Language Complexity

[HaloGeek]

So, with creating my first language, a proto language, I noticed that the declension systems are so complicated it makes me think how anyone could speak it.

Proto-Nevoran, the language I described above, has 8 cases and many other declensions for pronouns, verbs, and adjectives.

So I was wondering, does language complexity lower as they evolve? For example, take Latin: 7 (6) cases, four conjugations of verbs, six tenses I believe, three genders, and a whole host of other complexities.

Now take Spanish, a common ancestor of Latin: Lost all of its cases, has two genders, way less tenses, and lost a lot of other things from Latin.

We can even go with this to Almea. Verdurian and other descendants of Cadhinor are less complex than the ancestor language.

So, is it common for languages to less complex as they evolve? Is there a language that exists today that is harder and more complex than its ancestor?

[zompist]

I haven't had a chance to refer to the sci.lang FAQ for awhile. So start there, and then maybe do some searches on grammaticalization.

[CatDoom]

You can see some of this in English, where many grammatical particles have fused with content words in normal speech. "I will" to "I'll", "I am" to "I'm", "I have" to "I've", "I will have" to "I'll've", et cetera. From what I understand French is further along in the direction of polysynthesis; at least according to some analyses, a phrase like je vais le lui donner, "I am going to give it to him", is realized phonetically as [ʒ-vɛ-lə-lɥi-dɔ̃ne], which might be glossed as 1SG.SUB-FUT-3SG.DO.MASC-3SG.IO-give.

[Dē Graut Bʉr]

Keep in mind that having many inflections isn't quite the same as being complex. For example, Latin may have had way more inflections than modern Spanish does, but this really only means that Spanish has had to develop other ways (think of word order and prepositions) to express that that the Latin inflections used to do.

[HaloGeek]

Hmm.

I never really thought about it that way.

I suppose getting rid of the Genitive would force to put in a word for 'Of', but in essence they do the same thing.


Interesting.

[KathTheDragon]

Many languages express the genitive with neither an adposition or a case, but simply by placing the two nouns next to each other.

[Dē Graut Bʉr]

That's still a way of expressing it though, so it's really not that different from cases or adpositions.

[KathTheDragon]

I know; I was just saying that there're other options than case or adpositions.

[CatDoom]

Defining language complexity is fairly problematic. Since languages tend to have enough grammatical machinery to express basically any meaning (though sometimes only in a roundabout fashion), I tend to think that grammatical irregularity may be a more relevant measure of complexity than the inflectional synthesis of the verb or noun. A highly regular photosynthetic language would probably be easier to learn and speak correctly than a relatively analytic language with lots of suppletive forms or a complex inventory of grammatical particles with less-than-intuitive uses.

[Miekko]

Many languages express the genitive with neither an adposition or a case, but simply by placing the two nouns next to each other.

There's even more complicated systems in existence.

[KathTheDragon]

Many languages express the genitive with neither an adposition or a case, but simply by placing the two nouns next to each other.

There's even more complicated systems in existence.

I did not exclude that.

[Dē Graut Bʉr]

A highly regular photosynthetic language

Don't you mean "polysynthetic"?

[KathTheDragon]

Of course languages glean energy from sunlight. How could you suggest otherwise?

[CatDoom]

A highly regular photosynthetic language

Don't you mean "polysynthetic"?

...

Well, that's embarrassing. >_>

[Dē Graut Bʉr]

Of course languages glean energy from sunlight. How could you suggest otherwise?

Of course they do. If they had to get their energy from food instead, there'd be no food left for their speakers and thus they'd die out.

[Pole, the]

A highly regular photosynthetic language

Don't you mean "polysynthetic"?

...

Well, that's embarrassing. >_>

I'd like to see a language that is photosynthetic.

[Clearsand]

This gives me some interesting ideas...

[Trebor]

Defining language complexity is fairly problematic. Since languages tend to have enough grammatical machinery to express basically any meaning (though sometimes only in a roundabout fashion), I tend to think that grammatical irregularity may be a more relevant measure of complexity than the inflectional synthesis of the verb or noun. A highly regular photosynthetic language would probably be easier to learn and speak correctly than a relatively analytic language with lots of suppletive forms or a complex inventory of grammatical particles with less-than-intuitive uses.

A very good point. Quechua and Turkish are prototypical examples of the former; I'd like to know what some illustrative instances of the latter are.

[vec]

Well, English is full of strange suppletives and unintuitive analytical constructions.

[jmcd]

Many languages express the genitive with neither an adposition or a case, but simply by placing the two nouns next to each other.

Which languages are you thinking of? I know (the basilectal varieties of) Réunionese Creole do that but I am intereseted to know which languages would have that other than creoles.

[KathTheDragon]

I know Ancient Egyptian does. There is another construction using n, though.

[Curlyjimsam]

Concerning complexity of morphology - which is easier to quantify that the overall complexity of a language - it has certainly been the case that many languages have grown less complex over time: the Romance languages are an obvious example (and I presume the main influence in the Verdurian case), but we also see the same in English as elsewhere. But there are also ways in which languages can gain morphological complexity - of which grammaticalisation, mentioned by zompist above, is one of the main mechanisms (another might be sound changes that make paradigms less transparently regular). A language can have grammaticalisation of new morphology at the same time as losing old morphology - so whilst its true that Spanish has lost a great deal of the distinctions made in Latin, it (along with the other Romance languages) has also innovated a new type of future tense.

The nature of the interaction of the complexifying and simplifying forces is an interesting question - why do most Indo-European languages seem to have tended to lose morphological distinctions more than gain new ones, for example? It's not a question I feel qualified to answer; I'm not sure anybody is. But to answer your question (though keeping things to the morphological domain) - almost certainly it is possible for the overall morphological complexity of a language to increase over time, although I'm ashamed to say I can't actually think of a particularly good real-life example. The French case mentioned by CatDoom qualifies quite well. Suffice to say that, assuming the first languages didn't arise with complex morphology fully formed (though I suppose this possibility can't be discounted entirely), the morphology in question must have come from somewhere.

[vec]

I have a completely unresearched/unfounded theory that any creolization situation will lead to analytic structures, and I would posit that the Romance languages, as well as the Nordic languages, were subject to creolization amongst each other, resulting in dropping of entire inflection paradigms, not because of sound change but something else. North-Norwegian has much more inflection than the rest of the Scandinavian Nordic languages, and was much more isolated, while South-Norwegian has undergone massive loss of inflection, which I think is due to contact with Danish. This obviously happened to Yiddish too, and to English.

I see this happening in Icelandic in real-time, where contact with English is increasingly influencing the loss of inflection (English being the obvious culprit because the patterns are mostly the same). I recently spoke with some kids who have completely lost the subjunctive. Listening to them seems bizarre, like talking to foreigners, but they are native Icelanders with Icelandic parents. In addition, the genitive is disappearing from use amongst everyone, again, not through sound change but syntactic tweaks. People have even started using the preposition af (cognate with of) to compensate. It's jarring. But it's real. Our present tense is also disappearing in favor of að vera + infinitive constructions.

[Salmoneus]

But Latin was a language that millions of people had to adopt en masse, and had to cope with communication between people across a wide area, yet was morphologically complex. Why didn't it 'creolise' at that point, rather than centuries later, when there were far fewer second-language learners and speakers were more isolated from one another?

[zompist]

I think "creolization" is an unfortunate word choice here. A creole is a pidgin that's become a native language; I don't think any of the situations vec describes involved pidgins. (In general, there was a rush in the '70s and '80s, when pidgins and creoles received more attention, to call everything a creole; Thomason & Kaufman pretty much demolished the trend as being not well informed about either pidgins/creoles or the normal range of language contact.)

Inflections are lost all over, whether there's intense language or not; and they're also retained in areas that have experienced a lot of language contact, such as Russia. I can't say that there's no relationship, but I think it would be a lot of work to demonstrate satisfactorily.