11 ms·
New study measures grammatical complexity of languages
- jasfi 3y agoWhat about pronunciation? This can make a language especially difficult for some people, especially when they don't already speak a language that sounds like the one they want to learn.
- otabdeveloper4 3y agoThe fact that these people conflated "grammar" and "inflection affixes" makes me think they don't know what they are doing.
- tgv 3y agoThey know what they're doing: publishing and attracting attention. It's pop-sci-bait. You take some grandiose idea X, dumb it down, operationalize it in a way that doesn't really prove what you want to say, do a correlation, and publish "X is true" or "X is false".
- mcswell 3y agoI haven't paid a lot of attention to the debate, so I can't give you a citation. But there have been many claims over the years (some of them by real linguists) that language property X is correlated with physical/ social/ ethnic property Y. This "grandiose idea" that you're referring to was one of those, and not one made up by the authors of this paper.
- stinos 3y agoThey literally say it's an example, and 'along with many others'. Doesn't sound like actual conflation on their side, perhaps misreading or misinterpreting from your side?
- jcranmer 3y agoThe fact that is a PR about the research (and consequently "dumbed down" for a lay audience) and not the actual research paper makes it a bit hard to divine, but I'm pretty sure the authors didn't conflate grammar with inflection affixes. In fact, the section on "Defining complexity" in the article here does mention this: > "Many of the disagreements are down to differences in how 'complexity' is defined [...] In this study, we improved the methodology by teasing out two distinct measures: fusion (how many affixes verbs and nouns have) and informativity (how many distinctions are made)." It strikes me as more likely that they're focusing on easily quantifiable metrics of grammatical complexity. Without reading the paper myself, I'd guess that the metrics chosen here suffer from lamplight fallacy (they're focusing on what can be quantified because it can be quantified as opposed to working out what would be a useful guide).
- mumblemumble 3y agoI think you may be right about the lamplight fallacy. I just started learning Mandarin this year. It's somewhat famous for having a simple grammar, but that in no way means that it has lesser expressive capacity. It just means that the language has evolved different ways of expressing the same ideas. They don't happen to be ways that can easily be summarized as grammatical rules. But I'm not sure I'd take that to mean that it's a simpler language. I think it probably says more about the concept of grammar as defined by grammarians.
- otabdeveloper4 3y agoMandarin grammar is very complex. Mandarin inflection is quite simple. (Worse, Chinese language instructors say "grammar" but mean "inflection", which makes Chinese grammar textbooks an infuriating affair.)
- bdowling 3y agoUsually journalists don't actually understand what they are reporting on. Usually that's fine because most readers don't actually understand what they're reading.
- codr7 3y agoUntil you do, and realize they have no clue. And then you turn the page and again assume they know their shit.
- temporallobe 3y agoI’d like to see the breakdown of complexity for each language. The link to the DOI page is also broken.
- letmevoteplease 3y agoThe study does not include a breakdown, other than a map with unlabelled dots. https://www.science.org/doi/full/10.1126/sciadv.adf7704 https://www.science.org/doi/full/10.1126/sciadv.adf7704 You can browse the data they used here, though: https://grambank.clld.org/languages https://grambank.clld.org/languages
- meepmorp 3y agoTheir raw data is on github: https://github.com/grambank/grambank.git https://github.com/grambank/grambank.git
- nmarinov 3y agoeven more data, code, and materials related to the study: https://zenodo.org/record/8032086 https://zenodo.org/record/8032086 (linked at the bottom at the science article) also at github: https://github.com/OlenaShcherbakova/Sociodemographic_factors_complexity/tree/v2.0 https://github.com/OlenaShcherbakova/Sociodemographic_factor...
- pmontra 3y agoComplexity vs non native speakers could make sense but IMHO they must measure how many non native speakers a language had along its history and when the grammar complexity changed. A lot of non native speakers in the last ten years after one thousand years of isolation can't possibly change a grammar much yet.
- tanvach 3y agoAs a Thai, I’m somewhat proud to see that our gramma is quite simple, i.e. no tenses and genders, etc. These extra structures are superfluous if you think about it. This makes it easy for beginners to start forming simple sentences. However, the complexity gets pushed to the vocabulary. It tends to have different words that mean the same thing, but with differing levels of, say politeness. Not to mention a whole ton of slang and double meanings.
- fifilura 3y agoGood on you! For me the alphabet is the biggest hurdle. One time, many years ago, I ended up in a police station in Bangkok because of a robbery. They wanted my address but they didn't want to write it down in latin letters, so I had to pronounce my address and the woman wrote it in Thai script. I really wonder how they would ever use that address. I am pretty sure the swedish postal service would have an interesting time if they decided to mail me a document.
- tanvach 3y agoWorking with the Thai government can be a bit of a pain - forms must be filled in Thai only. Agree with the characters. Although it’s not super hard to learn (my non-Thai wife learnt the characters pretty quickly), I wish we officially have something like pinyin. Fun fact: written Thai in Latin characters is unofficially called ‘karaoke’ [1]. [1] https://blogs.transparent.com/thai/how-to-readwrite-in-karaoke/ https://blogs.transparent.com/thai/how-to-readwrite-in-karao...
- _a_a_a_ 3y agoKnowing a Thai lady, I've got some idea of how it differs from English. Bit of an eye-opener. But I don't conclude that tense and gender are superfluous; far from it. In fact the tenses in English are extensive and extremely useful. If they're unnecessary then a whole lot of other things might be considered so as well. Perhaps we should just say there are differences in languages, and not judge those.
- eru 3y ago
- endisneigh 3y agoInherently won’t you always need the same level of complexity to express the same meaning with equivalent preciseness? You either have complex grammar, vocabulary or both. Though I suppose in the far future sending thoughts to each others brains directly will probably require less energy and will be completely unambiguous.
- skummetmaelk 3y agoIMO complex grammar just creates a high barrier to being understood at all, while hiding the complexity in vocabulary allows a more gradual learning curve.
- endisneigh 3y agoWould it really be simpler if there were a language where the grammar were that you only say one word and one word only, but all expressions require you to memorize a dictionary of a billion words because they’ve decided they want their grammar to only be one word utterances?
- skummetmaelk 3y agoThe other extreme would also create a very high barrier to being understood at all.
- chc 3y agoAgglutinative languages say hi.
- mcswell 3y agoI think you meant polysynthetic languages. And you don't have to memorize all the forms of each stem, there are grammar rules that usually let you derive each form.
- abeppu 3y agoBut not all complexity that we see is needed for precise expression. Like, if romance languages like Spanish and French did not require nouns and adjectives to agree on gender, but instead only the noun needed to carry gender info, that would not obligate the vocabulary to get more complex. Similarly English has rules about the order of adjectives (Clifford is a big, red dog, not a red, big dog) and the presence of those rules didn't simplify the vocabulary.
- jmyeet 3y ago> Many linguists have claimed that languages with more non-native speakers tend to simplify their grammars I think this misses the confounding variable. What maintains grammatical complexity seems to be centralization. There's an age-old debate about whether grammar is prescriptive or descriptive. I'm in the latter camp, which is to say that if enough people say something "wrong" then it's not wrong anymore. Education however tends to be prescriptive. It teaches and maintains an existing way of using a language, so much so that in some places the colloquial and "official" versions of a language are esentially two different languages. The "many linguists" here ascribe this to a "society of strangers". I think the reality is that absence of a central authority means languages tend to change faster and thus lose unnecessary complexity. English is a great example of this. English is a Germanic language. Old English and Old German are very similar languages. After the Norman invasion in 1066, hte official language of the English court became French and remained so for ~300 years. This was the period of MIddle English. After this came early Modern English (eg Shakespeare). While Old English is essentially unintelligble to the modern English speaker, a lot of Middle English can be understood. But that 300 year period made English grammar much simpler. Most cases disappeared. Word order became fixed. Noun gender mostly disappeared (other than pronouns for people). The subjunctive mood mostly disappeared (eg "If I were king" instead of "if I was king" is about the extent of modern English subjuntive mood). The last vestige of genitive is mostly just adding "'s" to things (with a few exceptions of course). English spelling and phoentics is of course a complete mess. So are tenses. But absent central authority we stopped having to know that a table was "he" and a chair was "she" (made up examples; I don't know their gender in Old English). Isn't that interesting?
- codr7 3y agoYeah, we're still suffering from that mess in Swedish; ett bord (table), en stol (chair). And there are no viable heuristics, it's all arbitrary. I'm all for simply picking one and running with it, it's basically line noise.
- eru 3y agoGrammar in language can act like registers in a CPU: it allows you to use simple pronouns like 'he', 'she', 'they', 'we', 'it' to quickly refer to previously introduced entities. Grammatical gender gives you a quick way to assign the entities to different slots. If you want to optimally hash your entities to the different slots, assigning gender approximately random would be the efficient strategy. Btw, you might like Turkish: they don't have any grammatical gender. And in general it's a very regular and systematic language.
- smartmic 3y agoOff-topic, but I stumbled over the fact that the world map is Asia-centric. I would not have expected that from the Max Planck Society, a German research institution. The way it is displayed is very unusual in Germany and Europe (actually almost everywhere, isn't it?).
- eru 3y agoThey also had to cheat and make the Pacific proportionally a lot smaller compared to the continents.
- janaagaard 3y agoWhy are there more than one dot in the big countries like The United States, Brazil and Australia? I would have expected a single dot for each of these countries.
- meepmorp 3y agoThose countries all have plenty of languages which are native to their territory which aren't the dominant one in the country. Aboriginal languages in Australia, native American languages in the US, various tribal languages in Brazil.
- ecedeno 3y agoWikipedia has 17 indigenous languages for Brazil alone https://en.wikipedia.org/wiki/Languages_of_Brazil https://en.wikipedia.org/wiki/Languages_of_Brazil
- sverhagen 3y agoI'm a little weirded out that languages are plotted on a map. Where does English go? Or French? Or are these just plotted everywhere that language is spoken? (Probably not.) It reminded me of a hill I chose to (not really) die on, a long time ago, where the product folks in my organization were dead-set on indicating the language choice with a (country) flag. Not sure if we ever set on what flag to use for English (it was probably configurable). And the flag for Dutch was a weird abomination between the Dutch and Belgian flag. Weirdly, the French flag was used for French, while Belgium probably speaks as much French as they do Dutch (so why wasn't there a French/Belgian abomination too).
- meepmorp 3y agoGood point! Their dataset seems to have locations associated with the languages: https://grambank.clld.org/languages#map https://grambank.clld.org/languages#map spoiler: English goes in England, French goes in France
- eru 3y agoI like the little joke post that had a menu that let you choose between 'Traditional' and 'Simplified English'.
- sverhagen 3y agoNot to get entirely off topic here, but I heard a guest on the Dutch national radio this morning make a recommendation to all the Dutch folks coming back from vacation in France around this time, to prolong the vacation vibe by tuning in to the "français facile" radio station, with presumably a less waterfall-like experience of listening to the news in French, so that they now could finally understand what was being said.
- goodbyesf 3y ago[flagged]
- jstx1 3y agoReally? An unlabeled map scatterplot? The article copied only one of them from the original paper but the second one is just as useless. Isn't Science supposed to be a prestigious journal? The quality of the data presentation is embarrasing.
- eru 3y agoThe more prestigious the journal, often the more garbage the science they publish. Alas.
- DougMerritt 3y agoCitation please (preferably from a prestigious journal ;) But seriously, why is that, or why do you think so?
- eru 3y agoPart of the problem is that those prestigious journals only want to publish the most exciting science. And, alas, the most exciting science is more likely to be wrong and rubbish. See eg https://en.wikipedia.org/wiki/Publication_bias https://en.wikipedia.org/wiki/Publication_bias and the whole reproducibility crisis. (Though that doesn't explain why the graphics might be bad, because Nature and Science would still want polished content.)
- CollinEMac 3y agoIt would be nice if we could just see the list of languages ordered by complexity. I can't derive a thing from the map in the article.
- ur-whale 3y agoVery interesting topic, completely useless article content. Come to think of it, that's the definition of clickbait.
- materielle 3y ago> Language complexity is a hotly debated topic in linguistics, with many different opposing views. They bring this up, don’t address it, then proceed to the results anyways. The idea of grammatical or language complexity is very dubious. They seem to define it by number of distinctions. But, how is that inherently more complex? Wouldn’t we expect Russian children to start speaking later than, say, Chinese children? But we don’t actually see that, all children seem to learn their respective mother tongues at about the same rate. Now, if you’re an adult monolingual English speaker trying to learn Russian, then I’m sure the crazy conjugation and declination systems would appear complex. But it’s probably simply different rather than complex. 5 year old Russians seem to think that case declinations are the easiest thing in the world!
- veqq 3y agoRussian is an interesting situation, where the neuter gender is sort of dying in diaspora [1] communities or among young children. Currently, 13% of nouns are neuter, 41% feminine and 46% masculine. Before the education system and further exposure solidifies the standard, many children can form a 3 way distinction based on word ending. Consonant->masc., stressed -o->neuter, everything else -> feminine. In low proficiency speakers in heritage communities this often simplifies to consonant->masculine, everything else->feminine. [2] [1] https://d-nb.info/1205872779/34 https://d-nb.info/1205872779/34 [2] https://brill.com/view/journals/jlc/11/2/article-p233_233.xml?language=en https://brill.com/view/journals/jlc/11/2/article-p233_233.xm...
- pessimizer 3y ago> all children seem to learn their respective mother tongues at about the same rate. I need a reference for this. We don't even learn different features or sounds of our own languages at the same rate; some are harder and take longer, some are easier and we get them quickly. If this weren't true, there wouldn't be stereotyped ways that we have for imitating childish speech. This honestly sounds like the law of averages rearing its head again. There aren't many languages that five or ten years of immersion and few other responsibilities (like a child has) couldn't make easy, but that doesn't mean that all languages are equally difficult for their native speakers.
- 3y ago
- janalsncm 3y agoI found it interesting that the article made no mention of Sinitic languages even though they’re the most obviously different. Personally I have found that many Mandarin speakers will accidentally misgender people in English since the Mandarin words for “he” and “she” are homophones.
- mrwnmonm 3y agoDoes anyone know a good introduction for understanding languages in general, and modern linguistics too?
- kbrannigan 3y agoHere’s a high quality channel https://youtu.be/m3g6AOPvuPs https://youtu.be/m3g6AOPvuPs
- seo-speedwagon 3y agoIt’s not clear to me how good a measure of grammatical complexity “number of affixes” is. Like Japanese has a lot, English fewer, and the various Chinese dialects almost none. And if that’s one out of the two dimensions they’re using, it seems like a pretty incomplete representation.
- dmn322 3y agoIt’s not great. There are ~2 broad categories of language based on their use of affixes: analytic (which doesn’t make much use of affixes, instead relying on word order), and synthetic (which does use affixes). Synthetic has multiple sub-categories depending on how the affixes are used. This approach completely ignores analytic languages which put their complexity in word order instead of affixes, and it ignores the distinctions within synthetic languages such as agglutinating, which stacks up affixes vs fusional, which combines multiple affixes into a single new affix. It also misinterprets what people mean when they say languages which collide “reduce complexity” which really means they tend to become more analytic. Nobody expected the Scandinavian languages to have this characteristic because they’ve been in the process of diverging from a common language. The process they’re referencing applies to languages which come from the collision of distinct languages, such as pidgins and creoles. The hypothesized tendency is for languages which have had consistent boundaries and populations for a long period of time to be more agglutinating then eventually fusional, and eventually polysynthetic like some Native American languages, where a single word can be translated to a complete sentence in English for instance Edit: corrected for distinction between fusional and synthetic
- mcswell 3y agoThere's another category between analytic (no affixes) and agglutinating (potentially multiple affixes on a single word), namely inflectional (usually one affix per word). Many IndoEuropean languages like Spanish, German and to some extent English fall into that class, although there are nuances. Spanish verbs generally have a single suffix that encodes person, number, tense and mood, but you could analyze the Spanish future tense as having a future suffix followed by a person/ number suffix.
- frankfrankfrank 3y ago[dead]