I need help with words

[Furkan]

Hello, I'm designing a conlang and I need your help.

I am designing a conlang for the first time, my goal is to design a language that people can use to communicate in daily life.

I identified the letters in my alphabet and started to produce words. I wasn't very good at making up words so I wrote a simple program and now I can generate random words.

However, there is a problem. What will these words mean? My main language is Turkish and I plan to give the words a Turkish meaning. I actually have no problem with doing this. The program generates random words and I give them a meaning.

The real question is: What words do I need? I mean, what words are indispensable in a language? How did you decide which words to include in your language?
I've heard that you need to know 8000+ words to be able to speak a language. Should my language also have more than 8000 words so that people can speak? Is this a must?

In order to give meaning to my words, I tried to look at the list of the most used words in Turkish, which is my main language. Is this a good idea?

[Arayaz]

Hi Furkan, and welcome to the board!

This seems like an interesting project, especially considering that most auxlangs (auxiliary languages, designed for communication between people who don't share a common language) are based on English or a Romance language.

It's important to realize the distinction between how often a word is used, and how important it is to a language. For example, "a" is a very common word in English, but ... what would change if we removed it? The distinction between "fire" and "a fire"? Lots of languages get along fine without it.

Imagine if instead we removed the word "building." It's of course a rarer word, but now suddenly there's an entire concept that we'd have to substitute with some other name.

To answer your question, the fact that people need 8,000 words to speak a foreign language isn't very accurate. It's obviously possible to speak a language with fewer words. But I suggest that you take some roots, maybe a thousand or so, and assign them to basic concepts. Then create a system of derivation. As an example of how this works, let's say I have words for "snow," "to cook," and "dark green." I also have derivational methods meaning "place of," "very/a lot of," "one who is/does," and "-like." Suddenly, we can have words for "tundra," "kitchen," "forest," "blizzard," "cook a huge meal," "deep green/very green," "chef," "plant," and "cold." This is the best way to make it easy to learn new words.

You'll also need to decide on phonology (what sounds the language uses) and grammar. Remember than not all languages are the same! If English speakers suddenly had to communicate with, say, Turkish but with different words, society might collapse, because Turkish is wildly different from English in many ways.

Good luck, and I'd like to hear more about your language!

[Furkan]

Thank you for your help. As I understand it, I should start by creating root words. I once asked ChatGPT to list the most used words in Turkish for me. Thanks to this list I got from ChatGPT, there are 350+ words in my language now. Do these count as root words?

[Arayaz]

Thank you for your help. As I understand it, I should start by creating root words. I once asked ChatGPT to list the most used words in Turkish for me. Thanks to this list I got from ChatGPT, there are 350+ words in my language now. Do these count as root words?

A root is ideally a word with a simple, lexical (not grammatical) meaning. "Object," or "person," or "blue" would make a good root. "The," or "kraken," or "personal outlook on life" would not.

[Furkan]

I understand This means that there are words that can be root and not root in my language right now.

How will I know if I have produced enough words?

[Arayaz]

I understand This means that there are words that can be root and not root in my language right now.

How will I know if I have produced enough words?

Most languages have quite a few roots. Since you created the language, of course, you can add a root if you need one. But it's also important to think about how you might be able to derive the word you need from another word plus a derivational affix.

And remember to think about the grammar too!

[Furkan]

I understand. Grammar also seems to be challenging. I guess making a language is not as easy as learning a language that has already been made :D Syntax is SOV, SVO etc. right? So it's not grammar.
How do I create the grammar?

[Salmoneus]

I must correct Udj here. A 'root' is simply an important part of a word that contains non-grammatical meaning by itself, and can't easily be analysed in terms of more basic roots. [that's not an 'official' linguistic definition but I think it captures how most people use the word usually]. "Kraken", therefore is, in English at least, a root.

The reason why this concept is useful is that most languages do not only have roots - that is, they have some words that consist of a root plus something else.

The 'something else' can be 'derivational' (it creates a 'new word' with its own non-grammatical meaning) or 'inflectional' (it changes grammatical details about the word in a predictable way).

So, given the root noun "whale" we can add the inflectional ending "-s", indicating a plural, to create "whales". This will go under the same dictionary entry as "whale", because the concept in question is considered the same - we're still talking about whales, but now there's more than one of them.

Alternatively, we can add the derivational ending "-er", which creates new nouns usually relating to doers of actions, or implements by which actions are done. A "whaler" is a person who hunts whales, or a ship used in hunting whales.

We can also add an additional root noun in order to derive a new noun, which we call a "compound" noun. So from "whale" and "bone" we get "whalebone" (which isn't actually a type of bone, but maybe people thought it was when the word was invented, I don't know). Or from "kill" and "-er" and "whale" we get "killer whale", a name for the orca, which is actually a dolphin, not a whale.

[many derivational endings ultimately derive from old compounds where the original meaning of one root has been obscured and forgotten]

Some languages make a lot of use of derivational endings. Others use a lot of compounding. Either option allows you to increase the number of words your language has without having to invent more root nouns.

However, this is a bit of an illusion, because you still need just as many nouns. And although it may SEEM easier to just remember ten root words and ten derivational affixes, rather than having to remember one hundred nouns... you do actually still have to remember one hundred words. Because derivation - by affixes or compounding - is non-transparent: you often can't actually work out the meaning from the elements. From the words "whale", "kill" and "bone", plus "-er", you would not be able to just guess the meanings of "whaler", "whalebone" or "killer whale". From the words "black" and "bird", you might guess that a blackbird was a black bird... but you wouldn't know which one. From "vomit" and "-orium" (a rare suffix used to create words for places) you would not guess that a vomitorium was simply an exit passage - not a place for vomiting. You also have to learn which words CAN take which affixes and compounds, because not all the possible combinations will be used. So the apparent advantages of relying on roots and derivation are somewhat illusory (although it can be easier than a language with NO derivation, of course!).

[conceptually, there is no such thing as a 'simple' meaning, and 'simple' meanings aren't necessarily associated with roots.]

[Arayaz]

I understand. Grammar also seems to be challenging. I guess making a language is not as easy as learning a language that has already been made :D Syntax is SOV, SVO etc. right? So it's not grammar.
How do I create the grammar?

I'd include syntax under grammar. But here's a basic framework for a very easy-to-learn language:

- SOV word order (since it's the most common in the world)
- Adjectives precede nouns (what you'd expect from a language with SOV word order)
- Postpositions instead of prepositions

- All words have exactly one form, and particles (small grammatical words) indicated tense and number, and shouldn't be required.

- You speak Turkish and English, which are pretty different ─ try thinking of a way to do the grammar that people who only speak Turkish and people who only speak English would find easy.

And Salmoneus, I was merely saying that "kraken" probably shouldn't be a root in an IAL.

[Salmoneus]

On the subject of grammar:

'Grammar' traditionally consists of two things: syntax (the order of words) and morphology (how words change to reflect both syntactic function and regular alteration in meaning).

So, in English, knowing that the plural of "function" is "functions", and that the accusative of "I" is "me", is morphology. Knowing that it's "I ate the cat" and not "the cat ate I" is syntax. [the line between syntax and morphology is not always that precise and obvious, though].

However, syntax is much bigger than just the relative position of S, O and V.

Syntax is generally concerned with:

- identifying the roles of the participants in whatever you're talking about (S, O and V)

- identifying which things the speaker considers more important or less important, or which things the speaker expects the listener to already know about and which things will be new.

- linking together different clauses into larger sentences in different ways. [in English that's things like "I ate but the fish didn't", "I ate while the fish ate", "I ate the dog and ran", "I ate the dog that the fish saw", and so on].

[Salmoneus]

I'd include syntax under grammar. But here's a basic framework for a very easy-to-learn language:

- SOV word order (since it's the most common in the world)

That's not even true, Udj. At least, I've never seen anything proving it. In general, SOV and SVO are about equally common. And I'd argue that since most languages have V precede O, SVO is probably more intuitive for an "easy-to-learn" language. Before we even consider that nine of the eleven most-spoken languages, including the big two (English and Mandarin - plus Spanish, French, (most) Arabic, Russian, Portuguese, Indonesian and German) are all SVO!

- All words have exactly one form

Why would that make things easier!? I mean, it might be easier for speakers of Mandarin and English, sure, but not for most of the world!

And Salmoneus, I was merely saying that "kraken" probably shouldn't be a root in an IAL.

Why not? What are you going to say instead, "giant mythical underwater possibly-tentacle-y or possibly crab-y it's not really clear monster of some kind"? How is that easier than saying "kraken"?

[Arayaz]

I'd include syntax under grammar. But here's a basic framework for a very easy-to-learn language:

- SOV word order (since it's the most common in the world)

That's not even true, Udj. At least, I've never seen anything proving it. In general, SOV and SVO are about equally common. And I'd argue that since most languages have V precede O, SVO is probably more intuitive for an "easy-to-learn" language. Before we even consider that nine of the eleven most-spoken languages, including the big two (English and Mandarin - plus Spanish, French, (most) Arabic, Russian, Portuguese, Indonesian and German) are all SVO!

In fact, a wide range of resources can tell you that SOV is slightly more common than SVO. Though I do concede that perhaps SVO is the better choice in this specific context.

- All words have exactly one form

Why would that make things easier!? I mean, it might be easier for speakers of Mandarin and English, sure, but not for most of the world!

Observe creoles/pidgins, which are literally what happens when people try to simplify a foreign language so that it's easier for them to learn. They usually have much simpler, or even lack, inflection.

Also, in your last criticism, you noted that English and Mandarin were the "big two" and should count more than other smaller languages... please be consistent in your arguments, it makes it easier for both of us.

And Salmoneus, I was merely saying that "kraken" probably shouldn't be a root in an IAL.

Why not? What are you going to say instead, "giant mythical underwater possibly-tentacle-y or possibly crab-y it's not really clear monster of some kind"? How is that easier than saying "kraken"?

ahem...

sea + monster

And of course "kraken" isn't even a basic root in Proto-Germanic. It's derived from *krankaz- "twisted."

[Khemehekis]

And Salmoneus, I was merely saying that "kraken" probably shouldn't be a root in an IAL.

Why not? What are you going to say instead, "giant mythical underwater possibly-tentacle-y or possibly crab-y it's not really clear monster of some kind"? How is that easier than saying "kraken"?

My thoughts on this: 8,000 words is simply not enough to have all the specialized and technical vocabulary we'll need, especially for things like plants and animals, or monetary units, or music genres, or musical instruments. Esperanto, after all, has 77,000 words in its largest dicgtionary.

[Khemehekis]

ahem...

sea + monster

And of course "kraken" isn't even a basic root in Proto-Germanic. It's derived from *krankaz- "twisted."

I thought "sea monster" would be a sea serpent (Loch Ness Monster, Ogopogo, Champ, Mokélé-Mbembe, etc.)

[eldin raigmore]

I understand. Grammar also seems to be challenging. I guess making a language is not as easy as learning a language that has already been made :D Syntax is SOV, SVO etc. right? So it's not grammar.
How do I create the grammar?

Grammar = syntax + morphology.
Theoretically you could have all-syntax-and-no-morphology ; or all-morphology-no-syntax.
(I don’t know any natlang that does either of those extremes; but a few seem to come close!)
….
Morphology considers all the ways your words get inflected or derived (prefixed or suffixed or infixed or ablauted) to form related wordforms; such as different cases and/or numbers (and/or various other categories) for nouns, degrees-of-comparison (and/or other features) for adjectives (and, possibly, adverbs, if you have any); valence, person, tense, aspect, mood, polarity, voice, etc. for verbs; and whatever else for those or whatever other parts-of-speech you have.
….
I don’t want to risk discouraging you nor making you more confused. So I won’t say more until you ask!
What other folks have said here, seems good to me!

[Salmoneus]

The real question is: What words do I need? I mean, what words are indispensable in a language? How did you decide which words to include in your language?

There is no absolute list.

Four different things might be useful to consider here.

Firstly, there are the meanings that are most universal - that most languages have a specific, simple word for. These words also tend to be borrowed between languages less because they're so fundamental. But these words aren't necessarily what you might expect - they're not things like "object", they're things like "rope" and "louse". As a result, they're not necessarily very important for communication in business - you can talk to many business people, you can read many books, you can listen to many news reports or what many TV shows, and not need to learn the word for "louse" in that language.

Then again, if you're raising children in a place that speaks a language, you may have to learn the word for "louse" sooner rather than later. Which words are useful depends on what you're trying to talk about.

[a lot of the simplest, shortest, oldest, most fundamental words in English are of very little use in the modern world, because they refer to slightly different types of plant, or slightly different types of fish. Which would have been hugely important, centuries ago. But now many people probably can't tell an elm from an ash, a birch from a beech, even an oak from a larch. They certainly can't tell a perch from a pike or a char from a shad...]

----

Secondly, as you've thought of, there's the words that are most frequently used. But the problem there is that 90% of the average text uses only a few hundreds words... but 90% of the meaning is stored in the rare words. [hence that annoying feeling when reading a text in a language you've studied a bit but not enough - you can understand how the sentence works and all the unimportant bits, but don't know the one word that is actually important....

-----

Thirdly, there's the words that dictionary makers use to assemble dictionaries. There used to be lists of these, but I can't find one, I'm afraid. And of course the words that are useful in defining other words aren't always the words that are useful in a normal conversation.

----

And fourth: there have been many attempts to create a "Basic English", "Simple English", "Global English" or "Business English" vocabulary - a list of a small number of words that the creator feels will be particularly useful to learn in order to get by in English. Of course, other languages won't always have exact one-to-one translations of these words, and different languages will use some words more or less, depending on their approach to describing things, but it might be a starting place.

------------

I've heard that you need to know 8000+ words to be able to speak a language. Should my language also have more than 8000 words so that people can speak? Is this a must?

-----------

There are no 'musts' in conlanging, which is after all only to please your own impulses. Nobody will ever actually learn your language anyway (at least, it's incredibly unlikely).

It will vary somewhat depending on the language and what it considers a "word" to be.

In English and similar languages, a rule of thumb I've heard is that 2,000 words is enough to have some basic conversations and understand a lot of texts, but you'll need to guess a lot of words you don't recognise. 5,000 words is enough to understand most texts and conversations, provided they're not specialised, and is enough that you'll probably be able to guess accurately and learn new words from context. 10,000 words is enough to pass as a native English speaker in general conversation, though you won't stand out as having a good vocabulary and you may not feel comfortable with specialised texts, or literary texts that use a rich vocabulary. Someone with a big vocabulary might know 20-50,000 words. The OED has 600,000 words, though many will be archaic or extremely specialised.

[Khemehekis]

The Longman Defining Vocabulary:

https://www.pu-kumamoto.ac.jp/users_sit ... s/LDV.html

[Salmoneus]

I'd include syntax under grammar. But here's a basic framework for a very easy-to-learn language:

- SOV word order (since it's the most common in the world)

That's not even true, Udj. At least, I've never seen anything proving it. In general, SOV and SVO are about equally common. And I'd argue that since most languages have V precede O, SVO is probably more intuitive for an "easy-to-learn" language. Before we even consider that nine of the eleven most-spoken languages, including the big two (English and Mandarin - plus Spanish, French, (most) Arabic, Russian, Portuguese, Indonesian and German) are all SVO!

In fact, a wide range of resources can tell you that SOV is slightly more common than SVO. Though I do concede that perhaps SVO is the better choice in this specific context.

Citations needed. For instance, WALS does have SOV very slightly more common - 564 to 488 - but that's fairly meaningless given the tiny sampl size and the fact that "no dominant order" is 189 (as many of these languages will be classed as either SOV or SVO by other analysts).

- All words have exactly one form

Why would that make things easier!? I mean, it might be easier for speakers of Mandarin and English, sure, but not for most of the world!

Observe creoles/pidgins, which are literally what happens when people try to simplify a foreign language so that it's easier for them to learn. They usually have much simpler, or even lack, inflection.

That's not what pidgins are, and it's definitely not what creoles are. It's a lot more complicated than that. [simplistically: pidgins are attempts to communicate with others without learning their language, by using their words without learning their grammar. This process strips the imported words of grammatical distinctions that have no place in the grammatical matrix they're placed into]. In addition, these pidgins do not reflect the normal situation in language contact and language learning, but a very specific and relatively rare situation that's not fully understood yet. Many contact languages do not look like classical pidgins; some are even notable for their morphological complexity.

Nor do they reflect ordinary language learning conditions. Most English speakers, for instance, will find Chinese relatively difficult to learn.

Also, in your last criticism, you noted that English and Mandarin were the "big two" and should count more than other smaller languages... please be consistent in your arguments, it makes it easier for both of us.

There is no inconsistency here. There's a difference between "if a feature is as common as not, maybe a language looking for learners should make the same decision as widely-spoken languages" and "if a feature is in common betwee two widely-spoken languages, it must be more natural for all learners".

And Salmoneus, I was merely saying that "kraken" probably shouldn't be a root in an IAL.

Why not? What are you going to say instead, "giant mythical underwater possibly-tentacle-y or possibly crab-y it's not really clear monster of some kind"? How is that easier than saying "kraken"?

ahem...

sea + monster[/quote]

Not all seamonsters are krakens, and I think some would probably say that a kraken wasn't a "seamonster".

And of course "kraken" isn't even a basic root in Proto-Germanic. It's derived from *krankaz- "twisted."

"Kraken" wasn't a WORD in proto-germanic. All words are derived from something (you yourself suggested 'object', 'blue' and 'person', all of which are diachronically derived from other words. In English, 'kraken' is a root word with no synchronic derivation.

[Furkan]

I have read all the messages sent. I really learned a lot. First of all, I understood the concept of morphology and I think it will be much easier to build grammar than it was a few hours ago.

-----
- linking together different clauses into larger sentences in different ways. [in English that's things like "I ate but the fish didn't", "I ate while the fish ate", "I ate the dog and ran", "I ate the dog that the fish saw", and so on].
-----

I'm not entirely sure how I can do this. Will it be enough to complete the grammar or do I have to put in extra effort to make it possible?


-----
There are no 'musts' in conlanging, which is after all only to please your own impulses. Nobody will ever actually learn your language anyway (at least, it's incredibly unlikely).
-----

True, most likely no one will speak this language except me. I guess there is no right or wrong way to make languages.

-----
The Longman Defining Vocabulary:
-----

Thanks. I'm sure this will help.

[Khemehekis]

I'm not entirely sure how I can do this. Will it be enough to complete the grammar or do I have to put in extra effort to make it possible?

Completing the grammar means that there is a way to express any idea you want to express, provided your language has the available vocabulary.

For instance, does your language have a way to say:

1. Hand me that towel.
2. I am going to cry.
3. Pam was going to walk the dog, but she had an emergency this morning.
4. Steve would call some friends later that day.
5. Jack would go to the convenience store every Friday
6. Paul was drinking a soda.
7. The dog had already torn up my homework.
8. Do you like pizza?
9. Mike was arrested by the cop.
10. Let's go drink a few beers.
11. I got my house painted.
12. Sharon got her hair wet.
13. Tom has his eyes closed.
14. May you live for 100 years!
15. I believe that Putin is an idiot.