Evolving Nat-langs for a conlang/world

[Durakken]

So basically what I'm trying to do is to takes 4 natural languages and evolve them into other languages for use in a constructed world. Four of them to be exact, English, Japanese, French, and Arabic... This will obviously be problematic cuz I don't know Arabic at all and don't know French or Arabic well, but I'm leaving those aside for now as I don't need them just yet. How this comes about, as there is a reason, is irrelevant but these 4 languages are the roots of all languages in this world. They're largely isolated from each other so there is minor mixing on that front however there is supposed to be a little bit of mixture, just not much. The time span these languages evolve in is ~2750 years. The first 1000 of which the people are mostly nomadic and traveling over the same land, then the next 1000 any settlement that survives is mostly isolated from others, and finally the last ~750 years those settlements are able to expand. For simplicity sake and what it seems like from what I have seen language alter enough within 500 years, even if they're called the same thing, to be considered a different language so I've mapped out the path of the languages at ~500 year intervals...

As far as starting lexicon I'm using a word list that I found from a PDF on naming languages, Ogden's 1000 word list, my own look through meanings of names, and http://ielex.mpi.nl./meanings/ Once I get done adding and looking through all those I may add a few more that I think are needed, but are missing. For the basic rules of grammar I'm going to use Ogden's simplified English. I've also removed some words that simply didn't work... for example "industrious" from the naming language list which I don't see anyone from this culture ever using.

Soooo now for the questions...
1. I'm thinking of trying to remove all the weird complexities that I can for the transition between "English" and the "root" language, without throwing out sounds that I like. For example... I'd like to try to get rid dipthongs, but I like the sound from "out" ... so is there a way I could preserve that as a single vowel in some way? And is that unrealistic?

2. English benefits from having a lot of Suffixes and Prefixes, but those usually don't have stand alone meanings that would be preserved, I'd think, from the situation which is basically a culture having a simplified English magiced into them... and further the words that these affixes means are chunky and don't smoothly become an affix themselves without some mangling of the words... so what I'd like to do is somehow convert english back into, i don't know the word for, one of those languages that are additives so that you get things like the word for table being a combination like foursticksholdingupaflatsurface... and then somehow remove phoneme from that to become new affixes in later versions of the language, but not if that isn't realistic. Or does anyone have any suggestions for this...

3. Would it be a bad idea to get rid of the "taken for granted" aspirated phonetics so all descendant languages don't have to worry about them being there or not? It seems to me since I can't find a source that expressly states them that it would be better to just assume they've all been rendered as there or not there in all cases for the root language so that descendant languages can process a rule without worry that it's breaking an earlier rule, because I can't figure out which ones are /p^h/ and which ones are just /p/

Lastly... I'm using GA IPA pronunciation to make it easy on myself, despite not agreeing with all of it... and as soon as I get through getting all the words listed I'll have a link to the word list, pronunciation, and eventually it will have the steps from English to the final language. It is largely just for naming languages, but meh, I'll be using it at times in the web novel I'm writing, so if anyone's willing to help, I'd be grateful.

[CatDoom]

Soooo now for the questions...
1. I'm thinking of trying to remove all the weird complexities that I can for the transition between "English" and the "root" language, without throwing out sounds that I like. For example... I'd like to try to get rid dipthongs, but I like the sound from "out" ... so is there a way I could preserve that as a single vowel in some way? And is that unrealistic?

2. English benefits from having a lot of Suffixes and Prefixes, but those usually don't have stand alone meanings that would be preserved, I'd think, from the situation which is basically a culture having a simplified English magiced into them... and further the words that these affixes means are chunky and don't smoothly become an affix themselves without some mangling of the words... so what I'd like to do is somehow convert english back into, i don't know the word for, one of those languages that are additives so that you get things like the word for table being a combination like foursticksholdingupaflatsurface... and then somehow remove phoneme from that to become new affixes in later versions of the language, but not if that isn't realistic. Or does anyone have any suggestions for this...

3. Would it be a bad idea to get rid of the "taken for granted" aspirated phonetics so all descendant languages don't have to worry about them being there or not? It seems to me since I can't find a source that expressly states them that it would be better to just assume they've all been rendered as there or not there in all cases for the root language so that descendant languages can process a rule without worry that it's breaking an earlier rule, because I can't figure out which ones are /p^h/ and which ones are just /p/

Lastly... I'm using GA IPA pronunciation to make it easy on myself, despite not agreeing with all of it... and as soon as I get through getting all the words listed I'll have a link to the word list, pronunciation, and eventually it will have the steps from English to the final language. It is largely just for naming languages, but meh, I'll be using it at times in the web novel I'm writing, so if anyone's willing to help, I'd be grateful.

1. No two speech sounds are exactly alike... otherwise they would be the same sound. "Out" (in General American English) contains the diphthong [aʊ], so you're going to lose that sound if you get rid of all the diphthongs. The sound is extremely similar or identical to [a] followed by the semivowel /w/, however, so you might still have that sound if /w/ is allowed as a coda consonant, albeit not in the same words. That said, a vowel followed by a semivowel is a diphthong, phonetically, but it doesn't have to behave the same way as vowel would, phonologically, which diphthongs generally do. So, for instance, if a syllable could only have one coda consonant, you couldn't have a word like *awt (pronounced like "out"), while a diphthong like the one in the English word might be allowed before a coda.

2. I think what you're talking about is compounding, which English actually does a fair amount, mostly to form nouns like "armchair" or "sailboat." Shortening words is a pretty straightforward sound change, and has happened a lot in the history of English already, at least in part because English has a strong stress accent. It wouldn't be entirely unreasonable for unstressed syllables, or even syllables without primary stress, to gradually erode, such that, say, "sailboat" becomes something like [seɪlpt], or maybe [selpt] if you're getting rid of diphthongs. As meanings shift over time, you might eventually get something like "motselpt" for motor-boat, incorporating the old, eroded compound into a new compound word.

3. One thing to keep in mind is that sound changes affect sounds, not necessarily phonemes. Despite the fact that speakers don't consciously differentiate between [p] and [pʰ] in English, the two sounds would likely behave differently to some degree as the language evolves. It seems unlikely for the two sounds to simply merge, and the rules for when a stop becomes aspirated are fairly straightforward, so I wouldn't disregard the feature. Honestly, English (American English, at least) seems to be well on its way to becoming a language that uses aspiration as the primary distinction between stop series, with voiced stops being allophones of the plain voiceless series between sonorants. I, for one, have a lot of trouble distinguishing plain voiceless stops from voiced stops in word-initial position.

[k1234567890y]

I have something somewhat similar, although what I use is a West Germanic language ultimately evolved from (continental) Old Saxon(not Anglo-Saxon. so the Germanic language can be counted as a Low German dialect), a West Slavic language, Malay/Indonesian, Ainu, Japanese...

[Durakken]

1. No two speech sounds are exactly alike... otherwise they would be the same sound. "Out" (in General American English) contains the diphthong [aʊ], so you're going to lose that sound if you get rid of all the diphthongs. The sound is extremely similar or identical to [a] followed by the semivowel /w/, however, so you might still have that sound if /w/ is allowed as a coda consonant, albeit not in the same words. That said, a vowel followed by a semivowel is a diphthong, phonetically, but it doesn't have to behave the same way as vowel would, phonologically, which diphthongs generally do. So, for instance, if a syllable could only have one coda consonant, you couldn't have a word like *awt (pronounced like "out"), while a diphthong like the one in the English word might be allowed before a coda.

Well, see what I'm trying to do is make it easy on me, be realistic, and have it be a simplistic interpretation... Basicaly the culture listening to the words and then breaking them up into letter and I want 1 "sound" for each letter rather than multiple letters for a sound that doesn't necessarily mean that sound every time. So really I just want to be able to treat the dipthong as a single discreet unit rather than 2.

2. I think what you're talking about is compounding, which English actually does a fair amount, mostly to form nouns like "armchair" or "sailboat." Shortening words is a pretty straightforward sound change, and has happened a lot in the history of English already, at least in part because English has a strong stress accent. It wouldn't be entirely unreasonable for unstressed syllables, or even syllables without primary stress, to gradually erode, such that, say, "sailboat" becomes something like [seɪlpt], or maybe [selpt] if you're getting rid of diphthongs. As meanings shift over time, you might eventually get something like "motselpt" for motor-boat, incorporating the old, eroded compound into a new compound word.

yeah compounding is what happens but I'm pretty sure that's not what it is called.
For example "transform" the trans means something and that something is a discrete meaning, but it is not usable on its own as a word. I know it specific type of language name, but I cant remember what it is. Today, we make compounds, but in those languages you could have whole sentences as 1 words that were super complex and changing one syllable would change the meaning all together. The reason this is an issue is because names seems to be generally formed from those syllables so not having them 1 or 2 syllable words will look terrible imo, and really long names don't seem to be a thing either. Also, are there any good suggestions for which parts of a word will be lost? For example, is there any reason that in sailboat the "oa" would be dropped? Does it have something to do with making it monosyllabic and the second syllable is more likely to be dropped than the first or is there some other general concept there?

3. One thing to keep in mind is that sound changes affect sounds, not necessarily phonemes. Despite the fact that speakers don't consciously differentiate between [p] and [pʰ] in English, the two sounds would likely behave differently to some degree as the language evolves. It seems unlikely for the two sounds to simply merge, and the rules for when a stop becomes aspirated are fairly straightforward, so I wouldn't disregard the feature. Honestly, English (American English, at least) seems to be well on its way to becoming a language that uses aspiration as the primary distinction between stop series, with voiced stops being allophones of the plain voiceless series between sonorants. I, for one, have a lot of trouble distinguishing plain voiceless stops from voiced stops in word-initial position.

I agree, but at the same, I figure that since it is an evolved form of english that it wouldn't be all that much noticed if I did this, or is that just wrong?
My brain is not understanding what sonorants and such are. I've looked at several things to explain them and it looks simple, but something isn't clicking for me with it.


side question: Are there any lists of all sound changes that have happened... Let me be more clear... I don't mean a list of all sound changes between 1 language to another. I have that list. I mean as a base unit where all changes that a /b/ has been listed? I mean I'm pretty sure all /b/ > /ks/ would never be a valid change. I'm just curious if anyone has listed them and if anyone has figured out a frequency of change for each of those sound units. I ask because I might try to create a quick Javascript that could apply sound changes really quickly and easily in a realistic "random" way which would make this far quicker and easier ^.^

[Élerhe]

3. Would it be a bad idea to get rid of the "taken for granted" aspirated phonetics so all descendant languages don't have to worry about them being there or not? It seems to me since I can't find a source that expressly states them that it would be better to just assume they've all been rendered as there or not there in all cases for the root language so that descendant languages can process a rule without worry that it's breaking an earlier rule, because I can't figure out which ones are /ph/ and which ones are just /p/

In most English dialects the difference is mostly voiced stops <b d g>=unaspirated voiceless stops /p t k/, with the voiceless stops <p t k> being aspirated /p^h t^h k^h. Allophonically the stops are unaspirated in some positions (like after s- - Welsh, in fact, incorporates this into it's spelling system - 'spot' would be spelt 'sbot'

2. English benefits from having a lot of Suffixes and Prefixes, but those usually don't have stand alone meanings that would be preserved, I'd think, from the situation which is basically a culture having a simplified English magiced into them... and further the words that these affixes means are chunky and don't smoothly become an affix themselves without some mangling of the words... so what I'd like to do is somehow convert english back into, i don't know the word for, one of those languages that are additives so that you get things like the word for table being a combination like foursticksholdingupaflatsurface... and then somehow remove phoneme from that to become new affixes in later versions of the language, but not if that isn't realistic. Or does anyone have any suggestions for this..

The prefixes and suffixes can have their own meaning, or have another meaning in another language - trans, for example means beyond in Latin. You could calque the words, as in German - tele-vision became Fernsehen - both compounds mean 'far vision'. In which case just use native English words as materials for compounding - a TV could be a Farseer, for example.

Basicaly the culture listening to the words and then breaking them up into letter and I want 1 "sound" for each letter rather than multiple letters for a sound that doesn't necessarily mean that sound every time. So really I just want to be able to treat the dipthong as a single discreet unit rather than 2.

Then just create a single letter for that diphthong <ô> for example, now stands for /aw/. The alphabet you use (or more specifically the Romanisation) should reflect the phonology, not the phonology be constrained by the writing system. However it make it sound like you are just going to be creating languages with only 5 vowels if you're limited by the writing system...

Also, are there any good suggestions for which parts of a word will be lost? For example, is there any reason that in sailboat the "oa" would be dropped? Does it have something to do with making it monosyllabic and the second syllable is more likely to be dropped than the first or is there some other general concept there?

Unstressed syllables are more likely to be dropped - in the case of sailboat /seil'bout/ it would actually be the first syllable that's be likely to disappear /sil'pout/ > /siu'po:t/ > /Spo:t/ <shpôt> could be a possibility, along with some other phonetic/orthographic changes as an example. Of course you could change the stressed syllable.

side question: Are there any lists of all sound changes that have happened... Let me be more clear... I don't mean a list of all sound changes between 1 language to another. I have that list. I mean as a base unit where all changes that a /b/ has been listed? I mean I'm pretty sure all /b/ > /ks/ would never be a valid change. I'm just curious if anyone has listed them and if anyone has figured out a frequency of change for each of those sound units. I ask because I might try to create a quick Javascript that could apply sound changes really quickly and easily in a realistic "random" way which would make this far quicker and easier ^.^

we have that already - the sound change applier http://zompist.com/sca2.html from Zompist

and for examples of sound change, there are plenty of examples on this site or on Zompist's pages about his own language languages

[CatDoom]

Unstressed syllables are more likely to be dropped - in the case of sailboat /seil'bout/ it would actually be the first syllable that's be likely to disappear /sil'pout/ > /siu'po:t/ > /Spo:t/ <shpôt> could be a possibility, along with some other phonetic/orthographic changes as an example. Of course you could change the stressed syllable.

Sailboat was a bad example; I'm pretty sure that in most dialects the primary stress is on the first syllable, but the second syllable has secondary stress, and the difference between the two is minor to nonexistant. A better example would be something like "America" which, at least steriotypically, is already pronounced in some dialects as [ˈmɛɹɪkə].

[Durakken]

In most English dialects the difference is mostly voiced stops <b d g>=unaspirated voiceless stops /p t k/, with the voiceless stops <p t k> being aspirated /p^h t^h k^h. Allophonically the stops are unaspirated in some positions (like after s- - Welsh, in fact, incorporates this into it's spelling system - 'spot' would be spelt 'sbot'

Could you explain that with less linguistic terminology?
"Voiced" and "Voiceless" I can't identify enough to make that distinction when trying to figure out which is which.
"Stop" means what exactly?

Then just create a single letter for that diphthong <ô> for example, now stands for /aw/. The alphabet you use (or more specifically the Romanisation) should reflect the phonology, not the phonology be constrained by the writing system. However it make it sound like you are just going to be creating languages with only 5 vowels if you're limited by the writing system...

Plan on doing that, but knowing that we english speakers don't pronounce words correctly if there is any way to mess it up I am trying to make it as hard to mispronounce as possible ^.^ and easy for me to remember. I guess I'll figure something out, probably close to what you said.

Unstressed syllables are more likely to be dropped - in the case of sailboat /seil'bout/ it would actually be the first syllable that's be likely to disappear /sil'pout/ > /siu'po:t/ > /Spo:t/ <shpôt> could be a possibility, along with some other phonetic/orthographic changes as an example. Of course you could change the stressed syllable.

cool... Is there also rules for adding syllables and sounds?

side question: Are there any lists of all sound changes that have happened... Let me be more clear... I don't mean a list of all sound changes between 1 language to another. I have that list. I mean as a base unit where all changes that a /b/ has been listed? I mean I'm pretty sure all /b/ > /ks/ would never be a valid change. I'm just curious if anyone has listed them and if anyone has figured out a frequency of change for each of those sound units. I ask because I might try to create a quick Javascript that could apply sound changes really quickly and easily in a realistic "random" way which would make this far quicker and easier ^.^

we have that already - the sound change applier http://zompist.com/sca2.html from Zompist

and for examples of sound change, there are plenty of examples on this site or on Zompist's pages about his own language languages[/quote]

I know of that... and it isn't what I described.
What I mean is a program that selects several rules to apply and then applies them, rather than you making the rules.

Also I'm not looking for examples of rule changes, I'm looking for a list of rule/sound changes that have happened, not per language but in an absolute way.
There is a document that originates from here that that has something like 400 languages listed and all their changes, but I want the individual change and the likelihood of a change happening per sound...as well as which percentage chance of a sound becoming another sound.

For example... let's say there is a /b/ in the language I want to know is what percent chance will it change... say 50% chance... so the algorithm will roll a random number vs 100. if it's below 50 it won't change, but if it's above it will. If it changes what are the possible results of it change? And of those possible results what are each of their possible likelyhood. So let's say their is 5 possiblities /p/, /b[suph]/, /v/, /d/, and /g/ and each of them have 20% chances of happening then the algorithm will roll for which.

So you can see why I'd need those numbers, the last is not really needed, but without it you'd get less realistic results. But you can see that this is a different thing than that page, what i'm looking is a list, not examples in particular. If I'm really desperate I could just painstakingly go through that PDF and write down all of them. Seems a bit more work than I want to do, but I could do it.

[Durakken]

This is the current word list I'm working on... Please look through and see if you'd suggest any alterations...

https://docs.google.com/spreadsheets/d/ ... sp=sharing

animal types have been removed as have been metals
I'm trying to make it all present tense and non-compound with as few affixes in keeping with the idea of using Ogden's simple English as a base.
Still in the air about words like "Connection" and "Advertisement"/"Government" as I don't know how the affixes work in regular english to decide and have only briefly looked at the rules of OSE and I'm pretty sure it doesn't have those affixes.

And the words come from
Ogden's Simple English
A naming language tutorial PDF which came from a conlang site i can't remember
http://ielex.mpi.nl./meanings/
My own search through name meanings

The list on the right is from the 1000 most commonly used words which I then removed past tenses, metals, animals, and words already on the list. I have not decided if I'm going to use them all yet...

[Élerhe]

I really think you need to read the Language Construction Kit again, both for the basic vocabulary, and to get an idea of why completely random sound changes kind of defeats being a conlanger...

[CatDoom]

Applying statistically likely random sound changes is probably more complicated than you're imagining. For one thing, I'm not sure that good data exists on how common a given sound change is, and it would be hard to generate that data since sound changes are diverse and usually conditioned by features of both the individual word or utterance and the language itself.

For instance, /k/ > /tʃ/ is a sound change that has happened in plenty of languages, but it's highly unlikely to happen as an unconditioned change. In Old English, this change happened when k came before /i/ or /j/ and when after /i/. In other languages, however, the change has been known to happen around non-high front vowels as well, and sometimes it only affects velar stops on one "side" of the vowel. In some Athabaskan languages, on the other hand, every velar stop was palatalized, while the protolanguage's uvular stops shifted forward to become velar, regardless of the phonetic environment.

Which brings me to another potential issue, which is that sound changes feed into and influence one another. Although English generally doesn't have any uvular consonants, you could conceivably have a sound change that created uvular stops, and then have all of the back stops shift forward like in the Algonquian example. You could conceivably then have some of the remaining /k/s undergo palatalization to /tʃ/ in certain environments, but this might be rendered less likely (or more likely to be accompanied by changes to the existing /tʃ/), since /tʃ/ would have become a very common sound in the language and more semantic weight would be placed on distinguishing it from /k/.

[Durakken]

CatDoom,
I get what you're saying, but I don't think that is a problem. What you're explaining is just a nested structure of sound changes that rely on each other, that if x happens then y has some chance of happening as well. And the second issue isn't really an issue, because that rule change could happen with the original whether the prvious change happens or not, it just makes the both more pronounced than what it would be otherwise.

Elerhe,
Reading the same thing again won't help. Having it explained in a different way might help, but apparently you think not oh well.
"completely random" is not what I'm talking about and how I choose to construct a language doesn't "defeat being a conlanger". It's just a different way of doing things.


Lastly,
So no input on the word list?

[Mâq Lar]

I ask because I might try to create a quick Javascript that could apply sound changes really quickly and easily in a realistic "random" way which would make this far quicker and easier ^.^

What you're describing is not quicker, not easier, and not realistic. Because:

I know of that... and it isn't what I described.
What I mean is a program that selects several rules to apply and then applies them, rather than you making the rules.

Rules aren't random collections; they are interdependent. The clearest example is vowels: the vowels /a i u/ are the most different from each other in terms of mouth position and sound; they define a triangle. Other vowels fit between them, but they tend to be spaced out, nit all clustered in one part of the triangle. When one of the vowels undergoes a sound change, however, it moves to a different part of the triangle, causing it to become crowded, and as a result, other vowels in that area will change in response, moving away to reduce the crowding and increase the difference between the sounds. It's almost exactly like a crowd of people; if one area gets crowded, people move in response. Consonants do this as well as vowels. If you write a script that applies sound changes randomly, you will get impossible combinations of sound changes, like "all vowels change to /i/". If you wrote a script to model the distribution of people in a crowd by ignoring their interactions and just applying statistically common movements, you'd end up predicting that everyone would be standing on each other's shoulders in front of the most popular shop.
You need to just pick a few sound changes yourself that fit together, and apply them. Look through that 400 language list for one with changes you like, change it a bit and apply it.

For example... let's say there is a /b/ in the language I want to know is what percent chance will it change... say 50% chance... so the algorithm will roll a random number vs 100. if it's below 50 it won't change, but if it's above it will. If it changes what are the possible results of it change? And of those possible results what are each of their possible likelyhood. So let's say their is 5 possiblities /p/, /b[suph]/, /v/, /d/, and /g/ and each of them have 20% chances of happening then the algorithm will roll for which.

So you can see why I'd need those numbers, the last is not really needed, but without it you'd get less realistic results.

Also this is not even remotely close to realistic, I'm afraid. That isn't how sound changes work. Yes, /b/ can change into /p v w m d g/ and other things, but not at random in one language. The change depends on what sounds are before and after the /b/, what the other sounds in the language are, and what sound changes have just happened (/b/ might change as a response to one, or the target of its change might be in response to an earlier change. If /v/ just changed into /w/, /b/ might change to /v/ as there is a gap, but it less likely to change to /w/ as well since it would confuse lots of word pairs - /bot/ and /vot/ would both become /wot/.

[Mâq Lar]

CatDoom,
I get what you're saying, but I don't think that is a problem. What you're explaining is just a nested structure of sound changes that rely on each other, that if x happens then y has some chance of happening as well. And the second issue isn't really an issue, because that rule change could happen with the original whether the prvious change happens or not, it just makes the both more pronounced than what it would be otherwise.

No, I think you missed his point on the second one. You're confusing possibility with probability. The probability of any change happening changes as a result of each change that happens before. It's like that school question about taking coloured balls from a hat. If you take out a red ball and don't put it back, the chance of getting another red ball has changed from what it was before. This makes the probabilities harder to calculate. What CatDoom is describing is a much more complex version of this. It doesn't just make both changes more pronounced; it effectively rules out some changes completely, and makes some much more likely. How are you going to take this effect into account in your script? If you don't, and apply changes randomly, you will generate dozens and dozens of languages with impossible phonologies before you get it to give you one that could be a language. And even then, it won't be a phonology that could have realistically developed from the input language.

[Durakken]

Mâq Lar,

I get why you think it might not be possible, or at least not accurately, but that is really inconsequential in this case.
It's possible to write such a program. It might be hard for me to factor in things because I don't know about them, but certainly someone could do it provided the information in a format that isn't hard to get the data out of or if they knew the rules that you're so keen on.
But as I said, that's inconsequential to you're argument. A program such as I described would be more accurate than me in producing such evolution of language simply because it would be based on statistics while whatever I did without it would be based on nothing other than more or less randomly picking what I want to change anyways.

I'd guess there are other reasons for why people don't like this idea rather than the reasons presented, but whatever...

-----------------------------------------------
Still no comments on the word list?

[KathTheDragon]

A program such as I described would be more accurate than me in producing such evolution of language simply because it would be based on statistics while whatever I did without it would be based on nothing other than more or less randomly picking what I want to change anyways.

Actually, you have it backwards. If you browse a list of sound changes, they look pretty much like randomly picked sound changes. The only exception is what Mâq Lar mentioned. If you ask anyone who's actually derived a conlang by means of sound changes (including myself) they will all tell you to write out your sound changes by hand.

[Durakken]

If that's the only thing then it would be fairly easy to model sounds shifting like that and it would be much better to model all those sound and how changing them would impact other sounds than rely on an ear or person the can't distinguish pronounce them correctly consistently.

[M Mira]

The catch is that, if you can write a program to calculate the probability of a certain sound change happening with regards to enviornment and phonetic inventory, you probably can do that by hand anyway, unless what you write is an AI that trains itself with input data.

[Mâq Lar]

I'm afraid you haven't really understood the point, no. If you want to do it this way, that's your choice, but it won't be realistic. As has been pointed out, if your program can do it realistically, it would be be a very advanced AI. Realistic changes are often pretty random yes, but once a random change starts, it influences other ones non-randomly, but not in the general statistical way you are thinking. As well as a misunderstanding the nature of phonological change, you seem to misunderstand statistics too: your program won't be more accurate just because it is based in statistical trends, as I said, that's the same as saying that it's more accurate to conclude that in a busy shopping street you will find shoppers standing on towers on each other's shoulders in front of popular shops than it is to conclude that you will find people walking around in random but mutually influencing paths. It's a massive oversimplificication of both linguistic change and statistics, and will give you commensurately massive wrong results. You've neglected the connections between sound changes, which are what motivates them. You're concluding that it is plausible for trees to develop eyes simply because eyes have already evolved independently a large number (40) of times, without realising that it needs certain conditions, not simply statistical high frequency. Now if you want to do it this way, that's your choice, but understand that it will not give realistic results, and will be more complex and time consuming than the other ways which have been suggested to you.

It might be wise to consider why everyone else disagrees with you on the realism of your project. It's possible we have all misunderstood and we are all wrong, or it's possible that it's just you. You are apparently fond of following gross statistical trends; I would advise you to do so here.

[Durakken]

If you want to do it this way...

I might try to create...

You're fixated on this, not me. The last thing I asked about is the word list I made which hasn't been commented on.

The fact is, at this point, you're resistent to actually helping considering, even if I'm dead wrong about a program the list I'm asking for would be useful to have and it's a wonder why it hasn't been made or if it has, why it hasn't been shared.

[KathTheDragon]

The last thing I asked about is the word list I made which hasn't been commented on.

Probably because no-one's interested by word lists. I mean, what do you want to hear?

[Mâq Lar]

It's not a fixation; you misunderstand a basic feature of languages which is important if you are interested in conlanging. We were trying to help you. As for the word lists, I can't actually find a specific question you have asked about them; what exactly do you want to know?

[Durakken]

The last thing I asked about is the word list I made which hasn't been commented on.

Probably because no-one's interested by word lists. I mean, what do you want to hear?

Well let me go do a jig so that you might care about my broken leg.

It's not a fixation; you misunderstand a basic feature of languages which is important if you are interested in conlanging. We were trying to help you. As for the word lists, I can't actually find a specific question you have asked about them; what exactly do you want to know?

I need someone to go over it and see if they are anything that might be missing that would be used in a largely nomadic civ and to double check to see if there are any words that aren't present tense, can be uncompounded... also if I missed taking out/correcting words that have a suffix that may be like "-ing" and shouldn't be there, because there are words like "building" which is a different concept while "running" shouldn't be there. There are 1400 words there and i just want a second set of eyes to look at to check before moving forward because that is a good idea regardless, but also something that happened recently that i can't gaurantee isn't effecting my ability to correct things like that perfectly.

Also opinion on words like "government" vs "govern" do you think I should get rid of "government" and just consider that "ment" a suffix even though we don't really use it like that and we generally consider it as a completely separate concept.

[Salmoneus]

I need someone to go over it and see if they are anything that might be missing that would be used in a largely nomadic civ

Yes. You've only got 1400 words, so you're missing most of the words that would be used in a largely nomadic civilisation. [while including some odd choices - just looking at some of the early ones, I don't picture Mongols talking about advertisments, chemicals and automatic things - though of course (and this is the big problem with word lists) it does depend what you mean by these words and how much of the english nuance you are including]
To give a few examples, I don't see words for colt, filly, foal, suckling, weanling, yearling, gelding, mare, stallion, the verbs to foal, to wean, to geld, to spay... horse, horseshoe, farrier, halter, bit, crownpiece, throatlatch, noseband, bridle, hackamore, cavesson, frentera, bit, fiador, winker, shank hobble, phalara, sallong, martingale, cob, draft, liverpool curb, mosquero, mane, botfly, curry, thrush (the horse ailment), sulcus, fetlock, cannon (the anatomical term, not to be confused with artillery cannon, carom cannon or the upper and lower cannons of vambrace), hock, feather (in the horse sense rather than the bird sense), muzzle (in both senses), poll, crest, loin, croup, dock (and the verb), hack, caulkin, shabrack, barding, twitch, hipposandal, picket, hobble, lead, tack, rig, surcingle, quirt, gogue, terret, chambon, bosal, breeching, gaskin, ergot, stifle, colostrum...

and to double check to see if there are any words that aren't present tense, can be uncompounded... also if I missed taking out/correcting words that have a suffix that may be like "-ing" and shouldn't be there, because there are words like "building" which is a different concept while "running" shouldn't be there. There are 1400 words there and i just want a second set of eyes to look at to check before moving forward because that is a good idea regardless, but also something that happened recently that i can't gaurantee isn't effecting my ability to correct things like that perfectly.

Also opinion on words like "government" vs "govern" do you think I should get rid of "government" and just consider that "ment" a suffix even though we don't really use it like that and we generally consider it as a completely separate concept.

That's a lot of work I don't have time for, sorry. But also, these things aren't a matter of right or wrong, just your personal taste. Why do you consider 'building' "a different concept"? Why do you consider 'govern' and 'government' to be "completely separate concepts"? There's no matter of fact about which concepts are the same as each other, so it's purely what you happen to feel like doing.

[Durakken]

I need someone to go over it and see if they are anything that might be missing that would be used in a largely nomadic civ

To give a few examples, I don't see words for colt, filly, foal, suckling, weanling, yearling, gelding, mare, stallion, the verbs to foal, to wean, to geld, to spay... horse,

Most of those aren't there because "horses" don't exist. Same as cats, dogs, fish, etc.

horseshoe, farrier, halter, bit, crownpiece, throatlatch, noseband, bridle, hackamore, cavesson, frentera, bit, fiador, winker, shank hobble, phalara, sallong, martingale, cob, draft, liverpool curb, mosquero, mane, botfly, curry, thrush (the horse ailment), sulcus, fetlock, cannon (the anatomical term, not to be confused with artillery cannon, carom cannon or the upper and lower cannons of vambrace), hock, feather (in the horse sense rather than the bird sense), muzzle (in both senses), poll, crest, loin, croup, dock (and the verb), hack, caulkin, shabrack, barding, twitch, hipposandal, picket, hobble, lead, tack, rig, surcingle, quirt, gogue, terret, chambon, bosal, breeching, gaskin, ergot, stifle, colostrum...

I'll have to look those up to decide which of those to include as a good number of those i've never heard or don't have a working definition for.

That's a lot of work I don't have time for, sorry. But also, these things aren't a matter of right or wrong, just your personal taste. Why do you consider 'building' "a different concept"? Why do you consider 'govern' and 'government' to be "completely separate concepts"? There's no matter of fact about which concepts are the same as each other, so it's purely what you happen to feel like doing.

What i meant is that there is the verb "building" and the noun "building" and the noun "building" is separate from the root "build"
Likewise, Govern is a verb, while government is a noun... but the noun government is doing the verb governing. Same with Advertisement. I'm not sure whether to take that as a single word with multiple operators or 2 words with ment as some sort of compound.

The reason this is important is because I want a small list of words that will then be the basis... There are no other words that come from english other than these and these are the words that will evolve. It's not like what a majority of wordlists are which is that you're just adding as you go. If that was the case I wouldn't care, but when it is that these are all the words that will exist starting off I have to try to get a solid list first.

[Salmoneus]

I need someone to go over it and see if they are anything that might be missing that would be used in a largely nomadic civ

To give a few examples, I don't see words for colt, filly, foal, suckling, weanling, yearling, gelding, mare, stallion, the verbs to foal, to wean, to geld, to spay... horse,

Most of those aren't there because "horses" don't exist. Same as cats, dogs, fish, etc.

horseshoe, farrier, halter, bit, crownpiece, throatlatch, noseband, bridle, hackamore, cavesson, frentera, bit, fiador, winker, shank hobble, phalara, sallong, martingale, cob, draft, liverpool curb, mosquero, mane, botfly, curry, thrush (the horse ailment), sulcus, fetlock, cannon (the anatomical term, not to be confused with artillery cannon, carom cannon or the upper and lower cannons of vambrace), hock, feather (in the horse sense rather than the bird sense), muzzle (in both senses), poll, crest, loin, croup, dock (and the verb), hack, caulkin, shabrack, barding, twitch, hipposandal, picket, hobble, lead, tack, rig, surcingle, quirt, gogue, terret, chambon, bosal, breeching, gaskin, ergot, stifle, colostrum...

I'll have to look those up to decide which of those to include as a good number of those i've never heard or don't have a working definition for.

They're horse words. But no doubt your people would use equivalent terms regarding cows, goats, antelope, rhino, or whatever else it is that they herd.

That's a lot of work I don't have time for, sorry. But also, these things aren't a matter of right or wrong, just your personal taste. Why do you consider 'building' "a different concept"? Why do you consider 'govern' and 'government' to be "completely separate concepts"? There's no matter of fact about which concepts are the same as each other, so it's purely what you happen to feel like doing.

What i meant is that there is the verb "building" and the noun "building" and the noun "building" is separate from the root "build"
Likewise, Govern is a verb, while government is a noun... but the noun government is doing the verb governing. Same with Advertisement. I'm not sure whether to take that as a single word with multiple operators or 2 words with ment as some sort of compound.

...-ment is called a 'suffix'. In English, and in most languages in the world, one word can be formed from another by adding a 'suffix' or 'prefix' - in particularly, this is a common way to turn a noun into a verb or vice versa.
"Building" is either a participle ("the building tension") or a noun ("the new building"). In both cases it is derived from the verb "to build" with the suffix "-ing".

The reason this is important is because I want a small list of words that will then be the basis... There are no other words that come from english other than these and these are the words that will evolve. It's not like what a majority of wordlists are which is that you're just adding as you go. If that was the case I wouldn't care, but when it is that these are all the words that will exist starting off I have to try to get a solid list first.

But you haven't defined what a 'solid' list would be.