because I remember once changing my screen background to a word that meant "my computer" in Moonshine in the a priori script. However, as my notes for the script were on paper, I soon lost them and, just like with Manni, my knowledge of any notes I had made in that script were lost. (But I rarely wrote in the script.)

Before long I switched to a Roman script, although I refused to use digraphs, meaning that I had so many diacritics that I still had to keep my notes mostly on paper instead of typing them up. I used the spelling x for IPA /ʒ/, but spelled it ž just as often. I had taken the x idea from a Lonely Planet guide to Basque, which was apparently incorrect. Later, when I made the sound into IPA /ɣ/, I standardized x and considered it a variant of the Greek lowercase gamma; in handwriting I used a true gamma for the sound when it was functioning as a standalone word meaning "therefore", and used the x otherwise. I did something similar with /h/: it was spelled h within words but with a character somewhat like ɧ when functioning as a standalone word meaning "because". This too was only in handwriting. Later still, I flipped the gamma upside down (since it resembled an X just as much either way) and then began using the normal gamma as a standalone variant of /v̥/, which meant "using; instrumental".

To help me computerize my work I developed an alternate orthography that used digraphs even though I still preferred the cleaner "one letter, one sound" idea that had the sea of diacritics. In this script, only the acute accent was necessary for vowels, and many consonants were written with digraphs, although it was still necessary to use diacritics even then to disambiguate the four /m/'s and six /n/-like sounds. It looked vaguely like Irish Gaelic, but yet had a distinct look of its own, with words like bhaoṃí replacing the "clean" version ḅòṃī. (The dot below the M indicates that it is voiceless; I think I had to resort to some other symbol on the computer but I don't remember what it was.)

Moonshine grammar

The early grammar of Moonshine was heavily fusional, and I remember starting huge spreadsheets listing various inflected forms of words, with the goal of having so many forms of each word that I would need to look at my spreadsheet just to remember all of them. It was chaotic on purpose, and I remember listing "UNKNOWN" next to some forms to show that even the world's foremost authority on the language could not quite figure out what the Moonshine speakers were saying sometimes.


Moonshine was a mathematical language, meaning that new words were oftne derived by mathematically adding two words together. Each phoneme had a defined number, and this value never varied, but I never came up with a methodical way of adding words together. For example, k + vasabadas produced fasabadas, but pad "play" + i "not" produced pēd "work". That is, any letter within the word could change; I never had a rule that stated whether letters at the edge of a word or letters in the middle of a word were preferred and used both methods randomly.

The short, low-tone vowel /e/ had the value of 0, and was automatically inserted to break up certain consonant clusters, although I didn't apply this rule faithfully. The short, high-tone vowel /è/ had the value of -0, though I'm not sure if I ever developed a rule for which one to use.

The voiceless bilabial stop /p/ had the value of 0, and was automatically inserted to break up vowel sequences. The voiced bilabial stop /b/ had the value of -0, and was automatically inserted to break up certain vowel sequences. Here again, I'm not sure if I ever had a firm rule of which to use. I know that my word for boy was peb, consisting of three 0-valued phonemes in a row. For that matter, the entire bilabial row /p b ḟ ṿ m m̥/ all had the same value of 0. I dont think I ever used anything but the stops as automatically inserted consonants, though.

Even though the default vowel was /e/, short low tone /a/ was more common than by chance; I even had a word /vasabadas/, with 4 of them in a row. This was probably due to a subconscious preference for IPA /a/; beginning around 2000, /a/ was always the commonest vowel in my languages and many words similar to /vasabadas/ appear.

I used negation to form opposites, and some words could have more than one opposite. For example, k "God" had the value +4, and ġ "Satan" had the value -4. But there was another word in this family with the value of -4: "to worship". That is, since God is worshipped, his opposite is worshippers. The sound ġ was highly taboo to me for this reason, and I often substituted or (voiced labiodental) for it early on. Likewise, the creation of "fire" led me to immediately create "fire extinguisher" because I was afraid of fire at the time.

A few words were homophones. For example, a meant both "female; girl" and the agentive suffix. The agentive suffix was always followed by the nominative ending -s, producing -as, and to denote a female in a career I used -asa, which is an unusual choice since it is a noun that does not end in a nominative case marker. By coincidence, the a sound also had the value +4, which meant that the word for God did not have a vocalic form since it would have collided with these other words.

One longstanding problem in Moonshine was that words were mostly written with mathematically positive sounds, but voiceless sounds were positive and voiced sounds were negative (except /l/, which I added very late). Thus most nasals were negative and voiceless nasals occurred far more often than one would expect.

other gramnmar info

I had morphemes that served both as pronouns and independent words: ā "I", ē "you", a "she; woman", b "he; boy", etc. The only grammars I knew were those of English and Spanish, so I didn't have very many interesting ideas for Moonshine's grammar apart from the heavily fusional setup I abandoned early on.

Perhaps because of the mathematics, I ended up with a non-concatenative grammar, which could be considered fusional in a sense but wasn't at all what I had originally planned on. Furthermore, most morphemes didn't merge with the others around them.

I vaguely remember a morpheme /d/ or /ḍ/ that meant "a kind of". I would put words through a certain gradation that moved them up to a broader meaning, add the "d" morpheme, and then reverse the gradation, which because of the mathematics almost never ended up still having a "d" sound. The combination of these processes made words that meant "similar to". That is, a kind of thing that is a member of a larger set including the original object is something that is similar to the original object. Thus, word derivation became very opaque and many words arose from very few.


Although I had access to a computer, I kept my dictionary on paper, which meant that the dictionary was not in alphabetical order, but rather in the order that I came up with each word. In part, this was probably due to the large phoneme inventory of Moonshine, especially the nasals, of which, in the early form of Moonshine, there were ten phonemes sharing only two Roman letters: /m mʰ ṃ ṃʰ ṇ ṇʰ n nʰ ñ ñʰ/, where the underdots mark out labiodental and dental consonants. Later I moved the palatals to velars, but this would have been no easier to write on a computer.

Some years later, when I copied the English side of the paper wordlist to a computer, words that had been created on the same day in the original Moonshine ended up being next to each other alphabetically in the new language. This is why the words for "housewife" (wurop) and "urethra" (wupurop) resembled each other so much in the early wordlists for Pabappa and other languages that I derived from this old paper dictionary. (I believe these words have all been sifted out by now, though; not because they were in any way illegitimate but because most of the word-generator vocabulary turned out to be difficult to derive from the parent language, Play.)

Derivation of Moonshine words

I didn't seem to create compounds; I used the mathematical grammar to create new words which seemed to be atomic, even for complicated concepts. Thus, there was a basic word for "supermodel", "lab rat", and "brainchild that helps inventor". I also remember a word fdi for "to time travel". (This word is probably missing some diacritics, because I used this in a video game where I didnt have access to an expanded keyboard.)

It didnt occur to me at the time that in a fully developed language this kind of setup would be impossible (although one conlang, Yiklamu, tests it by having 90000 individual atomic roots). Any of these words that were on my paper wordlist were carried down into my later conlangs, which means that even Poswa more than twenty years later started out with atomic words for "supermodel" (plažalla) and "lab rat" (webiam) which are not visibly derived from other words in the language. However, I have since removed these because I've created unique Poswa words that I like even more than the inherited ones. (I seem to have thrown out the brainchild word at some point, even before Xap.)

I was fond of oddly specific words, particularly words that seem to have a missing piece that can only be filled in by forming a compound. e.g. as if English "bad hair day" were broken up as "bad" + "hair day" and then a new word meaning "hair day" were formed.

There were a few compounds in the language, such as lōpét "Tree of Life" (intending to be the tree in the Garden of Eden). But it could be said that the few compounds I had were all proper nouns.

Infixes and circumfixes

THere were a few circumfixes, such as n...h "composed entirely of" and s...s "can _ have a ..."? This is one feature I have copied into Poswa, although they are pseudo-circumfixes in Poswa formed by adding static suffixes to changeable suffixes.

Word families

I sometimes used the mathematical grammar to produce an ablaut-like process that formed word families. For example, I had pad "play" and pēd "work". Inversion also took place: another word family consists of pād "clean" and bát "dirty". Thus, every phoneme had an opposite. The two "pad" words were not related to each other, and I didnt consider them a pair .... when I later made a pun based on /pād/ I didnt draw /pad/ into it.

I started work on a classifier system, which could be considered technically a type of compound. The only classifier I used often was -m̃ (that is, a voiceless /m/) which meant both "body part" and "place". However, I think I remember deriving other "classifiers" such as , which may have meant either "army" or "country, nation, place of life" ... I dont remember. There was also h/ō "human". It could be said that there were no classifiers, just a series of one-phoneme words. I did once create mek "Heaven", and tūk "Bible; holy book", where k meant "God", showing that the placename classifer m could serve as the root of a word if it was padded with another classifier, and that that classifier could also be an independent word.

Earlier I had used dema for God, and insisted that "dema is not a word and cannot be inflected", and never really worked out a solution for how to use dema in a sentence without making the context unclear.


I liked inflection, but never really got around to deriving much of it for Moonshine. All nouns had a nominative suffix (after consonants) or -s (after vowels). The day I derived it (in gym class) I created an alternate form -us/-h which could be substituted for any reason at all, and was equally correct. (I liked making things optional to an extent not seen in natural languages; at one point there were two genitives, and à, also fully interchangeable.)

Because the nominative (+6) and the genitive í (-6) were mathematical opposites of each other, they canceled out in compounds and therefore nominal compounds had no suffix after the first word. I didnt follow this rule early on, and early compounds had two nominative endings at the end of the compound instead, which combined to form a single new ending -n̥ (voiceless /n/). I think this nominative case-stacking setup would be an interesting feature for a conlang to have, not necessarily related to math.

I originally intended Moonshine to be fusional, but apart from the mathematical combinatorics, it was actually agglutinating.

Aesthetics and goals

Despite the handicap of a paper dictionary (which my mom eventually threw out)[4] and a naive phonology, I considered Moonshine my best work for a long time, and it is the only one of my childhood conlangs that I've revived. I had such high standards, though, that the lexicon was never more than a few hundred words at its peak and it is lower than that now. In fact, when I derived words for #Asup, I had to first derive them for Moonshine and then carry them through. These were not sound changes, but rather a series of formulas meant to cipher one language into another. I still did not seem to know about true sound changes and would not for a few more years.

As I grew older, I associated Moonshine with my early teen years and came to look upon it with a similar type of nostalgia that I had acquired for my early childhood. I still referred to myself as a boy up until I was 17 or so and did not become comfortable calling myself a man until I was in college.

What set Moonshine apart from later conlangs was that Moonshine was a SpeedTalk-like conlang, where all words were as short as possible, and in some translation tests I managed to achieve 4X the compactness of the English translation. For example I had a word that meant "Power, the only boy and the laborer of the group, whispers his battle plan to the girls." In my later conlangs, the words had more room to grow, and I associated those long words with disobedience, and because Moonshine was a feminist society, I associated disobedience with male power structures poking through the feminist shield.

I did not understand sound changes at all at this stage of my development. I had only occasional access to the Internet, and this was long before Wikipedia and other such sites existed. If asked, I might have said that sound changes usually follow patterns but often are random. I even came up with an idea to turn Moonshine into an all-vowel language, saying all of the consonants had simply disappeared over the years, leaving only the vowels and tones.


I think I abandoned Moonshine in fall 1997. The last word I created may have been ṃùb "both".

I just remembered this language, which was meant as a foil for Moonshine. It was related to Moonshine, but was unlike Moonshine in almost every possible way. "ab ensp baonom" is one sentence I remember from Bé; it's the first three words of the Lord's Prayer.

My work on Bé was actually older than most of my work on Moonshine, and the Moonshine it was intended to contrast with does not really resemble the Moonshine of my later teen years. For example, Moonshine was originally intended to be a very conservative language, such that its speakers could understand words written by their ancestors of 5000 years ago, whereas Bé, like "Wamian", changed so rapidly that parents could barely understand their own children. However, as I moved towards developing Moonshine as a language for its own sake instead of representing a particular culture on a particular planet, I lost interest in the ideals of preserving the proto-language and essentially turned Moonshine into a timeless language with no parent language at all. Although I've revived Moonshine as a traditional language whose ancestor is Khulls in my newer work, my goal is to recreate a language that resembles the Moonshine of my late teenage years, not the Moonshine of my early teenage years.

Wild languages

North and east of Bé were languages that were only on the borders of being humanly usable. These were truly primitive human societies that were only just barely getting by, and I at one point had the superhero say that they were "less intelligent because they didn't think together", stopping just short of describing them as mentally inferior by nature. (This boy had been assigned a mission entailing the genocide of all the races of Manitans living within the southern half of planet Namma; his school was using the term "Manitan" incorrectly. As always, the child superheros only did bad things when adults pushed them into it; after a few hours on Namma, the boy decided not to start the genocide and instead he befriended two young children and came into conflict with two adults.)

I created a lot of these languages while playing outdoors, and never wrote them down. I associate each of them with particular experiences, e.g. playing with a ball (Loop & Pocati), or sitting by the sandbox (Ihhai). I was a teenager but we had kept all of the play equipment in our yard from when I was younger, and I still went there to get exercise.

Like Bé, Wamian, and "French", I consider most of these to be hatelangs. But some, like Bare, were originally intended to be "good" languages but I just never properly developed them.


This was a language in which all words, or at least all nouns, had to have 4 consonants and, at the time, also 4 vowels. It was thus extremely inefficient, the very opposite of Moonshine.

I've revived this idea for Repilian languages/Owl, though in Owl not every consonant pairs with a vowel and the surface realization of a word is often wildly unlike its spelling (e.g. /npwamtk/ could be [bɔŋkk] or possibly even [bɔŋk]), whereas in Umunisses my plan was to have the words pronounced as spelled.

It is possible that I will abandon the idea as narrowly stated, having been told by others that a so-called quadriconsonantal language is untenable (though other conlangs using it exist), but maintain the superficial appearance of a quadriconsonantal language with such languages as Palli and the other Trout languages, which attach both prefixes and suffixes to stems which typically contain two consonants.


A language in which all nouns, and possibly all words, had to have the sequence /jaččolo/ somewhere within the word. I was likely subconsciously drawing on Italian here, despite the spelling, as Italian had by this time already long been a target of my disdain for its inability to express thoughts as concisely as English or even Spanish.

I remember looking at a phrasebook and seeing a translation for "My friend is sick" in Italian and saying to my parents, "You might as well say My friend is dead, because he will be by the time you get through saying all that." The Italian translation was in fact only a few syllables longer than the English .... so Yaccolo is the parody of Italian that I was making fun of.


This was actually a revival of a language from 1994 or so that I forgot to write about up above. It was similar to Spanish and I abandoned it quickly for Moonshine. My memories of it involve my high school, meaning it would be a later creation than Moonshine, but perhaps this is just a false memory.


No inflection and verbs commonly require auxiliaries.


IPA /potʃati/. This was a sort of attempt to bring back the #Manni language but I never pursued it other than knowing I wanted it to have a simple phonology.


Not a proper tribal language, this was simply Moonshine distorted by shifting consonants further back in the mouth and allowing vowel sequences instead of always inserting epenthetic bilabials. I allowed double consonants as well. It was thus similar in this respect to the later #Asup, but my mental image of Ihhai was a cold-climate culture living in very primitive conditions, like the other languages in this section.

Ihhai was intended to feel like Finnish, or Uralic generally, with frequent use of length contrasts in both consonants and vowels, all stops being voiceless, and vowel sequences being permitted. But I never set it free from its Moonshine base and therefore it was never a proper language with a culture.

Ižda Mir

Im not sure where to put this because I cant quite pin down where I was when I created it. It may have been as late as 1998, by which time I had already started #Thaoa, but I think the names like Vembed were not originally part of the language, and that my memories of this language from 1998 were a revisiting of it and not the original draft. It may have been as early as 1995, but I doubt that as well since I remember that I spent a long period of time working on Moonshine and Moonshine only.

Culturally, Ižda Mir occupies the same position as Moonshine, and I think it was a sort of "liberated" Moonshine where I abandoned the strict mathematical derivation system and just put words together like ordinary languages do. I know that I allowed myself to be messy and to make words longer than they needed to be. There were unwieldy consonant clusters like /žz/, giving the language a "mouthful" feel that I have otherwise typically avoided. I never settled on a phonology because most or all of the words I used were from Moonshine, and I came up with new sounds only on an ad-hoc basis. I did not understand sound changes, and so, as with other languages before and after this one, every word evolved according to its own unique rules.

The only two words I remember from this language that are not proper names are ižda mir? "what country?", as said by a teenager asking their partner which country they will reappear in when they complete their time travel journey. (I don't remember if the boy asked the girl, or the girl asked the boy. The girl was typically in control but sometimes deferred to her boyfriend out of respect for his hard work.) I forgot this language so quickly that I ended up naming it after the only two words I remembered.

I revived the name of the language for the earliest draft of Poswa, because early Poswa had the same general feel to it that Ižda Mir had had ten years earlier. But Poswa has since become far more extreme and unlike even my other conlangs, and today that name would be starkly out of place in Poswa, which has very few words beginning with vowels, dislikes heterorganic clusters, has no /d/, and very rarely uses word-final /r/.

Southern Namman languages

By this time, I had decided that the southern half of Namma was so disorganized that it had thousands of languages, each spoken by just a few villages. Thus, the Pabappa-like language that the boys had learned after their missions to southern Namma would have been utterly useless in the other 99.9% of the planet. Loanwords passed through the many languages, however, and some even came from Moonshine. I envisioned the languages as being like Hawaiian if Hawaiian commonly loaned words from languages like Mandarin Chinese.

The actual word /paba/ originates from this stage; it was originally a loanword from an early and now-forgotten Moonshine word for Heaven, possibly /bón/, which ended up as /baa/ in one S Namman language.


Spoken at the South Pole, the people of Mrlu are smarter than and control the others to some extent. That their language name begins with three consonants may have been a subconscious way for me to indicate that they weren't as simple and babyish as their neighbors. (Though that's a syllabic retroflex approximant.)

Languages of the Southeast

There were languages spoken by other people in Southern Namma who were only distantly related to the tribe that the boys met up with, as though they had been separated for tens of thousands of years despite being geographically close by; this was a later idea, however, with a different map in my mind than what i had started with.


In 1997 I had a dream where was looking for a girl I'd spoken to on the Internet, and using AOL's member search function, I found that she was in the town of Ojajojojoojoijo, Finland. So I went there and met her and we had fun together. This dream was taking place in the mid-2000s, and I remember that only one of us had aged, but I don't remember which one. This girl was not the same girl who had sent me a tampon and I think she was only visiting Finland, not living there. In this dream, the js were pronounced as IPA /dʒ/.

This dream is important to me because it made me realize that it is possible to create a perfectly workable language where the consonants carry very little information. I soon worked on a language called Palli, and then Late Andanese, before finally creating Pabappa. Thus much of what makes my conlangs unique, and much of what makes them interesting to me, can all be traced to this one dream.

A year or two before this dream, I had had an idea for a language in which /b/ played a similar role (because /b/ was the most basic consonant in Moonshine), but I didn't think of it as a workable idea and abandoned the idea mere seconds after it occurred to me. It's possible that the dream was a subconscious recycling of that earlier idea.


One of a few names for a conlang I created in 1997 and played with for about a year afterwards. I also called it Gaze, but I think this may have been originally a name for their religion. It was rigidly structured, and good at succinctly expressing religious concepts. For example, the root word for church was an, and there were two roots for God: al and m. (I believe that I didnt capitalize them.) There were also monomorphemic roots for things such as "forgiven baby" (īpp) and "spirited man" (I dont remember this one though). Both of these were intended to be interpreted with meanings related to Christianity. There was also a word like lanītta meaning "land given up to God".

The phonology of Asup was not as babylike as that of Moonshine. However, I still do not consider it to have been a "harsh" guttural type of language. There were five vowels and an array of consonants somewhat smaller than that of Moonshine. There was also a syllabic m, which could be short or long (this was around the time that the song MMMBop came out).

All nouns ended in vowels or -y. The grammar was partly nonconcatenative, just as Moonshine's had been, but in a different way that I don't fully remember. The word for butterfly, I think yulīsse, was etymologically "beautiful animal that has its own refuge". I can only remember a few Asupian words. Dese "boy" and sta "girl" appeared in many people's names, because at the time I was mostly thinking of superhero names that people might give themselves rather than typical birth names. As I wrote, I changed the ages of my superheros to match my own, and even at age 15 I considered myself a boy, so that is the word I used to create the adventurers' names.


There was also a numeric system, and I ported over most of Moonshine 's vocabulary. The only rule was that the numeric values of the words needed to match. However, since Asup's numeric system was totally different from Moonshine's, the Moonshine words sounded nothing like the Asupian words. The classifier -m became -y. Boy was the word for outer space. I made a few exceptions to the rule of borrowing morphemes numerically when I decided to take over the Moonshine noun suffixes -o, -a, -(e) as such instead of deriving longer equivalents.

One major flaw in the math system of Asup was that it was based on prime numbers, so it was difficult to borrow short Moonshine particles that had had values like 6, 8, and so on. Even when the original word was prime, it often happened to convert to an Asup word that consisted of just a geminate consonant, which meant that I had to pad it with /e/ on both sides (/e/ was worth zero), and therefore had words like eyye which were much clumsier than the original Moonshine words. The geminates bothered me enough that I changed the pronunciation of some of them, such as making /tt/ into /þ/, but not all of them offered such convenient solutions.


Adjectives could appear as infixes. E.g. ābbe "baby", ābbie "small baby". The /ie/ sequence here may have been a mistake on my part, since I otherwise left final vowels to stand alone. Or perhaps I just wanted a girl's name in my language, in particular, someone who I had met.

Other than this, the grammar was very poorly developed, as I mostly used Asup as a naming language. I never even decided how to conjugate verbs ... I wrote some sentences where there were individual pronouns, and others where the verb was marked by a suffix. I think the individual pronoun idea was what I settled on, but by that time, it wasn't really the same language anymore.

Sound and feel

In some ways, I identify the spirit of the Asup language with Khulls, even though they look nothing alike on paper, nor do they sound much alike to the ear either.

Later evolution of Asup

Asupian was a very dynamic language, in the sense that I changed it rapidly as I worked on it. I went through a "Hawaiian" stage where I think all syllables were open and there were many long vowels. I replaced some consonants with sequences, which led to /m/ becoming /nan/ in some words. However, it still had a much larger phonology than Hawaiian. I think this stage of the language might have been the first time where /a/ was the null vowel rather than /e/. I think I also borrowed some words from Moonshine in the ordinary way rather than using the numerical values of the letters.

Eventually I shaded into Echo.


Echo was a language I created in 1998. It was the first proper language I made that consciously embraced ideas I disliked, such as a relatively small phonology, and a slow speech tempo that made sentences in Echo longer than in English quite often. I remember ovàvu "water", where previously my word for water had often just been a single letter. (I'm not including the hatelangs as proper projects.)

It was a "tropical" language as well, in the sense that when I was working on Echo I had fallen in love with tropical Africa and wanted to make the Camians somehow a warm-climate culture, despite their history of having always identified themselves with cold. This is why Camia was suddenly a racially diverse nation; I replaced the blonde boy I'd been using as a self-insert character with a boy I described as being halfway between white and black. I did not consider him mixed-race; I felt that Camia would have a separate identity for people like him that didn't depend on what his parents looked like. I seem to have considered dark-haired people to be part of this intermediate category, however, so many Mediterraneans would not have been white in this iteration of Camia.

Phonology of Echo

Echo was also the first language to have proper tones, as Moonshine's tones were created at a time when I didnt understand how tones worked. This was part of the "tropical" atmosphere of the language, as I was rejecting the Semitic-influenced "dry" phonology of the Asupian language in favor of the new, softer, "wet" phonology. However, it was a pitch accent language, with only one stressed syllable per word, and therefore not truly tonal. The bilabial consonants /b/ and /p/ were very common, but /w/ was rare. Neither vowels nor consonants had a length distinction, although towards the end of my time with Echo I began radically reworking the phonology.

The consonant cluster -bh- seems to have been common; my favorite name in the language was Yuyabhoybakopa, and another one was a similarly long word that I'm not posting because I might still be using it as a password somewhere. This was not a voiced aspirate; I just didnt seem to be bothered in having an /h/ follow immediately after a voiced stop. Yabhoy was the word for human.

Use of voiced stops

I used a lot of voiced stops in Echo, with words like dagʷomàbada. This is a common trait of the Bantu languages of sub-Saharan Africa, although I did not know this and was probably working subconsciously. I now hate this feature and many of my current projects, even minor ones, either have no voiced stops at all (Andanese, Palli, some descendants of Khulls) or have an incomplete set (nearly every other language). Of those with a full set, the voiced stops are always much less common than the voiceless stops. This is especially true in Khulls where the voiced stops are only just barely above being rare-place allophones of the nasals.

I have a vague memory that towards the end of my time with Echo, I dropped all voiced stops from the language, but I must have abandoned the language shortly afterwards because I don't remember ever looking at devoiced versions of personal names such as Pango and Batu. It's possible that this was actually a different language altogether and the memory has blurred in my mind.

I went through a stage with a lot of names that could be said to belong to no language at all, like "Simpax". This was where I began to use the spelling x to denote pharyngealized vowels; thus this name would be spelled simpâ in Leaper. It wasn't that I had regressed into my old 10-year-old self's manner of changing the entire language every day; I think I just was trying to create too many languages without bothering to create any words other than a few personal and place names.

Unstressed syllable reduction

Echo also introduced unstressed syllable reduction, a form of classifier suffixes. For example, the word for tree was something like debi, but the morpheme for tree as a classifier suffix was -di as in dabondi "apple tree". I believed I had borrowed this phenomenon from Dutch, as seen in koolsla "kale salad", but this Dutch word seems to be just an isolated irregular example. Something similar takes place in Poswa, but instead of involving classifiers, it generally involves inflectional suffixes, although content words can be contracted as well. See Poswa_phonology#Morphophonology.


Echo was my first language to use mandatory SOV word order, but it also used VO for passive sentences. I reverted to SVO for a few years, but around 2002 or so I switched back to mandatory SOV and have remained there since.

Echo may have been entirely isolating. I know I wanted fusional derivation, and developed a word for sunset that did not transparently have either the word for "sun" or any verb in it, but I don't remember any fusional inflections. The genitive was marked by a separate word, at one point chi (possibly IPA /çi/), even when translating the English possessive pronouns "my, your", etc. This was odd for me, since I love and have always loved fusional inflection systems, but I may have just been working beyond my means at the time and not known how to come up with a good inflectional system. After all, Moonshine's inflections were based on mathematics and thus not fusional in the traditional sense either.

Vocabulary of Echo

I still derived Echo nouns from Moonshine when I could. However, I still did not have ready access to the Internet, and had not yet learned how sound changes worked. So essentially every single word followed its own rules for sound changes. e.g. sometimes -e was dropped, sometimes not. Echo was also a very dynamic language, and never had a finalized phonology. I had no formal process for deriving tones, since it had evolved from a toneless language. Words with tones may have simply been those that I created out of thin air rather than those that had been maintained from the time when Echo had been #Asup.

This may have been the language where I first became worried about diphthongs crowding out words with single-vowel nuclei, because even though I was creating words by hand I felt that I needed to use all of the available syllable types, and I realized that if the nucleus permitted semivowels on either side of the vowel, statistically more of my words would have such glides than not.


This was the first language I made that used a conventional spelling for its tones: á was high, â was rise-and-fall, and à was low. This worked well because I did not have contrastive vowel length. In the early stages of Echo, I might have still retained the geminate consonants from Asup, but later on I threw them out, and I think at some point I started using geminate consonants to indicate either vowel length or a different tone on the preceding vowel. Thus my orthography became chaotic and, even for the few words I still remember, I would not be able to confidently give a pronunciation.

Lingering associations from English

I changed a few words to make them cuter: the verbal suffix -ay shifted to -ey, and the word boy, meaning outer space, shifted to bey. This was the first time I considered myself old enough to look back on my own childhood with nostalgia and see my younger self as cute, in contrast to my strong, liberated teenage self. But I still considered myself a boy, not a man, and therefore the English word "boy" was aggressively masculine to me in most contexts and I felt that removing it from the language would make the language cuter. In fact I associated the sequence /oi/ with "boy" and I believe I shifted all instances of /oi/ and /ai/ to /ei/ (spelled "ey"). Though the fact that I shifted /ai/ also leads me to believe that there was more to it than just dislike of the word "boy"; I think I may have wanted -ey because it appears in English on many girl's names.

I was still working on my original science fiction story at this time, though I had been stuck on Chapter 8 for four years and spent most of my time writing down plot ideas. I also made minor edits to the original story, and changed many characters' names. Some of the characters' surnames were still derived from English, but names based on native vocabulary were common, and these words were almost entirely original. There was just one loan in the language proper that came ultimately from English: sarabe "sun", where the sara- part is from my 10-year-old self's word for "(light) source" and the -be part is "outer space", because this is from a later stage in which bey had shifted further to be. I probably picked sara either because of how I'd carelessly left out the /t/ as a child (and thus made it "cute"), or because it was a girl's name (and thus also cute). I also worked the girl's name "Alley" into the language but later came to be satisfied with just having the -ey.

I had a few semantic coinages: snabbe (earlier snabbey) meant spaceship, because the spaceships I drew early on resembled men's underwear and the word for underpants was something like snap, snab or sná (there was a period around this time where I used geminate consonants to mark tones on the preceding vowel, so I can't be sure which geminates were real). Perhaps the derivation went the other way, with a word that originally meant spaceship shifting to underpants whenever the suffix was not added.

I only created joke words like this when I reached my late teenage years; as a child, I had always disliked potty humor and obscene jokes, which may have helped me create a story world in which the child superheros were much more mature and in control of their emotions than were the adults they interacted with. (My parents liked an obscene cartoon called Duckman that I hated. Any references to obscene concepts in my conlangs up until this point were for foil languages like French that I assigned to cultures that I hated.)

Thaoa (1998)

Thaoa was a language I created in 1998 and used in a computer game that I never published. I sometimes also called it Palli. Both of these names are being used for different languages in my newer work.

Early Thaoa

Early on, Thaoa was a very dynamic language, and there was no clear break between the last stages of Echo and the earliest stages of Thaoa. They belonged to different cultures, and I later on decided that they would both exist, but I think at the time I was transitioning from Echo to Thaoa I wasn't writing much, if at all, about the history of planet Teppala.

Probably during the winter of 1997-8, my working language had an extremely large phonology, including for example a four-way contrast among nasals between normal ones, voiceless ones, voiced nasal fricatives, and voiceless nasal fricatives. This was only a doubling of Moonshine's inventory, and I didn't believe it would be unnatural.

Thaoa was the last language I created that had words with initial /sC/ clusters, such as spap "fire" and sta "girl", the latter of which was a holdover from Asup. At the time, I didnt realize that those were found mostly in European languages and would thus be out of place in Thaoa. It was probably the first language with a distinct word for baby ... previously I had lumped the concepts of "young person" and "offspring" together (like English) and also not distinguished between children who were very young and those who were older. I needed the word for baby to name a magic spell ... that word was kuku, from onomatopoeia and a change from my earlier (and later) years where I would have chosen labials.

The name Mappamensam originated here, and I revived my Ižda Mir-style "open phonology" where I just added sounds on te fly without worrying whether I had too many. Thaoa had a deliberately slow pace, meaning that even though it had an enormous phonology, it took a lot longer to say things in Thaoa because Thaoa was very wasteful. At one point I changed the name of this woman to Mappamemmam, but perhaps that belongs to #Abapes instead.

Gradually I whittled the phonology down. The change of the name Tebbala to Teppala is an example of this, and though these are the names that have stuck in my mind, I later got to a stage where there were only three vowels and no closed syllables, so the name became Tipala.

Later Thaoa

Thaoa/Palli was the first language since 1993 with the explicit goal of imitating the speech of young children. However, I seemed to have no idea what I meant by this:

"Thaoa is the language to end all languages. What it is, is the speech of preschoolers, in some foreign language, brought up to a sophisticated adult level without changing the language. Thus, adult speech in Thaoa is only a more complicated version of children's speech, not a completely different thing."

Thus in a sense this could be seen as a prelude to Poswa and Pabappa, except that I had no clear idea of what I wanted. There was nothing special about the vocabulary or the grammar. I remember debating whether to base the phonology on babbling or not and ended up choosing not to. And without the distinctive phonology, there were no real babylike traits in the language at all, even if I thought I was creating something special. However, the sound /l/ was very common; for example, the word for human shifted between /ulila/, /lila/, and /lilal/ as I worked on the language. I had actually read in a parenting magazine that a baby's first consonants at some stage of development (presumably pre-babbling) were /h l ʔ/ and decided to make those consonants the most prominent in Thaoa, but then decided to focus on just /l/ specifically. At some point, I entertained using /r/ instead; this was sometime shortly after the Finland meetup dream.

I also wrote that Thaoa was actually a set of languages that were bound together by having identical grammatical rules. I believe that I also meant that they had identical phonologies. The concept of two languages having the exact same phonology and grammar is an idea I later reused with Andanese and Palli.

Thaoa had tones, but I was writing on a computer with no easy way to type diacritics and simply didnt bother to write them down, so that information is lost even from the small wordlist that remains. I still didn't understand how tones work, and I said that some vowels would occur only on certain tones. I remember the tones for only a few words, those being the ones that I remember pronouncing out loud. I also know that my work on this language overlaps with that of #Xap at least a little bit because of the one word /patali/ "shopping carriage".

Thaoa was messy on purpose. The name Tebbala arose from attaching a meaningless two syllable affix -(b)ala "of" to a one syllable content word, a loan from Moonshine. Like a toddler carrying her parents around on her back, Thaoa perversely made the most inefficient choices possible in its grammar. (I say the same thing about Japanese orthography.)

I've revived the name Thaoa for a different language that does not resemble the language I called Thaoa in 1998. The new Thaoa is my least favorite of the languages I'm currently working on, with very few interesting features. The only connecting feature between the two Thaoas that I can think of right now is the name Tebbala, which becomes Čebbala in the classical Thaoa stage of the language (Tebbala is explained as an exonym).

Culture of Thaoa

Thaoa was a language without a culture, although I did at one point write a story about a boy living in Deppam who spoke Thaoa. However, Deppam was a nation of immigrants, so the language would need to have originated elsewhere. At one point I remember calling Thaoa "the language of thirst", as its people had previously lived in a cool, humid, heavily forested area, similar to Moonshine's habitat, where food and water were easy to get and nature was never threatening. But the climate had somehow changed (perhaps they were exiled), and the speakers of Thaoa had found themselves living in a hot dry desert climate where food and water were difficult to reach and both animals and plants were dangerous, so full of sharp points and poisons that the humans spent their time trying to isolate themselves from nature as best they could.

In any case, the name Thaoa would have to be explained as an exonym if I were to revive this language, since it did not have aspirated consonants. However, I'm currently using the name Thaoa for an unrelated language where it makes more sense.


An extremely complicated language I worked on during high school. It was based largely on complex mathematics, the details of which escape me now. It was entirely CV, but had 15 tones and a large array of both consonants and vowels. Vowels could also be pharyngealized or nasalized or both. The 15 tones were actually broken down as 5 tones for vowels and 3 tones for consonants, with the intent that the "tone" of the consonant determines the starting point for the tone of the vowel, with the vowel's own tone merely being the ending point.

Unlike my previous two projects, this was a new ex nihilo creation and started with a bang rather than gradually appearing from what had once been Thaoa. (But Thaoa still survived on the back burner while I worked on Fojy.)

Due to the gigantic phonology, I had to regress to writing everything on paper, and I've lost my entire source material over the years and remember very little else. I think I later came up with a system that wrote the tones with vowel letters, so that, for example, "láykàa" could indicate a word where the two syllables start high and low, respectively, and end on the y and a tones, respectively, which are from a set of five rather than a set of three. There were no diphthongs since it was a purely CV language, so this system actually made sense.

Note: I seem to have used j to denote vowel hiatus.

Every consonant could be soft or hard. Generally, hard consonants were found in root words and soft consonants were found in infixes which marked inflections. Thus, a name like Mutuphijojewijygi is actually just a root word, I think Muti, with the infix -uphijojewijyg-.


In study hall I pored over my notes for hours, trying to figure out a way to combine two words without leaving a seam at the morpheme boundary. I eventually settled on mathematics, thus reviving the idea I had used for Moonshine five years earlier. Every word was literally a number. Many words began with fo- because this marked leading zeros in the number system. I made the mathematics much more complicated in Fojy than it had been in Moonshine so that each compound word would resolve to only a single pair of constituent words.

I intended to have a "smoothed" version of the language that would sound more like a typical human language, but I never got around to it.

I remember one place name, Uyxowaka, and I don't remember if the y is a consonant or a vowel. I seem to have assumed it was a vowel when I decided to use that name as the proto-form of the word for wolf in the Tapilula language. This word becomes Andanese ihuka and Pabappa pap.

This language was far more than I could handle, and this is one the few times I remember wishing I could add certain features to the language and not being able to because I couldn't figure out how. If I had succeeded in implementing all my ideas, the result would have been a language far more alien to human consciousness than even Lojban.


Fojy was the first language I created that I contemplated using for true interspecies communication; previously, the Mastermind boys had either been forced to learn aliens' languages or the aliens had been forced to learn human languages. Thus, I had an excuse for the wildly unrealistic grammar.

He pam, eʔoqaaniam?

I also briefly worked on a language with no name, but which was spoken in a place I called Kebanq. This was actually another regression to distorted English, at least some of the time, since I remember creating words like peyack for "five" and laika for "glacier". It was actually supposed to be the language of the young autistic boy named Sam Thompson, so deriving it from English made sense. Since Sam Thompson had autism, I decided that he must have a private language with which he communicates with his mother that is similar to English but has "private words". There were, however, some words in this language that were truly a priori. The name Kebanq itself may have been one of them, but it's similarly likely that this is a deformation of the name Kevin or of some word that I can't think of right now. I cannot remember any other a priori words right now except the sentence He pam, eʔoqaaniam?, which I never gave a meaning to.

This language may have been a sort of cool-down from my failed attempt to create Fojy, as they were both associated with planet Laba, but they did not occupy the same place in the story. Fojy was intended to be the language of Laba proper, but Kebanq's language was literally just for the boy and his mother, and any use outside that context was explained as the boy standing as an intermediary. (Thus, Kebanq was not the proper name of the place where he lived; he just called it that and he was powerful enough to make people play along.)

I also referred to the language as "Martian", because the nation in which it was spoken, although located on Earth, had territory on Mars as well and came to be associated with Mars by the other nations on Earth. Referencing Mars was a deliberate use of a well-worn trope, because this language was intended to be playful and not something I poured serious effort into creating. (I had never, even as a young boy, ever set stories on Mars or any other planets within Earth's solar system.) This made sense, because Sam Thompson was one of the few characters I created that I allowed to be "disobedient"; that is, to acquire personality traits and perform deeds that went against the direction I wanted the story to go; Sam's powers were the powers I denied to the child superheros I had created earlier, and they sometimes came into conflict; whenever I felt like the odds were in Sam's favor, I let him win even though it made my storyline less appealing to me.

The words I made up were an imitation of the distorted words I made up for numbers in my early childhood, such as "horty" for 40, "pyab" for five, and "nunner" for any sequence of two nines. The 99 word probably has nasal flaps, ... I just know that it sounded the same going up and going down so I could say it twice as quickly. I seem to have avoided IPA /e/, only using /æ/, so I did the same in Sam's language.


Tarise is a language idea I've tried to complete several times but have never succeeded. My goal when I first created Tarise in 1999 was to create a language that is simultaneously feminine and very unpleasant to my ears. I decided that the vowel letter y was feminine, perhaps because of its exotic appearance. I also added the voiceless ejectives /t_>/ and /k_>/ to the phonology, because to me ejectives somehow sounded both feminine and very sharp, as if representing a woman's long fingernails. All the stops were voiceless, and my original writeup included occasional voiceless vowels as well. The labial sounds I loved so much in Echo were very rare here; in particular, /p/ was almost entirely absent from the phonology, and the vowel /o/ had shifted to a schwa-like sound (the y vowel spoken of above).

Thus, Tarise was the opposite of the Danta language I had created six years earlier for the society of superhero children. In fact, I came to identify the Tarise language with the kids' cruel teacher, who was unwilling to help the kids when they faced unfair obstacles, and yet was able to push them around because the kids' home planet's ruling class only trusted adults. I realized that the teacher would not have a language all to herself, and so I made it also the language of the entire society that the kids grew up in, but I resisted the idea of having the kids actually speak Tarise or have names taken from it without consciously having a reason for it.


Though the superhero children were still canonically part of my storyline at this time, during the time I worked on Tarise I came to focus on a period of time about 500 years earlier, that is, around 3700 AD. The information below concerns that period:

Tarise's speakers were not feminists; they were women living in a society in which men were so weak that the concept of feminism was meaningless. Both the Tarises and their greatest enemies were female. The Tarises practiced abortion and would even kill newborn babies without any sense of guilt, but they devoted their greatest efforts to killing their enemies. They also killed small animals. Yet, the phonology I designed for Tarise was subconscious; I made no direct connection between baby talk and childbirth, even when I decided to allow the /p/ sound in the word for "pregnant" after all. I think I was working on abstractions and that the lack of /p/ was not due to the Tarises' neglect for their babies but because they were cold and aggressive in general.

Tarise was located in what had traditionally been a very cold area of Camia. I have always admired unrealistic climate setups, and so this area of Camia had always been colder than other areas at the same latitude and elevation, with a climate similar to Winnipeg or even Flin Flon despite being centered around 45N. But during the period I was working on the language I contemplated turning it into a hot desert climate instead, something like the inner Sahara, without moving its location any closer to the Equator.


Tarise had multiple roots. I remember saying that it was a successor to the failed project #Fojy, but I don't know if I intended it to actually derive from Fojy historically. The /s/ infix that marked the genitive was an old trait from #Abapes. (The name of the territory was Taryte.) Some words seem to have come from #Thaoa. I know that much of the core vocabulary, including the language's name, came from #Echo. I still did not understand sound changes by this time, so the few sound changes I used for Tarise's history are mostly unrealistic, with some based on moving letters around in the alphabet. For example, I universally shifted all /r/ to /s/ in all positions with no exceptions, and then shifted all /p/ to /r/ with no exceptions, saying that they had simply switched the letters they used to write their words. Thus the word popró, the name of a religion, shifted to ryrysy.


Tarise inherited the vocabulary of #Echo but it was distorted by the script-based sound changes. I had recently heard of Láadan, and took the idea of a feminine-centered language as my own, but adapted it to the violently anti-male and anti-child culture of Tarise. The only similarity with Láadan was a vulgar one: there were many words for abortion and menstruation, but no words for sexual intercourse because sexual intercourse leads to children.

Tarise's main contribution to my present-day conlangs is that all of the languages I have created in this century start with a basic vocabulary that distinguishes between tampons and menstrual pads, and usually has more than one word for each of those.

Tampons in particular are an element in my imagination not because of Laadan, but rather a dream I had in 1996.


Probably the main reason I have not been successful with developing Tarise is that it has the very sort of phonology I most dislike, and therefore the only way that I can succeed with Tarise is to make it, to me, a very ugly language.


This language could be considered to be as one with Andanese, since it shares the same phonology and alphabet, and has a similar grammar. But Xap's lexicon was still of the type where the language did not have a proper history, and I flooded the dictionary by dumping in thousands of words from old, abandoned conlangs that didn't make sense when mixed together. On the other hand, the freestyle approach to lexicon building allowed me to make oddly specific words such as:

  • kutaka "mayor sign in the middle of the road"
    This word indicates a perversely counter-productive advertisement: a mayor who approved a project to repair a major road ensured he'd get credit for it by placing a sign with his name on it in the middle of the road, distracting drivers and making travel more difficult, leading people to blame him for the resulting mess.
  • kahalaca " locked in jail with the hands and head stuck outside the cell"
  • yukana " relying on intelligence and physical strength only "
  • It's difficult for me to describe this coherently; it was the ideology that the superhero kids believed in, and which I came to associate with science.
  • nihilalaka "without any divine intervention"
  • This word is here because it's what I thought intrinsic meant when I was younger.
  • luala "to behave like a "mascot", trying to sweep away conflict by making jokes"

Some words had private definitions that, even with explanation, would still be difficult to explain to other people because they reference events that happened privately to me:

  • muhua " uipila abandon all hope ye who enter here"
  • kutau "Beethoven"

The language also had enormous piles of synonyms, with no difference in meaning whatsoever between many words. For example, the words ahahama, kumihanana, kuna, kunana, lanahala, and luna all mean "moon" and each can substitute for any other with no change in meaning. (Luna is not a loanword from any Earth language; even here, I stuck firmly to a priori instincts. The pre-Andanese form of the word was something like lónay.) I think these six are just six of a much larger set of Andanese words of which most have not survived, because I remember once saying that Andanese/Xap had "nine words for cat and no word for pants" whereas in the dictionary I have now I can "only" find seven words for cat.

The most extreme example I can find is that Andanese has 43 words for "love". Although some are given specific definitions, more than half of them are defined simply as "love" and therefore can substitute for each other with no change in semantics.

Some words were taken from the names of characters (mostly female) in my earlier work. For example, a girl named Kaiciti gave her name to the emotion of being "in awe at the power before her".

History of Xap

Xap actually began as a language called Abapes. This was an attempt to create a language with extremely fusional polysynthetic inflection, such that the entire stem of a word could change by adding an inflection. For example, the name Abapes itself was simply the genitive of Baeba, the name of the city where it had originated. Though I don't remember it, I must have begun work on this in the late 1990s because a few scraps of Abapes appeared in #Tarise and #Thaoa.

Xap is the same name after a series of sound changes, although the sound changes were not of the proper historical type; I was still thinking of the language as being detached from culture and time and therefore my sound changes were random and unrealistic. The name "Xap" actually merges several words, not just Abapes. One of them is Zebes, the name of the planet in the first Metroid game. Normally, I've thrown out all borrowed names like that, but in this case I kept it because it's a triple etymology (Zebes, Abapes, and Peepa), where Peepa was a name I had created for myself earlier on.

Xap was the last language I had that featured script-based sound operations, although these were not diachronic sound changes but rather grammatical operations. That is, an inflected form of one word could be formed by moving its final syllable forward in the syllabary by some number, or by adding a digit and then regrouping (i.e. 59 + 1 --> 591 but spelled as 5[91]).


Xap continued the downward drift of Thaoa and ended up with just 30 syllables. (Sometimes I consider that to be Thaoa's final phonology too, because there was no clear border between the languages, but I later came to distinguish between Thaoa (with a larger phonology) and Palli (with 30 syllables)). I had stopped thinking of the language as being for toddlers, and thus ironically arrived at a phonology well suited for toddlers without meaning to. However, the abundance of labial consonants was not very strong here; the commonest consonants were in fact /l h k/.

Early on, I allowed consonant clusters by elision of /a/, but I had no rules for when the /a/ was elided and when it was not.

I considered the consonants to have tones, because while all five vowels could occur with all three tones, the tones were affected by the preceding consonant and I think there may have been some forbidden consonant/tone combinations up until the later stages of the language where consonants fell together and sequences like /uhú/ collapsed to just /ú/. Thus, while this situation may be typologically unusual, I think this is one of the ideas I got right. There was a stage I called Middle Andanese where only some consonants still had tones, and they were all gutturals, so I had an unbalanced phonology with more than half of the consonants being some form of /k x ɣ/, even though the language was not particularly guttural overall.


I used á for a high tone, ā for a mid tone, and à for a low tone. At some stage, I remember not having tones at all, and using the macron to indicate stress. But I must have reintroduced them, because the tones were the last feature to drop as the language reduced itself to its final inventory of just 30 syllables.


The word order was initially SVO, then flexible, then mandatory SOV. All languages that I've made since then are mandatory SOV as well except those of Dreamland, but those are side projects.

Sound changes in Xap

Although I still derived some of Xap's evolutionary changes from its script, here for the first time I also used proper sound changes. They were still very naive, but not impossible, and most sound changes were conditional.

Lexicon of Xap

I have 4427 words in the 37th edition of the Grand Unified Dictionary of the Xap language (although the title labels it as Andanese). Xap became a depository for all of the words from all of the conlangs I had previously created. This led to massive tables of synonyms, as above with the six words for "moon" and 20+ words for "love", and also massive piles of homonyms, such as hauaka, which means both "tree" and "human".

For a long time I planned to rescue the Xap words by carefully selecting only the ones that I know could not have possibly been also retained in another language, but I've decided not to do this, as it would be a great effort for very little gain.


Xap was the last language I worked on during my science fiction phase, and it was intended for a culture who had lost some basic human traits. I pictured the speakers as young boys, because they were born from incubators, and had lost the need for wombs and even for adult bodies. There were other societies that had done much the same thing but had chosen to become all-female instead, and the Xap speakers saw these societies as friendly rivals, not as enemies. I also said that they would have partly mechanical bodies, and that they would not grow at all once they were able to walk, as the natural growth process would interfere with their mechanical parts. Thus they were not merely boys, but toddlers, even as they became the rulers of the known universe.


I used the simple 30-syllable phonology to make ciphers of Xap that sounded totally unlike standard Xap but were grammatically identical. One contained the phrase "nupy nutu". This cipher was a six-vowel, four-consonant setup, spoken only by children, that in some ways resembled Pabappa (which I had not yet started work on). I only call it /nupy nutu/ because that was the phrase I remember one boy saying to another boy who was not interested in listening. Another one had 14 consonants and 2 vowels and was intended to imitate #Tarise. Yet another kept the three-vowel inventory but changed the consonants ... I think I added /pʰ tʰ/ and removed at least one of the two dorsal stops, and perhaps both.

Xap II

For a while I worked on Xap II, a language unrelated to Xap but similar in some ways. It was a "sequel", not a daughter language, and it was intended for interspecies use. It could be said that Xap II had no associated culture or that it was an artificial language even on Tebbala.

The phonology of Xap II varied as I worked on it, but it always had tones, a large vowel inventory, and a small consonant inventory. This is where I switched to using the consonant order /p m s l k r t n/, instead of the Xap-Andanese order /Ø h l k m c n s p t/.

As with Xap I, the consonants conveyed less information than the vowels; all eight of the basic consonants were gender markers, meaning that there were eight genders. I was still hanging on to this idea in 2018 but I'm not actively working on any language with a consonant-based gender system any more.

The genders, to the best of my memory, were:

  • p babies; pregnant women; men and women together. Sometimes called epicene
  • m adult women
  • s all genders combined, including inanimates
  • l babies, children, and inanimates. When used of babies, this gender implies greater independence than /p/.
  • k possibly young boys
  • r inanimate; no gender
  • t adult men
  • n young girls; maiden

There was at some point a fourth feminine gender, intended for women who have passed childbearing age, marked by a /t/-like sound, possibly IPA /t_>/. But I dont think Xap II ever had ejectives, so I either was intending this for a different language or I meant that it would be pronounced /t/ but spelled with its own symbol.

The genders were used metaphorically for objects in nature; for example, tornadoes were feminine.


These people were ordinary humans with two sexes, born from wombs, with full adulthood and men taller than women. I had not abolished Xap-I flying toddler society, but merely lost interest in it, and I didn't consider Xap II to have conquered or replaced Xap I.


Xap III was similar to Xap II, but was actually even more complicated, and I remember very little of it today. Like Xap I and II, it was a logical language, so it didn't bother me that it had a lot of unrealistic features. All syllables of content words began with /p/, and therefore any syllable not containing /p/ was a function word. Despite the presence of /p/, it was not a babylike phonology because the syllable structure was very complex.

I remember building a website with outer space in the background, and I thought of Xap III as belonging to outer space, not even to a planet.

Xap III was the last language I worked on before beginning Pabappa.

Later conlangs

Pabappa (2004)

All of the languages I've created since beginning Pabappa in 2004 have been genetically related to Pabappa, and therefore are attempts at naturalistic languages rather than logical or artistic languages. Pabappa today doesn't resemble the Pabappa of 2004 very much except for its phonology, which I have never changed at all.

Design goals

Here I returned to my goal of creating a language imitating the speech of toddlers, but by this point I had realized that what I had wanted all along was a language with a phonology based on babbling, and so I followed the path I had rejected six years earlier with Thaoa. I realized I had liked the ciphers of Andanese that used labial consonants but that they were very difficult to work with, and that what I really needed was a fresh new language.

As with Thaoa, I also debated other things, such as building a vocabulary in which concepts young children are familiar with are the only basic morphemes, and anything beyond that must be derived from compounds or other morphemes. I rejected this idea, since the speakers were biologically normal humans and therefore had the same adult mental capacity and adult emotions as anyone else. I abandoned my science fiction story shortly after beginning Pabappa, and have in some sense the opposite situation I used to have .... a baby's language in the mouths of adults, when I had previously had a society of eternal children with the minds and all the abilities of adults. And yet the two worlds share much in common because each would see Earth as Earth sees the world on the opposite side.

I think I have a bias towards playful concepts ... e.g. having an atomic word root for "playground", along with several words for different types of dolls and other toys, but this is true of all of my well-developed languages, so Pabappa doesn't stand out.

Today Pabappa suffers from the fact that it isn't my favorite language, and has no redeeming features that aren't also in Poswa.


Leaper, also known as Khulls, is a revival of the "powerful" feeling I used to have with Moonshine, but the language feels different. I imagine the Khulls speakers as spanning all the way from the tropics to the poles, whereas Moonshine even at the peak of its power was confined to very cold climates. In my new world, Moonshine is a daughter language of Khulls, but shares little in common with the Moonshine of 1994: essentially the only similarity is the phonology, and even that has differences, although privately I consider it a continuation because I say I would have made the same changes eventually if I had kept working on the old language all along.

Khulls right now is little more than a phonology and a simple noun inflection system, as I have had to remove almost everything else in order to reconcile it with the shared proto-language of Khulls and Poswa. In other words, whenever I come up with a new idea for Poswa or Pabappa, I'll do it even if it breaks Khulls. The only things in Khulls that cannot be affected by changes I make to the Pabap family languages are the phonology and the noun declension system ... this is because I have no plans to ever change the phonology of the shared parent language, and because the noun inflection system is extremely conservative in both branches.

Spelling of Khulls

This is where I revived my idiosyncratic tone spelling, where both length and pitch are indicated by each individual tone diacritic. Here á indicates a long rising tone (but level in some environments), ā indicates a long falling tone (again, level in some environments), à indicates a short high tone sometimes followed by a glottal stop, â indicates a long, low, pharyngealized tone, and ă indicates a short mid tone with no other features. All of these diacritics also mark stress, because the tone of unstressed syllables is derived from their surroundings.

This is similar to the setup I had used for Moonshine in 1994.

Thaoa (2011)

Thaoa, by contrast, has no shared features at all with the Thaoa of 1998, other than its name and the word Tebbala, neither of which fit nicely into the list of sound changes I have up and thus might need to be declared as exonyms. Thaoa is my least favorite language of those I'm currently working on, and I've made essentially no progress on it since 2011.


  1. This may be a mistake, actually, because there were peoples of widely varying height across the planet. It may be that they were so tall that other tribes' men could only reach their hips, but they were nonetheless far taller and stronger than their own men as well.
  2. This might be a later insertion, actually; if so, I no longer remember the original name.
  3. This assumes that "Manni" and "Manitan" are interchangeable, which I'm not sure of. I may have intended Manni to be the language of just one tribe of Manitans, but I never really got far enough to even consider creating other languages for the Manitan peoples.
  4. My parents were not against conlanging; she merely mistook the dictionary for cartoon porn