Did big languages generally have a net loss of inflectional morphology in the past 1-3 millennia and small languages the other way round?

Question

a. R. M. W. Dixon (1998) theorizes that languages normally evolve in a cycle from fusional to analytic to agglutinative to fusional again like a clock. There are two opposing forces: one reduces morphology, one increases morphology or otherwise the languages today would be all isolating or polysynthetic. Languages at the extreme ends of isolating/polysynthetic are less likely to go even more extreme, just like offsprings of geniuses/very tall people are less likely to become ultra-geniuses/ultimately tall. When there is no proof that the direction of evolution MUST move clockwise/anticlockwise, languages in the middle could move either direction (gain/lose morphology) and there are still many unknown factors.

b. Paper1 Paper2

I found quite a few popular papers online about the relationship between grammar complexity (or morphological complexity), phoneme inventory size, etc. While some of the results are contradictory, the result is that the greater the population size of the language/the number of L2 speakers, the simpler the grammar (in terms of irregularity and morphological complexity).

It is hard to measure the overall "morphological complexity" of a language. As far as I remember, some of the papers measure selected features on WALS, e.g. WALS 22A: the number of categories of inflectional synthesis of the verb as a measurement of morphological complexity.

c. Most “classical languages” were written/spoken 1–3 millennia ago. They are the ones we have the most records of today.

With respect to its classical form, there are three kinds of modern languages :

1. Obvious net loss of inflectional morphology

In the last 1–3 centuries, most Indo-European languages have had a net loss of inflectional morphology and become more analytic with respect to their old (classical) form(s). The change is so obvious that no tallying up is needed (you name it: Greek, Italian, Spanish, Hindi…), and everyone is asking “why is ancient Greek grammar more complicated than modern Greek grammar?” and not the other way round (though grammar complexity not necessarily entails morphological complexity).

Other examples in this group: Japanese, all Arabic varieties, Turkish.

2. Hard to measure

There should be many languages the net gain/loss of inflectional morphology of which is hard to measure.

3. Obvious net gain of inflectional morphology

Except perhaps some creoles, there is no language that has obviously gained so much morphology that everyone says "the classical form of the language used to have much fewer inflections".

As far as I have read on the internet, Chinese, Hungarian and Finnish have gained some morphology as a whole in the past 1–3 millennia. Perhaps gaining morphology takes much longer time, but are they special cases?

---

Many linguists still claim it is not true that languages lose morphology as a whole over time (or how could they gain morphology at the first place?). Although none of the widely spoken languages today are on the extreme end of morphological complexity (polysynthetic + extremely irregular grammar like verbs in Navajo), many of them, on the other hand, are analytic and have streamlined grammar.

Here is a hypothesis based on this answer on Quora:

Damon Siefert's answer to Why do most languages have the tendency to simplify their morphology, grammar, and phonology over time, while some do not?

Languages might have generally gained morphology as a whole 3 millenniums ago. Human used to live in small tribes and according to point b. above, small languages tend to have larger force of gaining/ retaining more morphology (Most languages in the “residual zones". Small languages are more on the extreme ends of morphological complexity (Hawaiian, Rapa Nui, languages of the Caucasus, Greenlandic, Navajo etc).)

In the last 1–2 millenniums, more and more people became multilingual. The extinction of small languages with lots of morphology have accelerated and large metropolises + larger languages have dominated the world. Those large languages are less likely to have lots of inflectional morphology and many small languages have assimilated grammars of the large languages (eg. Nahuatl -> Spanish grammar). As a result, most languages left today have much less morphology. And in the past millennia, those languages which have gained lots of morphology are the small ones that we do not know much about.

I wonder if there are some related researches about this hypothesis/ the factors of gaining and losing morphology. Or I am biased that large languages did not generally lose morphology as a whole (please give me examples).

Answer 1

My take on this is that the common ancestor of Sanskrit, Classical Greek, Latin, and other languages, but not of Hittite and the Anatolian language was almost at peak inflection and had no direction to evolve but to loose some of its inflections (at least). The amount of loss of inflection, however, was very different in the individual languages, and some languages also acquired new inflections not present in the proto-language (e.g., the new synthetic future tense of the Romance languages).

I don't see a specific correlation between the loss of inflection and the number of speakers (I assume, this is meant by "large language" in the original question), Russian has retained a huge amount of inflection despite being "large" in this sense, and English became an isolating language when still being "small", i.e., before acquiring an overseas empire.