How to make a Fictional Language (Conlang)

(yes I'm making this as an opportunity to gush about linguistics. I'm a big nerd)
Going to get into a bit of linguistic jargon in this so just in case...

root word / word stem - the basic form of a word with no conjugations added
suffix - ending added onto a root word
prefix - beginning added onto a root word
infix - added into the middle of a root word
circumfix - added in 2 parts to both the beginning and end of the root word
adposition - words like "in, at, on, by, under, towards, before, after, of, for"
preposition - adpositions but they come *before* the noun (like English does)
postposition - adposition but they come *after* the noun
IPA - International Phoenetic alphabet, catalogues every sound humans can vocalize consistently

I'm going to be using a lot of examples from Turkish cause it's the language I've studied the most. Apart from English and romance languages, I'm most familiar with agglutinative language grammar systems like Turkish, Hungarian, Finnish, etc.

I’ll give a lot of examples for what other languages do (mostly Turkish lol I’m sorry in advance) so you can have an idea of what options you have to play around with in the traits and grammar of your language so you don’t make it just like English.

Like with any kind of good design, designing a language starts with *intent* / with a *goal* in mind.

If you're making a language for an alien species or fantasy culture you have already come up with, then you need to think of how that culture would express itself through language, and what words would be most important to them based on their environment. Like how Finns living in a cold winter environment lent to them inventing different words for different specific types of snow.

For example, Old English never had the word "jungle" because there are no jungles in England. If it's not something within your fictional culture's area or environment, they're unlikely to have an original word for the thing. If they do encounter the thing later, they will likely borrow a pre-exisiting word for it from another language (like English borrowing the word "jungle" from the Sanskrit word "jangla").

Idioms and expressions based on culture are also important. In English we say "fits like a glove", in Italian they say "fits like a shoe painted on" (suggesting Italy's artistic culture), but in Finnish they say "fits like a fist in the eye" (Finns like to punch people??? lol)

Don't set out to make a perfectly logical and consistent language, unless that's your specific goal (like the people who speak your language are robots). Perfect languages are unrealistic. Your language *should* have irregularities, complications, etc.

Also whether or not you stick to more common language traits or uncommon ones is up to how realistic or rare you want your language to sound. Languages like Klingon were intentionally created to sound awkward and unnatural compared to human languages.

Step 1. Picking Sounds

Referencing the IPA chart is useful for this.
http://www.ipachart.com/
(you can click on sounds on the chart to hear the audio)

If you have a language you're inspired by (take Greek for example), you can just look up the IPA chart/alphabet for Greek and base your languages sounds around that.

Whatever sounds you choose to include in your language, you should come up with a way to type out those sounds in the basic US standard keyboard and avoiding special letters. For example the hard "h" sound in languages like Farsi, Arabic, Hebrew, etc. can be written like "kh". You might write it as 2 or 3 letters, but in your language it will still represent 1 sound (like "sh", "th", "ch", "ck", and "tch" sounds in English).

In the IPA there are some boxes with a set of two sounds inside. The sound on the left is the unvoiced version, the sound on the right is the voiced. Natural languages can have just the unvoiced (left) sounds, but usually if you want your language to have the voiced (right) sound, the language will have its unvoiced version as well.

For example, you can have the "ts" sound on its own in a language, but if you want to have the voiced version of it "dz", then usually a language will have both the "ts" and "dz" sounds.

As far as I know every language on Earth has the "n" sound.

Avoid having too many sounds in your language that sit alone in a vertical row on the IPA chart. Exception to this can be the "h" sound. Usually sounds in a language will come in a series of sounds clustered in a few of the vertical categories/rows.

Vowels:

Not every language has the same 5 vowel system like English does (a, e, i, o, u). You can pick how many or how few vowel sounds you want for your language (again they're all available on the IPA chart). Arabic has just 3 vowels (a, i, u) with long versions of each (aa, ii, uu), while Turkish has 8 vowels (a, e, i, ı, o, ö, u, ü).

Ultimately with sounds, less is more. Try to narrow down how many sounds you have in your language to at least less than 30, maybe less than 20 sounds ideally. The less sounds you have, the more cohesive and unique your language will sound.

Intonation:

Decide the intonation rules for your language. English generally puts emphasis on the first syllable of the word, like Hungarian. But English also changes the emphasis on words like "REcord" and "reCORD" to clarify that one is a noun, and the other is a verb.

Japanese generally has no emphasis, generally all syllables are pronounced with the same level of emphasis.

Turkish puts the emphasis on the last syllable on the word (some exceptions though).

Latin generally puts the emphasis on the second-to-last syllable of a word.

Come up with a general rule, but remember you can break this rule if for some words you think the change in intonation sounds better (because real languages change things to sound better/smoother all the time)

Step 2. Starting Vocabulary

Plan out what kind of pattern you want your language's words to form.
C = consonant
V = vowel

Your word pattern could be like Japanese's "CV" pattern for syllables (where a vowel always ends the syllable), only cases in Japanese of there being a "CVC" pattern (where a consonant ends the syllable) is when the last consonant is the letter "n", or when the starting consonant of the next syllable is doubled/extended, like the doubled "k" in "gakkou" (this is called 'gemination' btw).

Your word pattern can have consonant clusters like English, in words like "check" or "swatch" being CCVCC pattern. You can go even farther with consonant clusters until you have patterns like Russian, other Slavic languages, or Georgian.

Make a simple set of vocabulary, common and natural words like "person", "water", "tree", "sun", etc. and common verbs like "to walk", "to be", "to have", "to eat", etc.

Make a set of pronouns (more on that in a minute) and simple particle words like "in", "at", "on", "to", "for", "of", "after", "before", "until", "many", "few", "very", etc.

Pronouns:

In English, 3rd person singular pronouns are the only ones we have with genders - "he/she".

In Spanish and French, 3rd person plural pronouns also have genders - "they" (masculine) and "they" (feminine).

In Russian and other Slavic languages they will also have a neutral gender. "It" and "they" are neutral gender in English.

In Hindi, "you" has masculine and feminine versions.

In Japanese, "I" has masculine and feminine versions. (masc: "boku", "ore" -- fem: "aatashi" -- neutral: "watashi")

In Turkish there is no gender for 3rd person. The 3rd person pronoun is just "o", which can mean "he/she/it/that". Hungarian and Finnish also do this, but Finnish differentiates between a gender-neutral "he/she" (hän) for people, and an "it" (se) for objects.

There is also formality in pronouns, most European languages will use the plural "you" (which English doesn't have) when addressing someone politely ("vous" in French, "vosotros" or "usted(es)" in Spanish). Languages like Hungarian will refer to who they're talking to in 3rd person when they're being polite (even if you're talking directly to your boss, you would ask "How is he/she?" when you're greeting them in Hungarian).

Not every language has the words “a” and “the”, or has both.
Russian and Mandarin have neither.
English, German, and Spanish have both (and variations for gender/case).
Arabic just has a word for “the” but no “a”
and Turkish just has the word for “a” but no “the”

Step 3. Grammar

It's important for you to have your basic set of grammatical words like "many", "in", "before", etc before proceeding with this step.

A lot of suffixes/prefixes/etc. in languages come from previously separate words being condensed together. The plural ending in a lot of languages was at some point condensed into an ending from an old form of the word for "many" or "multiple".

This happens with verb tenses too. Like the "-miş" (reported past-tense ending) in Turkish originally came from the word "imiş" (meaning something like "was reportedly"). So the phrase "git imiş" (he went, reportedly/allegedly) was condensced down into "gitmiş", and so the -miş ending eventually became its own verb tense.

(btw reported tense or reported past tense is a tense in Turkish you use when talking about something you didn't witness yourself, or talking about knowledge you know secondhand from someone else. Phrases like "she told me he was at the party" or "apparently he was at the party" (but you didn't see him there yourself) would use the -miş ending. It's also used for telling fairtytales, stories, some news reports, and telling jokes "the skeleton walked into a bar..." obviously you've never actually seen a skeleton walk into a bar (or I hope not).)

Based on the simple words you created, "in", "after", "before", "until", etc you can condense these words into cases and verb tenses to create your language's grammar.


Head-Intial (Left-Branching) VS. Head Final (Right-Branching)

Grammar is basically made up of a *subject* and a *comment* on that subject.

For example, a noun is a subject, and an adjective describing that nouns is the comment.

"A grey car." = ("car" is subject, "grey" is comment).

The subject is the "head".

"Head-final" means the subject comes *after* the comment (like English: "A grey car.")

"Head-initial" means the subject comes *before* the comment (like French adjectives coming after the nous: "Une voiture grise.")

A language that follows a mostly head-initial pattern with its grammar is called a "Left-Branching Language" (ex: Turkish) because most or all the comments branch off onto the left side of (after) their subjects.

A language that follows a mostly head-final pattern with its grammar is called a "Right-Branching Language" (ex: English) because most or all the comments branch off onto the right side of (before) their subjects.


Head-Initial Order

noun -> adjective

verb -> object (English: "I kicked the ball" (the ball is the object, the verb comes before it))

noun -> adposition

possessee -> possessor



Head-Final Order

adjective -> noun (English: "a grey car" (car is subject))

object -> verb

adposition -> noun (English: "to the store" (the store is subject, "to" is the comment), "before the party" ("before" is the comment))

possessor -> possessee (English: "the man's dog" (dog is subject, "the man's" is the comment))

As you can see, English is not perfectly head-final. When it comes to the order of verbs vs. objects English takes a head-initial pattern. Your language doesn't have to follow the strict order of either. For each of these 4 sets of grammar scenarios, pick out if you want it to be the head-initial pattern or head-final pattern for your language.

It looks like this for Turkish.

Head-Initial Order
noun -> adjective
verb -> object
noun -> adposition (Turkish)
possessee -> possessor

Head-Final Order
adjective -> noun (Turkish)
object -> verb (Turkish)
adposition -> noun
possessor -> possessee (Turkish)

The object comes before the verb in Turkish because verbs go at the end of the sentence (like Japanese, Korean, Persian, and Hindi as well).

Adpositions come after the noun because Turkish has cases and postpositions. "To the grocery store" is "bakkala" where the final "-a" at the end is the case ending for "to".



"Before the party" = "Partiden önce".

"Önce" means "before".

"Before" in English comes first, so it's a preposition, but in Turkish "önce" comes afterwards, so it's a postposition.

"Parti" = party (no surprise)

"-den" ending is ablative case meaning "from". Turkish will use certain cases to match with certain postpositions. The literal word order is "Party-from before".



"Until 6pm" = "Saat altıya kadar."

"-a" at the end of "altıya" being dative case "to" again. "Kadar" is a postposition means "until".

The literal order is "Hour 6-to until".



"grocery store" = "bakkal"

"to the grocery store" = "bakkala"... literally "grocery store-to"

"at/in the grocery store" = bakkalda... lit. "grocery store-in/at"

"from the grocery store" = "bakkaldan"... lit. "grocery store-from"

"of the grocery store" = "bakkalın"... lit. "grocery store-of"

This is just some examples of cases and postpositions you could use for your language if you want ideas to do it differently than English does.

Based on whether the arrangement of your language's words is head-final or head-initial in certain scenarios will dictate how your words will merge into affixes.

If your language's adpositions come before the noun and adposition word eventually gets merged into a case conjugation, it will probably become a prefix onto the beginning of that noun (a lot of Bantu-African languages use prefixes, like Swahili for plural puts an "m-" at the start of the word, whereas English we usually put an "-s" at the end). Like "marafiki" (friends) in Swahili if you recognize the word "rafiki" (friend) from the name of the Lion King character.

To sum it up:

If your comment comes *before* the subject it will likely be condensed onto the beginning of the subject's root word, into a prefix or a beginning.

If your comment comes *after* the subject, it will likely be condensed onto the end of the subject's root word, into a suffix or an ending.

Step 4. Verbs

Most of the world's largest languages have verb conjugations based on person. Swedish, Norwegian, Japanese, Korean, and Mandarin don't have it.

A lot of other languages like French, Spanish, Italian, German, Arabic, Hindi, Persian, Turkish, Hebrew, Russian, Polish, etc. do have it.

English sorta has a trace of the conjugations Old English used to have.
"I am"
"You are"
"He/she is"

The verb "to be" changes a lot on English based on conjugation, but in other verbs 3rd person still adds that "-s" onto the end (he/she sees, he/she walks vs. I see, you walk)

If you choose to have verb conjugations like this you can conjugate them at the end, beginning, middle, or BOTH beginning and ending (so the root of the verb is in the middle, called a circumfix).

Zulu conjugates with beginnings/prefixes.

The "Ngi-" prefix means "I".

"hamba" means "to go".

so "Ngihamba" means "I go".

once you have conjugations for a verb by person/pronoun, the pronoun itself can be irrelevant. Like in Spanish instead of saying “yo camino” (I walk) all the time, you can just say “camino”. This is “pronoun drop” or just “pro-drop”. The pronoun will only be added back in sometimes for emphasis.
Or your language could be like Japanese where they don’t conjugate verbs but still drop pronouns anyway, and then you have to guess based on context who is doing the verb in the sentence.


Casually we use "verb tense" as a broader definition than what it actually is.

A verb tense is actually just about the time the verb takes place in, so there are only 3 options (past, present, future).

Things like continuative (-ing at the end of English verbs), completed, conditional, perfect, imperfect, subjunctive are all called verb aspects and moods, not tenses.

You can have as many or as little verb tenses, moods, and aspects you want.

Hungarian just has a simple past, present, and future.

Turkish has 20 verb tenses including the "-miş" tense that is unique to Turkish (although Albanian slang borrowed it and uses it too sometimes).

Mandarin doesn't use verb tense, but they have aspect conjugations to describe if the action is still ongoing or has been completed. "I am (still) eating" vs. "I have finished eating".

Mandarin also conceives of time in a different pattern than English does. Westerners generally think of time as being "backwards to forwards" but in Mandarin they express time as being "down to up." (xia -> shang).

Negation / Negative Verbs

English conjugates negative verbs with "do not" or "don't", using the auxiliary verb "to do"

Spanish, Polish, Russian, Hungarian, etc... just use the word "no" and "not" in front of the verb.

Spanish - "no quiero" = I don't want / "no" also means "no"

Polish - "nie chce" = I don't want / "nie" also means "no"

Hungarian - "nem akarok" = I don't want / "nem" also means "no"

but Turkish conjugates negation as an ending
"istemiyorum" = I don't want
"iste" = is the root word for "want"
"m" = is the negation
"iyor" = continuative present tense
"um" = I

So play around with how your language does negation, it doesn't have to be the same kind of pattern a lot of languages do, you can come up with something that's uncommon among real-world languages.

——————

some languages like Russian often drop the verb “to be”. It’s enough to say “I hungry” in Russian than it is to say “I am hungry.”

some will conjugate it instead of using the verb “to be” so the noun or adjective itself is like a verb (except for 3rd person usually it just goes to “he hungry” and “they hungry”)
acım (I’m hungry)
acın (you’re hungry)
o aç (he/she/it’s hungry)
acımız (we’re hungry)
acınız (you (pl.) are hungry)
onlar aç (they are hungry)

Step 5. Building Vocabulary

Compound words:

Take basic words you've already come up with in your language and combine them to make new words. "Back" and "yard" make "backyard" in English.

"Zhongguo" means "China" in Mandarin, and "ren" means "person", so "Zhongguoren" means "Chinese" (Chinese person).

Switching Verbs, Adverbs, Adjectives, etc:

Have affixes or context differences to change the meaning of words. "Picking" is a verb, but in the context of "berry-picking" it becomes a noun that describes a pass-time. Also in expressions like "my choosing", "choosing" is a verb being used like a noun based on context.

"Slow" is an adjective, but add a suffix like "-ly" and now you have the adverb "slowly".

So you can reuse already existing words like verbs, and change them slightly or add endings to change them into a different form of word.

Adding the ablative case to "ne?" (what?) in Turkish becomes "neden?" which literally means "from what?" but eventually came to have the meaning "why?".

Contractions:

English has contractions like "they're" and "we're" being contractions of "they are" and "we are."

But usually other languages don't use appostrophes for contractions like this, for example the Turkish word "nerede?" (where?) is a contraction of the phrase "ne yerde?" (at what place?).
”nereye” (to where?) comes from “ne yere?” (to what place?)
“nereden” (from where?) comes from “ne yerden” (from what place?)

“öyle” means “such, such as...”
“bu öyle” becomes “böyle” (such as this)
“şu öyle” becomes “şöyle” (such as that)

Consonant Root System:
another idea you could use for forming words could be something similar to Arabic and Hebrew’s tri-consonantal root system.
Arabic root words are mainly based around (usually 3) consonant sounds that have a common meaning, but no vowels. You have to slot these consonants into an arrangement of vowels to clarify the meaning.

One set of consonants is “k,t,b” which vaguely refers to books.
“kitab” means book
“kutub” means books
“kutayib” means booklet
“aktub” means write
“maktaba” means library
all 5 share the same “k,t,b” consonants in a different arrangement of vowels

save 1

save 2

save 3

Agglutinative Languages vs. Fusional Languages:

Languages that have conjugations like Spanish, Russian, Latin, German, etc. are called Fusional languages because of the way they conjugate words is a bit messy and irregular generally. They will add endings onto words but the endings have morphed into part of the word in irregular ways. This usually involves the language defining grammatical genders in order to categorize the irregular ways different words can conjugate.

More Agglutinative languages (a-"glue"-tinative) conjugate and add endings more cleanly and logically, where the ending doesn't get morphed or blended into the root word, and generally the ending stays the same no matter what word it's added onto. For example Hungarian, Finnihs, and Turkish are highly agglutinative languages. These 3 languages don't have any genders, but in order to make endings work they will use sound change rules like vowel harmony and consonant degradation to make the ending "match" or "rhyme" with the word it's glued onto.

Vowel harmony works by splitting the vowel sounds in the language into categories, front vowels, middle vowels, and/or back vowels. This is based on where in your mouth the vowels are pronounced.
If the last vowel of the root word is a back vowel, the vowel(s) in the ending should change into a back vowel as well in order to "rhyme" and make the whole conjugated word sound smoother.

Consonant degradation means if there is a hard consonant at the end of the root word, it might get turned into a similar but softer consonant when the ending is added after it, just to make the root word flow better into the ending.

(Personally I like agglutinative languages better because you can think through vowel harmony and consonant degredation logically if you forget the right ending, rather than having to remember or look up which gender the word is).

Cases

"I kicked the ball from the line to Susan in the net."

General cases are...
nominative - the subject of the sentence, the one doing the verb, which in this case would be "I"
accusative - the direct object of the sentence, the one directly affected by the verb, in this case "the ball"
dative - the indirect object of the sentence, means "to", "towards", "at", "for", in this case "to Susan"
ablative - means "from"
locative - location, means "in", "on", "at", in this case "in the net"
vocative - a command
essive - means "as a ___"
instrumental - means "with ___" or "using ___" (he eats with/using a spoon)...(she goes to the store with Todd)

"Ben ağ'da Susan'a çizgiden topu tekmeledim."

I kicked the ball from the line to Susan in the net."

nominative
accusative
dative
ablative
locative
verb

a lot of languages take this information that takes a lot of little words (to, at, in, etc) in English and make them into endings (or in some languages prefixes)

Possession (or Genitive Case)

a simple way to do possession is to just have a word for “of” like “the dog of the man”

or like English, the possessor (the man) will have an affix (‘s) = “the man’s dog”

Turkish (and Hungarian and Finnish) has affixes for both the possessor (the man) and the possessee (the dog). “Adam” (the man) becomes “adamın” (the man’s) + “köpek” (the dog) becomes “köpeği” (“his dog”). Adamın köpeği.

some languages can also conjugate nouns based on who owns them (Turkish, Hungarian, and Finnish again lol)
köpek (dog)
köpeğim (my dog)
köpeğin (your dog)
köpeği (his/her/its dog)
köpeğimiz (our dog)
köpeğiniz (your (pl.) dog)
(again with agglutinative languages, consonant degradation happens where the hard “k” at the end gets changed to a soft “ğ” (which is basically just a silent letter) when certain endings get added on.)

i'm looking back and i can't believe I wrote all of this wow