Grammatical Philosophy: Is mine good for conlanging?

[caters]

I am thinking of having a language in which writing is based off of the Arabic script except full on letters are added for vowels that in Arabic are just represented by dots and their position and number. At the same time I was thinking of having the grammar be not all that similar to English or Arabic(some similarity like with double negatives, I will accept but I want to limit this grammatical similarity).

I just think that if I make a conlang with the Arabic script and Arabic grammar that I would really just be inventing a dialect of Arabic and not a conlang whereas if I go the other direction and try to make it as similar to English as possible, I will get English with a weird writing system.

Now I am not sure I would want genders and cases and conjugations and declensions.

I mean yes if you take a verb in English and inflect it, you get conjugations but those are not conjugations from grammatical case+gender like in say, Latin. Same with inflections of nouns and adjectives in English being declensions but not from grammatical case+gender. Arabic has gender and case so this would be a place where more English similarity would be desired.

I mean with gender, how do you decide which words are neuter, which ones are masculine, and which ones are feminine when most do not have to do with the biological sex being known or unknown.

For example, it makes sense that these words translated into a conlang would be feminine:
Beauty
Pregnant
Milk
Breast
Garden(my fictional world has way more female gardeners than male gardeners)
etc.

And it makes sense that these would be masculine words:
Muscle
Strength
Hunting(my fictional world has way more male hunters than female hunters)
etc.

And that these would be neuter words:
Baby
Infant
Toddler
Neutral
etc.

That last list is mainly because in my fictional world, there is domination of 1 fictional species and that species happens to be externally gender neutral but internally obvious until 5 years after birth when it becomes obvious externally.

Anyway back to grammar, most words do not have to do with children or age or females or males. But with gender they can't all be neuter otherwise gender would comprise a minority of words as far as grammar. But with something like chair or computer or yarn, how would you decide the gender? Statistics(like for example how many females knit compared to males)? Randomness?

Similar logic goes for grammatical case.

And what about verb tense? How would I decide if I want it to be tenseless with time expressions used next to the verb, past vs non-past, future vs non-future, past, present, and future, or multiple degrees of past and/or future depending on the time scale?

[AzureWings]

Disclaimer: I am very primarily an English speaker, so my own knowledge of grammatical gender, etc. is relatively limited. That said - from what I am aware of, my impression in at least some languages with grammatical gender is that it does not always correlate with actual gender at all for many words, so in many cases a random assignment might be appropriate. Fact-checking myself with quick lookup (just Wikipedia research, so take with a grain of salt) suggests that the question of how to decide the grammatical gender of words in a conlang that does use grammatical gender is another distinct decision you need to make about the language, as it seems how strictly real languages correlate grammatical gender with the semantic meaning of words varies quite a bit.

So you could try to tie the gender of words to their meanings for all words via statistics and so on as you considered (although personally I think after a certain point that would be exhausting ), but some (or even a lot) of random assignments of gender for many words that don't make a semantic relationship immediately obvious would also probably be fine (and from the looks of it that would be similar to a lot of real-world languages with grammatical gender). In short, it doesn't seem like there would be any problem with randomness in assigning gender to words for a conlang - the impression I have at least is that it's normal for grammatical gender to not always have a connection to word meaning. Even within the more directly-related words you listed as examples, plenty of exceptions seem to exist within real languages with grammatical gender, so you're probably pretty free to assign grammatical genders to words freely, as you please, and it probably won't seem out of place next to real languages in that regard at least.

As for writing - distinct languages can certainly share scripts, characters, or parts thereof without being considered the same language or dialects thereof (English and French use the same characters for example (barring a few diacritics), but I'm sure there would be plenty of people upset if one were to call either one just a dialect of the other!). If the grammar is also shared it might get a little bit more nebulous though - but in principle, just using the same script/characters doesn't seem like it would make the conlang so related as to just be a separate dialect (although that is more my opinion than anything). That said, there is a meta-concern of the fact that using an existing script will probably cause readers of the conlang to associate it with the real language that shares the script on some level, and you'd need to consider if the feel of that association is something you want your conlang to have or you'd rather it have a cleaner break with real-world languages.

For verb tenses and time scale, as with the other issues above, a lot of it comes down to what you want the language to be, as it could use a variety of methods of marking time and in different ways. If you really want to get into it, I imagine you could take into account the thought of how the culture this conlang is intended to have arisen in might have influenced it and thus factors like said constructed cultures' perspectives on time (like being an M-time or P-time culture could potentially be a factor in how the language indicates time, although I'm sure such a relationship has vast complexities I'm completely unaware of) could play a factor, depending on how much detail you've envisioned said culture in.

I hope that's helpful - all I've really given you are some perspectives on the same questions, or in some cases more questions, rather than answers I think for many of these questions, you're the only one who can really determine what the 'right' answer is within your conlang, as real languages are so diverse.

[krinnen]

The first question that comes to mind is: What do you want it for? If you want a naming lang, i.e. a language to put names of places and rivers and such in your maps, then you don't need to stress about case, gender, tense, mood, etc. Just decide on a phonology, some phonotactics, compounding and a bit of derivational morphology, and off you go.
For example, you may decide on a small word list including things like {mouth, river, lake, water, hill, mountain, rock, range, red, blue, green, forest, birch, oak, pine, big, small, lonely, cluster, man, hand, plain, valley, etc}, and how does words compound; noun-adj, adj-noun, possessor-possessed, possessed-possessor. So if you have the words bilq 'birch' and holo 'valley', holobilq could be name of Birchvalley town. Or something along those lines.

Now, if you want to create a full conlang, and not just a naming lang, I'd suggest you read Mark Rosenfelder's Language Construction Kit and maybe visit his conlanging forum, the zbb

[Falconius]

In Hebrew, and very likely Arabic given how closely related they are, all nouns have a gender, verbs generally speaking don't. Verbs, well some of them, do get modified by the subject the verb is describing, ie "she wants" and "he wants" the word want is gendered differently. Some words have both gender forms, for instance there are masculine and feminine forms of all the numbers used depending upon the gender of the subject (ie for money, a masculine word, you'd use the masculine form of the number).

[seusomon]

If you the actual Arabic writing system (same association of symbols to sounds, at least for consonants), I think this would require explaining in terms of the fictional world. What is the connection between this world and ours, so that one of our writing systems came to be used there? If it is a fictional world separate from our own, then I would prefer to use a writing system *inspired* by Arabic (written right to left, calligraphic letter forms, etc.) without using the actual Arabic alphabet.

Grammatical gender in natural languages is a complex structure that evolves over time. Often the gender is determined by the phonological form of the word, rather than any obvious connection with biological sex. On the other hand, associations with the society's concept of male and female can also play a part. What a gender system arises from and how it works are choices to be made when developing the language; there is no single answer.

[RobEm]

This sounds really challenging, but really fun. I had a couple of ideas that I thought might be worth sharing. If you don't want to invent an alphabet of new letter forms, symbols and glyphs, it might be worth considering something like da Vinci's mirror writing to reuse the alphabet to create something that looks quite different and reads right to left. Given the complexity of assigning gender to nouns which don't have a gender another alternative might be to appropriate the set of gender mappings from an existing language, if you chose French then "boat" might be male gendered based on "bateau". If the setting is high fantasy, then the language could have a species based "gender system" that means either an elven sword is distinguishable from a dwarven sword by the article and/or having a different form (ledrows le, dudrows ed, etc), or that buildings and swords are human, trees and bows are elven and rocks and axes are dwarven.

[Tiana]

Note that this thread is a year old and the OP has posted 6 times, I doubt they'll be back though I don't think it's a bad discussion to continue, specific critiques will likely be time wasted. The idea of species based high fantasy conlang gendering is amazing and shouldn't be abandoned to a potentially dead thread, what a great idea, RobEm.

[RobEm]

The idea of species based high fantasy conlang gendering is amazing and shouldn't be abandoned to a potentially dead thread, what a great idea, RobEm.

Thanks Tiana, I hope it helps someone!

[Galrion]

A side note i feel obligated to share in my minor studies of linguistics gendered mood don't have to make sense as to why they are gendered the way they are. In many languages that use genders those that use the word don't really think of something like a fork as female or a spoon male. Logically you might think it's backwards, but it does happen and it's never really something to be concerned with.

If generous is wintering you want to do away with you can look at why English doesn't have genders. It did at one point. The theory is that when England was invaded by the Saxon (or someone else... it happened a lot and my memory is rusty) the invaders integrated with the locals. They didn't impose their language on them but started to learn the local language. When you have a a bunch of people learning a language second hand at an older age one of the first things to drop is gendering. They are simply trying to communicate. Gender doesn't help this. The locals may have chuckled if they thought it funny, but then this generation starts passing its learned form of English on to its children and it starts to spread.

English is a strange language with a strange history that makes it what it is.

Tldr: your historical timeline can change what you really want to be in your language.