Morphological complexity

[Kuchigakatai]

I personally think this idea is highly dubious — there’s been plenty of complex languages spoken by large communities (e.g. Latin, Mayan, Quechua, Turkish, Fillipino, Swahili…). And anyway, how do you define complexity? To my understanding, most languages are about as complex as each other.

Quite a number of tribal languages seem genuinely a lot more morphologically complex than any of those six languages though, in terms of number of lexical stems used in inflections and/or total number of inflectional forms (including regular very productive derivations). I'm thinking about most languages from the north-eastern quarter of North America and also Papua. I remember reading a paper that had used records from Nunavut's Inuktitut-speaking parliament as data and found that the texts effectively violated Heaps' law showing linear growth instead, to give you an idea of just how complicated its morphology is. Inuktitut morphology is so complicated and essential to express anything that a speaker is basically constantly making up new words with it, in much the same way we add words to a verbal phrase in novel permutations. And Inuktitut seems relatively tame next to Ojibwe and still more so Navajo or Wichita...

I don’t think it’s particularly good either (I’m particularly suspicious of ‘fusional’, which seems to apply to IE and nothing else), but it’s certainly useful for getting a good idea for what a language might be like.

The examples usually given for fusional languages are typically European (Spanish 1SG.PRET -í, Latin 1SG.PRES.ACT -ō), yes, but the categories can apply to any language. For example, Inuktitut -ara, which is a verbal suffix for a 1st person singular subject doing something to a 3rd person singular specific object in a statement. If you turn it into a question, then it becomes -igu.

[Frislander]

Welcome! The gherkins and tea will be arriving shortly...

I was wondering whether we still did that! I’m glad we still are.

You can actually get quite far with a language without figuring out its context — for my most well-developed language, I still know absolutely nothing about the world it’s spoken in. The context only really affects the lexicon, semantics and pragmatics of a language — the rest is more or less orthogonal to context.

I’m not entirely sure why you would ask this… as I said, most of the basics of language is more or less orthogonal to culture. There isn’t really any commonality between the cultures which speak agglutinative languages, for instance.

Small quibble, but this may not necessarily be true for naturalistic languages if the view held by Trudgill and Mcwhorter and co that smaller more intimate social networks tend to accumulate complexity/irregularity more easily is correct (and I tend to think that this is probably true to some degree), in which case more complex morphology with more irregularity (of whatever kind) should probably restricted to smaller groups of people, i.e. not large empires. This of course is not the same as saying that smaller tribes speak universally fusional languages - even highly morpholgically transparent agglutinative languages can be highly complex/irregular in other ways.

I personally think this idea is highly dubious — there’s been plenty of complex languages spoken by large communities (e.g. Latin, Mayan, Quechua, Turkish, Fillipino, Swahili…). And anyway, how do you define complexity? To my understanding, most languages are about as complex as each other.

If someone could split this into the Miscellany thread that would be great but here's the follow-up. Basically the fullest picture can be gained by reading Trudgill's Sociolinguistic Typology if you can get a hold of it, but the main points are the following.

Firstly, "complexity" in this instance is tied mainly to irregularity and redundancy, which comes in several varieties. Firstly there's irregularity of form, that is the degree both to which forms change depending on context (especially things such as arbitrary inflection classes) as well as the degree to which forms supplete. Secondly there's irregularity of lexicon, the degree to which you can predict the meaning of a form from its components. This point partly ties into the third for example with cases where you might have a suffix -i but it signifies plurals with one group of nouns and genitives with another and so on, but also includes instances where otherwise agglutinative languages show irregular derivation, such as bipartite verb complexes in language families such as Na-Dené and Caddoan. In essence these points make the task of acquisition difficult for a non-native speaker. Finally there's also a third less important type which is complexity of the lexicon, which is mainly about homonymy (which is what allows languages such as Hawai‘ian to be "complex" for these purposes, since the small number of sounds results in a much higher amount of homonymy than in languages with a more average sound inventory).

All of these are essentially functions of the community and a shared context, especially the latter - where you are familiar with most of the people you are likely to be talking with, you can afford to have much higher amount of redundancy with possibility for ambiguity because the people you are talking with are much more likely to be able to disambiguate what you are saying with other information they have access to. By contrast, if more of the people you speak with are relatively unfamiliar to you, ambiguity is bad because you can't rely on your interlocutors having access to shared information to the same extent, and if those people have a different linguistic background to you then redundancy is bad because it makes the process of acquisition harder (because for something to be redundant it by definition must be more than strictly necessary, i.e. people learning languages with higher amounts of redundancy in their morphology have more to learn). As a result in the latter situation of non-intimate contact the tendency will be for the people learning the dominant language to somewhat pidginise the dominant language to remove redundancies and native speakers of the dominant language will begin to acquire those pidginisations over subsequent generations, though this depends on how large and self-contained the dominant language's speakers are, hence why English in America is much less strongly influenced by Native American languages than Indian English is by Indo-Aryan ones (and more extreme examples exist).

The above languages then are not actually particularly complex according to this definition, since their morphologies (with the exception of Latin) are quite transparent and regular. It's interesting you bring up Latin because the development of Vulgar Latin actually fits with this theory very well, because the majority of people in the Empire outside of the immediate environs of Rome itself did originally speak Latin and settlement of ethnic Romans in the rest of the empire was sparse enough to the point that they couldn't indirectly impose this irregularity on the Latin spoken by the conquered peoples (especially in the absence of overt instruction), so much of the most irregular parts of Classical Latin were already lost among the speech of most Roman subjects before the Empire even fell. A similar thing has also to some extent happened in Quechua where varieties outside of the mountain heartlands simplify much of the more complex and idiosyncratic person-marking morphology.

Furthermore let's put to be this idea that "all languages are equally complex" because they're really not. Mandarin Chinese doesn't have "syntactic classes" in its verbs to compare with Latin's verb conjugations, much less anything even remotely close to the mess of Navajo verbal morphology. The closest you get is the classifiers and even then it's easy to imagine a version of Chinese without them of with far less of them and a language with a lot more inflection with them. Indeed we may not even have to imagine it given that many of the more uncommon classifiers appear to be being lost in modern vernacular Mandarin, as well as much of the idiosyncratic selection of classifiers by particular nouns.

I dunno… I’d say those options are pretty mutually exclusive, even if it is possible to combine them to a small degree. But I do get what you’re saying; there do exist other options which can only be combined and balanced in many ways.

I also quibble this, but more because I think the fusional-agglutinative-isolating trichotomy is bad pop-typology that actually hinders understanding of any language whether natural or constructed.

I don’t think it’s particularly good either (I’m particularly suspicious of ‘fusional’, which seems to apply to IE and nothing else), but it’s certainly useful for getting a good idea for what a language might be like.

I can sort of see that, but I will note that "fusional" is not solely IE, it has been used for Semitic languages as well I've noted, but that only further problematises the term because IE and Semitic morpholgy are fundamentally very different in how they are organised, reducing the utility of the word "fusional" as a classification.

[bradrn]

I personally think this idea is highly dubious — there’s been plenty of complex languages spoken by large communities (e.g. Latin, Mayan, Quechua, Turkish, Fillipino, Swahili…). And anyway, how do you define complexity? To my understanding, most languages are about as complex as each other.

Quite a number of tribal languages seem genuinely a lot more morphologically complex than any of those six languages though, in terms of number of lexical stems used in inflections and/or total number of inflectional forms (including regular very productive derivations). I'm thinking about most languages from the north-eastern quarter of North America and also Papua. I remember reading a paper that had used records from Nunavut's Inuktitut-speaking parliament as data and found that the texts effectively violated Heaps' law showing linear growth instead, to give you an idea of just how complicated its morphology is. Inuktitut morphology is so complicated and essential to express anything that a speaker is basically constantly making up new words with it, in much the same way we add words to a verbal phrase in novel permutations. And Inuktitut seems relatively tame next to Ojibwe and still more so Navajo or Wichita...

I’m struggling to figure out how to respond to this, mainly because I have no idea why this is relevant. Are you trying to say that ‘complexity’ here is basically the number of morphemes per word? If so, then yes, tribal languages tend to be more complex than widely-spoken languages, but only because the most widely spoken languages are disproportionally drawn from language families which generally have less morphemes per word. But I’m not sure why the Inuktitut verbal system should be considered any more complex than e.g. the English verb phrase.

I don’t think it’s particularly good either (I’m particularly suspicious of ‘fusional’, which seems to apply to IE and nothing else), but it’s certainly useful for getting a good idea for what a language might be like.

The examples usually given for fusional languages are typically European (Spanish 1SG.PRET -í, Latin 1SG.PRES.ACT -ō), yes, but the categories can apply to any language. For example, Inuktitut -ara, which is a verbal suffix for a 1st person singular subject doing something to a 3rd person singular specific object in a statement. If you turn it into a question, then it becomes -igu.

Yes, I agree that many languages have at least a couple of fusional morphemes. But usually the term is used few languages are as thoroughly fusional throughout as IE (and Semitic) are. To my understanding, the vast majority of Inuktitut inflectional morphemes have only one meaning and are unfused (polypersonal agreement being a prominent exception in many languages), whereas the vast majority of IE inflectional morphemes are highly fused. I don’t think it really makes sense to describe a language such as Inuktitut as ‘fusional’ on the basis of a couple of morphemes alone.

If someone could split this into the Miscellany thread that would be great…

I agree. (Maybe it should even get its own thread.) This thread isn’t really the place for this discussion.

Firstly, "complexity" in this instance is tied mainly to irregularity and redundancy, which comes in several varieties. Firstly there's irregularity of form, that is the degree both to which forms change depending on context (especially things such as arbitrary inflection classes) as well as the degree to which forms supplete. Secondly there's irregularity of lexicon, the degree to which you can predict the meaning of a form from its components. This point partly ties into the third for example with cases where you might have a suffix -i but it signifies plurals with one group of nouns and genitives with another and so on, but also includes instances where otherwise agglutinative languages show irregular derivation, such as bipartite verb complexes in language families such as Na-Dené and Caddoan. In essence these points make the task of acquisition difficult for a non-native speaker. Finally there's also a third less important type which is complexity of the lexicon, which is mainly about homonymy (which is what allows languages such as Hawai‘ian to be "complex" for these purposes, since the small number of sounds results in a much higher amount of homonymy than in languages with a more average sound inventory).

Thanks for clarifying! This makes more sense now.

All of these are essentially functions of the community and a shared context, especially the latter - where you are familiar with most of the people you are likely to be talking with, you can afford to have much higher amount of redundancy with possibility for ambiguity because the people you are talking with are much more likely to be able to disambiguate what you are saying with other information they have access to. By contrast, if more of the people you speak with are relatively unfamiliar to you, ambiguity is bad because you can't rely on your interlocutors having access to shared information to the same extent, and if those people have a different linguistic background to you then redundancy is bad because it makes the process of acquisition harder (because for something to be redundant it by definition must be more than strictly necessary, i.e. people learning languages with higher amounts of redundancy in their morphology have more to learn). As a result in the latter situation of non-intimate contact the tendency will be for the people learning the dominant language to somewhat pidginise the dominant language to remove redundancies and native speakers of the dominant language will begin to acquire those pidginisations over subsequent generations, though this depends on how large and self-contained the dominant language's speakers are, hence why English in America is much less strongly influenced by Native American languages than Indian English is by Indo-Aryan ones (and more extreme examples exist).

This concept seems vaguely suspicious to me, but I can’t see anything which would make it actually wrong.

The above languages then are not actually particularly complex according to this definition, since their morphologies (with the exception of Latin) are quite transparent and regular. It's interesting you bring up Latin because the development of Vulgar Latin actually fits with this theory very well, because the majority of people in the Empire outside of the immediate environs of Rome itself did originally speak Latin and settlement of ethnic Romans in the rest of the empire was sparse enough to the point that they couldn't indirectly impose this irregularity on the Latin spoken by the conquered peoples (especially in the absence of overt instruction), so much of the most irregular parts of Classical Latin were already lost among the speech of most Roman subjects before the Empire even fell. A similar thing has also to some extent happened in Quechua where varieties outside of the mountain heartlands simplify much of the more complex and idiosyncratic person-marking morphology.

That’s pretty interesting actually — I hadn’t known this before. I can definitely accept that a rapid expansion of a language can lead to simplification — but I find it hard to believe that that complexity cannot be restored in the centuries afterwards.

Furthermore let's put to be this idea that "all languages are equally complex" because they're really not. Mandarin Chinese doesn't have "syntactic classes" in its verbs to compare with Latin's verb conjugations, much less anything even remotely close to the mess of Navajo verbal morphology. The closest you get is the classifiers and even then it's easy to imagine a version of Chinese without them of with far less of them and a language with a lot more inflection with them. Indeed we may not even have to imagine it given that many of the more uncommon classifiers appear to be being lost in modern vernacular Mandarin, as well as much of the idiosyncratic selection of classifiers by particular nouns.

True, Mandarin Chinese is far less complex than Latin or Navajo — if you only look at morphosyntax. But, as you said above yourself, that isn’t all there is to complexity! Mandarin Chinese has a huge number of homophones, which definitely counts towards total complexity by your definition above. And even if you want to restrict yourself to morphosyntax, let me quote zompist from the ALC: ‘Whenever someone suggests that languages without inflections are easy, I like to hit them with a copy of Mandarin Chinese: A Functional Reference Grammar (by Charles Li and Sandra Thompson, 1981), which is nearly 700 pages and leaves quite a welt.’.

I don’t think it’s particularly good either (I’m particularly suspicious of ‘fusional’, which seems to apply to IE and nothing else), but it’s certainly useful for getting a good idea for what a language might be like.

I can sort of see that, but I will note that "fusional" is not solely IE, it has been used for Semitic languages as well I've noted, but that only further problematises the term because IE and Semitic morpholgy are fundamentally very different in how they are organised, reducing the utility of the word "fusional" as a classification.

This is exactly the point I was making!

[zompist]

Admin note: I split this off, but some of the earlier discussion is still in this thread because it was tied to answering the OP's question.

[bradrn]

Admin note: I split this off, but some of the earlier discussion is still in this thread because it was tied to answering the OP's question.

Thanks zompist! This was getting a bit long.

[zompist]

Over the years I've heard a lot of arguments about complexity, but I've never heard any any refutation of the related ideas that a) most humans learn their native language adequately in about a ten-year period, and b) really good language grammars are of roughly equal size.

As Dixon puts it, it's certainly possible that some languages are twice as complex as others, and no one really maintains that we know that languages are less differentiated than this in complexity. Do the critics seriously maintain that some languages are, say, ten times as complex as others? E.g., that they take ten times as long for a native child to learn?

I really think a lot of people, including unfortunately a few linguists, approach the problem primarily thinking about teaching the written language in a college course. When the average non-linguist thinks about complexity, that's what they mean too: 2L acquisition with a focus on the difficulties of the first year, or even the first month. (I expect most English speakers would think "Russian is hard" because you have to learn an alphabet, not because you have to learn all the cases.) In that sense, there's certainly a lot more to do in the first year of Latin than, say, Swedish. If you look at things like "when can you read an ordinary newspaper article", the answer might be "four months" for French and "six years" for Mandarin.

But again, is it really maintained that, say, a Swedish child will learn the language ten times faster than a Roman child? Things like Romance verbal morphology are hard for the college student. And they're hard for the toddler, too! But they have the time and energy for it, and can't complain effectively about it, and morphology is really dwarfed by all the lexicon they have to learn anyway.

Subjectively, some languages do really seem intimidating. I have Valentine's Nishnaabemwin grammar, and it scares me. But that's not a linguistic judgment, and it's arguably just provincialism. The complexity is not what I know from other areas. Plus, honestly, Valentine is not the best writer.

There are a bunch of caveats, of course, which I'll briefly go over to avoid straw men. There's an obvious increase in complexity due to writing systems, which are not all equal. There's a huge increase when to be educated you have to learn a second language-- most obvious for (say) Akkadians who had to learn Sumerian, but also for medieval Europeans who had to learn Latin, or pre-1905 Chinese who had to learn Old Chinese. There's also arguably a lot of complexity in technical terms, and in languages with a long, preserved history. Most of it is arguably not linguistic, but some is. That's one reason a 12-year-old, though fluent, may still have an awful lot left to learn. Plus of course pidgins and conlangs may be less complex. Possibly creoles, though I seldom hear what actual creole speakers think about this.

Finally, my point (b) above is of course highly informal. Brad already quoted my standard invocation of Li & Thompson. But I think people still routinely underestimate the complexity of English syntax. I've written a 300-page book on it, and believe me, it's a primer. McCawley's textbook is 800 pages and many of his sections are bare overviews of the phenomena they discuss. Compare how many pages it takes to list, say, all the irregularities of Latin morphology.

(Once when I brought this up, I think someone said that we don't know that Latin syntax wasn't equally complex. That may be so! But the advocates of particular-language complexity can hardly just make the assumption that, oh, all syntax is equally complex. It's reasonable, though hardly proven, that the work not done by a language's morphology has to be done by its syntax.)

Ah, and I should add that I haven't read Trudgill's book and I'm not trying to address it. I have nothing against the idea of sociolinguistic barriers to 2L acquisition; in fact I think it's a big part of why the major standard languages retain features that are hard to learn.

[akam chinjir]

Finally, my point (b) above is of course highly informal. Brad already quoted my standard invocation of Li & Thompson. But I think people still routinely underestimate the complexity of English syntax. I've written a 300-page book on it, and believe me, it's a primer. McCawley's textbook is 800 pages and many of his sections are bare overviews of the phenomena they discuss. Compare how many pages it takes to list, say, all the irregularities of Latin morphology.

The Huddleston and Pullum Cambridge Grammar of the English Language is around 1800 pages, fwiw. I suspect a grammar (for any natural language) could be longer than that without wasting words, and the lengths of actual grammars don't say much about the complexity of the languages. I mean, a grammar of English can be ridiculously long in part because so many people, including so many native speakers, have devoted a lot of time and intelligence to exploring its nooks and crannies.

(For the record, I'm not sure whether it makes sense to talk about the complexity of a language as a whole, and I'm all but certain that it's nonsense to say that all languages are equally complex.)

[bradrn]

(For the record, I'm not sure whether it makes sense to talk about the complexity of a language as a whole, and I'm all but certain that it's nonsense to say that all languages are equally complex.)

Surely those two sentences are contradictory? (I do agree with the rest of your post though.)

[akam chinjir]

(For the record, I'm not sure whether it makes sense to talk about the complexity of a language as a whole, and I'm all but certain that it's nonsense to say that all languages are equally complex.)

Surely those two sentences are contradictory? (I do agree with the rest of your post though.)

Not at all, the first sentence entails the second. (If you can't measure the complexity of a language of a whole, then there's no way to say that it is or isn't equal to the complexity of another language.)

[bradrn]

(For the record, I'm not sure whether it makes sense to talk about the complexity of a language as a whole, and I'm all but certain that it's nonsense to say that all languages are equally complex.)

Surely those two sentences are contradictory? (I do agree with the rest of your post though.)

Not at all, the first sentence entails the second. (If you can't measure the complexity of a language of a whole, then there's no way to say that it is or isn't equal to the complexity of another language.)

Sorry, looks like I misinterpreted your second clause. (I misread it as ‘all languages are not equally complex’ rather than ‘it is nonsense to say that all languages are equally complex’.)

[akam chinjir]

Sorry, looks like I misinterpreted your second clause. (I misread it as ‘all languages are not equally complex’ rather than ‘it is nonsense to say that all languages are equally complex’.)

Ah, yes, maybe I should've been clearer, the idea was that the question might be a bad one, with no correct answer.

[Raphael]

The Huddleston and Pullum Cambridge Grammar of the English Language is around 1800 pages, fwiw. I suspect a grammar (for any natural language) could be longer than that without wasting words, and the lengths of actual grammars don't say much about the complexity of the languages.

I might be wrong on this, but I've got the impression that for well-established natural languages, you could theoretically write 100-volume grammars without repeating yourself.

[Raphael]

(For the record, I'm not sure whether it makes sense to talk about the complexity of a language as a whole, and I'm all but certain that it's nonsense to say that all languages are equally complex.)

Surely those two sentences are contradictory? (I do agree with the rest of your post though.)

Not at all, the first sentence entails the second. (If you can't measure the complexity of a language of a whole, then there's no way to say that it is or isn't equal to the complexity of another language.)

Then again, if you can't measure the complexity of a language of a whole, then you can't really say that one language is more complex than another, either.

[Ares Land]

Iroquoian languages, for instance, are morphology complex in the sense that the morpheme-to-word ratio is high.

And just like Nishnaabemwin, grammars are intimidating. But when it comes down to it, I think say, Cherokee is not really worse than Latin, in terms of ease of learning. There's about an equivalent number of verb classes, affixes are actually pretty regular. The polypersonal marker are difficult, but in terms of learning arbitrary affixes, no worse than noun and adjective declension.
For that matter, you can treat a fair bit of the affixes as derivational and just learn the derived forms.

The intimidating part is stylistic, I believe. It stems from an exhaustive treatment of morphology, including rare forms, derivations, with constant references to proto-Iroquoian, and little concern for putting things together in neat, rhyming tables like rosa, rosa, rosam,...
Plus, the restricted phonology means that everything kind of looks identical.

Let's take a case study, Ancient Greek is another poster child for morphological complexity. And the reputation is justified. I mean, every verb is irregular in its own way...
Given the right incentives, though people learned it just as they learn English now. They say the Koine Greek of the Gospels is really awkward, so they didn't learn it perfectly, but hey, my English's awkward too. Koine Greek spoken in a Syrian marketplace was probably kind of a pidgin, but the English you hear in business meetings today would make Strunk & White cry in despair.

[Raphael]

They say the Koine Greek of the Gospels is really awkward, so they didn't learn it perfectly, but hey, my English's awkward too. Koine Greek spoken in a Syrian marketplace was probably kind of a pidgin, but the English you hear in business meetings today would make Strunk & White cry in despair.

One German Writer, Peter Littger, has written several books on the mistakes Germans tend to make when trying to speak or write English. One example I can think of right now is when he met a man who worked as a consultant, and apparently thought he was the one doing the consulting, rather than the one being consulted.

[zompist]

I might be wrong on this, but I've got the impression that for well-established natural languages, you could theoretically write 100-volume grammars without repeating yourself.

This is kind of wrong, and kind of right, for interesting reasons.

As an expression for the holy grail, the Absolutely Complete Grammar of English, I think it's in the ballpark. That's using grammar in the fullest sense: lexicon, all the syntax, all the pragmatics, all the idioms, all the quirks of combining these things that don't get put into dictionaries.

On the other hand, you don't need a hundred books to learn a foreign language. You will do very well if you read (and absorb!) a textbook, a dictionary, a book of pitfalls, a book on slang, a style guide, and (to refine your style and understanding) five books written in the language.

To get a degree in the language, you'd probably read somewhat more than 10 and less than 100 books on or in it.

And yet, again, the morphology of even the quirkiest language can fit into one thick book.

(I don't think it's a contradiction that it's also the case that even a very stupid person can also learn the same language if they grow up speaking it, and without reading a word. What you can learn over a 15-year period is pretty immense!)

[Raphael]

On the other hand, you don't need a hundred books to learn a foreign language. You will do very well if you read (and absorb!) a textbook, a dictionary, a book of pitfalls, a book on slang, a style guide, and (to refine your style and understanding) five books written in the language.

To get a degree in the language, you'd probably read somewhat more than 10 and less than 100 books on or in it.

Hm, from my own experience learning English, I don't really remember how many books I had read by the time I was half-way fluent - I guess the important part was that after a while I had developed a kind of instinct for how things are usually done in English. Then again, I've read English texts written by German speakers who have probably been exposed to English for as long as I have been, but still use recognizably German syntactical features and idioms, so I guess developing an instinct for a language takes longer for some people than for others.

[Zju]

I personally think this idea is highly dubious — there’s been plenty of complex languages spoken by large communities (e.g. Latin, Mayan, Quechua, Turkish, Fillipino, Swahili…). And anyway, how do you define complexity? To my understanding, most languages are about as complex as each other.

Quite a number of tribal languages seem genuinely a lot more morphologically complex than any of those six languages though, in terms of number of lexical stems used in inflections and/or total number of inflectional forms (including regular very productive derivations). I'm thinking about most languages from the north-eastern quarter of North America and also Papua.

Isn't that just due to chance, though? Throughout history there were just two, maybe three dozen lingua francas - depending on how you count them. And how many languages or linguistic areas are as or more morphologically complex as NE North America and Papua? Probably Caucasus, maybe also parts of the Andes. Other than in those areas, about any language is less morphologically complex, with the odd exception here and there. I'd rather attribute it to chance that there was never so morphologically complex language spoken by a large community or as lingua franca.

Alternatively, one could look at how languages develop in terms of morphology once they gain a sizeable amount of speakers, but even then there's the question of causality.

[aporaporimos]

The claim that all languages are equally complex has always seemed implausible to me on its face. If it were actually true that 1) complexity of languages can be measured and quantified and 2) doing so reveals that all languages are equally complex, then this would surely be a monumental discovery, the fundamental theorem of linguistics. But in practice people use it to mean the opposite, that language complexity can't be quantified, and so any assessment that one language is more complex than another (usually on the basis of how many inflectional categories it has) is specious. Which is fair enough as far as it goes, but doesn't justify the much stronger claim that complexity in one part of a language will always be perfectly counterweighed by simplicity in another part.

[zompist]

The claim that all languages are equally complex has always seemed implausible to me on its face. If it were actually true that 1) complexity of languages can be measured and quantified

This seems to me to fundamentally misunderstand why Linguistics 101 books say things like this. It is not because there is some sort of claim that there is a complexity metric.

It's because non-linguists are obsessed with which languages are better than others, and complexity is part of that. They want to hear that French is more logical, Italian is more beautiful, Arabic is God's language, Phrygian is the first language, etc, etc. They want to hear that the standard languages are better than dialects. They want to hear that primitive cultures speak primitive languages. (Think of the native characters in, oh, any old movie set in the West or in Asia.)

And it's not just Americans... when I was in Iquitos, my wife's uncle really wanted to hear that Spanish was the hardest language in the world. (I forget if I was nice enough to not tell him that US high school students think it's the easiest.) If I recall correctly, there's a mini-literature in Japan about how Japan is different (and of course better than) other countries, including in its language.

Everyone here is past the Ling 101 stuff, but linguistics professors all run into it and get tired of it and throw in some stuff to combat the myths.

the much stronger claim that complexity in one part of a language will always be perfectly counterweighed by simplicity in another part.

There is no such claim.

The annoying bit is that the people who criticize the Ling 101 tidbit, as I'm afraid you're doing, are the ones who seem to think there is a complexity metric.

What is it? If it's so important, what exactly is your measurement by which you know that "all languages are equally complex" is wrong?

All of the above is why I prefer Dixon's way of putting it: that it seems pretty evident that languages don't vary by an order of magnitude in complexity. We don't have a world where Primitivese is mastered by toddlers at 4, while Complexish takes people till 40.