7 ms·

The trouble is the near-impossibility of controlled studies. Has there EVER BEEN a study where adults not only had immersion but went through the child's experi

The trouble is the near-impossibility of controlled studies. Has there EVER BEEN a study where adults not only had immersion but went through the child's experience of having some parent-figures and teachers care for all their needs while constantly correcting and helping them with every bit of language and reading picture books or otherwise engaging with the most basic nouns and verbs (perhaps otherwise playing and engaging with mainly peers who are going through the exact same process) over about 2-3 years before being expected to grasp any really abstract concepts??? I'm pretty damn certain that I could go anywhere in the world and learn all the names for stuff in a house and basic actions and syntax in any language if I got to role-play as an 18-month-old 24-hours a day for a year, and I could go on to have 4-year-old near-native ability (even in pronunciation) in just 2-3 years and on and on. Except there's no way I could ever get that experience and unlikely I would ever commit to it anyway. The only people who could ever even approach this are already outliers in massive ways. But it's always seemed to me that people are crazily overconfident about what we know about child vs adult learning.

gridlockd 7y ago
An 18-month old isn't just learning a language, they're acquiring language itself. At four-years-old, they aren't at "native speaker" level either. A few years later, they acquire literacy, if taught. It's well-established that acquiring literacy at an older is more difficult. Young brains have higher plasticity. Is it impossible to acquire literacy as an adult? Of course not. But it's far more difficult, to the point of being practically impossible for many people. In context, the fact that 1000 out of several hundreds of thousands of samples managed to do something highly unlikely isn't that remarkable. It doesn't support the headline at all.
- cletus 7y ago
  While I agree with you about plasticity there's a pretty good counterexample to your claim about literacy: in 1929 Turkey switched from the Arabic alphabet, which never quite fit the Turkish language phonetically, to a Roman script variant that they still use today. This was largely credited to the massive increase in literacy over the next decade (random searches suggest it went from 5-10% to 80%). Consider in comparison SE Asian languages (eg Mandarin) that are _incredibly_ difficult to learn to read let alone write. I learned some German years ago and the one thing I appreciated was that it is phonetic. You can probably teach someone to pronounce German (badly) in a day or two. Likewise if you hear a word you can probably spell it with very minor variations.
  - gridlockd 7y ago
    > While I agree with you about plasticity there's a pretty good counterexample to your claim about literacy: in 1929 Turkey switched from the Arabic alphabet, which never quite fit the Turkish language phonetically, to a Roman script variant that they still use today. This was largely credited to the massive increase in literacy over the next decade (random searches suggest it went from 5-10% to 80%). Claims about the success of the Ataturk's reforms are probably about as credible as Soviet claims about worker productivity. Other sources claim a literacy rate of 68% in 1975: https://data.worldbank.org/indicator/SE.ADT.LITR.ZS?end=2016&start=1970 https://data.worldbank.org/indicator/SE.ADT.LITR.ZS?end=2016... Of course it can't be denied that school reforms had a massive impact on literacy, but it's not so clear how previously illiterate adults fared when learning to read and write for the first time.
  - CydeWeys 7y ago
    > You can probably teach someone to pronounce German (badly) in a day or two. Likewise if you hear a word you can probably spell it with very minor variations. If they're already familiar with the Latin alphabet, anyway (which admittedly is a good chunk of the world at this point; even if you're, say, Japanese or Korean, you know the Latin script because it's used pervasively). If not, then just learning the characters is gonna take a bit.
    - sumedh 7y ago
      > If not, then just learning the characters is gonna take a bit. It would still be easier than learning 1000+ characters in Mandarin.
  - Mvandenbergh 7y ago
    At the same time though, the new Turkish state pushed universal literacy as a policy goal and invested enormously in it in a way that the Ottoman empire didn't at all. It's very difficult to extract from data of that period (which may also have been manipulated for political purposes) how much of that change as due to the script vs other factors. While I don't think it's controversial that phonetically written languages are easier to learn at first, nor that the Turkish Ottoman alphabet based on Arabic was not ideal for writing Turkish, I don't think it was uniquely difficult. I think it's about as difficult as writing English correctly using our alphabet, which is to say it is harder than Spanish but much easier than writing Chinese languages. Note that literacy rates in Arabic speaking parts of the Ottoman empire - most of it - where the script and the language matched was also low. You can make the argument that Arabic diglossia was responsible for some of that but still)
3131s 7y ago
Look up the "conditioned head turn procedure" as a means of testing phonetic perception in infants. It is pretty well established that infants have a much greater ability to distinguish phonemes compared to adults. What you said is true the other way around too. Infants can't look something up in a dictionary, or ask a bilingual person to clarify the finer points of a word or phrase in a language they already know. Infants aren't just learning the language but also the concepts behind the language, which would seem to make what they're doing more exceptional than adult language learning.
- gerbilly 7y ago
  > Infants aren't just learning the language but also the concepts behind the language Some people say that children don't just learn grammar, but that they sometimes also create the grammar rules of emerging languages through a process called creolization.
- quadrangle 7y ago
  I don't doubt those differences, but as someone with a music background and an awareness of sound (and having seen others pick up accents and phonemes much less readily than myself), I still see so much we can't control for. Maybe nobody has succeeded at getting adults as close as they can get to the infant state. How do we know the difference between adults being incapable of hearing phonemes similarly to infants versus failure to express to the adults what state they need to be in to listen? Obviously, most adults are filled with thoughts and cultural interpretations of any psychological study context and don't have the same open mind as an infant. The differences are stark in practice between adult and infant learning and behavior. But what are the underlying factors that create the distinctions? I think we have very little practical capacity to tease them apart.
mikekchar 7y ago
I live in Japan and speak Japanese reasonably well (not native level by any stretch of the imagination, but I can have casual conversations in virtually any normal context). I know a lot of expats who live near me and just can't seem to learn Japanese. One of the biggest things I noticed they had in common was a desire to avoid being treated like a child. My level is like that of an 8 or 10 year old (though I lack fluency is some normal grammar which most children have no trouble with). I'm very comfortable acting like an 8 or 10 year old. When I go to the doctor for example, I don't expect to understand what he's saying. The "explain it to me like I'm five" thing is my ordinary lifestyle. But I find that some people just can't accept the step down. If they are 30 years old, they want to go out with native speakers and interact as a 30 year old. Then they get frustrated and humiliated when they can't understand what's going on. They want to watch the news on TV and get frustrated when they can't understand any of it. They refuse to watch things (like cartoons) that they can understand, because it is too childish. It's a bit of a blocker, I think. Admittedly, though, it's frustrating that people always have to make an effort with you. If you go out with a group of native speakers to a bar for an evening, there's a good chance that the only conversation you understand is when someone makes a special effort to speak to you. There is nobody in your peer group that has the same communication level as you, so you can get really lonely. I always had the strategy of coming up with one-liner jokes. Just sit back and listen and listen and listen until you finally understand something and then think of a quick funny, but irrelevant thing to insert in the conversation. It's doubly funny because everybody thinks you have no idea what's going on. But that was my thing for years -- I never said anything meaningful in a conversation in a group. Just like a kid... But all that aside, it's actually pretty easy to show that adults can learn languages as fast as children. Children acquire about 3 "word families" per day on average from the age of 1.5 up to 15 (and continue as long as they continue formal education). A "word family" is a bit like a piece of vocabulary, but includes all the inflections and combination words (so if you know "police" and "station", then you also know "police station" -- it's not counted as a new piece of vocabulary). This works out to about 1000 word families a year (although the rate is very variable between the ages of 1.5 and 5 -- the average child ending up with 5000 word families by the age of 5). That kind of stuff is very well understood. You can learn faster than that as an adult. It's not even very hard. The trick is understanding that it's going to take up to 5 years to speak like a 5 year old. Most people forget what it's like to be 5 and how crappy your language skills are. They think, "Oh these 5 year olds can play easily together and jabber on constantly" -- but put them in an adult environment and the vast majority will clam up or be just as frustrated as we are.

gridlockd 7y ago

An 18-month old isn't just learning a language, they're acquiring language itself. At four-years-old, they aren't at "native speaker" level either. A few years later, they acquire literacy, if taught. It's well-established that acquiring literacy at an older is more difficult. Young brains have higher plasticity. Is it impossible to acquire literacy as an adult? Of course not. But it's far more difficult, to the point of being practically impossible for many people. In context, the fact that 1000 out of several hundreds of thousands of samples managed to do something highly unlikely isn't that remarkable. It doesn't support the headline at all.

cletus 7y ago
While I agree with you about plasticity there's a pretty good counterexample to your claim about literacy: in 1929 Turkey switched from the Arabic alphabet, which never quite fit the Turkish language phonetically, to a Roman script variant that they still use today. This was largely credited to the massive increase in literacy over the next decade (random searches suggest it went from 5-10% to 80%). Consider in comparison SE Asian languages (eg Mandarin) that are _incredibly_ difficult to learn to read let alone write. I learned some German years ago and the one thing I appreciated was that it is phonetic. You can probably teach someone to pronounce German (badly) in a day or two. Likewise if you hear a word you can probably spell it with very minor variations.
- gridlockd 7y ago
  > While I agree with you about plasticity there's a pretty good counterexample to your claim about literacy: in 1929 Turkey switched from the Arabic alphabet, which never quite fit the Turkish language phonetically, to a Roman script variant that they still use today. This was largely credited to the massive increase in literacy over the next decade (random searches suggest it went from 5-10% to 80%). Claims about the success of the Ataturk's reforms are probably about as credible as Soviet claims about worker productivity. Other sources claim a literacy rate of 68% in 1975: https://data.worldbank.org/indicator/SE.ADT.LITR.ZS?end=2016&start=1970 https://data.worldbank.org/indicator/SE.ADT.LITR.ZS?end=2016... Of course it can't be denied that school reforms had a massive impact on literacy, but it's not so clear how previously illiterate adults fared when learning to read and write for the first time.
- CydeWeys 7y ago
  > You can probably teach someone to pronounce German (badly) in a day or two. Likewise if you hear a word you can probably spell it with very minor variations. If they're already familiar with the Latin alphabet, anyway (which admittedly is a good chunk of the world at this point; even if you're, say, Japanese or Korean, you know the Latin script because it's used pervasively). If not, then just learning the characters is gonna take a bit.
  - sumedh 7y ago
    > If not, then just learning the characters is gonna take a bit. It would still be easier than learning 1000+ characters in Mandarin.
- Mvandenbergh 7y ago
  At the same time though, the new Turkish state pushed universal literacy as a policy goal and invested enormously in it in a way that the Ottoman empire didn't at all. It's very difficult to extract from data of that period (which may also have been manipulated for political purposes) how much of that change as due to the script vs other factors. While I don't think it's controversial that phonetically written languages are easier to learn at first, nor that the Turkish Ottoman alphabet based on Arabic was not ideal for writing Turkish, I don't think it was uniquely difficult. I think it's about as difficult as writing English correctly using our alphabet, which is to say it is harder than Spanish but much easier than writing Chinese languages. Note that literacy rates in Arabic speaking parts of the Ottoman empire - most of it - where the script and the language matched was also low. You can make the argument that Arabic diglossia was responsible for some of that but still)

3131s 7y ago

Look up the "conditioned head turn procedure" as a means of testing phonetic perception in infants. It is pretty well established that infants have a much greater ability to distinguish phonemes compared to adults. What you said is true the other way around too. Infants can't look something up in a dictionary, or ask a bilingual person to clarify the finer points of a word or phrase in a language they already know. Infants aren't just learning the language but also the concepts behind the language, which would seem to make what they're doing more exceptional than adult language learning.

gerbilly 7y ago
> Infants aren't just learning the language but also the concepts behind the language Some people say that children don't just learn grammar, but that they sometimes also create the grammar rules of emerging languages through a process called creolization.
quadrangle 7y ago
I don't doubt those differences, but as someone with a music background and an awareness of sound (and having seen others pick up accents and phonemes much less readily than myself), I still see so much we can't control for. Maybe nobody has succeeded at getting adults as close as they can get to the infant state. How do we know the difference between adults being incapable of hearing phonemes similarly to infants versus failure to express to the adults what state they need to be in to listen? Obviously, most adults are filled with thoughts and cultural interpretations of any psychological study context and don't have the same open mind as an infant. The differences are stark in practice between adult and infant learning and behavior. But what are the underlying factors that create the distinctions? I think we have very little practical capacity to tease them apart.

mikekchar 7y ago

I live in Japan and speak Japanese reasonably well (not native level by any stretch of the imagination, but I can have casual conversations in virtually any normal context). I know a lot of expats who live near me and just can't seem to learn Japanese. One of the biggest things I noticed they had in common was a desire to avoid being treated like a child. My level is like that of an 8 or 10 year old (though I lack fluency is some normal grammar which most children have no trouble with). I'm very comfortable acting like an 8 or 10 year old. When I go to the doctor for example, I don't expect to understand what he's saying. The "explain it to me like I'm five" thing is my ordinary lifestyle. But I find that some people just can't accept the step down. If they are 30 years old, they want to go out with native speakers and interact as a 30 year old. Then they get frustrated and humiliated when they can't understand what's going on. They want to watch the news on TV and get frustrated when they can't understand any of it. They refuse to watch things (like cartoons) that they can understand, because it is too childish. It's a bit of a blocker, I think. Admittedly, though, it's frustrating that people always have to make an effort with you. If you go out with a group of native speakers to a bar for an evening, there's a good chance that the only conversation you understand is when someone makes a special effort to speak to you. There is nobody in your peer group that has the same communication level as you, so you can get really lonely. I always had the strategy of coming up with one-liner jokes. Just sit back and listen and listen and listen until you finally understand something and then think of a quick funny, but irrelevant thing to insert in the conversation. It's doubly funny because everybody thinks you have no idea what's going on. But that was my thing for years -- I never said anything meaningful in a conversation in a group. Just like a kid... But all that aside, it's actually pretty easy to show that adults can learn languages as fast as children. Children acquire about 3 "word families" per day on average from the age of 1.5 up to 15 (and continue as long as they continue formal education). A "word family" is a bit like a piece of vocabulary, but includes all the inflections and combination words (so if you know "police" and "station", then you also know "police station" -- it's not counted as a new piece of vocabulary). This works out to about 1000 word families a year (although the rate is very variable between the ages of 1.5 and 5 -- the average child ending up with 5000 word families by the age of 5). That kind of stuff is very well understood. You can learn faster than that as an adult. It's not even very hard. The trick is understanding that it's going to take up to 5 years to speak like a 5 year old. Most people forget what it's like to be 5 and how crappy your language skills are. They think, "Oh these 5 year olds can play easily together and jabber on constantly" -- but put them in an adult environment and the vast majority will clam up or be just as frustrated as we are.