Recently I came across a short text on Language Log briefly discussing a phenomenon which seems to affect certain languages. The author noticed that loss or heavy weakening of inflection during language's histories tend to be concomitant with certain political or social contexts. As for example in the case of English - the language of Germanic invaders of Britain or Persian - for a long time a regional lingua franca. In both circumstances the language in question was a language that was adopted by adults (either by direct force or social pressure, or willingly), natively speaking in a different language (the invaded Celtic and maybe other unknown non IE population in Britain or later incoming Old Norse speaking people for Old English and merchants or travellers for Persian). This population would then acquire the language imperfectly (being a L2 for them), for example by losing or contaminating certain language features (the inflectional endings in this case), and being such a large part of the new community of that language's users, it would permanently change it.

That scenario would oppose a situation of for example West and East Slavic languages which, without a history of intensive conquests etc., preserved the complex both noun and verbal inflection.

What in my opinion is crucial in these cases is that the cause of loss of a certain feature is not the lack of this feature in the L1 of the acquisitor (Old Celtic or Old Norse languages, morphologically speaking, were not any simpler than Old English) but the sheer fact of acquisition of the language in question as a L2.

On the other hand, there is a counterexample that came to my mind, namely, Koine Greek that despite being a common lingua franca on a vast area for quite a long time did not considerably simplify its inflection (of course, comparing it to the Classical Greek it was simpler, but on a scale uncomparable to the case of Old English and Modern or Early Modern English).

My question is, does anybody have any more information on that topic, some publications, projects or their own work or ideas?

These scenarios, would be different from just creolisation, since as far as I know creolisation involves merging of features of two languages or assimilating one to another, while the cases that I am writing about involve, just as I wrote before, changing by imperfect, unnatural acquisition.

  • 1
    I don't know much about the topic, but I think Language Complexity as an Evolving Variable discuss it: global.oup.com/academic/product/… . I think specifically Peter Trudgill has worked on the topic; check out Sociolinguistic typology: social determinants of linguistic complexity . Sep 12, 2016 at 19:17
  • 2
    Incidentally, I don't know much about Greek either, but at a guess, I would wager Koine is an exception because it was a written, prestige literary language. Literary languages tend to be highly conservative; according to Roger Wright, Latin morphology had long been simplified in speech while they kept writing all the cases (as evidenced by writing errors, educational materials, and informal texts). Sep 12, 2016 at 19:27
  • 1
    That is a point but Koine was also used as a spoken tool of communication between people of different cultures and L1's and on the other hand Old English also had strong tradition of literature.
    – czypsu
    Sep 12, 2016 at 20:20
  • 1
    I've read arguments that what Wright says about Latin is true of Old English too, though (based on similar evidence) -- i.e. the morphologically simpler Middle English emerged much more gradually than the textual record seems to suggest, and wasn't for example a sudden change driven by social dislocation. Can't recall a specific source, though, sorry...
    – Matt
    Sep 12, 2016 at 23:32

2 Answers 2


I think you need to start with language change, rather than a question of complexity. One might "lose" some lexical complexity on the one hand, but introduce ambiguity which brings more cognitive complexity on the other hand.

I hope these general pointers are helpful: From what I have learned, the reasons for language change are many. Some drivers are easier to identify because they have been documented. The introduction of a new technology, for example, can lead to changes in format, therefore in message time/contraints/functional needs. Equally, it can lead to an increase in contact with another language or another dialect. Socio-economic influences are also well documented drivers of language change.


First, I don't think English really did get simpler, it just moved its complexity into other parts of the language. Less inflectional morphology, but more complicated word order rules, and many more complicated constructions (e.g., the various similar but not at all identical resultative constructions). And if you look beyond just syntax, look at the tense/lax vowel system and the destressing rules that go with it; that's certainly not a simplification.

These scenarios, would be different from just creolisation, since as far as I know creolisation involves merging of features of two languages or assimilating one to another, while the cases that I am writing about involve, just as I wrote before, changing by imperfect, unnatural acquisition.

What you're describing isn't creolization, it's mixed languages. Briefly:

  • A mixed language is something like Ma'a, with (roughly) Bantu grammar and Cushitic vocabulary. They tend to arise when the mixed community has a lot of bilingual speakers.
  • A pidgin is something like Pidgin Hawaiian (as opposed to Hawaiian Pidgin, which is the misleading name of the creole that's supplanted it…), with a defective grammar that can't express things like relative clauses, causatives, etc. They tend to arise when there's a mixed community where most of the L2 learners don't get a chance to learn the lexifier language.
  • A creole is something like Haitian, with a grammar spontaneously invented by the children of pidgin speakers.

So, your proposal is actually a lot closer to creolization than you think it is, which is why people keep suggesting that English really is a creole. But English is clearly missing the two-step part, going through a pidgin, which is why most people reject the creole-English theory. But still, the process you suggest and the pidgin/creolization process have enough parallels that it wouldn't be surprising if there were similar effects.

Also, unlike mixed languages, which tend to have a grammar very close to one of their parent languages, most creoles have highly analytic grammars, no matter what any of the parents had. There are many theories that try to explain why this might be true; it's at least plausible that whatever explains creole grammar also explains English moving in the same direction.

Meanwhile, English is also often used as a paradigm example of the step toward analytic (or grammaticalized or isolating or whatever) in "grammaticalization cycle" theories.

Within languages, there are many cases of individual words where, e.g., a content word becomes a function word, which becomes a clitic or particle, which becomes part of the inflection system, and eventually it disappears.

Sometimes, a language is turning lots of things into morphological systems; at other times, it's losing lots of morphological distinctions and creating lots of new function words or syntactic constructions to make new distinctions. This can drive a language (or family) around in a circle from, e.g., analytic to agglutinative to fusional and back to analytic.

According to some people, this analytic->agglutinative->fusional->analytic cycle, or something else similar, is common to most languages. They have different theories on why it's true and what drives it. One thing they all have in common is using English as a prime example of the ->analytic step.

So, if there really is some general process that drives language change in this way (and English really is a good example and not just a coincidence), that would imply either that your theory is wrong, or that your theory is actually one of the common drivers of the ->analytic step.

I don't know of anyone who's proposed the latter, but it does seem like an interesting possibility, and definitely worth looking into.

On the other hand, the article you linked seems to be (indirectly) arguing against any such general process even existing. (After all, if only English and the Romance languages ever made the ->analytic change, then any general theory that predicts that most languages should go through such a stage must be wrong…)

But really, I think you need more than two examples (English and Persian) and one counterexample (Greek) to even start asking what the general pattern is.

Especially since in some ways English and Persian went in different directions. For example, Persian simplified its vowel system. And, while they both complicated their word order constraints, Persian's is a lot less like the typical creole than English's. And so on.

The article you linked seems to be suggesting that there really are no examples besides English and the Romance languages. But you can find plenty of other people offering other examples. Japanese's inflection system is much simpler than Old Japanese, and it's turned its non-inflection affixes into particles.

On the other hand, it might also be worth looking into the examples that people find harder to fit into universal theories, like the developments from Classical Chinese to Mandarin, or Ottoman Turkish to Turkish. I don't know much about the details of why these are hard to fit into people's theories, but, given the Mongol and Manchu influence in China and the Persian and Arabic in Turkey, it seems like they might be very relevant to your question.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.