As a turk I sadly have to say that the ural-altaic language family is taught in our schools as factual information and it has become a general fact in the public, so much so that when I mention that the theory is actually discredited in the general linguistics circles I get a lot of side eyes, even from the teachers.
I agree with this as a turk, i also want to add that many people in turkey still believe hungarians and turks as brothers because of how some teachers taught students about huns in history class, and uralic-altaic language family theory seems so cool to them. As a fun fact, also in history class, it's still taught that scythians are turks because just they're nomads live in central asia, and it was surprising for me to learn that they're iranic nation and there are lots of evidences about it.
as a slavic language native speaker, i like that the interslavic exist, it may be a nice hobby to learn to speak it someday. people le-learning latin is also cool
I feel the same about Lingua Franca Nova as a native Romance speaker, really easy both to understand and learn if you speak a Romance language (and the former doesn’t require the latter).
I'm from bosnia and when I first heard it I thought one of the creators was from "our" countries (croatia/ bosnia/serbia) and thus put an outsized influence on interslavic from our language
I fully agree with you regarding Chomsky and natural language, but the importance of his research for formal languages and the development of computer science should still not be underestimated.
Great video. I actually have some opinions on these issues 1. While I don't currently believe that the Altaic family exists, that's really due to a lack of evidence more than anything. As more information is found and more comparative studies are done accounting for all the niche languages in the family, it could well be that the Altaic hypothesis is proven 2. The case/adposition debate is mildly annoying but ultimately not worth fighting about imo. I've seen multiple analyses that treat Tagalog particles as "case markers", even though they come before the entire noun phrase and are literally separate words lmao. Ultimately what's more important is that the analysis allows us to understand the inner workings of the lang, no need to argue about the actual terminology 3. The selection of a truly global lingua franca is probably more fantasy than reality, but if such a language was going to chosen, I find it pretty clear that the best choice would be Indonesian, or some creole of indonesian with a large amount of loans from other languages. BI fits all the requirements (limited phonemic inventory, lack of consonant clusters, isolating structure, lack of case and gender marking) and is right now being used as a second-language lingua franca by like 100 million people 4. I mostly align with whatever you said about UG, GG and their derivatives. It's probably way above my pay grade to argue about this, but I find variations of Functional Grammar to be much closer to the "truth" than those on the UG side. Specifically Systemic Functional Grammar is the one I've read most deeply into and I find very little to criticise about it
I love Esperanto so much but I don't think it's a good interlang. Instead, I love it for the culture that's built up around it-it's a fun community for people who love languages and share slightly related mindsets towards peace and such. Esperanto is worth it for the culture, not for international communication
First I want to say, great video overall! However, as the title predicted, I do have some issues with the UG section-I think they're very valid complaints....about an outdated model/understanding of generativism. Please don't read this as an "umm ahckshually" bit, but simply as additional information from someone who went really deep into syntax and about to start a PhD on it: One point is that modern generative grammar is a more broad school than just orthodox Chomskyism, and while a basic framework is rooted in his hypotheses there is a very strong contingent of scholars within generativism who take most of his proclamations with a metric ton of salt. Most generativists I know do not subscribe to an orthodox view of UG as decreed by Chomsky but see the goal of generative linguistics as trying to formulate predictive models which can be tested against natural language in order to discern the true nature of UG, i.e. whatever the actual structural inherences of human language are. At the end of the day, "generativism" =/= "Chomskyism, " it's more about making formalized, predictive models which could theoretically "generate" a structure on their own, rather than solely being descriptions of structure. Another point is that rather than traditional X'/Government & Binding theory, the predominant framework in contemporary generativism is the Minimalist Program, the specifics of which are highly up for debate. But the point is... they did make the models more or less simpler in the 90s. Most of the specific problems you mentioned disappear in the more streamlined modern framework or even more modernized X'-hypothesis reimaginings; for example, all subjects are understood as originating in a functional position below the TP but above VP which typically encodes voice (vP). It explains VSO word order as occurring in languages with (1) mandatory V-to-T movement and (2) a prohibition on Spec-TP being occupied, rather than an EPP obligating it be filled, as well as explaining English double object constructions since both objects have space to exist in the VP. As a bonus, it neatly describes passive-active distinctions crosslingusitically (passive v's just don't have a subject), especially when you take additional contemporary hypotheses like Distributed Morphology into account which seek to better integrate morphemes into the syntax. Determiners appearing on either side of a noun? DPs are the top-level and NPs are within them, not the other way around (the indicators of which also hold up pretty robustly crosslinguistically, unless the language doesn't have DPs at all in which case the point is moot) while linear order establishes post-syntax, not concurrently. Minimalism is just not typically taught in many undergrad curriculums because the specifics of how components interact is still fiendishly complicated without context of X'-theory, so better to start with a pre-MP incarnation and work towards it; there's still a lot of room for improvement in simplifying the framework while keeping it flexible enough to handle any language. As for the general complaints about Anglocentrism, unfalsifiability, and overall convelutedness? Yeah that's pretty spot on. However, I would say that most generativist scholars I know and have worked with recognize these as problems and want to find solutions. They use generative models on a predictive basis and test against natural language, adjusting for contradictions when they come across them and retesting. There's also a massive push to reassess the framework in order to better incorporate non-IE languages. I also know a lot of generativist psycholinguists who are doing the best they can to test processing and production against predictive models and adjusting theory accordingly. As I mentioned before, Minimalism is neither perfect, homogenous between advocates, or even particularly straightforward, and it has no shortage of ridiculous, unfalsifiable proposals that are entirely theoretical. However, in a class I took in my undergrad with Jorge Hankamer (a fairly Chomsky-critical generativist), he noted that theoretical linguistics as a whole is still in the "Dark Ages" and that existing models will be probably be discarded at some point, but the basic principle that generativism is simply about making formal, predictive rules is what's important. We just have to wade through a lot of bullcrap to actually get there. I would call myself a generativist on these principles but I am also quite critical of a lot of popular theory, for example I don't really buy into "covert movement" or that coordinated structures are entirely binary branchings. Again, you did a great job on this video and I think you actually captured the popular debate about UG really well, it's just in my experience that the parts of UG most of the popular debate centers on are actually either a few decades out of date or assume that Chomsky speaks for all generativists, let alone UG supporters. **edit for some typos
Considering the topic and the title of the video, this could very easily degrade into a flame war. That said, I disagree with you on Universal Grammar (UG), but of course the comment section is not the place to have a discussion. Because of that, I'd like to mention the following, not as a rebuttal to the arguments exposed, but as an alternative explanation to anybody who might be interested. To anybody reading, even if you disagree, I hope you'll at least get an idea as to why a heck of a lot of linguists at least consider UG OR use X-bar theory or some sort of tree analysis in their papers (which aren't the same thing at all). Also, please consider the origin of these theories and how they compared both to other existing theories at the time and to other possible theories. A very compelling feature of the current syntactic theories within UG is that, although the theory is complex, it should, in theory, explain all languages, and not just one or two. I feel like that level of analysis is a lot more compelling than having to learn whatever arbitrary categorizations old men have come up with to explain some rule in a language that are typical of "traditional" grammars, but more on that down below. Anyway, the first thing to note is that UG is (to simplify) a theory stating that all languages in the world share certain similarities. It originates in part as a critique of the American behavioural psychologist Skinner's view of language. According to Skinner, language was just behavioured, and it was the result of operant conditioning and such. This theory made some predictions which just didn't add up (if it's just conditioning, can we teach animals language? See "Why Koko (probably) couldn't talk" on youtube for a GREAT explanation on that topic, including some generativist theory). UG originated from some observations about language acquisition and syntax, but the idea behind is that "language" is a human phenomenon, and has such, it has features that are universal, regardless of which language is being considered. Discrediting the theories of syntax that proponents of UG use would not disprove UG, although the new hypothesis/theory that it's probabilistic learning which enables language acquisition might. As an aside, although many people associate UG with Chomsky, and he's been a very influential linguist, many people dispute how central he has been to the "cognitive revolution" in linguistics (Pollum has an essay on that, although I don't know what to think of it). X-bar theory and successive tree-analysis systems for syntax within the generative tradition exist as something between a framework and a hypothesis. They have been modified considerably over time, back from a simple rule-based grammar that attempted to describe grammar through something like a logical system (something that a computer could read) up to "Government and Binding" and nowadays "Bare-Phrase structure". If you google any of these, you'll find that they are associated with UG, of course. However, calling "my dog" a DP (determiner phrase) doesn't necessarily mean adopting all of UG's assumptions. In the past, people used "subordinating conjunction" or "adverb" as labels, and that didn't mean they subscribed to a particular theory of language. Again, although some other theories of language have their own terminology, I find that a lot of people just use these terms because it's a good way to communicate, and a lot of the time the terms are more accurate and meaningful than arbitrary groups like "subordinating conjunction". When I learned linguistics in college, that was one of the first things we had to learn, and I hated it with a passion! Consider that people have been writing about grammar and languages, if not linguistics proper, for at least two millennia already. More modern definitions of "conjunctions", such as UG's "complementizer", are at the very least better justified and explained. I can use a "syntactic test" to determine whether a word is a pronoun or not by using it in different sentences (some grammatical, some ungrammatical), and that's something that many linguistic frameworks just don't give a lot of importance to, so I don't think it's unwarranted that many people would use UG terminology in papers that might be more interested in just describing the grammar of a language. Anyway, UG and UG's theories of syntax are not exactly the same. But let's consider they were, and address the problem of "updating" the theory instead of discarding it. Is that scientific? I'm not very acquainted with T. Kuhn's ideas about what science is or should be, but I vaguely remember that it works a bit like Piaget's ideas about assimilation and accommodation: you come up with an explanation about the world, you amend it in light of new evidence, and when the theory is beyond repair, you come up with a new one, with a new framework and a different set of assumptions, a "scientific revolution". I think it's safe to say that most people agree that's what happened between Skinner's idea of language as behaviour and the "cognitive revolution" was a scientific revolution. Some might even argue that there has been another revolution within generativists with the Minimalist Program, as it introduced the notion that there might be some principles to the computation of syntax that might be at the core of UG. There's a lot more to say about the history of UG and of syntax, but I think the former is a good starting point. Although that selection of information kinda argued against Watch your Language's view, I tried to keep it just informational. I'd like to add my opinion on a few things: -Yes, generativist linguists were and are mostly European. That doesn't disqualify them more than it would disqualify de Saussure to talk about languages. I think the syntactic theory proposed by generativists is just a starting point to be updated by more information gathered from other languages. We could take these European ideas and (a) offer them as a framework for linguists of different languages to use, hopefully contributing to a global database that might be used to study "language" and not just "a language" or (b) toss it in the trash at the first sign of problems and just stick to learning seemingly inexplicable rules for any given language. Other frameworks, particularly comparative linguists, have attempted to do something similar, but their ideas are much less ambitious in scope (pointing similarities is different from trying to explain the emergence of syntax and syntactic rules!) -Evidence in favour of UG back in the day created a furor around the enterprise. However, more research is needed. Evidence against UG should be followed up on as well. Although there are some central arguments in favour of UG, such as "Plato's problem" or the observation that almost every human child will learn language, there is no single argument on which UG relies on, so just disproving the existance of recursion wouldn't be enough, although it would require a major redesign of the theory, if not a scientific revolution. As regards the evidence for that and other things, many things have been published that just didn't hold up. Many studies on teaching languages to animals seemed promising at the time, but in the end they all ended up proving Chomsky right. On that topic, he said that if an actual fully-fledged language was possible for animals, we wouldn't need to teach it, it would simply happen, just as language simply emerges between people even when there is no common language (or no language at all, see Soup Emporium's video on Hellen Keller). If recursion was indeed a feature absent from some languages, we would have a bit more evidence than just two people's potentially biased account of one single language.I think time will tell. -I do believe that UG is a bit at odds with scientific thought, and that might be okay. For one, science is not a perfect institution, and the study of scientific revolutions can only get us so far. If I've understood Kuhn's ideas correctly, scientific revolutions should override prior theories, but in many fields of research we have many different "paradigms" that don't quite rule out each other (see psychology for example). -I think that UG being questionable is a good thing. Other grammatical descriptions of languages are just that: someone's description of a language. There is no science to speak of, just a description of rules and features. UG is more ambitious, yes, and will very often make questionable claims, but at the very least it has set up a framework to test out these ideas (grammaticality judgements mainly). Of course, some ideas that originated among generativists are no longer exclusive to them (again, grammaticality judgements), and now many linguists try to write down their ideas in a more critical, less arbitrary fashion, trying to provide tests to determine whether some is grammatical or not. If anything, linguists who have tried to study languages in a more standardized way (among them, generativists) have made a positive impact on the discipline in that regard. -Generativists' accounts of language, particularly syntax, have potential applications in computer programming, regardless of how scientific they are. -Many concepts within UG are interesting by themselves, regardless of what the fate of the rest of the field. For example, the idea that syntax is not linear, a fact proved by, for example, garden-path sentences, and a very important aspect of UG and its theories of syntax. It's very interesting to consider silent elements such as tense, or "features" that "move", or null elements, like some complementizers. I think these are great alternatives to other theories of syntax that seem simpler at first glance, but are actually either inconsistent, or riddled with exceptions, inaccuracies and a lot of unnecessary rules.
I remember being fascinated by Esperanto and learning it when i was 12. That was until one day i realised that in order to to that i'm using English (which is not my native language) which already is, you know, the de facto world language now.
I have some big problems with a universal language. How long would it be before several major groups diverge and become incomprehensible to each other? How could we provide education to everyone, because I feel that privileged countries can afford to teach their people and join in while poor countries (or uneducated people in general) would be left out. How would a universal language be implemented? Even if it catches on in one region, others might be uninterested. For example in most of South and Central America, Spanish is already a lingua franca and the people there might see no need to learn any other language besides those directly related to their interests, probably English or Portuguese I like your videos :)
"Major groups diverging" You can design a language to avoid common phonological changes. For example, Toki Pona phonotactics forbids "VNNV" where two nasals are next to each other, which often turns into gemination. "How can we provide education to everyone" Linguistics can't solve global disparity and neo-colonialism. But simple auxlangs are _much_ faster to learn than natural ones. Esperanto learning is 4x faster than French for some speakers (and that's the badly-designed auxlang). If you need to learn a language to find work - you'd benefit from having a faster one. "People with regional lingua-francas would be disinterested" I personally think that auxlangs would only work by starting in areas with no regional language, then spreading out. Esperanto itself started in such an area for that exact reason. As it grows in speakers, it becomes a greater benefit for other speakers with lingua-francas (on top of speed, and not being the language of your colonizer). Even if it doesn't grow, it's still created a cultural network between multiple regions with no lingua franca, which is valuable in itself. _EDIT: I don't believe that an international language will catch on, but it's an interesting idea_
You definitely can make a conlang big enough to express any concept you could ever want. It might take a while, but it's doable. The problem is doing that, while also making it universally easy to learn.
if I’m not mistaken, there’s increasing evidence that most pastoral “Altaic” languages might have started out as early Neolithic farmers that actually switched to pastoralism, which could support the idea of a common geographical homeland tying them to Korean and Japanese which would have remained agricultural. Ultimately I have no idea if this lends more support to a sprachbund or not.
>says he won’t include fringe theories >first one is altaic I jest of course, I know it still has a significant following in East Asia. I don’t think it’s as big as these other controversies, but the IE Urheimat debacle has always interested me. I’m a Kurganer myself (LIKE YOU ALL SHOULD BE!) but the Anatolian hypothesis has picked up some steam again recently due to some nasty geneticists misconstruding some data!
I find your final argument against conlangs as lingua francas a little weird, because, like... if the problem is the lack of lexicon, just create more? Even natlangs create new words all the time, it's a normal thing for any language. Sure, for conlangs, everything is created, but if someone is dedicated enough, a large enough lexicon could be created and the gaps can then often be filled with more creation from scratch or just using pre-existing roots like natlangs do. I totally agree that toki pona isn't great as a lingua franca because there are too few words, but that's a problem of toki pona, not necessarily of conlangs as a concept. (Not saying I necessarily support conlangs as lingua francas, I just found that particular argument a little strange)
Yeah - for example in Esperanto, Zamenhof would coin new words and add them to his regular newsletter for Esperanto speakers. So he could grow the vocabulary really naturally.
Most of linguistic debates are because if you have a definition "definition1" that describes a thing "thing1" having properties A, B, C, D and you find in other languages phenomenon or thing "thing2" that has properties A, B, X, Y, Z, instead of the simple approach of putting a new a definition for a thing that has properties A, B and say thing1 and thing2 are examples of it, you try to describe thing2 with definition1 + lots of modifications and exceptions which is a more needlessly complicated way to go about it. That's why I like construction grammars the most they don't assume universals which are mostly language specific things squeezed to fit as an underlying principle for all human languages. describe everything in its own terms and you will see many things are simpler than expected. less assumptions is better.
Also, I was thinking about cases, and I feel like I almost think they don't even exist? Or maybe they always exist? Aren't cases just specific instances of giving information about a noun's role in the sentence? Maybe, not sure :D
First video I've watched in this series. I loved learning years ago that Turkish and Korean were related. But now I understand why speakers of those languages were always surprised to learn this contested fact.
6:58 Regarding the "Germanic possessive s" in English, is it true to say that the word "of" expresses possession, as suggested? Or at any rate that it expresses the 'direction' of possession i.e. possessor and possessed, which the possessive 's' definitely does. Personally, "A friend of my mum" and "A friend of my mum's" have slightly different meanings. The former could indicate a mutual friend, or it could be an unrequited friendship: someone who likes my mum but she secretly doesn't really like them back, perhaps just tolerates them. But the latter form with the possessive 's' indicates that the direction of possession is that "my mum" is the one who possesses the friendship, which in the case of friendship means she is the one who definitely likes them and thinks of them as a friend, whether or not they happen to return the feeling. That's all a bit metaphorical in the case of friendship, and perhaps other people might not feel the same distinction as I do. But I think it's perhaps easier to see in another example. I would contend that: "A shield of Brian" would be generally understood to refer to something or someone that protects Brian, while: "A shield of Brian's" would conjure the image of Brian in mediaeval garb ready for his historical re-enactment society's performance of the Battle of Towton. So the "of" alone in that first sentence doesn't indicate the same possession that the possessive 's' indicates in the second, even if you think they might both be seen as some form of possession (whether physical or conceptual). Off the top of my head, is the "of" in "A shield of Brian" more akin to the "of" in "A pie of cherries" perhaps? Even if you say the pie 'possesses' a nature of cherryness, it's definitely not the cherries that own the pie in the same way that Brian owns his shield in "A shield of Brian's".
Can we talk about optimality theory? I've only taken one course on it last semester, and while I think it has many merits, part of my mind is always screaming that it's kinda just unfalsifiable and that you could model any process that way. I never got to speak any of those concerns and the storm has just been raging internally lol
Nice to have the pro-Altaic and Anti-Altaic arguments summarized. I did not know about the way Sprachbunds impact word order! And I find the sprachbund hypothesis quite believable. Given the Turkic urheimat is the one that's actually around the Altai mountains, it requires a sprachbund that only stretches from there to central Manchuria to include Tungusic- not very far by steppe nomad standards. I am impressed by how YOUNG that makes the constituent families though, they cannot be traced further back than the first few centuries BC. Where did they come from? The eastern steppes are weird. As for the others I had no idea those arguments about morphology and universal grammar were even were a thing. Thanks for this look down the rabbit hole! Hot take regarding interlangs: Esparanto is a French-Polish pidging and Toki Pona is too restrictive (having to simplify your thoughts with it is NOT praiseworthy!). Given the current status quo if you gave English some spelling reform (and/or reversed the great vowel shift) you'd me most of the way to a working AuxLang. Fairly analytical so a beginner can get their ideas across without thinking much about grammer, plenty of flexibility for more nuanced communication, no need to rebuild centuries of linguistic baggage from scratch like an a priori clon. English language and culture for better or worse is already a prestige language of the world, might as well work with it.
One argument against an Altaic family is this : The farther back you go in analyzing. samples of the Turkic, Mongolic and Tungusic families. , the less similar they are . This would not be the case if they stemmed from one proto language .
The biggest problem with Altaic is that the family’s second largest branch, Mongolic, can only be reconstructed to the time of Genghis Khan. Thankfully, Khitan is opening a window into prehistoric Mongolian but even by 1000 AD, Mongolic and Turkic peoples have been interacting for millennia, meaning that both language families share a lot of vocabulary.
Although our host exiled all fringe theories at the start of his video, I'd like to ask in this forum: Does anyone know of an Oklahoma-based professor from about the 1980s or Nineties who theorized the human race originated in the Americas, not Africa, and his evidence was a series of language families that classed English with Chinese, French with Arabic (if memory serves), etc -- all so he could argue the American Indian nations have always lived in their historical homelands and were not migrants via Beringia. I didn't xerox his paper then, and I can't remember his name now. His beliefs were so bizarre, I find it kind of fun.
I don't know Arabic, but as far as I understand states of a word are supposed to be a grammatical quality/property/whatever which occurs together with cases, gender and number (if any). So, while Syriac nouns have 3 states (or 4: 2 being identical) formed with endings, but no cases at all (what makes their states to be cases in practice, so the name "state" is employed for historical reasons, I guess)... I think Arabic and/or other semitic languages do have both states and cases appearing simultaneously in a word. Thus, you would find a feminine plural word in nominative and construct, or in dative and construct, etc... The other states, with different names in every language's grammatical tradition, usually are supposed to tell wether the word is definite or indefinite. In Arabic these states, if I'm not mistaken, are often also alternatively explained as presence or absence of a definite article, since they are marked by adding or lacking the prefix -ال. In my point of view, cases and states are basically the same thing in function, but if languages have both of them you rather make the difference since you can combine each from each category between them, in the same way you don't call nominative singular and plural different cases; otherwise you would have not a single construct case for these languages, but a subject-possesor case, object-possesor case, etc.
I don't understand why people have double standards about global lingua franca. People often list flaws of conlangs as lingua franca while somehow accepting those exact same flaws for the current global lingua franca, English. 1. "Esperanto is too Eurocentric", but the current lingua franca is English which is worse than just being Eurocentric, it's Anglocentric. My native language is Indonesian, so both English and Esperanto are equally foreign to me (and to most people globally). 2. "Esperanto has to many phonemes, with 22-23 consonants and 11 vowels", meanwhile English has 24 consonants and 20-25 vowels in RP, 14-16 vowels in General American. Not to mention, English phonemes don't match one to one with their graphemes. Learning to pronounce English words as a kid was a nightmare. Not only I had to learn weird sounds, but I also had to learn weird spelling. 3. "Esperanto grammar is not that easy", I don't think English grammar is easy either. At least Esperanto grammar is more predictable than English. Sure, Zamenhof had a bias because of the languages he speaks, but most people don't have any bias towards either English or Esperanto. If I had a choice on which language I had to learn as a kid between English and Esperanto, I would've chosen Esperanto. 4. "Natural languages have richer vocabulary than conlangs". I think this is a weakness of natural languages, not a strength. The purpose of a global lingua franca is for international communication. Most people around the world (including me) learned English not because we're interested in Anglophonic culture or literature or poetry. We just want to be able to speak with people from other countries (mostly for work related purposes). But somehow, we were forced to learn all these BS literary and bougie words in order to pass English exams. Global lingua franca does not need vocabulary that rich, it only needs enough words to have basic conversations, business conversations, and scientific conversations. If you're interested in literature or poetry, you can just learn a language whose culture you're genuinely interested in, not a global lingua franca. We could've spent minimal effort learning a lingua franca and put more effort learning a language we're genuinely interested in if we were forced to learn an easier lingua franca than English. A conlang does not have to be perfect to be a lingua franca, it just has to be better than the current one (English).
I strongly disagree with the sentence "Conlangs can never have the sheer amount of vocabulary that makes natural languages operational". Conlangs have proven to be operational. That's not even a question anymore. As you said. There's even native speakers Esperanto. Toki Pona is the extreme example. Other full fledged conlangs actually in use typically have much larger vocabulary list to utilize. The key is context. As long as you know the topic, what someone is talking about, and understand what they are saying about that topic. You understand what they are saying. You're communicating. Context is extremely important in natural languages too, just slightly more in conlangs. With conlangs able to use context more directly as a tool to drastically reduce the vocabulary needed to be taught to someone. Even though Toki Pona isn't the best at more complex topics. It's still what I'd consider usable for more complex topics, you can still build up the context. Additionally I'd say real life communication is actually where Toki Pona excels as an interlang. You can easily get by 99% of your day to day life just fine with Toki Pona with minimal effort. It doesn't mean you need 6 different words to describe a Banana. If you don't just want to describe it as a fruit. You can use 2. kili jelo. Or Fruit that is Yellow. If there happens to be a lemon near it. Maybe throw in a 3rd, or point. You can do a hell of a lot with 123 words. And for that reason I think it's a great interlang. Especially due to the barrier of entry being so low. However. I wouldn't want it to be the language of education, math, science, etc. However. Conlangs still have significant advantages inside those fields. You can design a language to be more regular and predictable. Natural languages have ALL the issues you typically critique certain interlangs for. Too many difficult to pronounce consonants, too Euro centric. Isn't regular enough, etc. Ignoring the reason to compare and nick pick the variety of interlangs there are to choose from. They certainly beat out natural languages dramatically. Especially from the standpoint of having a language relatively easy for any person in the world to learn despite their native language. You don't NEED gendered words. You don't NEED tones. You don't NEED 10 direct synonyms for the same word. You don't NEED formal and informal speech. Sure. You can dislike a conlang based on how artificial or non-elegant something sounds. But most conlangs don't really have that issue.
About point 2, Portuguese uses elisions and sandhi for our most common prepositions that mark the direct object like "a/a" for direct object, "pro/pra" which is a fusion of per + our definite article for indirect object, "de/da" for our genitives, etc.. So, if our preposition's elisions could one day be reanalyzed as prefixes, as they can't occur on their own unlike the words they were based of, wouldn't "d'" be a genitive case? "pr'" our dative case? "n'" our locative case? Given the fact that despite our word order being SVO, Portuguese can be somewhat flexible with word order(Many of our phrases are in SOV like "Eu te amo"), but we need the prepositions to mark the object of the sentence. If case markings don't necessarily need to be suffixes, our elisions of our most used prepositions could easily be reanalyzed as cases, and Portuguese would have at least 6 or 7 of them. Also, if we consider they fuse with the articles instead of the object nouns when the object doesn't start with a vowel, these could be thought as them inflecting for number and gender since the prepositions that are most ellided are also the ones we commonly fuse with our articles, and our articles are literally just vowels or a vowel+nasal and s which on normal speech, is fused into the following word through sandhi. They could be easily thought as just gender/number inflections for the "case markings" instead of the articles they are.
I count myself an amateur linguist but am only interested in diachronic linguistics, that is explaining origins and describing changes. In most cases these languages are unattested and cannot be analyzed in the way modern ones can for lack of any speakers. But that is the beauty of this field for me. 😅
The question of what is or isn't a case has been on my mind again. Years ago I was learning Turkish. I learnt its cases of course. Only, the word "case" was never uses in my lessons, this despite the mutual language of me and my tutor also having similar cases! I learnt that the suffix -de means the word is where something is/happens, -le means with the thing, -e means out of/away from the thing, etc. Turkish cases work similar to Hungarian, and I wonder: If the languages did not have vowel harmony, would we still identify them as cases? Looking back, those case endings matched more with prepositions in my head with how I learnt them. So I wonder, how closely can an adposition "glue" to a word before it's a case? I pulled this conlang my assfrom, this conlangin all adpositions Englishfrom moved the othersideto the wordof they modify and no longer have a space. Does this monstrosity have cases now? A dedicated vid on the subject would be cool!
Utter amateur here. I think it would make the most sense to take the seemingly ridiculous route, and actually _do_ consider prepostions cases, reworking the whole system in the process. Cases would be then simply split into the standard categories: analytic (adpositional, prepositional) agglutinative (suffixed, infixed), synthetic and mixed. The borders between them would then be casted by already existing linguistic definitons. Examples: Into the house - analytic A friend of mine - analytic, synthetic Poszczuć psem (ins.) - synthetic Iść z psem - analytic, synthetic Während der Kreuzigung - analytic, synthetic Kreuzigung des Verbrechers - synthetic Etc. (Entschuldigen Sie bitte mein deutsch. Wenn ich einen Fehler irgendwo gemacht habe, korrigieren sie mich bitte)
Just a question since Altaic is formed partly by genetics. Does altaic make sense without Turkic? I’ve asked some of my friends who know linguistic and he say grammatically Turkic is the odd ones out. From learning about genetics myself, Turkic seems to have two possible roots which are Altaic and non Altaic, thus if we cut Turkic out maybe it can make more sense
Alright, I'm totally nitpicking here, but I am studying linguistics so this video has already made me wanna fight something. Hi, I am a native Hungarian speaker! "A követelményeitek hülyék" is not what I think should be there. It means that the requirements (those being living, breathing things) are stupid. If I am correct, it should have said that the requirements (unalive, impersonal things) are stupid, which should be root hülye + "ság/ség" (vowel harmony!) + pronunciation aiding vowel (courtesy of vowel harmony) + "k" to mark the plurality that ties back to the requirements being several. So the whole sentence should be "A követelményeitek hülyeségek". Or I might be in the wrong. Somebody correct me if I am! Thanks!
This is interesting, that your native intuition is telling you that Hungarian adjectives work differently based on animacy. Is this just for this adjective, or is it a Hungarian grammar rule I should know? Also, isn’t -ség a noun suffix? How exactly does that work in this case? Is it saying literally that “Your specifications are stupidity?” (I’m not challenging, just asking, you know more than me)
@@watchyourlanguage3870 Since today is a national holiday in Hungary, everybody's home and doing nothing, so I did a miniature empirical study on the topic. My father got confused by my question, but he also thinks in Hungarian/English in a sort of mutually interchangeable way. My mother, by profession a grammar and literature teacher, told me the sentence is ungrammatical, since a "követelmény" will always be an inanimate list of requirements. I still wouldn't be surprised to hear a small child or a drunk person say such a sentence, but I will not mix in spontaneous native speech, because we don't have time for all that. As for why it didn't even occur to me that " A követelményeitek hülyék" is ungrammatical.... I am a polyglot, my mother is not. I have the tendency to say sentences with Sanskrit grammar, then get angry with myself when I can't untangle the mess I've made and that makes my ears bleed. You should see the emails that get exchanged between me and my teachers. Although we use English in lectures and seminars, emails are usually in Hungarian, and the struggle is real from both parties. As for -ság/-ség being a noun suffix. It is, specifically one of the largest groups of suffixes: an affix. In Hungarian we have a strictly fixed order in which suffixes can follow a root, and affixes come right after the root without exeptions. The general rule for ság/ség we learn in school is as follows: noun/adjective root + ság/ség = abstract adjective. For example "barátság" - friendship. "Your specifications are stupidity" is kind of on point. Maybe put stupidity in plural. When I was learning Spanish and English in highschool, I suffered a lot because I could not yet switch my brain to think in the target languages and I struggled to express exactly what I wanted to say. Most other languages feel restricted to me in a way. Now I have finally learnt to let go of this. Regardless, Hungarian has 18 nouncases or so, and if I think about the number of ways I can alter a simple noun with just suffixes, I stop around 25 but I could go on. The closest I got to think in a natural way is with Turkish, one of the languages I study at uni. Lithuanian is a close second. I think I might make Sapir and Whorf rotate sleeplessly in their graves. All the above madness surrounding linguistic relativity is also the reason why I can't finish my BA thesis on Ted Chiang's Story of Your Life. All in all, Hungarian adjectives depend on the subject of the sentence. "A követelményed hülyeség", but " a pingvin hülye" if you are in an argument with your professor on his final exam requirement, and you two ran into each other at the zoo, near the penguin enclosure, with a particularly dumb penguin in it. As for suffixes, we have a gazillion for nouns, adjectives, verbs, and so on. I think Wikipedia has a nice little chart under "Szóképzés" in Hungarian. Lastly, I think somewhere along the line somebody mistranslated some linguistic terms from Hungarian to English, hence half the confusion for foreign linguists. But this is yet to be proved.
Altaic hypothesis is interesting, but there is no hard evidence to prove or disprove it. One of the reasons we could establish PIE is because we have written evidence pretty close to the split. We have Greek, Latin and Sanskrit texts (not just inscriptions, but actual texts) from time that is roughly similar to English-German-Swedish split. Where the languages already lost mutual intelligibility and went their separate ways, but there is little doubt that their similarities are not a coincidence. And we have relatively "pure" languages, that don't have a lot of outside influence. Latin is Latin. It doesn't randomly have 70% of X vocabulary that replaced all the native words except for the most basic concepst and things. Few Greek loans? Yes. But just a few. With "Altaic" languages we don't have that. All the evidence is relatively new so you can neither properly establish what is and isn't a "core" word nor can you see words at a relatively close point in their history. It is as productive as comparing Spanish directly to Swedish. There clearly are a lot of same-ish words and some grammar overlaps, but you can neither find exact cognates (because a lot of them were simply lost or mutated beyond recognition) nor establish good sound correspondances beyond "there is L in leche and mjolk" A problem with "Universal Grammar" is the fact that these restrictions are of computation coplexity origin, not of "innate grammar" one. You will not find a single language that doesn't keep its articles as a part of a noun phrase, simply because it takes too much brain power to do so. Even with additional measures, like grammatical gender. "iThe bMan lives in iHouse" is a chore to parse through even at such a small scale and where all the nouns quite clearly belong to different genders. And when you pack everything together there is no real reason to encode certain information in a specific way. "in house" is as easy to parse as "house-in". Just as there is no real reason to select "green house" over "house green" (unless we are talking about more complex interactions).
Do you have a list somewhere of all the languages that you have learned/are learning, and your level of proficiency in each? Because every video, it seems like you speak three more than I knew before. Based on your language overview videos, your Germanic correspondences video, and this one, I know that you speak English (duh), Dutch, Danish, Swedish, Icelandic, and Yiddish (but not German); MSA, Levantine Arabic, French, Spanish, and Hebrew, that you're currently learning Hungarian, and that you don't speak, but plan to learn Turkish, Korean, and maybe Japanese. Please let me know what other ones are missing.
Do you think a language made so that any Germanic speaker could communicate with it would work? Some would argue English already serves this purpose, but we have so many loanwords from Latin and French that English wouldn’t be able to accurately represent all Germanic languages.
That exact fact that English borrows so much from Romance languages actually does hinder Germanic auxlangs. They’ve tried, Jan Misali reviewed the conlang known as Folkspraak th-cam.com/video/KegIeZwXUDk/w-d-xo.html , which does exactly this, but it seems to me like a version of Anglish (even for other Germanic language speakers)
About the construct; I think you can say that it's not a case, beacuse Arabic (for example) has an entirely separate case system, which also extends to modifiers and such. So a noun can be in the construct state, and itself can be accusative. Calling the construct a case doesn't make sense unless you allow nouns to stack cases.
Currently since it’s my last semester I’m just in Applications of Linguistics, it’s my capstone and it’s basically just “remember everything from earlier classes? here’s how to use it irl” I’m enjoying it. Thanks for the words of praise!
That was so interesting!!! Great video! However, I have to disagree about conlangs as international languages, similarly to what @Mercure250 already said. I feel like you are employing the naturalistic fallacy (since were talking about logical fallacies) here, there is nothing that says a conlang couldn't be as good, or better, with as good, or better, lexicon, and as good, or better, phonology, it's just that none so far have achieved that. I agree though on Esperanto, even though I speak it and love it as a language, no way it is good for truly global communication, as like you said, it is very eurocentric.
My opinion on cases is that if it's an affix it's a case, if it's an adposition it's not (but you can still use case terminology like "accusative" or "genitive" to describe it if you want, because it's practical), and if it's a clitic uhhh, I dunno... Also I feel like you didn't really present any good argument in favor of natural languages being better IALs than conlangs and I do remain absolutely convinced that conlangs are a better alternative: Lexicons always have to be continually enriched by usage anyway so I don't see why that would be a specific problem for conlangs as opposed to like Pirahã or Sanskrit which also lack words for a lot of concepts that we use daily, and I find criticizing conlangs for being ethnocentric while advocating for natlangs to be quite self-contradictory. The main problem for conlangs is probably that there'll always be some issues with them so it's hard to settle down on one conlang as an IAL instead of continually trying to make better ones. For instance I think esperanto is a great candidate but if I had it my way I know I would absolutely change its official grammar to implement the "-iĉo" suffix as well as a "hi" masculine 3PS pronoun, a distinct 2PS pronoun, replace all instances of /ʒ/ by either /dʒ/ or /j/ and of /x/ by either /k/ or /h/, rework the orthography to get rid of the diacritics, and maybe simplify the use of the accusative marker a bit as well. By then it would probably be a pretty different language and I guess this is why new esperantidos kept popping up and nowadays new unrelated auxlangs do.
But in Arabic, the construct state itself can have different cases, for example, the word queen in Arabic: malikatu (nominative construct) malikata (accusative construct) malikati (genitive construct) Why is it a case if the construct state can be in different cases?
Here's my takes on this: [1] - Not much to say. I'm sceptical of any new language families, and so I'd incline to be an Anti-Altaicist. I'm not a linguist however. [2] - I agree with the end part - a better definition of case is needed. [3] - Mostly agree. Conlangs aren't great as possible IALs. Latin, and English, have both served as de facto IALs in the past. Toki Pona speakers, from personal experience, also really wouldn't like their language becoming an IAL. (you only need to spend 3 minutes on the largest discord server for toki pona to get a free lobotomy) [4] - I think Occam's Razor is enough to cut down UG at this point because of how complicated it is. It killed geocentrism, it'll probably kill string theory, it's gonna kill UG. It's sad because Chomsky is a guy that heavily influenced my political views, which are still 85% Chomsky, even today.
dbi like "2 cases", there are only 1 because nominative is unmarked and the other is the simplest case known to man, even languages without case inflection have accusative marking in one way or another like just dont beg it please its one case
@@watchyourlanguage3870 ancient Semitic speakers forming a substratum for Celtic languages in the Bri'ish Isles is not the theory I was ready for today. Is there any other theory besides Venemaan's on this? also, I'd watched this video when it came out, but have since forgotten completely, my mind is like a sieve xd
Thankful that my department is decidedly on the side of sociolinguistics so I've never had to suffer the way you have with UG! Re lingua franca, I don't believe in any lingua franca. The eurocentrism of synthetic languages is not overcome with the status quo where we leave English as the lingua franca. If using translators everywhere is cumbersome, it is worth it to me in order to preserve the equality of languages. In fact, as a matter of state policy, I don't believe a government should have an official language and should be responsive to the languages spoken by communities (in education, public health campaigns, etc.). Inclusion in the public sphere should not come at the expense of your language and culture.
I was never impressed by Chomsky in ANY intellectual endeavour, since he gives every evidence of being dishonest (e.g., claiming he never said things that he said and altering republications of his texts to cover it up), and in general seems to be a pompous windbag. I too was told that UG was a fantastically important achievement in linguistics --- until I read his statement that it was unnecessary to be familiar with actual languages to think it up. When he proclaimed, on the basis of pathetically uninformed misunderstanding of genetics, that language didn't exist until a few tens of thousands of years ago and first appeared in Europe ----- failing to explain how, for example, Australian aborigines who had been isolated from the main body of humans for a much longer period ended up having language like everyone else, and somehow managed to navigate in groups across a treacherous 300km expanse of ocean without possessing language ----- well, at that point I threw in the towel.
@@thevilmoron It's f*cked that he denied the Cambodian genocide. As far as I know, his other politics have been criticizing US interventions and coups. For example in my country, our first president was assassinated and replaced by a US-friendly dictator, and the US govt admits to it. Yet most US nationals aren't aware of even 5% of these interventions. And when learning about it, they'll even try to justify it. Even if they criticize distant-past colonialism, they claim that "It's so sad there's corruption in developing nations" without questioning where this corruption comes from. So you can criticize Chomsky on the points where he's very wrong (including linguistics), without throwing out the very accurate ones.
Don't Hungarian case endings have to be bound morphemes? Since they can't be written as prepositions, they can't really be read as prepositions. Simples.
I think your opinion of natural languages being better lingua franca's can potentially be harmful to smaller cultures and has historically killed off languages with fewer speakers. Real-world examples of this include Tagalog in the Philippines, Russian in Eastern Europe and the Caucasus and English... everywhere. Meanwhile, with auxlangs, everyone has to learn a new language, and simple enough that people would prefer to use their native language in contexts where they can. I believe auxlangs albeit unlikely to work on a major scale, have their place
Serious question: Isn't descriptivism paradoxically a Form of prescriptivism? Isn't judging "rules" for language as bad inherently a judgment/rule itself? Hypocrisy.
@@winnebagotrout1997 There's definitely some people who say that. I know someone who thinks standard languages is imperialism and purposely talk and spell things wrong to "combat" it
My quippy take on the Anglo universal grammar tradition is that it's wrong, but not as wrong as you'd think. Like, there's pretty clearly some sort of built-in neurological basis to human language acquisition, and this should constrain the sorts of languages you can get. But the actual grammatical operations they come up with can't possibly be what the brain is actually "computing", any more than the tables of forms you find in classical grammars exist in a native speaker's head. Squishy neural networks simply don't operate like that. They're mistaking the map for the territory.
As a turk I sadly have to say that the ural-altaic language family is taught in our schools as factual information and it has become a general fact in the public, so much so that when I mention that the theory is actually discredited in the general linguistics circles I get a lot of side eyes, even from the teachers.
I agree with this as a turk, i also want to add that many people in turkey still believe hungarians and turks as brothers because of how some teachers taught students about huns in history class, and uralic-altaic language family theory seems so cool to them. As a fun fact, also in history class, it's still taught that scythians are turks because just they're nomads live in central asia, and it was surprising for me to learn that they're iranic nation and there are lots of evidences about it.
istg i'm under a week away from my Everett/Chomsky video and then this comes out. great video!!
Thanks! And… whoops sry, I look forward to watching that video!
as a slavic language native speaker, i like that the interslavic exist, it may be a nice hobby to learn to speak it someday.
people le-learning latin is also cool
I feel the same about Lingua Franca Nova as a native Romance speaker, really easy both to understand and learn if you speak a Romance language (and the former doesn’t require the latter).
They say speaking Interslavic is like being able to speak in tongues to other Slavic people
I'm from bosnia and when I first heard it I thought one of the creators was from "our" countries (croatia/ bosnia/serbia) and thus put an outsized influence on interslavic from our language
@@redhidinghood9337 i didnt know that. wasn't he "just" from poland? djesi brt
Altaic!
Edit: before even reaching the Altaic section of the video, I didn't expect you to actually mention it
"If UG is God, I'm agnostic,"
Amen, brother, me too.
I fully agree with you regarding Chomsky and natural language, but the importance of his research for formal languages and the development of computer science should still not be underestimated.
Great video. I actually have some opinions on these issues
1. While I don't currently believe that the Altaic family exists, that's really due to a lack of evidence more than anything. As more information is found and more comparative studies are done accounting for all the niche languages in the family, it could well be that the Altaic hypothesis is proven
2. The case/adposition debate is mildly annoying but ultimately not worth fighting about imo. I've seen multiple analyses that treat Tagalog particles as "case markers", even though they come before the entire noun phrase and are literally separate words lmao. Ultimately what's more important is that the analysis allows us to understand the inner workings of the lang, no need to argue about the actual terminology
3. The selection of a truly global lingua franca is probably more fantasy than reality, but if such a language was going to chosen, I find it pretty clear that the best choice would be Indonesian, or some creole of indonesian with a large amount of loans from other languages. BI fits all the requirements (limited phonemic inventory, lack of consonant clusters, isolating structure, lack of case and gender marking) and is right now being used as a second-language lingua franca by like 100 million people
4. I mostly align with whatever you said about UG, GG and their derivatives. It's probably way above my pay grade to argue about this, but I find variations of Functional Grammar to be much closer to the "truth" than those on the UG side. Specifically Systemic Functional Grammar is the one I've read most deeply into and I find very little to criticise about it
I love Esperanto so much but I don't think it's a good interlang. Instead, I love it for the culture that's built up around it-it's a fun community for people who love languages and share slightly related mindsets towards peace and such. Esperanto is worth it for the culture, not for international communication
Tute samopinie, kara!
Any thoughts on "a language is a dialect with an army and navy"?
It’s a very good explanation for what people casually consider languages, is what I think about it
you talk at a speed that my brain loves
First I want to say, great video overall! However, as the title predicted, I do have some issues with the UG section-I think they're very valid complaints....about an outdated model/understanding of generativism. Please don't read this as an "umm ahckshually" bit, but simply as additional information from someone who went really deep into syntax and about to start a PhD on it:
One point is that modern generative grammar is a more broad school than just orthodox Chomskyism, and while a basic framework is rooted in his hypotheses there is a very strong contingent of scholars within generativism who take most of his proclamations with a metric ton of salt. Most generativists I know do not subscribe to an orthodox view of UG as decreed by Chomsky but see the goal of generative linguistics as trying to formulate predictive models which can be tested against natural language in order to discern the true nature of UG, i.e. whatever the actual structural inherences of human language are. At the end of the day, "generativism" =/= "Chomskyism, " it's more about making formalized, predictive models which could theoretically "generate" a structure on their own, rather than solely being descriptions of structure.
Another point is that rather than traditional X'/Government & Binding theory, the predominant framework in contemporary generativism is the Minimalist Program, the specifics of which are highly up for debate. But the point is... they did make the models more or less simpler in the 90s. Most of the specific problems you mentioned disappear in the more streamlined modern framework or even more modernized X'-hypothesis reimaginings; for example, all subjects are understood as originating in a functional position below the TP but above VP which typically encodes voice (vP). It explains VSO word order as occurring in languages with (1) mandatory V-to-T movement and (2) a prohibition on Spec-TP being occupied, rather than an EPP obligating it be filled, as well as explaining English double object constructions since both objects have space to exist in the VP. As a bonus, it neatly describes passive-active distinctions crosslingusitically (passive v's just don't have a subject), especially when you take additional contemporary hypotheses like Distributed Morphology into account which seek to better integrate morphemes into the syntax. Determiners appearing on either side of a noun? DPs are the top-level and NPs are within them, not the other way around (the indicators of which also hold up pretty robustly crosslinguistically, unless the language doesn't have DPs at all in which case the point is moot) while linear order establishes post-syntax, not concurrently. Minimalism is just not typically taught in many undergrad curriculums because the specifics of how components interact is still fiendishly complicated without context of X'-theory, so better to start with a pre-MP incarnation and work towards it; there's still a lot of room for improvement in simplifying the framework while keeping it flexible enough to handle any language.
As for the general complaints about Anglocentrism, unfalsifiability, and overall convelutedness? Yeah that's pretty spot on. However, I would say that most generativist scholars I know and have worked with recognize these as problems and want to find solutions. They use generative models on a predictive basis and test against natural language, adjusting for contradictions when they come across them and retesting. There's also a massive push to reassess the framework in order to better incorporate non-IE languages. I also know a lot of generativist psycholinguists who are doing the best they can to test processing and production against predictive models and adjusting theory accordingly. As I mentioned before, Minimalism is neither perfect, homogenous between advocates, or even particularly straightforward, and it has no shortage of ridiculous, unfalsifiable proposals that are entirely theoretical. However, in a class I took in my undergrad with Jorge Hankamer (a fairly Chomsky-critical generativist), he noted that theoretical linguistics as a whole is still in the "Dark Ages" and that existing models will be probably be discarded at some point, but the basic principle that generativism is simply about making formal, predictive rules is what's important. We just have to wade through a lot of bullcrap to actually get there. I would call myself a generativist on these principles but I am also quite critical of a lot of popular theory, for example I don't really buy into "covert movement" or that coordinated structures are entirely binary branchings. Again, you did a great job on this video and I think you actually captured the popular debate about UG really well, it's just in my experience that the parts of UG most of the popular debate centers on are actually either a few decades out of date or assume that Chomsky speaks for all generativists, let alone UG supporters.
**edit for some typos
I thought for sure Sapir-Whorf would be in this video
Considering the topic and the title of the video, this could very easily degrade into a flame war. That said, I disagree with you on Universal Grammar (UG), but of course the comment section is not the place to have a discussion. Because of that, I'd like to mention the following, not as a rebuttal to the arguments exposed, but as an alternative explanation to anybody who might be interested. To anybody reading, even if you disagree, I hope you'll at least get an idea as to why a heck of a lot of linguists at least consider UG OR use X-bar theory or some sort of tree analysis in their papers (which aren't the same thing at all). Also, please consider the origin of these theories and how they compared both to other existing theories at the time and to other possible theories. A very compelling feature of the current syntactic theories within UG is that, although the theory is complex, it should, in theory, explain all languages, and not just one or two. I feel like that level of analysis is a lot more compelling than having to learn whatever arbitrary categorizations old men have come up with to explain some rule in a language that are typical of "traditional" grammars, but more on that down below.
Anyway, the first thing to note is that UG is (to simplify) a theory stating that all languages in the world share certain similarities. It originates in part as a critique of the American behavioural psychologist Skinner's view of language. According to Skinner, language was just behavioured, and it was the result of operant conditioning and such. This theory made some predictions which just didn't add up (if it's just conditioning, can we teach animals language? See "Why Koko (probably) couldn't talk" on youtube for a GREAT explanation on that topic, including some generativist theory). UG originated from some observations about language acquisition and syntax, but the idea behind is that "language" is a human phenomenon, and has such, it has features that are universal, regardless of which language is being considered. Discrediting the theories of syntax that proponents of UG use would not disprove UG, although the new hypothesis/theory that it's probabilistic learning which enables language acquisition might. As an aside, although many people associate UG with Chomsky, and he's been a very influential linguist, many people dispute how central he has been to the "cognitive revolution" in linguistics (Pollum has an essay on that, although I don't know what to think of it).
X-bar theory and successive tree-analysis systems for syntax within the generative tradition exist as something between a framework and a hypothesis. They have been modified considerably over time, back from a simple rule-based grammar that attempted to describe grammar through something like a logical system (something that a computer could read) up to "Government and Binding" and nowadays "Bare-Phrase structure". If you google any of these, you'll find that they are associated with UG, of course. However, calling "my dog" a DP (determiner phrase) doesn't necessarily mean adopting all of UG's assumptions. In the past, people used "subordinating conjunction" or "adverb" as labels, and that didn't mean they subscribed to a particular theory of language. Again, although some other theories of language have their own terminology, I find that a lot of people just use these terms because it's a good way to communicate, and a lot of the time the terms are more accurate and meaningful than arbitrary groups like "subordinating conjunction". When I learned linguistics in college, that was one of the first things we had to learn, and I hated it with a passion! Consider that people have been writing about grammar and languages, if not linguistics proper, for at least two millennia already. More modern definitions of "conjunctions", such as UG's "complementizer", are at the very least better justified and explained. I can use a "syntactic test" to determine whether a word is a pronoun or not by using it in different sentences (some grammatical, some ungrammatical), and that's something that many linguistic frameworks just don't give a lot of importance to, so I don't think it's unwarranted that many people would use UG terminology in papers that might be more interested in just describing the grammar of a language.
Anyway, UG and UG's theories of syntax are not exactly the same. But let's consider they were, and address the problem of "updating" the theory instead of discarding it. Is that scientific? I'm not very acquainted with T. Kuhn's ideas about what science is or should be, but I vaguely remember that it works a bit like Piaget's ideas about assimilation and accommodation: you come up with an explanation about the world, you amend it in light of new evidence, and when the theory is beyond repair, you come up with a new one, with a new framework and a different set of assumptions, a "scientific revolution". I think it's safe to say that most people agree that's what happened between Skinner's idea of language as behaviour and the "cognitive revolution" was a scientific revolution. Some might even argue that there has been another revolution within generativists with the Minimalist Program, as it introduced the notion that there might be some principles to the computation of syntax that might be at the core of UG.
There's a lot more to say about the history of UG and of syntax, but I think the former is a good starting point. Although that selection of information kinda argued against Watch your Language's view, I tried to keep it just informational.
I'd like to add my opinion on a few things:
-Yes, generativist linguists were and are mostly European. That doesn't disqualify them more than it would disqualify de Saussure to talk about languages. I think the syntactic theory proposed by generativists is just a starting point to be updated by more information gathered from other languages. We could take these European ideas and (a) offer them as a framework for linguists of different languages to use, hopefully contributing to a global database that might be used to study "language" and not just "a language" or (b) toss it in the trash at the first sign of problems and just stick to learning seemingly inexplicable rules for any given language. Other frameworks, particularly comparative linguists, have attempted to do something similar, but their ideas are much less ambitious in scope (pointing similarities is different from trying to explain the emergence of syntax and syntactic rules!)
-Evidence in favour of UG back in the day created a furor around the enterprise. However, more research is needed. Evidence against UG should be followed up on as well. Although there are some central arguments in favour of UG, such as "Plato's problem" or the observation that almost every human child will learn language, there is no single argument on which UG relies on, so just disproving the existance of recursion wouldn't be enough, although it would require a major redesign of the theory, if not a scientific revolution. As regards the evidence for that and other things, many things have been published that just didn't hold up. Many studies on teaching languages to animals seemed promising at the time, but in the end they all ended up proving Chomsky right. On that topic, he said that if an actual fully-fledged language was possible for animals, we wouldn't need to teach it, it would simply happen, just as language simply emerges between people even when there is no common language (or no language at all, see Soup Emporium's video on Hellen Keller). If recursion was indeed a feature absent from some languages, we would have a bit more evidence than just two people's potentially biased account of one single language.I think time will tell.
-I do believe that UG is a bit at odds with scientific thought, and that might be okay. For one, science is not a perfect institution, and the study of scientific revolutions can only get us so far. If I've understood Kuhn's ideas correctly, scientific revolutions should override prior theories, but in many fields of research we have many different "paradigms" that don't quite rule out each other (see psychology for example).
-I think that UG being questionable is a good thing. Other grammatical descriptions of languages are just that: someone's description of a language. There is no science to speak of, just a description of rules and features. UG is more ambitious, yes, and will very often make questionable claims, but at the very least it has set up a framework to test out these ideas (grammaticality judgements mainly). Of course, some ideas that originated among generativists are no longer exclusive to them (again, grammaticality judgements), and now many linguists try to write down their ideas in a more critical, less arbitrary fashion, trying to provide tests to determine whether some is grammatical or not. If anything, linguists who have tried to study languages in a more standardized way (among them, generativists) have made a positive impact on the discipline in that regard.
-Generativists' accounts of language, particularly syntax, have potential applications in computer programming, regardless of how scientific they are.
-Many concepts within UG are interesting by themselves, regardless of what the fate of the rest of the field. For example, the idea that syntax is not linear, a fact proved by, for example, garden-path sentences, and a very important aspect of UG and its theories of syntax. It's very interesting to consider silent elements such as tense, or "features" that "move", or null elements, like some complementizers. I think these are great alternatives to other theories of syntax that seem simpler at first glance, but are actually either inconsistent, or riddled with exceptions, inaccuracies and a lot of unnecessary rules.
(11:40) That might explain why I haven't learned about these trees if they fail basic Swedish.
I remember being fascinated by Esperanto and learning it when i was 12. That was until one day i realised that in order to to that i'm using English (which is not my native language) which already is, you know, the de facto world language now.
As someone who's reading Everett's books at the moment, I appreciate your take on UG. Shit was cash, keep up the good work.
more like gnome chompski amirite
I have some big problems with a universal language. How long would it be before several major groups diverge and become incomprehensible to each other?
How could we provide education to everyone, because I feel that privileged countries can afford to teach their people and join in while poor countries (or uneducated people in general) would be left out.
How would a universal language be implemented? Even if it catches on in one region, others might be uninterested. For example in most of South and Central America, Spanish is already a lingua franca and the people there might see no need to learn any other language besides those directly related to their interests, probably English or Portuguese
I like your videos :)
"Major groups diverging"
You can design a language to avoid common phonological changes. For example, Toki Pona phonotactics forbids "VNNV" where two nasals are next to each other, which often turns into gemination.
"How can we provide education to everyone"
Linguistics can't solve global disparity and neo-colonialism. But simple auxlangs are _much_ faster to learn than natural ones. Esperanto learning is 4x faster than French for some speakers (and that's the badly-designed auxlang). If you need to learn a language to find work - you'd benefit from having a faster one.
"People with regional lingua-francas would be disinterested"
I personally think that auxlangs would only work by starting in areas with no regional language, then spreading out. Esperanto itself started in such an area for that exact reason.
As it grows in speakers, it becomes a greater benefit for other speakers with lingua-francas (on top of speed, and not being the language of your colonizer).
Even if it doesn't grow, it's still created a cultural network between multiple regions with no lingua franca, which is valuable in itself.
_EDIT: I don't believe that an international language will catch on, but it's an interesting idea_
You definitely can make a conlang big enough to express any concept you could ever want. It might take a while, but it's doable. The problem is doing that, while also making it universally easy to learn.
"I will not be covering fringe hypotheses"
*Immediately starts talking about Altaic*
if I’m not mistaken, there’s increasing evidence that most pastoral “Altaic” languages might have started out as early Neolithic farmers that actually switched to pastoralism, which could support the idea of a common geographical homeland tying them to Korean and Japanese which would have remained agricultural. Ultimately I have no idea if this lends more support to a sprachbund or not.
Proto-altaic respect UG, therefore, UG exist. However, proto-altaic probably never existed, but we still need to create it as a conlang.
Great comparison with epicycles! And powerful point about Chomsky's theory not being refutable.
>says he won’t include fringe theories
>first one is altaic
I jest of course, I know it still has a significant following in East Asia.
I don’t think it’s as big as these other controversies, but the IE Urheimat debacle has always interested me. I’m a Kurganer myself (LIKE YOU ALL
SHOULD BE!) but the Anatolian hypothesis has picked up some steam again recently due to some nasty geneticists misconstruding some data!
I find your final argument against conlangs as lingua francas a little weird, because, like... if the problem is the lack of lexicon, just create more? Even natlangs create new words all the time, it's a normal thing for any language. Sure, for conlangs, everything is created, but if someone is dedicated enough, a large enough lexicon could be created and the gaps can then often be filled with more creation from scratch or just using pre-existing roots like natlangs do.
I totally agree that toki pona isn't great as a lingua franca because there are too few words, but that's a problem of toki pona, not necessarily of conlangs as a concept.
(Not saying I necessarily support conlangs as lingua francas, I just found that particular argument a little strange)
+1
Yeah - for example in Esperanto, Zamenhof would coin new words and add them to his regular newsletter for Esperanto speakers. So he could grow the vocabulary really naturally.
Most of linguistic debates are because if you have a definition "definition1" that describes a thing "thing1" having properties A, B, C, D and you find in other languages phenomenon or thing "thing2" that has properties A, B, X, Y, Z, instead of the simple approach of putting a new a definition for a thing that has properties A, B and say thing1 and thing2 are examples of it, you try to describe thing2 with definition1 + lots of modifications and exceptions which is a more needlessly complicated way to go about it. That's why I like construction grammars the most they don't assume universals which are mostly language specific things squeezed to fit as an underlying principle for all human languages. describe everything in its own terms and you will see many things are simpler than expected. less assumptions is better.
Also, I was thinking about cases, and I feel like I almost think they don't even exist? Or maybe they always exist? Aren't cases just specific instances of giving information about a noun's role in the sentence? Maybe, not sure :D
First video I've watched in this series. I loved learning years ago that Turkish and Korean were related. But now I understand why speakers of those languages were always surprised to learn this contested fact.
I think most linguists would definitely say that they AREN’T related.
The construct state is defenitely a case, but not an "inverse genitive"; it's a thing of its own
NOSTRATIC
BOREAN
PROTO-HUMAN LANGUAGE
6:58 Regarding the "Germanic possessive s" in English, is it true to say that the word "of" expresses possession, as suggested? Or at any rate that it expresses the 'direction' of possession i.e. possessor and possessed, which the possessive 's' definitely does.
Personally, "A friend of my mum" and "A friend of my mum's" have slightly different meanings. The former could indicate a mutual friend, or it could be an unrequited friendship: someone who likes my mum but she secretly doesn't really like them back, perhaps just tolerates them. But the latter form with the possessive 's' indicates that the direction of possession is that "my mum" is the one who possesses the friendship, which in the case of friendship means she is the one who definitely likes them and thinks of them as a friend, whether or not they happen to return the feeling.
That's all a bit metaphorical in the case of friendship, and perhaps other people might not feel the same distinction as I do. But I think it's perhaps easier to see in another example. I would contend that:
"A shield of Brian"
would be generally understood to refer to something or someone that protects Brian, while:
"A shield of Brian's"
would conjure the image of Brian in mediaeval garb ready for his historical re-enactment society's performance of the Battle of Towton.
So the "of" alone in that first sentence doesn't indicate the same possession that the possessive 's' indicates in the second, even if you think they might both be seen as some form of possession (whether physical or conceptual). Off the top of my head, is the "of" in "A shield of Brian" more akin to the "of" in "A pie of cherries" perhaps? Even if you say the pie 'possesses' a nature of cherryness, it's definitely not the cherries that own the pie in the same way that Brian owns his shield in "A shield of Brian's".
That’s interesting, but I, and those who I know, use both forms to mean the same thing
Can we talk about optimality theory? I've only taken one course on it last semester, and while I think it has many merits, part of my mind is always screaming that it's kinda just unfalsifiable and that you could model any process that way. I never got to speak any of those concerns and the storm has just been raging internally lol
Has any metalinguistic theory ever tried to be falsifiable?
@@pawel198812true dat
Nice to have the pro-Altaic and Anti-Altaic arguments summarized. I did not know about the way Sprachbunds impact word order! And I find the sprachbund hypothesis quite believable. Given the Turkic urheimat is the one that's actually around the Altai mountains, it requires a sprachbund that only stretches from there to central Manchuria to include Tungusic- not very far by steppe nomad standards. I am impressed by how YOUNG that makes the constituent families though, they cannot be traced further back than the first few centuries BC. Where did they come from? The eastern steppes are weird.
As for the others I had no idea those arguments about morphology and universal grammar were even were a thing. Thanks for this look down the rabbit hole!
Hot take regarding interlangs: Esparanto is a French-Polish pidging and Toki Pona is too restrictive (having to simplify your thoughts with it is NOT praiseworthy!). Given the current status quo if you gave English some spelling reform (and/or reversed the great vowel shift) you'd me most of the way to a working AuxLang. Fairly analytical so a beginner can get their ideas across without thinking much about grammer, plenty of flexibility for more nuanced communication, no need to rebuild centuries of linguistic baggage from scratch like an a priori clon. English language and culture for better or worse is already a prestige language of the world, might as well work with it.
Just putting it out there but a video of you and Connor Quimby discussing these topics would be very entertaining.
One argument against an Altaic family is this : The farther back you go in analyzing. samples of the Turkic, Mongolic and Tungusic families. , the less similar they are . This would not be the case if they stemmed from one proto language .
The biggest problem with Altaic is that the family’s second largest branch, Mongolic, can only be reconstructed to the time of Genghis Khan. Thankfully, Khitan is opening a window into prehistoric Mongolian but even by 1000 AD, Mongolic and Turkic peoples have been interacting for millennia, meaning that both language families share a lot of vocabulary.
Although our host exiled all fringe theories at the start of his video, I'd like to ask in this forum: Does anyone know of an Oklahoma-based professor from about the 1980s or Nineties who theorized the human race originated in the Americas, not Africa, and his evidence was a series of language families that classed English with Chinese, French with Arabic (if memory serves), etc -- all so he could argue the American Indian nations have always lived in their historical homelands and were not migrants via Beringia. I didn't xerox his paper then, and I can't remember his name now. His beliefs were so bizarre, I find it kind of fun.
I don't know Arabic, but as far as I understand states of a word are supposed to be a grammatical quality/property/whatever which occurs together with cases, gender and number (if any). So, while Syriac nouns have 3 states (or 4: 2 being identical) formed with endings, but no cases at all (what makes their states to be cases in practice, so the name "state" is employed for historical reasons, I guess)... I think Arabic and/or other semitic languages do have both states and cases appearing simultaneously in a word. Thus, you would find a feminine plural word in nominative and construct, or in dative and construct, etc... The other states, with different names in every language's grammatical tradition, usually are supposed to tell wether the word is definite or indefinite. In Arabic these states, if I'm not mistaken, are often also alternatively explained as presence or absence of a definite article, since they are marked by adding or lacking the prefix -ال. In my point of view, cases and states are basically the same thing in function, but if languages have both of them you rather make the difference since you can combine each from each category between them, in the same way you don't call nominative singular and plural different cases; otherwise you would have not a single construct case for these languages, but a subject-possesor case, object-possesor case, etc.
I don't understand why people have double standards about global lingua franca. People often list flaws of conlangs as lingua franca while somehow accepting those exact same flaws for the current global lingua franca, English.
1. "Esperanto is too Eurocentric", but the current lingua franca is English which is worse than just being Eurocentric, it's Anglocentric. My native language is Indonesian, so both English and Esperanto are equally foreign to me (and to most people globally).
2. "Esperanto has to many phonemes, with 22-23 consonants and 11 vowels", meanwhile English has 24 consonants and 20-25 vowels in RP, 14-16 vowels in General American. Not to mention, English phonemes don't match one to one with their graphemes. Learning to pronounce English words as a kid was a nightmare. Not only I had to learn weird sounds, but I also had to learn weird spelling.
3. "Esperanto grammar is not that easy", I don't think English grammar is easy either. At least Esperanto grammar is more predictable than English. Sure, Zamenhof had a bias because of the languages he speaks, but most people don't have any bias towards either English or Esperanto. If I had a choice on which language I had to learn as a kid between English and Esperanto, I would've chosen Esperanto.
4. "Natural languages have richer vocabulary than conlangs". I think this is a weakness of natural languages, not a strength. The purpose of a global lingua franca is for international communication. Most people around the world (including me) learned English not because we're interested in Anglophonic culture or literature or poetry. We just want to be able to speak with people from other countries (mostly for work related purposes). But somehow, we were forced to learn all these BS literary and bougie words in order to pass English exams. Global lingua franca does not need vocabulary that rich, it only needs enough words to have basic conversations, business conversations, and scientific conversations. If you're interested in literature or poetry, you can just learn a language whose culture you're genuinely interested in, not a global lingua franca.
We could've spent minimal effort learning a lingua franca and put more effort learning a language we're genuinely interested in if we were forced to learn an easier lingua franca than English.
A conlang does not have to be perfect to be a lingua franca, it just has to be better than the current one (English).
its very hard to argue that japanese is part of an altaic language family, most of the given similarities are superficial at best
Love the channel, can't wait for more content! Congrats on ur last semester :)
I strongly disagree with the sentence "Conlangs can never have the sheer amount of vocabulary that makes natural languages operational".
Conlangs have proven to be operational. That's not even a question anymore. As you said. There's even native speakers Esperanto.
Toki Pona is the extreme example. Other full fledged conlangs actually in use typically have much larger vocabulary list to utilize.
The key is context. As long as you know the topic, what someone is talking about, and understand what they are saying about that topic. You understand what they are saying. You're communicating. Context is extremely important in natural languages too, just slightly more in conlangs. With conlangs able to use context more directly as a tool to drastically reduce the vocabulary needed to be taught to someone.
Even though Toki Pona isn't the best at more complex topics. It's still what I'd consider usable for more complex topics, you can still build up the context. Additionally I'd say real life communication is actually where Toki Pona excels as an interlang. You can easily get by 99% of your day to day life just fine with Toki Pona with minimal effort. It doesn't mean you need 6 different words to describe a Banana. If you don't just want to describe it as a fruit. You can use 2. kili jelo. Or Fruit that is Yellow. If there happens to be a lemon near it. Maybe throw in a 3rd, or point. You can do a hell of a lot with 123 words. And for that reason I think it's a great interlang. Especially due to the barrier of entry being so low. However. I wouldn't want it to be the language of education, math, science, etc.
However. Conlangs still have significant advantages inside those fields. You can design a language to be more regular and predictable. Natural languages have ALL the issues you typically critique certain interlangs for. Too many difficult to pronounce consonants, too Euro centric. Isn't regular enough, etc. Ignoring the reason to compare and nick pick the variety of interlangs there are to choose from. They certainly beat out natural languages dramatically. Especially from the standpoint of having a language relatively easy for any person in the world to learn despite their native language. You don't NEED gendered words. You don't NEED tones. You don't NEED 10 direct synonyms for the same word. You don't NEED formal and informal speech. Sure. You can dislike a conlang based on how artificial or non-elegant something sounds. But most conlangs don't really have that issue.
About point 2, Portuguese uses elisions and sandhi for our most common prepositions that mark the direct object like "a/a" for direct object, "pro/pra" which is a fusion of per + our definite article for indirect object, "de/da" for our genitives, etc.. So, if our preposition's elisions could one day be reanalyzed as prefixes, as they can't occur on their own unlike the words they were based of, wouldn't "d'" be a genitive case? "pr'" our dative case? "n'" our locative case? Given the fact that despite our word order being SVO, Portuguese can be somewhat flexible with word order(Many of our phrases are in SOV like "Eu te amo"), but we need the prepositions to mark the object of the sentence. If case markings don't necessarily need to be suffixes, our elisions of our most used prepositions could easily be reanalyzed as cases, and Portuguese would have at least 6 or 7 of them.
Also, if we consider they fuse with the articles instead of the object nouns when the object doesn't start with a vowel, these could be thought as them inflecting for number and gender since the prepositions that are most ellided are also the ones we commonly fuse with our articles, and our articles are literally just vowels or a vowel+nasal and s which on normal speech, is fused into the following word through sandhi. They could be easily thought as just gender/number inflections for the "case markings" instead of the articles they are.
I agree about UG. Chomsky only knew the USA dialect of English, so he assumed ALL languages are USA English with bits tacked on
I count myself an amateur linguist but am only interested in diachronic linguistics, that is explaining origins and describing changes. In most cases these languages are unattested and cannot be analyzed in the way modern ones can for lack of any speakers. But that is the beauty of this field for me. 😅
What about the Sapir-Whorf hypothesis?
That’s a good one
The question of what is or isn't a case has been on my mind again. Years ago I was learning Turkish. I learnt its cases of course. Only, the word "case" was never uses in my lessons, this despite the mutual language of me and my tutor also having similar cases! I learnt that the suffix -de means the word is where something is/happens, -le means with the thing, -e means out of/away from the thing, etc. Turkish cases work similar to Hungarian, and I wonder: If the languages did not have vowel harmony, would we still identify them as cases?
Looking back, those case endings matched more with prepositions in my head with how I learnt them. So I wonder, how closely can an adposition "glue" to a word before it's a case? I pulled this conlang my assfrom, this conlangin all adpositions Englishfrom moved the othersideto the wordof they modify and no longer have a space. Does this monstrosity have cases now?
A dedicated vid on the subject would be cool!
Objection: Philosophy is good (it is about computable models), universal grammar is bunk.
ona li toki lon toki pona!!
I think it's very useful internationally _if_ you give up on the purist philosophy and pick up loan words in shiploads
The more I learn about Chomsky, the more I wonder how he got so popular.
Utter amateur here. I think it would make the most sense to take the seemingly ridiculous route, and actually _do_ consider prepostions cases, reworking the whole system in the process. Cases would be then simply split into the standard categories: analytic (adpositional, prepositional) agglutinative (suffixed, infixed), synthetic and mixed. The borders between them would then be casted by already existing linguistic definitons. Examples:
Into the house - analytic
A friend of mine - analytic, synthetic
Poszczuć psem (ins.) - synthetic
Iść z psem - analytic, synthetic
Während der Kreuzigung - analytic, synthetic
Kreuzigung des Verbrechers - synthetic
Etc. (Entschuldigen Sie bitte mein deutsch. Wenn ich einen Fehler irgendwo gemacht habe, korrigieren sie mich bitte)
Just a question since Altaic is formed partly by genetics. Does altaic make sense without Turkic? I’ve asked some of my friends who know linguistic and he say grammatically Turkic is the odd ones out. From learning about genetics myself, Turkic seems to have two possible roots which are Altaic and non Altaic, thus if we cut Turkic out maybe it can make more sense
Alright, I'm totally nitpicking here, but I am studying linguistics so this video has already made me wanna fight something. Hi, I am a native Hungarian speaker! "A követelményeitek hülyék" is not what I think should be there. It means that the requirements (those being living, breathing things) are stupid. If I am correct, it should have said that the requirements (unalive, impersonal things) are stupid, which should be root hülye + "ság/ség" (vowel harmony!) + pronunciation aiding vowel (courtesy of vowel harmony) + "k" to mark the plurality that ties back to the requirements being several. So the whole sentence should be "A követelményeitek hülyeségek". Or I might be in the wrong. Somebody correct me if I am! Thanks!
This is interesting, that your native intuition is telling you that Hungarian adjectives work differently based on animacy. Is this just for this adjective, or is it a Hungarian grammar rule I should know?
Also, isn’t -ség a noun suffix? How exactly does that work in this case? Is it saying literally that “Your specifications are stupidity?” (I’m not challenging, just asking, you know more than me)
@@watchyourlanguage3870 Since today is a national holiday in Hungary, everybody's home and doing nothing, so I did a miniature empirical study on the topic. My father got confused by my question, but he also thinks in Hungarian/English in a sort of mutually interchangeable way. My mother, by profession a grammar and literature teacher, told me the sentence is ungrammatical, since a "követelmény" will always be an inanimate list of requirements. I still wouldn't be surprised to hear a small child or a drunk person say such a sentence, but I will not mix in spontaneous native speech, because we don't have time for all that. As for why it didn't even occur to me that " A követelményeitek hülyék" is ungrammatical.... I am a polyglot, my mother is not. I have the tendency to say sentences with Sanskrit grammar, then get angry with myself when I can't untangle the mess I've made and that makes my ears bleed. You should see the emails that get exchanged between me and my teachers. Although we use English in lectures and seminars, emails are usually in Hungarian, and the struggle is real from both parties.
As for -ság/-ség being a noun suffix. It is, specifically one of the largest groups of suffixes: an affix. In Hungarian we have a strictly fixed order in which suffixes can follow a root, and affixes come right after the root without exeptions. The general rule for ság/ség we learn in school is as follows: noun/adjective root + ság/ség = abstract adjective. For example "barátság" - friendship.
"Your specifications are stupidity" is kind of on point. Maybe put stupidity in plural. When I was learning Spanish and English in highschool, I suffered a lot because I could not yet switch my brain to think in the target languages and I struggled to express exactly what I wanted to say. Most other languages feel restricted to me in a way. Now I have finally learnt to let go of this. Regardless, Hungarian has 18 nouncases or so, and if I think about the number of ways I can alter a simple noun with just suffixes, I stop around 25 but I could go on. The closest I got to think in a natural way is with Turkish, one of the languages I study at uni. Lithuanian is a close second. I think I might make Sapir and Whorf rotate sleeplessly in their graves. All the above madness surrounding linguistic relativity is also the reason why I can't finish my BA thesis on Ted Chiang's Story of Your Life.
All in all, Hungarian adjectives depend on the subject of the sentence. "A követelményed hülyeség", but " a pingvin hülye" if you are in an argument with your professor on his final exam requirement, and you two ran into each other at the zoo, near the penguin enclosure, with a particularly dumb penguin in it. As for suffixes, we have a gazillion for nouns, adjectives, verbs, and so on. I think Wikipedia has a nice little chart under "Szóképzés" in Hungarian. Lastly, I think somewhere along the line somebody mistranslated some linguistic terms from Hungarian to English, hence half the confusion for foreign linguists. But this is yet to be proved.
@@annalevai5512 Fascinating, thanks for writing all that. Good luck with your BA thesis!
Altaic hypothesis is interesting, but there is no hard evidence to prove or disprove it. One of the reasons we could establish PIE is because we have written evidence pretty close to the split. We have Greek, Latin and Sanskrit texts (not just inscriptions, but actual texts) from time that is roughly similar to English-German-Swedish split. Where the languages already lost mutual intelligibility and went their separate ways, but there is little doubt that their similarities are not a coincidence. And we have relatively "pure" languages, that don't have a lot of outside influence. Latin is Latin. It doesn't randomly have 70% of X vocabulary that replaced all the native words except for the most basic concepst and things. Few Greek loans? Yes. But just a few. With "Altaic" languages we don't have that. All the evidence is relatively new so you can neither properly establish what is and isn't a "core" word nor can you see words at a relatively close point in their history. It is as productive as comparing Spanish directly to Swedish. There clearly are a lot of same-ish words and some grammar overlaps, but you can neither find exact cognates (because a lot of them were simply lost or mutated beyond recognition) nor establish good sound correspondances beyond "there is L in leche and mjolk"
A problem with "Universal Grammar" is the fact that these restrictions are of computation coplexity origin, not of "innate grammar" one. You will not find a single language that doesn't keep its articles as a part of a noun phrase, simply because it takes too much brain power to do so. Even with additional measures, like grammatical gender. "iThe bMan lives in iHouse" is a chore to parse through even at such a small scale and where all the nouns quite clearly belong to different genders. And when you pack everything together there is no real reason to encode certain information in a specific way. "in house" is as easy to parse as "house-in". Just as there is no real reason to select "green house" over "house green" (unless we are talking about more complex interactions).
link to "On Problems with X-Bar Binary Branching"?
Portuguese language overview please 👉👈🥺
Hi, your videos are so great! Thanks a lot! Could you maybe make a video on the Ithkuil language? I think its soo interesting
Prescriptivism vs Descriptivism is also a funny one...
Do you have a list somewhere of all the languages that you have learned/are learning, and your level of proficiency in each? Because every video, it seems like you speak three more than I knew before. Based on your language overview videos, your Germanic correspondences video, and this one, I know that you speak
English (duh), Dutch, Danish, Swedish, Icelandic, and Yiddish (but not German); MSA, Levantine Arabic, French, Spanish, and Hebrew, that you're currently learning Hungarian, and that you don't speak, but plan to learn Turkish, Korean, and maybe Japanese.
Please let me know what other ones are missing.
Actually I studied Hungarian last year, currently I’m working on Telugu
I’ve also studied Portuguese, Italian, Russian, Amharic, Catalan, and Polish
@@watchyourlanguage3870 I honestly thought I'd said Russian. Thanks for the reply!
Do you think a language made so that any Germanic speaker could communicate with it would work? Some would argue English already serves this purpose, but we have so many loanwords from Latin and French that English wouldn’t be able to accurately represent all Germanic languages.
That exact fact that English borrows so much from Romance languages actually does hinder Germanic auxlangs. They’ve tried, Jan Misali reviewed the conlang known as Folkspraak th-cam.com/video/KegIeZwXUDk/w-d-xo.html , which does exactly this, but it seems to me like a version of Anglish (even for other Germanic language speakers)
About the construct; I think you can say that it's not a case, beacuse Arabic (for example) has an entirely separate case system, which also extends to modifiers and such. So a noun can be in the construct state, and itself can be accusative. Calling the construct a case doesn't make sense unless you allow nouns to stack cases.
Also, I need to watch your videos at least two times to actually fully understand them
I love ur channel so much I feel like ur videos were made for me lol
what course are you taking at college? do you recommend taking it?
Currently since it’s my last semester I’m just in Applications of Linguistics, it’s my capstone and it’s basically just “remember everything from earlier classes? here’s how to use it irl” I’m enjoying it. Thanks for the words of praise!
That was so interesting!!! Great video! However, I have to disagree about conlangs as international languages, similarly to what @Mercure250
already said. I feel like you are employing the naturalistic fallacy (since were talking about logical fallacies) here, there is nothing that says a conlang couldn't be as good, or better, with as good, or better, lexicon, and as good, or better, phonology, it's just that none so far have achieved that. I agree though on Esperanto, even though I speak it and love it as a language, no way it is good for truly global communication, as like you said, it is very eurocentric.
My opinion on cases is that if it's an affix it's a case, if it's an adposition it's not (but you can still use case terminology like "accusative" or "genitive" to describe it if you want, because it's practical), and if it's a clitic uhhh, I dunno...
Also I feel like you didn't really present any good argument in favor of natural languages being better IALs than conlangs and I do remain absolutely convinced that conlangs are a better alternative: Lexicons always have to be continually enriched by usage anyway so I don't see why that would be a specific problem for conlangs as opposed to like Pirahã or Sanskrit which also lack words for a lot of concepts that we use daily, and I find criticizing conlangs for being ethnocentric while advocating for natlangs to be quite self-contradictory. The main problem for conlangs is probably that there'll always be some issues with them so it's hard to settle down on one conlang as an IAL instead of continually trying to make better ones.
For instance I think esperanto is a great candidate but if I had it my way I know I would absolutely change its official grammar to implement the "-iĉo" suffix as well as a "hi" masculine 3PS pronoun, a distinct 2PS pronoun, replace all instances of /ʒ/ by either /dʒ/ or /j/ and of /x/ by either /k/ or /h/, rework the orthography to get rid of the diacritics, and maybe simplify the use of the accusative marker a bit as well. By then it would probably be a pretty different language and I guess this is why new esperantidos kept popping up and nowadays new unrelated auxlangs do.
2:22 due to the Altaic language family!
Not sure why the Germanic /s/ possessive can't just be treated on a language-by-language basis?
But in Arabic, the construct state itself can have different cases, for example, the word queen in Arabic:
malikatu (nominative construct)
malikata (accusative construct)
malikati (genitive construct)
Why is it a case if the construct state can be in different cases?
Here's my takes on this:
[1] - Not much to say. I'm sceptical of any new language families, and so I'd incline to be an Anti-Altaicist.
I'm not a linguist however.
[2] - I agree with the end part - a better definition of case is needed.
[3] - Mostly agree. Conlangs aren't great as possible IALs. Latin, and English, have both served as de facto IALs in the past.
Toki Pona speakers, from personal experience, also really wouldn't like their language becoming an IAL.
(you only need to spend 3 minutes on the largest discord server for toki pona to get a free lobotomy)
[4] - I think Occam's Razor is enough to cut down UG at this point because of how complicated it is. It killed geocentrism, it'll probably kill string theory, it's gonna kill UG. It's sad because Chomsky is a guy that heavily influenced my political views, which are still 85% Chomsky, even today.
what languages do you speak and what do you plan on learning?
Eurocentric and Piraha are two words that immediately shut my brain off.
"I will not, however, be covering fringe hypotheses"
*proceeds to cover Altaic, a fringe hypothesis*
dbi like "2 cases", there are only 1 because nominative is unmarked and the other is the simplest case known to man, even languages without case inflection have accusative marking in one way or another like just dont beg it please its one case
dont beg it literally "if its supposed to be relatively simple, why are there cases?" 🤓🤓🤓🤓🤓🤓 its one case dbi please
2:28 HOL UP, Semitic influence on Celtic languages?? HOW? You can't just include a passing note like this and then not explain it any further!
Langfocus has a great video on the subject: th-cam.com/video/OAAmwtdP1bE/w-d-xo.html
@@watchyourlanguage3870 ancient Semitic speakers forming a substratum for Celtic languages in the Bri'ish Isles is not the theory I was ready for today. Is there any other theory besides Venemaan's on this?
also, I'd watched this video when it came out, but have since forgotten completely, my mind is like a sieve xd
Thankful that my department is decidedly on the side of sociolinguistics so I've never had to suffer the way you have with UG!
Re lingua franca, I don't believe in any lingua franca. The eurocentrism of synthetic languages is not overcome with the status quo where we leave English as the lingua franca. If using translators everywhere is cumbersome, it is worth it to me in order to preserve the equality of languages. In fact, as a matter of state policy, I don't believe a government should have an official language and should be responsive to the languages spoken by communities (in education, public health campaigns, etc.). Inclusion in the public sphere should not come at the expense of your language and culture.
1) Koreanic: *hǎtǎn-nǎlh "single blade" > Japanese: kata-na "single-edged blade"
2) Tungusic: *kulïn-ki "snake" > Koreanic: *kusïn-ki "snake" > Japanese: kusana-ki/kusana-gi "grass-cutter" (lambdacism /l/ > sigmatism /s/ in Korean something like in Proto-Turkic lambdacism /lʲ/ > Shaz-Turkic sigmatism /ʃ/)
3) Tungusic: *xürgü-či "tailed" > Old Japanese: *worö-ti "tailed" > Japanese: oro-chi "snake"
4) Tungusic: *sele-me "iron-y" > Mongolic: *seleme > Turkic: *seleme/*selebe/*sebele/*seble > Old Hungarian: *szeble + szab- "to cut cloth" > Hungarian: szablya > Polish: szabla > German: Sabel > Dutch: sabel > Japanese: saber-u (cognate to Russian dialectisms: сулеба/сулема "saber", compare to Manchu: selemu and very possibly to Finnish säilä "saber" that can be a mix of native säilä "splinter" + sapeli "saber" a Tungusism)
Non-Altaic sprachbund borrowings in Japanese
1) Sinitic: nn̄g-chat-kùn "two-segmented cudgel" > Japanese: nunchaku
2) Austronesian: duduk "stabber" > Old Japanese: turuk-i/turug-i > Japanese: tsuruk-i/tsurug-i (Indo-European analogy to duduk with Celtic *kladiwos "stabber" that related to *kladeti "to stab" > Latin: gladius "sword"; Slavic: *kladenьcь "sword", *kladivo "hammer")
I was never impressed by Chomsky in ANY intellectual endeavour, since he gives every evidence of being dishonest (e.g., claiming he never said things that he said and altering republications of his texts to cover it up), and in general seems to be a pompous windbag. I too was told that UG was a fantastically important achievement in linguistics --- until I read his statement that it was unnecessary to be familiar with actual languages to think it up. When he proclaimed, on the basis of pathetically uninformed misunderstanding of genetics, that language didn't exist until a few tens of thousands of years ago and first appeared in Europe ----- failing to explain how, for example, Australian aborigines who had been isolated from the main body of humans for a much longer period ended up having language like everyone else, and somehow managed to navigate in groups across a treacherous 300km expanse of ocean without possessing language ----- well, at that point I threw in the towel.
I was familiar with his politics before his linguistics, which, suffice it to say, predisposed me to disliking UG
@@thevilmoron It's f*cked that he denied the Cambodian genocide.
As far as I know, his other politics have been criticizing US interventions and coups.
For example in my country, our first president was assassinated and replaced by a US-friendly dictator, and the US govt admits to it.
Yet most US nationals aren't aware of even 5% of these interventions. And when learning about it, they'll even try to justify it.
Even if they criticize distant-past colonialism, they claim that "It's so sad there's corruption in developing nations" without questioning where this corruption comes from.
So you can criticize Chomsky on the points where he's very wrong (including linguistics), without throwing out the very accurate ones.
i'm so here for the x-bar slander, afaik i was the only one in my class mad about how anglo-centric it is
Lingua? more like ligma!!!!!
Don't Hungarian case endings have to be bound morphemes? Since they can't be written as prepositions, they can't really be read as prepositions. Simples.
I think your opinion of natural languages being better lingua franca's can potentially be harmful to smaller cultures and has historically killed off languages with fewer speakers. Real-world examples of this include Tagalog in the Philippines, Russian in Eastern Europe and the Caucasus and English... everywhere. Meanwhile, with auxlangs, everyone has to learn a new language, and simple enough that people would prefer to use their native language in contexts where they can. I believe auxlangs albeit unlikely to work on a major scale, have their place
Altaic is not a language family
Yes, English will dominate the world but in doing so, it will disappear just as Latin and the Altaic language 👀
Meanwhile Latin speakers: I didn't hear no bell!
Serious question: Isn't descriptivism paradoxically a Form of prescriptivism? Isn't judging "rules" for language as bad inherently a judgment/rule itself? Hypocrisy.
descriptivism isn't "judging", it's attempting to describe the general patterns that language has
@@Jerald_Fitzjerald It's saying at least implicitly that prescriptivism is fascist and all those other fun buzzwords.
@@seronymus what the fuck are you talking about
@@seronymus Nobody says that...? You're literally just making a person up so you can get mad lol
@@winnebagotrout1997 There's definitely some people who say that. I know someone who thinks standard languages is imperialism and purposely talk and spell things wrong to "combat" it
My quippy take on the Anglo universal grammar tradition is that it's wrong, but not as wrong as you'd think. Like, there's pretty clearly some sort of built-in neurological basis to human language acquisition, and this should constrain the sorts of languages you can get. But the actual grammatical operations they come up with can't possibly be what the brain is actually "computing", any more than the tables of forms you find in classical grammars exist in a native speaker's head. Squishy neural networks simply don't operate like that. They're mistaking the map for the territory.
Why is philosophy a slur? It's what all good linguists do but don't want to admit
Hehehe I don’t know this stuff, but I like it hehe