Should I include morphology in my proto language?

Question

Most, if not all natural languages have some form of morphology, such as noun case or poly-personal agreement. However, these features are usually evolved from prepositions and auxiliary verbs, and require time for them to appear. This was my original route for creating languages, to create proto languages without morphology and then to evolve them later on. I then started researching proto-indo-european and proto-afroasiatic, and I found out they had morphology such as noun cases, even though they were supposedly the original languages for their regions.

The main question is: Would it be naturalistic to insert morphology into proto-language(s)?

Answer 1

I am not at all certain why you would assume that protolanguages must have no morphology. PIE is just a language like any other before or since. If we review the relevant article, the Font of All Knowledge explains that morphology is the study of words, how they are formed, and their relationship to other words in the same language.

Whether your language has affixes, adpositions, or some other means of distinguishing roles all of that falls under morphology. Thus, I'd argue that it is "naturalistic" to insert morphology into your protolanguage. After all, it's already there! Else, your daughter languages most likely wouldn't have the concept at all. (And that kind of speculative invented language would be very interesting indeed!)

Standard examples, consider Latin & English:

john bit the dog
the dog bit john
the dog bit john
john bit the dog

johannem momordit canis
canem momordit johannes
canis momordit johannem
johannes momordit canem

Morphology tells us how the words are related to each other. Just because English got rid of almost all of its endings doesn't mean we did away with morphology! We just tucked it away somewhere else, in this case, word order. This is why Latin can put the agent or the patient before the verb, but English, except in unusual circumstances, can not. We expect that whatever noun comes before the verb is the agent, be it the dog or be it John.

Morphology has even been proposed for Nostratic, the protolanguage from which PIE descended.

Answer 2

Proto-Indo-European (PIE) and Proto-Afroasiatic (PAA) are just the earliest ancestors we can reconstruct with reasonable certainty for their respective language families. That does not at all mean that they were the first languages of their regions or the earliest languages that existed.

They didn’t materialize out of thin air. They, too, had ancestors and were influenced by neighbouring languages that didn’t leave any other trace. Humans probably had been speaking for (at least) hundreds of thousands of years at the time of PIE (possibly somewhere between 4500 BCE and 2500 BCE) or PAA (possibly somewhere between 16000 BCE and 10000 BCE) and there are proposed larger families and attempts at reconstructing proto-languages that would group language families like Indo-European and Afroasiatic together (e.g. Nostratic), though they are controversial.

Unless time-travel gets invented, we will never know how exactly the first language(s) came to be and what they were like, if they were immediately fully-formed or if they lacked features we now consider universal (possibly subordinate clauses?). Maybe they weren’t even fully oral but rather gestural with accompanying sounds.

That does not mean, however, that your proto-language cannot be the first language of its world. If that’s what you want to create, it’s up to you to speculate what a language like that could be like. The origin of language is so uncertain that anything plausible can be counted as naturalistic.

If you just want to create a naturalistic proto-language for a family of languages, that isn’t the(/a) first language to exist, there really isn’t any difference to creating any other naturalistic conlang.