Are some languages more "complex" than others?

[Bomb#20]

I didn't notice that this thread had been bumped!

Sentiment seems to be that language complexity is too ambiguous to quantize easily. (Of course this makes the "equal complexity" claim itself undecidable or even meaningless.)

A working definition might be based on how long it takes someone to become conversant in the language.

Which someone? I.e., how could you control for the many extraneous factors that might affect how long it takes, such as how motivated a given learner is, how much help he gets, which languages he already knows, how much the native speakers he talks to switch languages in response to his clumsy attempts, and so forth?

 

[Swammerdami]

I didn't notice that this thread had been bumped!

Sentiment seems to be that language complexity is too ambiguous to quantize easily. (Of course this makes the "equal complexity" claim itself undecidable or even meaningless.)

A working definition might be based on how long it takes someone to become conversant in the language.

Which someone? I.e., how could you control for the many extraneous factors that might affect how long it takes, such as how motivated a given learner is, how much help he gets, which languages he already knows, how much the native speakers he talks to switch languages in response to his clumsy attempts, and so forth?

IOW you agree with me and others in the thread (except perhaps Politesse) that any definition will be fuzzy, difficult, and/or ambiguous.

Don't preach to the converted. Try to bring Politesse around.

 

[Don2 (Don1 Revised)]

I'm not involved in the current discussion and this isn't an addendum or related to it....

I agree with most that talking about complexity of a language is ambiguous. It seems like if we narrow down the entity to components of a language, at least the complexity of the component begins to become less ambiguous but we might also have to provide some context.

Here's an example. Suppose we are discussing phonology. We might say that a language with more phonemes and more interactions between them creates more complexity or something like that....at least it starts to sound more reasonably implemented than taking a broader question of a whole language. It still doesn't sound low-level enough to know what the computations would be and so maybe you would need to look at something more specific phonologically, like with rules, patterns, or possible words.

But an additional problem I have with this is that saying/writing a thing in theory can be infinite. So imagine we are discussing how language#1 has a bigger vocabulary than language#2. Language#1 has one word for airplane and likewise a lot of specific single words for ideas. Language #2 calls airplanes flying-demon-that-we-first-saw-at-the-sun-moon-festival-with-four-wings-and-pooping-smoke. What does this actually say about complexity numerically, if anything at all?

 

[Politesse]

I didn't notice that this thread had been bumped!

Sentiment seems to be that language complexity is too ambiguous to quantize easily. (Of course this makes the "equal complexity" claim itself undecidable or even meaningless.)

A working definition might be based on how long it takes someone to become conversant in the language.

Which someone? I.e., how could you control for the many extraneous factors that might affect how long it takes, such as how motivated a given learner is, how much help he gets, which languages he already knows, how much the native speakers he talks to switch languages in response to his clumsy attempts, and so forth?

IOW you agree with me and others in the thread (except perhaps Politesse) that any definition will be fuzzy, difficult, and/or ambiguous.

Don't preach to the converted. Try to bring Politesse around.

Interesting turn of phrase.

 

[Politesse]

I didn't notice that this thread had been bumped!

Sentiment seems to be that language complexity is too ambiguous to quantize easily. (Of course this makes the "equal complexity" claim itself undecidable or even meaningless.)

A working definition might be based on how long it takes someone to become conversant in the language.

Which someone? I.e., how could you control for the many extraneous factors that might affect how long it takes, such as how motivated a given learner is, how much help he gets, which languages he already knows, how much the native speakers he talks to switch languages in response to his clumsy attempts, and so forth?

Motivation, hrs per day, etc is easily controlled for, but the most important factors are familiarity and the language learning context. It is much quicker and easier to learn a language whose phonology and syntax are very similar to yours, and/or that use the same written script. Hence why English speakers pick up German and French readily but will take much longer to learn Chinese or Hmong (but Chinese speakers pick up Hmong/Miao fairly quickly despite those languages being from rather different language families). Manner of learning tends to obviate those differences a bit, as a student in a full immersion program learns languages pretty quickly no matter how different those languages might be. Previous exposure also matters. Czech and Slovak are structurally distinct but for many residents of the region, surprisingly mutually intelligible languages due to exposure of one to the other growing up - tv, music, signage, etc. It makes them quick learns for locals who have an interest in picking up the other.

 

[Bomb#20]

I didn't notice that this thread had been bumped!

Sentiment seems to be that language complexity is too ambiguous to quantize easily. (Of course this makes the "equal complexity" claim itself undecidable or even meaningless.)

A working definition might be based on how long it takes someone to become conversant in the language.

Which someone? I.e., how could you control for the many extraneous factors that might affect how long it takes, such as how motivated a given learner is, how much help he gets, which languages he already knows, how much the native speakers he talks to switch languages in response to his clumsy attempts, and so forth?

IOW you agree with me and others in the thread (except perhaps Politesse) that any definition will be fuzzy, difficult, and/or ambiguous.

Don't preach to the converted. Try to bring Politesse around.

I'm not following. What was your point in proposing time-to-learn as a working definition, if not to resolve the ambiguity problem? And aren't you claiming Thai is less complex than other languages?

 

[Swammerdami]

A working definition might be based on how long it takes someone to become conversant in the language.

Which someone? I.e., how could you control for the many extraneous factors ...

any definition will be fuzzy, difficult, and/or ambiguous.

I'm not following. What was your point in proposing time-to-learn as a working definition, if not to resolve the ambiguity problem? And aren't you claiming Thai is less complex than other languages?

Even though definitions may not be crisp, easy, and/or clear, one can still hope for one of the better choices from a bad lot.
(One could desist from ever assigning a simplicity attribute to a language taken as a whole. But that ship has already sailed.)

- - - - - - - - - -

I doubt we could pull it off, but as a group exercise we could vote on whether certain traits suggest "simplicity" or suggest "complexity."

What about these examples? :

• a Thai sentence ("heard in the wild") consisting of a pronoun (he/she/they) followed by 12 action verbs. Does this suggest simplicity or complexity?
• very few obligatory markers. Very few distinct non-obligatory marking words.
• a large set of pronouns dependent on social status.
• a plethora of two syllable words which map to a pair of simple one-syllable words. (Interestingly, many/most of the THREE syllable words are not compounds, but e.g. words borrowed 3 syllables at once from e.g. Sankskrit.)
• several sentence-ending particles to show mood, etc.. This may be considered complexity. But as we noted, English also has several words or interjections that play a similar role.

And so on.

I'm not following. What was your point in proposing time-to-learn as a working definition, if not to resolve the ambiguity problem?

As I just tried to explain a function that "evaluates" a language for simplicity requires
. (a) precision of definition
. (b) scores based on agreed measures of "simplicity."
ANY solution will be "fuzzy", and very hard to agree on. The "time-to-learn" seemed logical: A multitude of features (a frequent sign of complexity) takes more time to learn than fewer features.

And aren't you claiming Thai is less complex than other languages?

That's how it seems to me; and in this thread I've tried to advocate for that stance.
But agreeing on any specific measure of simplicity will certainly be impossible in this thread.
Arguing a specific linguistic point might be of interest. Generalities on non-linguistic matters are out of place here.