-
"How many languages do you speak?"
Jan 11 ⎯ If you’re anything like me - openly interested in foreign languages from a very young age (or not), and occasionally overheard speaking a few of them - you’ve probably ended up being known, willingly or not, in your small circle of family or friends, at work, or in your tiny rural town, as ✨a polyglot✨. Whether that’s a good thing or not really depends on what you plan to do with that flattering label. Because, come to think of it, it really is flattering. Most people, fuelled by traumatising memories of language lessons from their school years, are very aware of how intense, long and tedious learning a language is. So the mere idea that you haven’t done it just once, not twice, but several times? Surely that must mean you’re some kind of genius! It’s hard not to enjoy being on the receiving end of exaggerated praise about the supposed depth of your mind. So when you’re introduced as “the person who speaks tons of languages”, a sentence you mentally footnote immediately if you’re the humble and self-conscious kind, you don’t really know how to respond. Well, actually, you kind of do, because you’re already anticipating the question that almost inevitably follows: “Which languages?” You’re asked with wide eyes and genuine, dazzled interest. And unless you’re either extremely confident in your skills or mildly delusional about them, your next instinct is usually to launch into a brief explanation of fluency levels and why they’re much harder to define than people think. That is usually when you lose people’s interest. Because people don’t really want to hear you explain the intricacies of what speaking a language actually means (unlike you, my dear reader. I know you will happily read on). No, what people want is a demonstration. A proof. A performance. A visible token of your brilliance. In their minds, you are now a jukebox, and they are ready to toss the coins. You are a circus monkey doing hula-hoop on a walking globe rolling up onto a platform and back down again. (Something that I mention not for dramatic effect, but because that was the proud final act on my first year of circus school, so I can tell.) Because what is very likely to happen, if you agree to show off just a little bit, is that someone slightly cheeky will ask you something very specific: oh yeah, you speak Zulu? Then how do you say ‘Government is most stable when its decisions are mistaken for necessity’. Or something equally twisted, pretty much designed to put you on the spot, that son of a b…. And then there is, of course, the risk that someone around you has a complete command of the language you’ve just claimed to speak, whether they’re native or not, and simply wants to switch and chat with you. Not out of malice. Just out of genuine pleasure. This is where your credibility can collapse dramatically if you’re not quite as good as you may have implied, or as people have generously assumed. You lose instant aura points, as Gen Z would say, and there is no real recovery from that. So better safe than sorry: the safest bet is to never brag about your language skills to begin with, and to not show off when prompted even if you are pretty good. You don’t have to prove anything to anyone, you little monkey. Even if you’re just asked to say something relatively easy, like “I tried to study and I ate instead.” Speaking on demand is a strange enough concept to make anyone freeze, even when they genuinely can speak. And there is always the added risk that stress will make you stumble over pronunciation, which immediately casts doubt on everything else you’re saying. People start wondering whether you actually speak the language or are just babbling something vaguely foreign-sounding. There is, unfortunately, a strong correlation between pronunciation and the level of mastery people think they’re witnessing. Which is deeply unfair, when you consider that someone can know a language in depth - its vocabulary, its grammar, its expressions - and still never quite sound the part. Pronunciation carries authority in a way it probably shouldn’t (even though I have argued in an article that aiming for a good pronunciation is paramount to make real progress). Conversely, people with excellent pronunciation are often trusted more, and assumed to speak better than they actually do - a dynamic on which many internet polyglots have built entire careers. And when you stop to think about it (and thank you for sticking with me, dear reader: I promise I’m about to land this plane), what does it actually mean to speak a language? At what point do you decide that you can speak it? This is a question I ask myself every time I update my CV and have to compress my language skills into painfully narrow categories, usually ranging from “elementary” to “native”, or from “school level” to “professional”. What do those terms even mean? Because what usually comes after “native” is “fluent”, and what comes after “advanced” is often “intermediate”. So if you happen to have a wide range of linguistic abilities across several languages, where exactly do you put them? Am I only “advanced” in English if I’ve organised my life in such a way that I breathe English every minute of it, even though I wasn’t born into the language and still make occasional pronunciation or wording mistakes? Is my Spanish merely intermediate if I understand everything I listen to and read, yet wouldn’t be able to write articles like this one? Is my Italian only elementary if I intuitively understand most of the content I consume, but hesitate when it comes to actually contributing substance to a conversation, simply for lack of practice? And when it comes to “school level”, what exactly is implied here? Personally, after three years of high school, with around three hours a week of almost private Russian classes (because the rest of my class was uninterested and most people simply didn’t show up to class), I had reached a level good enough to travel freely in Russia for nine months after high school—and to obtain a TRKI-2 level at the end of that period, which roughly corresponds to a B2. That was quite a measurable progress. Although, if I’m being honest, I don’t think I had actually reached that level—but that’s material for another article, because proficiency tests like the CEFR are, in my opinion, deeply flawed and not especially representative of one’s real abilities. And finally, “professional” is probably the level that makes the least sense to me. Your job is likely so niche even in your native language that many of your fellow citizens wouldn’t fully understand your jargon and what the heck you are doing everyday. So the idea that you could do the same job, in another language, without friction (if that’s what “professional” is meant to imply) feels slightly absurd. Try asking a bilingual lawyer or a doctor to perform the very same tasks in another language. It’s wildly unrealistic. Even translators and interpreters, whose sole job is to work between two languages they are supposed to master completely, still have to learn new vocabulary every single day, depending on the situation: a conference, a scientific article, a novel. There are also things you can probably say in one language that you simply can’t in another, regardless of your official level in either. My Turkish is much more lived in than my Spanish, simply because I have (literally) lived in Turkey and never in a Spanish-speaking country. I have also been in a loving relationship in Turkish, fully integrated into the family, and I haven’t experienced that in Spanish. As a result, I comprehend the first on a much deeper level than the second, even though my Spanish is objectively much stronger in academic terms: a very broad vocabulary, a solid command of grammar, but very few cultural references and little emotional attachment to truly feel connected to it. Once I spent a full week enjoying Baselang’s first free week of unlimited classes in Spanish, where you can book 30-minute lessons with speakers from all across Latin America whenever you like (and I cannot recommend it enough). The first day alone was a blur of back-to-back half-hour sessions in which I struggled constantly and stalled mid-sentence. And yet, by the start of the second day, I was comfortable again and ended up being placed at a C1 level and constantly congratulated on my “neutral” accent and my neat speaking abilities, despite having almost never practised Spanish conversationally in my life. What really struck me, though, was the realisation that I had very little to say. I knew almost nothing about the cultures of the people I was talking to. I’d seen maybe three films in Spanish in my entire life, at best, and I’d never really listened to music from their countries. I was genuinely stunned to be able to understand and speak freely, using fairly intricate grammar and words I didn’t even know I knew; yet still have so little substance to bring to the conversation. I’d experienced something similar a few months earlier with my Latin American and Spanish friends. We spoke English at first, because we were in Australia and it simply made more sense - but also because whenever they switched to Spanish, they were usually talking about cultural references I didn’t have access to. My attempts to join in felt awkward to me, and probably just as awkward to them. What this leaves me with is the feeling that fluency isn’t about how much you can produce, but about whether the language feels like a place you can exist in comfortably. And once you see it that way, the urge to impress people with it mostly disappears.
-
Learning a language is like moving houses
Jan 11 ⎯ Learning a foreign language often feels like moving houses, except no one warns you how many emotional phases are involved, and there is no checklist that actually applies, no method that truly fits your linguistic profile. At first, it’s pure excitement. You haven’t moved in yet, but you’re already mentally hosting dinner parties. You walk through empty rooms in your imagination, assigning them functions with absolute confidence and zero prior measurements. This will be the living room. This will be me when I speak Swahili. The language looks like a room: open, bright, full of potential. You don’t know where the switches are, or whether the water pressure works, but that feels like a detail you’ll handle later. Enthusiasm is high, and you’re eager to start packing. Then packing begins. Or maybe you misread panicking. The words are close for a reason. Suddenly, everything you own is your problem. You’re sorting through drawers asking deeply philosophical questions like why do I have this and will I ever need this again. In language terms, this is when you start wondering what you use every day in your native language, and what you never do, so you begin bargaining with yourself: I don’t care about farm animals, so I don’t need to learn them. The subjunctive is too complicated, so I’ll just find ways around it. You realise how much you already have in your native language, and how little of it transfers cleanly. Everything needs to be boxed first. Everything needs a label. That already feels daunting, but the world has been informed that you are moving into this new house and that you will soon be hosting those kick-ass dinner parties in Swahili. So you can’t give up. And of course, you shouldn’t give up. The moving truck has been ordered already. By the time you’re ready to leave, you’re tired but far too invested to quit. Your new place isn’t ready, and you’re living among stacks of things that technically belong to you but are completely unusable. This is the linguistic no-man’s-land: you know it’s time to stop relying on your native language because it’s holding you back, but you can’t express yourself in the new one just yet. You are linguistically homeless, surrounded by structures and rules you know are in the boxes somewhere, but you don’t know exactly where. Then the moving truck arrives. Instant relief. Something external is finally happening. Things are being lifted. Progress looks visible. This is often when comprehension suddenly improves, and you remember why you decided to move in the first place. Everything feels promising again. You start recognising patterns, understanding more than you expected, and you even catch yourself thinking or dreaming in your new language. Simple thoughts, sure, but thoughts nonetheless. You think, Yeah, I can see my new home. I’m half-way there. You are not remotely half-way there. You’re now standing in the new place, surrounded by boxes, none of which contain what you urgently need. You don’t know where to start. Every decision feels monumental. Kitchen or bedroom? Vocabulary or grammar? Pronunciation drills or syntax you’re not mentally ready for? You open one box, get distracted by another, and then somehow end up watching videos about how to fastest and most effective way to unpack, instead of unpacking anything at all. And you can stay surrounded by half-opened boxes for a very long time. Long enough to forget what “finished” is supposed to look like. Long enough to feel stuck despite being surrounded by everything you need. When it comes to your target language, you don’t actually need to add anything new at this point, well, at least not to make it through those dinner parties with dignity intact. You already have the material in your boxes. It just isn’t organised. Words don’t talk to each other yet. Sounds haven’t settled, and structures keep collapsing like badly assembled shelves. The task now is not learning more, but making what you already have coherent. You actually have to unpack now, rearrange your place, and finally get rid of those cardboard boxes. And then, slowly, without ceremony, things begin to work. A chair gets assembled. A light turns on. You find your toothbrush. The space becomes livable not because you acquired anything new, but because what you already had finally found its place. The language works the same way. Not at the moment of arrival, not when the truck pulls up, but after the long, slightly chaotic work of unboxing, rearranging, reassembling, and accepting that this, too, was part of the move.
-
The Cost of Speaking Too Early
Jan 10 ⎯ This is it. You have decided, at last, to learn that language, for reasons that made perfect sense at the time and are now slightly blurry. You have been diligent. You followed the first lessons in the book. You did some of the grammar drills. But more importantly, you are on an impressive Duolingo streak, which is basically fluency gamified. You are, by any reasonable metric, doing great. Even more crucially, you have added yet another A1 flag to your Reddit profile. Identity has been updated. The world has been informed. After watching approximately a gazillion beginner-friendly videos—each one promising you will “LEARN [insert language] IN 30 MINUTES”—you start catching words you vaguely remember seeing once. Progress feels alarmingly fast. Clearly, you are absorbing the language at an unprecedented rate. You know more than you did yesterday, far more than you did last week, and it doesn’t feel that hard just yet. Your motivation is high, which keeps you in the flow. You got it, you genius! You happily spend at least an hour a day studying, probably even more because you are really into it, and frankly, time flies by. You feel confident that all these efforts will soon pay off and that you’ll be able to have short conversations that will justify all that hard work. Well I hate to break it to you, but you are in fact on the rising curve of the Dunning-Kruger effect, and soon you’ll be collapsing. But for now, you aren’t aware of it and you feel that nothing can stop you from there. Words come to your mind quickly, you can already conjugate in the present, probably even in the past and future tenses, and when you throw in some adjectives and adverbs here and there, you feel invincible. Grammar bows before you; syntax fears you. Surely this is is easy-peasy and you are so bright that you will nail that language in no time. Why most people spend years learning their target languages seems like a mystery. You are standing right at the peak of Mount Stupid (not my term), very confident in your ability to progress exponentially. After all, why shouldn’t your habits continue to pay off? And this is when you begin to gradually fall off the top of the hill. (Told ya.) You’ve been so eager to put into practice everything you’ve learned so far - to communicate and/or to be praised (usually both)- that you have sped up the natural pace your brain needs to form and solidify those neural connections. (Turns out the brain does not respond well to being pressured). In doing so, you have already heard yourself make a bunch of mistakes with approximate pronunciation, excused yourself for lacking vocabulary, and tripped over grammar in the broadest sense of the term. And along the way, those early mistakes can settle in, becoming habits that are hard to undo later; a process known as fossilization. In short, you have tried to run before you could walk, and now you are beginning to be afraid to toddle. And unless you have tried to talk with a tutor you were paying, or with a loved one who doesn’t mind the lack of substance of your exchanges (bless them), you have most likely bored everyone to death… if they haven’t already tried to switch to English, or simply disengaged altogether if what you’re learning is English. Two reflexes seem especially hard to shake in language learning. One is the impulse to rush ahead before the ground is ready, fueled by the brief exhilaration of hearing oneself produce words in a language that still feels new. A linguistic sugar rush, really. The other grows right out of it: a kind of attachment to that rush, to the pleasure of being heard, noticed, sometimes even praised, instead of staying with the quieter work of listening to how native speakers actually speak, and paying attention to what they are saying rather than to the sound of one’s own voice. I know, that may sound a bit controversial, and mildly unflattering, but that is the feeling I get from so-called polyglots, both online and offline. Of course, it is motivating to know that progress is happening, that all that tedious work is not going to waste. It is also reasonable to want to check whether we are on the right path, and practicing a bit certainly helps with that, or at least in theory. Because real-life conversations, or even conversations with AI bots (something you can now do with apps like Langua), are high-stress environments. They force you to speak about something fairly specific, within a limited amount of time, to keep the exchange going. But linguistic ping-pong gets tiring fast when the ball keeps coming back faster than expected. When the language is still so new that it hasn’t settled into long-term memory, retrieving the right word, syntax, and pronunciation all at once becomes extremely difficult. It’s not even necessarily easy in one’s native language when asked to speak on a precise topic, which is essentially what language tutors make you do even in the early stages of learning, to give you a chance to make use of all the words and structures that you have learned. If stressful environments aren’t particularly kind to proficient speakers, you can easily predict the fiasco in the making when the learner knows very little of the language. As the linguist Stephen Krashen explained in his famous video on language acquisition in the 80s: “We acquire a language in one way and only one way, when we get more comprehensible input in a low anxiety environment.” And I will elaborate on the first part of his argument some other day. I keep coming back to how much modern language learning distrusts the natural phases required to acquire a new language. There is a rush toward speaking, toward being heard. Silence is treated as hesitation, or worse, as avoidance. If you are not producing anything, you must not be learning. But that doesn’t quite match how the mind behaves when it is actually absorbing patterns. Perception seems to need time on its own, without the pressure to perform. It rearranges itself quietly. As I discussed in another article, when you focus your energy on listening without feeling the urge to participate, intonation settles before words do. Rhythm arrives before accuracy. You notice how sentences breathe, where they tighten, where they loosen. You overhear conversations you don’t fully understand and still come away with a sense of how they moved. It’s a bit like listening to music from another room: the melody reaches you even when the lyrics don’t. Something registers anyway. You are not wasting your time. You are sowing the seeds for a strong tree, whose branches will be able to grow exponentially once the roots are sound. Children are allowed this. They listen for years, accumulating sound without being asked to demonstrate much. And they aren’t expected to, because they simply aren’t capable of using their vocal cords properly yet. Their early speech is sparse, sometimes clumsy, but it rests on a dense foundation of familiarity. Adults, by contrast, are pushed straight into output. The result is speech that appears quickly but carries very little weight behind it. The accent lingers, the rhythm resists the natural flow. Sentences feel assembled rather than grown, and roots fail to take hold the way they should. I don’t think this is a failure of effort. It’s more a question of reference. Without having heard enough, correction floats in the air. You’re told something is off -but off from what, exactly? The language has no internal anchor yet. Listening provides that anchor slowly, almost imperceptibly. Patterns repeat. Structures reappear. At some point, you stop noticing them consciously, which is usually when they start to work. Listening a lot instead of speaking right away is like preparing for an exam rather than winging it and hoping for the best. Extended listening and reading do something strange to time. You don’t feel productive while you’re doing them, but later you realize that expressions and words come to mind before you actively search for them. Because you just know them. Somehow. For some reason. You anticipate turns of phrase. You recognize what is likely to come next. Speech, when it finally shows up, feels less like construction and more like recognition, as if you were stepping into something already prepared. The first time I ever spoke English in a real-life situation, I was a few months shy of turning seventeen. My first class had been when I was around six, but aside from learning colours, animals, vegetables, and a few isolated words like window, I wouldn’t say I really learned anything until I was around eleven, when I had more formal classes three to four hours a week. Back then, there was almost no opportunity to practise or even listen to English. The internet as we know it didn’t exist, and films and series were only available on TV and were dubbed. So I read, and read, and read everything I could find - from whatever I could get my hands on online to newspapers made for young English learners. I devoted a lot of time to consuming English without worrying about whether I was wasting my time or not. I was doing it with pleasure, and to this day I don’t remember ever learning lists of vocabulary. I learned in context, through intense exposure. And more importantly, I never doubted that I was progressing. So when an elderly English couple asked me what time it was near a campsite, I answered with confidence. I knew, internally, that I could, even though I had never spoken to anyone “real” before. I remember distinctly that they congratulated me on my command of their language and my accent after the brief conversation that followed. That alone gave me all the motivation in the world to keep going. Within a week, I had made friends - mostly Dutch - and we had seamless conversations using words I had no idea where I had learned. So the takeaway I am trying to share here is simple: take your time. If you don’t, you may completely jeopardize your learning for years, as happened to me with another language. (I will write about it some day). I know that the slow approach doesn’t look impressive from the outside. It doesn’t reward quick wins. It offers very few visible milestones. It asks for presence without display, attention without immediate reward. But eventually, speaking does emerge - and it does so differently. Not urgently, not defensively. Sentences move with fewer interruptions. Pronunciation still needs work, of course, but it bends more easily. Grammar feels familiar, not because you can explain it, but because you’ve encountered it many times before. Pauses no longer signal confusion; they feel more like listening continuing inside the speech itself. There is always pressure to show progress, to prove that learning is happening. Silence makes people uneasy. But the brain seems unmoved by that discomfort. It keeps responding to repetition, to time spent close to the language, to the slow accumulation of sound and structure. The effects remain hidden until they don’t. None of this feels passive to me. It feels patient, which is something else entirely. Not doing nothing, just not rushing in and messing with it too soon. Letting the language settle where it settles, without poking it every five minutes to see if it’s working just yet. The work happens quietly, without proof, without anything you could really point to, which is mildly unsettling but also kind of the point. And when speech finally comes out, it carries this odd sense of recognition, as if the language had already been there for a while and you’re only now catching up.
-
What makes a text feel off, even when it's grammatically correct
Jan 10 ⎯ Something caught my attention last night, as I was inspecting the automatic translation Google made of my last article into my native French. An early 90s kid, I was raised during the toddler years of the internet, when little existed yet, and most of what did was in English. Back in the early 00s, translations were awkward, to say the least, and nobody would have suspected they would become so good one day, because languages can’t really be translated word for word - they are lived, and interpreted at best. In fact, this very observation encouraged me to take up languages early on, and I soon found myself browsing the early internet in the very limited English I had back then, just to be able to find actual information rather than weird translations of it. I never used Google to translate any page, because I didn’t really ever need to. But this writing platform allows me to automatically translate my articles into several languages, so I thought: of course, why not let native speakers of other languages read me? My Spanish is advanced, but not native-like, and I wouldn’t be able to judge the quality of an Italian text, let alone of all the other languages offered, so I could only go mistake-hunting in French. I read each sentence carefully as it unfolded, thinking to myself that I don’t know whether I would have been able to phrase them so accurately, because I have long since lost the habit of writing in my native language. Yes, the sentences unfolded almost suspiciously well: the grammar held, the vocabulary behaved. The text marched forward with suspicious discipline. Nothing jumped out as wrong, or even slightly crooked. Except… there was, at times, that faint sensation of displacement. Not enough to stop the reading, not strong enough to name immediately. Just a soft awareness that something in the phrasing didn’t quite belong where it had landed. Like guests who followed the instructions on the invitation, but still showed up overdressed. Or in pijamas. Either way. What struck me was that I couldn’t correct them in any obvious way. There was no error to fix, no rule to invoke. The problem, if it even deserves that word, was not one of correctness but of likelihood. These were sentences that could exist, but probably wouldn’t, for some reason. I was particularly attentive to this because, just an hour earlier, I had gone through a translation task for a gig, where I basically had to provide French equivalents for marketing sentences (what’s called “copywriting'“). I found myself noticing that, although such sentences were easily and directly translatable (a good half of the English vocabulary being derived from French and Latin), that’s just not really how we would say it. We don’t say “des termes et conditions s’appliquent” for “Terms and conditions apply” (notice how transparent the words are). We’d say “Offre soumise à conditions” (“offer subject-literally submitted-to conditions”). The first would be understood, but it would sound unnatural. The fact that these subtleties are well known is the reason why human translators are still (although less and less) in demand. Because industries localizing their content are still aware that machines aren’t (yet) able to interpret messages as well as natives do. What translators are doing right now is helping train the systems on the only thing they can’t figure out themselves, and that will eventually replace them. I’ve been thinking about this since I got up this morning. We tend to think of language in terms of permission: what is allowed, what is grammatical, what passes inspection. But living languages don’t really operate on permission alone. They run on habit, preference, repetition, avoidance. On things people say because they’ve heard them said a million times before, and on things they never say—not because they’re forbidden, but because no one ever quite reaches for them. I think that’s what I was hearing in my own text: phrases that had arrived by logic rather than by use. They made sense, they were even elegant in places, but they hadn’t passed through the soft filter of everyday speech. They hadn’t been worn down by mouths. This feeling comes back very clearly when watching dubbed films—which I never do, but sometimes overhear when I’m at someone’s place who doesn’t speak a second language. To me, dubbed movies can’t offer a truly immersive experience, unless we’re talking about cartoons, where voices are performed by expressive actors and the text doesn’t have to fit precisely the movements of the characters’ lips. No - movie voices sound slightly inflated, almost theatrical, yet still oddly flat, with a certain breathiness to them. There’s the sense that everyone is enunciating for a room that doesn’t exist. And then there’s the strange constraint hovering over every line, making it such a hard job for dubbing writers: the need to fit words to mouths, syllables to lips, timing to faces that were never meant to produce those sounds. And Brad Pitt ends up sounding uncanny, his talent as an actor compromised, because what he says—and the way he says it—just isn’t it. But even if you look past all of that—even if you generously suspend disbelief—the unease persists. What the characters say simply isn’t how people speak. Not because it’s grammatically wrong, but because it’s unfamiliar in a deeper way. The sentences feel imported. You can sense, almost physically, that they began life elsewhere, under different pressures, with a different tolerance for explicitness, for length, for rhythm. They’re sentences that survive translation but lose their social camouflage. You don’t hear them in real conversations. Not at the table, not in arguments, not in moments where people hesitate or overshare or choose the wrong word and live with it. They sound complete in a way everyday speech rarely is. That’s also why, in my opinion, dialogue and monologues in movies—spoken by actors in their native languages—often feel a little off as well. The lines are just too good. Too impactful. Too witty. Too long, at times. Too… too much. I remember how hard it was for me to understand movies in English without relying on those little wheels that are subtitles. I had passed my C1.2 level long ago and could speak with natives, across accents, with ease. But movies were still difficult to follow. And for the longest time, so couldn’t quite understand why. Until I did. And that was because, beyond the occasional cultural reference I was missing, the scripts were just too intense - for lack of a better word. They felt too performative, too clean, too cookie-cuttered to fit the situation wittily. To this day, even though watching movies has become a walk in the park (on the topic, I wonder how this expression will be translated!), I often think that films are the ultimate boss level of a language. You hear in movies the most intricate sentences that a regular person, with a regular brain, would never come up with in a real-life situation. But back to our initial topic. What really makes a translated message blur into something oddly translated is what linguistics calls non-idiomatic phrasing. It seems to live exactly there, in that narrow gap between sense and use. It doesn’t announce itself loudly, and it doesn’t even break comprehension. It just carries a quiet accent—technically correct, but phrased in a way a native speaker wouldn’t have chosen. You can smooth, adjust, nudge, but some part of it remains intuitive, resistant to explanation. You end up thinking, it’s correct, but there has to be another way to say it. Which leaves me wondering—though not conclusively—whether fluency is less about mastering rules and more about intuition carved through intense cultural exposure. And whether mastery in a foreign language is even achievable at all. About absorbing not just structures, but preferences. About learning, slowly but surely, which sentences a language seems to avoid, and trusting that avoidance as much as its rules. I’m not sure what it takes to sound unmistakably natural. But I suspect that this awkwardness, the kind that doesn’t break anything, is where languages reveal what they are most protective of. And having said that, I’ will (ironically) check my own text for unnatural phrasing before posting it. [Note: a few sentences were very slightly adjusted, which, incidentally, illustrates quite well the point I was trying to make!]
-
On pronunciation fossilization
Jan 09 ⎯ There is a moment, often easy to miss, when pronunciation stops feeling tentative and starts feeling settled, not because it has reached some ideal form, but because it has become familiar, almost backgrounded, like a piece of furniture you no longer notice even though you still navigate around it every day, and I find myself returning to that moment when I try to think about why poor pronunciation tends to fossilize rather than slowly soften with time and exposure. What is pronunciation fossilisation? It is the process by which a learner’s pronunciation patterns become stable and resistant to change, even after continued exposure to the language. In the beginning, things usually feel open, and sounds are approached with a certain attentiveness and curiosity, but fairly quickly the practical need or desire to communicate fast begins to dominate, and communication seems to reward speed and approximation far more consistently than caution and precision. So the body learns that being mildly understood is enough - and once that lesson has been absorbed, it does not easily loosen its grip. The mouth discovers ways of moving that work well enough, the ear relaxes its demands, and something narrows, not suddenly, but gradually. I often think of pronunciation less as a skill to be acquired than as a path that forms through use, the way grass slowly bends where people repeatedly walk across a field. At first, there are many possible ways of crossing, but one route becomes slightly easier, then slightly clearer, until eventually it no longer feels chosen at all. By that point, stepping elsewhere feels unnecessary, even a little uncomfortable, and poor pronunciation can begin to settle in this way. Not through neglect, but through repetition that quietly stabilises what was once, albeit for a very limited time, flexible. Listening is unfairly treated as a passive activity, when in reality it is a crucial stage of the learning process and should be treated as a cornerstone of one’s learning journey. Listening is shaped by expectation, and once certain sound categories and timing patterns have taken hold, they begin to guide what is noticed and what fades into the background. At that stage, a learner may still be listening attentively, but the ability to hear itself has changed, because the ears have been trained in a certain way. What reaches awareness is already filtered, already adjusted to fit familiar patterns, and imitation begins to mirror not the external language, but the internal version that has gradually formed. And there seems to be little to no way back from there, because what ends up happening is that you are trying to mimic sounds and rhythms that you can no longer perceive in the first place. Articulatory habits, once repeated thousands of times, tend to settle into muscle memory in much the same way posture does. Changing them later can feel less like learning something new and more like trying to alter how one stands or walks - an effort that requires sustained attention and often slips back the moment that attention softens. Not a task that is particularly enjoyable either, especially for those who are less interested in having an authentic pronunciation than in being able to converse quickly. (Is that where the line between introverts and extroverts is drawn?) This may be one reason pronunciation is so often addressed through explanation, as if understanding where the tongue should go might gently persuade it to move there under real conditions, even though explanation belongs to a different layer than execution, and the two do not always meet. Over time, it becomes possible to accumulate quite detailed knowledge about sounds without any corresponding change in how those sounds actually emerge in spontaneous speech, and this gap, once established, can begin to feel normal rather than troubling. Not to mention the abysmal lack of energy spent teaching prosody (the melody of a language), which really is the missing piece in the realm of pronunciation and which feels decidedly impossible to unlearn once poorly learnt - or perhaps awkwardly mimicked, for lack of formal instruction on the topic. (Prosody is one of my passions, so I’ll write more about it, because there is so much that needs to be addressed.) I also find myself wondering whether fossilisation has a social dimension, shaped by the moment when a speaker becomes recognisable through their accent. Once others begin to identify you through a particular sound pattern, that pattern acquires a kind of stability that goes beyond technique. It becomes part of how you are heard, and perhaps even how you hear yourself. Changing it can feel subtly disorienting, as though one were altering a long-established handwriting style or the way you cross your legs when sitting. I know that, by default, I can’t really help but speak like a little girl in Turkish, because I mainly learned it in the context of my past relationship, where sounding cute was still relatively acceptable for a 23–25-year-old, to the point where speaking like the adult I am (sigh) in Turkish still feels unnatural to this day. But back to our fossils. None of this seems to happen because learners are indifferent or resistant, and it does not feel accurate to frame fossilisation as a failure of motivation, since many people care deeply about pronunciation (I know I do) while remaining unable to shift it in meaningful ways. Care alone, however, does not appear to reopen pathways that have been reinforced through repeated use, especially when everyday communication continues to confirm that existing habits are sufficient. What I keep circling back to is the thought that poor pronunciation may persist because speaking is generally attempted too early, because it solves the immediate problem of being understood. Once a solution proves reliable, the nervous system seems inclined to preserve it rather than revise it. Improvement then appears to require something more than better input or clearer explanation, perhaps a temporary willingness to sound unstable again, to unsettle what has already settled. And that is not a demand most learning environments acknowledge explicitly. And I still wonder whether old habits can truly be unlearned and eradicated, and whether everyone is physically and cognitively capable of accurately imitating the sounds of a foreign language at all. It’s a topic I’ll take pleasure in exploring someday. I remain cautious about approaches that promise simple corrections (although I would most definitely like to work on an approach to unlock good prosody), but it feels important to notice that fossilisation is neither accidental nor mysterious, and that it emerges naturally from the interaction between perception, movement, repetition, and usefulness over time. Seen this way, it becomes harder to locate the problem solely in the learner, and easier to see it as a quiet consequence of how language learning often unfolds from the very beginning.
- accent
- fossilisation
-
The Power Of Predictive Listening
Jan 01 ⎯ When the brain becomes familiar with the sounds and rhythm of a language, listening starts to work ahead of time. Speech is no longer heard as a chain of separate sounds that need to be decoded one by one. Instead, it is followed as a movement that already points in a certain direction. The ear begins to expect what is likely to come next, guided by patterns that have settled through repeated exposure. Vowel harmony is a clear example of this process. In languages such as Turkish and Finnish, vowels within a word are not chosen freely. They follow consistent patterns based on features like frontness, backness, or rounding. To someone unfamiliar with these languages, the changes between vowels can seem random (or mildly unhinged). With enough exposure, they begin to feel natural and expected, and it takes less and less cognitive effort to understand and construct sentences built upon such phonetic principles. Your brain eventually stops protesting, lets it go and says a big “amen”. In Turkish, e is followed by e or i; a with a or ı; o with a or u; ö with ö or ü, as in the following examples: Gel-e-cek-tim and not Gel-u-cek-tum or Gel-a-cak-töm, and so on Al-a-cak-tım and not Al-ö-cak-tam Ol-u-r-um and not Ol-i-r-em Öl-ü-r-üm and not Öl-e-r-um How does that even work? At first, a listener simply notices variation. Over time, that variation becomes organised. Certain vowels start to suggest which vowels are likely to follow. A back vowel prepares the ear for a similar continuation, while a front vowel sets a different expectation. The listener no longer waits until the end of the word to recognise its shape. The word is anticipated as it unfolds. See and try to analyse: ev-ler-i-niz-den-miş-siniz büyü-t-ül-ü-yor-muş mu-y-dunuz? Those are one and two words, respectively. (I know, intense.) It takes a certain amount of time to get used to it (and a brief crying session or two), but once settled in the brain, it simply feels right this way, and wrong otherwise, which is half of what you need to know to be able to construct a very long word on the spot, as illustrated above. This shift happens through listening, not through memorising rules. Rules can describe vowel harmony, but they do not make it automatic. They mostly sit there looking important. What changes perception is repeated exposure. As words are heard again and again, harmony becomes part of the overall sound of the language. The brain does not apply a rule; it follows a pattern. The next vowel feels predictable before it is heard. Prosody strengthens this effect. Stress patterns and syllable timing offer additional clues about how words are built and how endings attach, like subtle signposts you didn’t know you were following all along. In Turkish, suffixes follow phonetic paths already established by harmony. In Finnish, case endings settle into place in the same way. The listener develops a sense of how a word will grow, based on familiarity with its sound rather than conscious analysis. As this familiarity increases, sentence formation becomes easier. Words stop feeling like separate units. They connect through shared sound patterns. Endings come more easily because their form has already been anticipated. Speaking follows listening. The speaker naturally reaches for forms that fit the sound environment already in place. This ability to anticipate extends beyond individual words. Vowel harmony contributes to rhythm and flow across longer stretches of speech. It helps the listener keep track of structure over time, like a subtle metronome ticking away in the background. Comprehension improves as well. When the brain expects certain vowel patterns, it can separate words more efficiently. Long forms are easier to follow, and fast or reduced speech becomes less difficult because expectation fills in the gaps. Vowel harmony shows how attention to sound reshapes how language is processed. What begins as simple exposure gradually becomes orientation. Prediction develops without effort. The brain learns to follow the internal logic of the language as it unfolds, guided by consistency in sound, which turns out to be more persuasive than explanation. With sustained listening, these patterns settle into memory and perception. Building sentences becomes less about assembling pieces and more about following familiar paths. Sound supports structure. Expectation guides expression. In this way, attention to phonetics and rhythm helps understanding and speaking develop together, carried by patterns the ear has learned to recognise and trust. With enough exposure, the brain stops decoding step by step and starts predicting what is likely to come next, because the language has recurring constraints and habits that “pull” speech in certain directions. To move away from the vowel harmony illustration, let’s consider patterns that you have almost certainly never consciously thought about in English, yet have probably absorbed intuitively. You can have words starting with str, but not srt: street, strong, strike. This is what is called phonotactics, and many speakers of languages that do not allow the same groups of sounds have a hard time with such clusters and pronounce them as “estr” (Spanish speakers) or, exaggeratingly, “soturu” (Japanese speakers). In English, a noun and a verb can often be differentiated by stress placement: on the first syllable for nouns and on the second for verbs. Compare: a project and to project; a comment and to comment. Granted, this is something foreign learners may struggle to notice without guidance, but as a native speaker, you most likely sense where to put the stress, because it simply feels intuitive and right. Along the same lines, Russian speakers—both native and non-native—quickly notice when o is pronounced as “a” (when unstressed) and as “o” (when stressed). So хорошо will be pronounced “kharasho” and not “khorosho”, and водка as “vodka” and not “vadka”. The English expression “I am looking forward to” will be followed by a verb in the gerund (-ing), and German temporal adverbs like “Morgens” will push the verb before the subject, as in “Morgens gehe ich” and not “Morgens ich gehe”. Once the rule is known, it becomes unnatural to say it otherwise. Verb conjugations and noun declensions in case-marking languages follow this predictive logic too. Without it, intuitive conjugation would be impossible, and every form would have to be memorised individually. You know that -AR verbs in Spanish follow the “o, as, a, amos, áis, an” pattern, with virtually no exceptions. And that to form the subjunctive, you simply replace a with e: “e, es, e, emos, éis, en”. Semitic languages like Arabic and Hebrew rely almost exclusively on patterns. It takes considerable training to intuitively predict how words will change, but with sustained exposure and practice, it becomes relatively natural. For instance, the root k-t-v in Hebrew relates to writing, and from it emerge words like kotev, katav, ktiva, and mikhtav (compare with the Arabic root k-t-b, as in kitāb, “book”). You may not be aware of it, but everything you say in your native language, and in the languages you have learned, relies on such patterns, whether or not you consciously recognise or acknowledge them. Success in speaking a foreign language largely depends on making these principles automatic, in order to lower the cognitive load required to produce longer and more complex sentences without effort. And your brain can finally stop micromanaging every syllable like an overcaffeinated supervisor.