The most popular thing I’ve written

I’ve mostly written about writing on this blog, so you’d assume that my most popular post would be about writing, right? As is turns out, that isn’t right at all. My most popular post is, in fact, about conlanging.

It is this post: Creating a basic naming language.

I imagine that some of the popularity comes from the fact that it is #3 on Google when you search “naming language”.


I genuinely never expected to make it to the front page of Google. In fact, I never expected to make to any page of Google. But I actually get a lot of traffic from search engines, including Google – in fact, about three times as many people come here from various search engines than they do through the WordPress reader. Again, that’s not something I expected.

I’m curious if letting the domain name expire will have an effect on this. Guess I’ll know if traffic to that post suddenly drops in 2023. I won’t be too upset if it does. I imagine most people who find that post get disappointed that there aren’t more posts about conlanging on this blog and leave anyway.

But seeing this one post get more hits than anything else has had me thinking for the past few months. What if I write a sequel to it – a post on how to expand a naming language into a “full” conlang?

I think that’s what I’ll be doing in December this year.

Fèdzéyí, the language of too many verbs


Fèdzéyí is one of my older conlangs. It was created in 2015, and the original ideas for the conlang are very different than how it turned out.

The then-unnamed conlang was one of three languages in a language family, all three of which were spoken by a fictional people who lived in and around a canyon. At that point, there were some influences from Romanian, of all languages. I’d already decided that the language strucure would resemble Japanese (CVN syllables), but with tones.

At this point, I decided that the original idea for the conlang needed to be scrapped. I scrapped the language family idea entirely and decided to take inspiration from Mandarin & Navajo, two languages I was looking into at the time. I had a very poor understanding of both of those languages, and it showed.

Eventually, Fèdzéyí moved away from those incluences and became something of its own. Fèdzéyí means “circle language”. When I named it, I still had the idea of making it be spoken by a fictional group of people. After some time, I abandoned that entirely and kept the name.

Fèdzéyí is primarily based on verbs. All verbs have a “base” form that is monosyllabic, with grammatical particles attching to them. Most nouns are derived from verbs with an accompanying class suffix. There are very few “true” adjectives; most of the time, verbs are used in their place as descriptors.


Fèdzéyí has a syllable structure of CVN, where:

  • C = any consonant or /kw/
  • V = any vowel
  • N = /n/ or /ŋ/

/m/ was originally included as a final, but I dropped that when I realized that would create too many base words.

Consonant inventory:

  • Nasals: /m/, /n/, /ŋ/
  • Stops: /p/, /b/, /t/, /d/, /k/, /g/
  • Fricatives: /f/, /v/, /s/, /z/, /ɕ/, /ʑ/, /h/
  • Affricates: /tɕ/, /dʑ/
  • Approximants: /l/, /j/, /w/
  • Trills: /r/

I spent some time deliberating on whether I wanted a voicing or aspiration distinction on the stops, and also whether or not I wanted to include /f/ and /v/, and whether I wanted to delete /r/ entirely and only have /l/.

Vowel inventory:

  • Front: /i/, /u/
  • Central: /ə/, /a/
  • Back: /u/, /o/

There was originally a distinction between long and short vowels, but again, it created too many base words.

Fèdzéyí has three tones: level, rising, and falling. The rising tone is indicated with an acute accent, and the falling tone is indicated with a grave accent.

Verb Construction

A verb is constructed in the form Modifier-Adverb-Stem-Mood-Tense.

“Modifiers” are particles that indicate positivity (vùn) or negativity/negation (éng). You can use verb with vùn prefixed to indicate that you completed an action (such as if you were asked if you already took out the trash). Éng is used in largely the same way.

Adverbs are just verb stems that are used to modify the main verb. There’s nothing particularly unique about them.

Fèdzéyí has four moods: indicative (keun), subjunctive (dzo), conditional (), and imperative (). The three tenses are past (heu), present (da), and future (dzí).

I once considered marking person and number on the verbs since the conlang is so verb-heavy to begin with, but decided against it as it would make the words too long for my own liking.

Noun Construction

This is where things get fun.

The vast majority of nouns are based on verb stems. They’re constructed in the form Determiner-Verb Stem-Class-Number/Numeral-Case-Adjective.

There are proximal, medial, and distal determiners. They’re either singuar or plural, and this is a bit redundant with the numeral being marked elsewhere, but you know what? Redundancy is a part of most (all?) natural languages, so I have no problem putting it into this conlang.

  • Proximal: dà, dàho
  • Medial: tú, túho
  • Distal: néng, néngho

There are ten noun classes, and they’re used to create nouns out of verb stems:

  • People: wu
  • Animals: lu
  • Plants: ò
  • Heavenly objects, weather: é
  • Body parts: neu
  • Human-made objects (tools): mùng
  • Vehicles, methods of transport: ngèù
  • Human-made inanimate objects: kwí
  • Naturally-occurring inanimate objects: ye
  • Abstract inanimate concepts: dzé

Let’s look at the name of the language, Fèdzéyí:

  • = verb, to enclose/encircle
  • dzé = noun class, an abstract inanimate concept
  • = noun, a language

+dzé creates the noun “circle”. is one of the few nouns that isn’t based on a verb, and means “language”.

(Since there’s a word for “encircle”, I decided to create verbs for other shapes – en-rectangle, en-triangle (èn), en-square, en-parallelogram – because why not?)

Let’s look at a couple more nouns I’ve created:

  • do “to gift ” + kwí “human-made inanimate object” = dokwí “a gift”
  • fo “to swim”+ vàng “to be scaly”  + lu “animal” = fovànglu “fish”
  • mi “to meow + lu “animal” = milu “cat”
  • “to count” + dzé “abstract concept” = nèdzé “number”
  • ngí  to fly” + ngèù “vehicle” = ngíngèù “airplane”
  • shé “to flow” + é “heavenly object” = shéé “water”
  • tsà “to know” + dzé “abstract concept” = tsàdzé “knowledge”

There are a couple of numbers: singular (not marked), “some” (meun), a general plural (ho), and a mass plural (kún). If an exact number is needed, you can just put a numeral in that slot instead.

There aren’t as many cases in Fèdzéyí as in some of my other conlangs, primarily because…I didn’t want to deal with very many. That’s the whole reason.

Fèdzéyí cases:

  • Nominative: unmarked
  • Genitive: ú
  • Accusative: shí
  • Dative:
  • Instrumental: yún
  • Locative: vi
  • Adessive: ní
  • Inessive: dzu

Like adverbs, adjectives are verb stems that are used to modify the noun.

Sentence Structure

Naturally, word order is Verb-Subject-Object. Due to the number of cases, it can actually be more flexible than that, but there’s really no reason to use anything other than VSO.

If the sentence is a question, then the question word comes at the beginning of the sentence. Question words are:

  • dèn – when, at what time
  • eng – who, what person
  • – what
  • gè nèdzé – how many
  • – why
  • úng – where, at what place
  • yu – a generic question word that’s used when none of the others apply

Following are examples of each question word:

Dèn lódzodzí waho?

  • When do we leave?

Eng tú?

  • Who is that?

Eng ténkeunda milushí?

  • Who owns the cat?

Gè tú?

  • What is that?

Gè nèdzé zhekeunda ingho wínkwídzu?

  • How many people are in the house?

Kí éngdzekeunda ngóho?

  • Why aren’t you sleeping?

Úng zhekeunda ngó?

  • Where are you?

Yu túlu milushí?

  • Is that (animal) a cat?

Sänirtäm, a personal language


The first time I tried to create a personal conlang was in 2016. It went through a few changes and was, at one point, named läin höuror. Sänirtäm is the (hopefully) final version of this personal conlang.

When I first started working on Sänirtäm, I was fairly interested in creating a language with the aesthetics of Sanskrit. I was also interested in Finnish, and toyed with the idea of having vowel harmony. This idea was scrapped when I decided that I also wanted to be able to pronounce this language.

Sänirtäm is nominative-accusative and agglutinative, because I really like agglutination. It also has a lot of “paired” verbs, where changing the final vowel from /a/ to /u/ changes the meaning from passive to active.


Sänirtäm’s consonants are fairly “normal” by European standards:


  • Syllable-final /h/ → [x]
  • [ŋ] is a rare sound that only occurs before /g/ and /k/

Certain consonants are palatalized before (and sometimes after) front vowels:

  • /n/ → [ɲ]
    • non /ˈnon/ vs. nisvah /ˈɲiɕ.wəx/
  • /t/ → [tɕ]
    • tovnas /ˈtou.nəs/ vs. täm /ˈtɕæm/
  • /d/ → [dʑ]
    • dawa /ˈdɑ.wə/ vs. däras /ˈdʑæ.rəs/
  • /k/ → [c]
    • kavara /ˈkɑ.wə.ˌrɑ/ vs. kera /ˈce.rə/
  • /g/ → [ɟ]
    • gunan /ˈgu.nən/ vs. geskerrada /ˈɟeɕ.cer.ˌrɑ.də/
  • /s/ → [ɕ]
    • sohres /ˈsox.reɕ/ vs. sän /ˈɕæɲ/
  • /h/ → [ç]
    • hanem /ˈhɑ.ɲem/ vs. hinnas /ˈçiɲ.ɲəs/
  • /w/ → [ʋ]
    • varjas /ˈwɑr.jəs/ vs. vihki /ˈʋiç.ci/

The /w/ shifting to [ʋ] was a result of me wanting both sounds in the language, but not wanting them to be represented by separate letters. I don’t actually know if this happens in any natlang.


Vowels are not particularly unusual. The only allophone is [ə], which is what /ɑ/ becomes in unstressed syllables.


/ɑ/ and /o/ followed by syllable-final /w/ become diphthongs:

  • /ɑw/ → /ɑu/
  • /ow/ → /ou/

Diphthongs do not exist otherwise.

Syllable Structure & Stress

Syllable structure is (C)V(F):

  • C = any consonant
  • V = any vowel
  • F = nasals, fricatives, & /w/ (never word-finally)

Primary stress falls on the initial syllable, and secondary stress falls on every other syllable.



With the exception of <ä>, zero letters use diacritics. This was a deliberate decision, as the US-International keyboard doesn’t allow you to type every diacritic on every letter. I did not want to continually have to copy/paste letters.

/w/ is represented by <v> specifically because I wanted Sänirtäm to look more like Sanskrit. I don’t particularly think I succeeded in that endeavor.

Demonstrative Articles

Sänirtäm has four demonstrative articles: proximal, distal, medial, and one that means “on/at”:

  • tokairo – “here, on/at me”
  • kairo – “this, close to me”
  • muru – “that, close to you”
  • vere – “that, far from us both”


Sänirtäm doesn’t have any pronouns, because I wanted to see if I could make a language without them.

In place of first-person pronouns (I, me, we, us), Sänirtäm uses the demonstrative tokairo “this, at/on me” + tären “person”. Names are used instead of second-person pronouns.

Proximal/media/distal demonstratives are used in place of third-person pronouns:

  • kairo tären “this person” – the person close to me
  • muru tären “that person” – the person close to you
  • vere tären “yonder person” – the person far away from the both of us

Nouns & Adjectives

Nouns are either animate or inanimate, and inflected for case, number, and gender.

  • Cases: nominative, accusative, genitive, dative, ablative, allative
  • Number: singular, dual, plural
  • Gender: animate, inanimate

Animate nouns end in consonants in the nominative singular and are declined as such:


Inanimate nouns end in vowels in the nominative singular and follow this declension:


Adjectives come in front of nouns and agree in case, number, and gender:

  • Animate: han hinnas “red star”
  • Inanimate: hana kerra “red paper” – since kerra is an inanimate noun, han gains a final “a” to match the noun’s gender


Verbs are conjugated for tense and mood, and are grouped into two categories: active and passive.

  • Tenses: present, recent past, far past, near future, far future
  • Moods: indicative, subjunctive, conditional, optative, imperative

Passive verbs are primarily verbs that the agent has little or no control over:


Examples of passive verbs:

  • dänna – to be moved, to be acted upon by a force
  • illa – to exist
  • kässora – to unknowingly tell a lie

Active verbs are verbs that the agent has full control over:


Examples of active verbs:

  • dännu – to move (physically), to act
  • illu – to live
  • kässoru – to deliberately tell a lie

Example Sentences

Baruh tevna maunos.

  • The dog sees the cat.
  • dog.NOM.SG see.PRS.IND cat.ACC.SG

Sur maun tevnudem kelos baruhos.

  • The black cat recently looked at the yellow dog.
  • black.NOM.SG cat.NOM.SG look.RPST.IND dog.ACC.SG

Tokairo tären litaradem es Sänirtamos savniyunesu.

  • I was recently inspired, so I’m currently working on Sänirtam.
  • here person.NOM.SG inspire.RPST.IND thus Sanirtäm.ACC.SG build.PRS.COND

Creating a basic naming language

A naming language is a very basic conlang: a phonology, an orthography, a small dictionary, and enough grammar to name things: people, places, deities, etc. Naming languages aren’t intended used for full sentences or passages, so there’s no need to create detailed grammars.

Of course, you can always do that. Your conlang just wouldn’t be a strict naming language anymore.

Step 1: Phonology

The first thing I think about when creating a naming language is (aside from the culture that will be speaking it, of course) the phonology. A phonology is the sounds that comprise a language and how those sounds fit together – syllables, morae, stress, tone, etc.

What consonants do I want the language to have? What vowels? Do I want consonant clusters? Diphthongs? Vowel harmony? All those decisions are made in this stage.

Looking at the phonology of a real-world language can be helpful. If I want a language to sound Finnish, or Japanese, or Irish, I look at the sounds of that particular language.

In my amateur conlanger opinion, Wikipedia is actually a pretty good reference for phonology. Here are the Wikipedia pages for Finnish phonology, Japanese phonology, and Irish phonology.

Picking phonemes

Now, let’s pick out some vowel sounds: /a/, /e/, /i/, /o/, /u/

And some consonants: /p/, /t/, /k/, /m/, /n/, /ŋ/, /s/, /ʃ/, /x/, /j/, /ɹ/

I like to organize this information in tables:


Some of these sounds may look unfamiliar to those unfamiliar with linguistics:

  • /ŋ/ is the sound in sing
  • /ʃ/ is the sound in shoe and ash
  • /x/ is sound in Scottish loch and German acht
  • /ɹ/ is the non-trilled, non-rolled, non-tapped, r-sound used in English words such as red, right, and wrong
  • /j/ is the consonantal sound in yes and you

If you want to familiarize yourself with IPA, here’s an IPA chart with sounds.

For this example conlang, I’ve decided against having diphthongs or vowel harmony. Maybe another time!

Deciding on a syllable structure

Syllables can have relatively simple structures of CV, like in me /mi/ or more ridiculous structures like CCCVCCCC, like in strengths /stɹɛŋkθs/. Yes, that’s three consonants, followed by one vowel, followed by four more consonants, all in the same syllable.

Let’s use a simple CV syllable structure, where C = any consonant and V = any vowel. This produces syllables such as:

  • mi, mu, me, mo, ma
  • pi, pu, pe, po, pa
  • si, su, se, so, sa


This probably isn’t entirely necessary for a naming language, especially if you never actually plan on pronouncing any of the words you create. I like to do it anyway.

Let’s put the stress on the first syllable of a word, because that’s pretty easy to remember:

  • /ˈ
  • /ˈ
  • /ˈ

Step 2: Orthography

What letters are you going to use to represent the sounds of your language? I generally like to stick with something naturalistic. There’s no reason, of course, that you can’t use <c> to represent /t/, or <X> to represent /a/. However, a language that did that would be obnoxious.


I generally arrange my orthography like this, with separate tables for consonants and vowels. I find it’s easier to keep track of things this way.

Step 3: A Dictionary

This is where I start creating the words I need – names & placenames. I generally find creating words to be pretty tedious, so this is my least favorite part of the process.

Usually what I do is create a word and assign an meaning to it. Here are a couple of examples using the words I created earlier:

  • momo – noun – bird
  • nanaka – noun – king
  • kasajani – adjective – red-colored

This continues until I have all the words I think I need. Most of my planning & worldbuilding happens before I start writing, but it’s inevitable that I miss something and end up having to come up with a couple more names later on in the story.

Step 4: Basic Grammar

Sometimes I skip this part if all I need are names. Okay, I usually skip this part for a naming language. But it may be worthwhile to work out how adjectives modify nouns or how the verbs in your language work. This usually requires deciding which cases, genders/classes & numbers a noun/adjective is marked for, and tense (at the very least) for verbs.

Two naming languages I created for The Book of Immortality

Meitsung soré: the main language of the elves and humans in the Meitsung Empire.

  • Lisel = bigleaf hydrangea
  • Tsensung = heavenly overseer
  • Suli = purple
  • Showakelu = stone river
  • Meitsung  =splendid
  • soré = language
  • Hengshal = north borderland
  • Shensi = east homeland
  • Kelutshélin = river plateau
  • Shihun = great lake
  • Ménghun = south lake
  • Méngwing Nang = southwestern mountain range
  • Wingshal = west borderland
  • Tengming = central capital city

Mayu lháni: the language of the tigers. It has the sound /ɬ/ <lh>, which I think is most commonly known as the Welsh <ll> sound.

  • Muhánquri = mountain tribe
  • Hirúka = shining sun
  • Qursin = luck + belief
  • mayu = language
  • lháni = tiger
  • Arquhin = true origin
  • Símaqágu = great forest