Input vs output in language learning (with Adele Goldberg)

in the world of linguistics there are very few people with a resume as impressive as Adele Goldberg she has had a major influence on the way that we understand and study language in this very special interview we talk about vocabulary learning input versus output and her language acquisition research this is an edited version of our interview if you would like to listen to the full version you will find a link down in the description box I hope you enjoy it adele goldberg thank you very much for taking the time to talk to me today well thanks for having me the first question i wanted to ask you is actually about where you work because although you're a linguist and a lot of your work is related to linguistics you're actually in the department of psychology and could you explain why that is i yeah i I actually find findings in psychology are hugely important to understanding language and how it's learned and how its processed linguists have always cared about psychology and principle so when modern linguist was linguistics was created people talked in terms of it being part of cognitive science and these days more and more people are really fulfilling that promise by studying psychology and language I also wanted to ask you actually about something which you replied to me in the first email I sent you you said thank you very much for spreading the word that language depends on context and communication and and actually I thought it was kind of a little bit of a strange thing to say because what other way is there a viewing language if it's not about context in the communication well now you're opening Pandora's box in the world of linguistics so there's been a big chasm in the field between those of us who want to emphasize meaning and communication and other linguists who follow a more Chomsky and perspective which has held sway in some departments since the 1950s or 60s and they argue that it's advantageous to study language in purely formal terms purely in terms of of syntactic properties and remove it from communicative situations I know that a lot of your your your work especially your recent work has been to do with vocabulary learning and especially vocabulary learning in babies and children and so what what does the science sort of tell us about what do we know about the way that babies and children alone vocabulary well that's a really good question we know surprisingly little so it's clearly in principle a challenging task to learn words because babies hear a continuous stream of sounds and they have to attach meanings to these two at first they have to break up the speech stream into units and then they have to assign meaning and often there are multiple meanings for each word but it's clear that children are actually very good at this task and they learn not only individual meanings but multiple meanings for words relatively quickly so by the time there are three or four they're learning nine new words a day so in one study we did that you're alluding to this was run by my graduate student Sammy Floyd she looked at toddlers they were two-year-olds and she was looking at whether they recognized multiple familiar meanings for English words so did they recognize for example that the word cap could either refer to a baseball cap or a pen cap and she did that by following their eye gaze when the pictures of one of those meanings on a foil were put on the screen and children did reliably look at both of those individual meanings for the English words and then remarkably she also tested possible potential that didn't exist in English so in English if you look at a Tupperware container and you pull off the top that's not a cap that we would call that a lid right but in Spanish they would use the same word tapa for this and the babies when they heard the word cap and one of the choices was a lid or a foil a different object they also looked at the at the lid indicating that they used the meanings that they knew to extend to extend what they understood cap to mean in English but there was a difference in so so they were more they looked more clearly at the two english meanings then at the related Spanish meaning and that seems to indicate that they remembered the English meanings they remember both of them um the Spanish meaning isn't just the same but they're able to get there from the English meanings so it shows that we have this strong memory for what we've experienced even very young children at - they're know they're learning multiple meanings for words and we're able to spontaneously extend these meanings wow it's pretty amazing and does it tell us something about maybe vocabulary learning in general in the sense you know if I if I'm an adult and I want to learn and you were should I try to learn the word in context or you know in isolation or should I maybe somehow try and relate it to a category I already have it in my mind great question so absolutely you want to learn it in context so we all know cases where we've we've tried to learn vocabulary using flashcards but then we have no idea how to use those words and the the context is usually very very important right so there's subtle differences if you just learned for example that the word bachelor means unmarried man which it does you might incorrectly decide that the Pope is a bachelor right with which isn't appropriate because you have to know how the word is used and then you're right attaching the new meaning to it familiar meaning is usually helpful and we usually do that so when we have new when we encounter new entities like when the computer was invented and we had to decide what to call the files and the garb the trash and the folders we used familiar words that we already had we just extended those meanings to the non-physical element and that helps on both speakers because you already know how to say those words and you know what they mean and it also helps listeners because they can figure out what's meaning what's intended flash cards type learning systems and also especially like spaced repetition systems are enormous ly popular amongst a lot of second language learners and personally I don't understand why because I can't think of anything more boring myself and and there's kind of an obsession there's an obsession with you know if I get 10,000 words I'll have fluency yeah and right so I don't think that can work because you have to know how to eat what context to use the words in and it you know so native speakers know which words go together and if you ever want to sound like a native speaker you have to learn that so we talked about arousing suspicions we don't talk about rising suspicions if we would know what it means it would mean the same thing but it just doesn't sound right but if you learned each word independently you wouldn't know that these words go together and those words don't so I don't think there's any way around ultimately if you want to sound like a native speaker to learn language in context so with we've talked a little bit about our that kind of input so about you know it's important to take in language you know in in context and and actually what's really interesting is a lot of pretty much everybody that I've spoken to including my polyglots and and other other kind of people who work in the world of English learning they always advocate very strongly for huge amounts of input like input is is everything if you want to learn a language but I'm wondering what's your sort of opinion on on also the importance of output of if expression of the things that you know well that is important and we know it is important because there are lots of children who grow up with around parents who know another language and the parents speak to each other and the child understands but those children often say that they're unable to speak because they haven't practiced so it would it is easier for them to learn to speak than it is for a child who doesn't understand the language but there is another hurdle there is another piece to the puzzle in learning to use it at the same time there is some controversy about learning about using a language too quickly because it's possible that the mistakes you make as an early learner will be reinforced by your own repetition of those errors so you know in in first language acquisition when you're learning your native language you don't actually say much for the first you know two and a half years so there's a lot of listening before you're producing much and it's possible that that plays some advantageous role yeah and and also when when you do start to speak you you make a lot of mistakes right I mean I have a son who's five years old and even now he some of the irregular verbs he will use the e D ending in the past like he'll say I go to the park for example but it seems that they these things seem to they go away without without me correcting them oh absolutely absolutely and there have been studies done that show that overt Corrections aren't needed so some parents do correct their children over at Lee but a lot of them don't and that makes sense right if your son said me loves you daddy you're not likely to correct his English you're more likely to give him a hug right and so we all respond to the content more than the form and and but yet we do learn to speak like those around us ultimately those the the over generalization like using saying goad it follows a pattern in English we're all of like any new word would get that ending for the past tense so those mistakes are indications that your son has recognized implicitly that generalization one thing that I wonder is well let me let me change the example slightly to a case where meaning is important so young kids also over generalized word meanings so young kids will often refer to any animal as a dog and but if you show them a picture of say a cow and a dog and ask them which one is the dog they reliably get it right and so that seems to indicate that they know what dog means but they don't have a better word at the moment of speaking and I wonder this hasn't been tested but I suspect that your son if you ask him which sounds better I went to the store or I go to the store I bet he might know that I went to the store is better but at the moment of speaking gold was more accessible well and this is something that's kind of surprised me in the limited reading I've done about about human memory is that actually humans human memory is pretty incredible right like we underestimate our ability to subconsciously absorb things is that fair oh it's so true and people for decades underappreciated that especially in language but now we know that we have this vast memory where it's not that we can recall things that will I for one have a terrible ability to remember people's names or remember where my shoes are but implicitly so there was an experiment done where people had to look at a list of 2,500 different images of random objects and then then they were shown after looking at these images for about 6 seconds each over a period of hours they were shown the same objects in a different position and out and also other kinds of objects that they hadn't seen and they were asked did you see this image before this exact image and people were accurate at at identifying the exact image they'd seen as opposed to say a briefcase that's closed versus a briefcase that's open or half of a cantaloupe versus a quarter of a cantaloupe at a level of 90% people were amazingly accurate yeah and and we've done some studies in our own lab showing that verbatim memory for language is also very strong and so it's not that if I ask you repeat the sentence I just said back to me that you could do it but if forced to identify it did I say this or did I say this you would be way above chance in getting that right interesting and this actually sort of brings me on to the next you want your next piece of work which is non-native speakers do not take competing alternative expressions into account could you sort of just explain that that work a little bit yeah sure so um this is related to the idea that for native speakers oh well yeah for native speakers we don't generally get corrected for mistakes we make and so the question is how do we ever learn to avoid them how does your son ever learn to use went systematically instead of God and the answer seems to be that you you he might anticipate God but he's going to systematically hear went in that context that is went and goat compete compete with one another and this idea that you might expect something but here's something else and that that process of getting an error will help you shape your your language so I I'm convinced that this is exactly how native speakers learn to avoid things that make sense but native speakers just don't say so things like explain me this that's a title of a book I just published and it sounds odd to native speakers even more odd when I say I'm going to explain you this but how do we know that well I would say the reason we know that is because we've we might have expected to hear explain you this there context where it's appropriate and we know what it meet what mean but instead we systematically here explain this to me so what we did with non-native speakers was we looked to see whether they seem to seem to take into account whether there exists a better alternative so we all say novel things sometimes and and we know they're novel but they're they're still acceptable so you know the crumbling asst of this cookie for example we might not have heard crumble enos before but we know it's fine and so it's not novelty in itself that makes something unacceptable it's I think that we have a better way to say it but non-native speakers in in the tests that we've done except at the highest levels of proficiency they seem to have trouble taking into account whether there is a better way to say something so they know what they've heard they know what sounds familiar they do know when something sounds novel but they're not treating the things that have a better alternative a conventional way to say those things as as less acceptable so and in fact they do tend to say explain me this and and and believe that that's reasonably acceptable whereas me don't like it and but how much of that can be explained by it maybe interference from their native language well that's a really good question and we we actually were able to look at that in the in one of these newer studies and that is so we tested two different constructions and in one when was the this construction like explained me this and another was that he forced that she leave and it turns out that Spanish makes exactly the same contrast that English does in the forced cases but they don't make the difference between explain and give the way English does so they have one distinction and they don't have the other but it was sort of a surprise to us that have played no role they were more generous in both cases that is transfer effects or interference from your first language or your more dominant language no doubt play a role they absolutely play a role but I think there's something above and beyond that and the way we think of it is when you're speaking a less dominant language your second language you have to inhibit your first language and that that is your first language competes with your second language but because you have to inhibit your first language it may mean that competition within the second language isn't as effective I say so it's and what's interesting is you said it's not because maybe they don't have enough examples stored in their mind it's not because of that it's just because they don't have the cognitive the kind of horsepower to process the alternatives maybe because it's busy it's busy trying to stop them from speaking Spanish that's really the idea I mean I will say at high at the highest levels of proficiency they can they do learn these things so I don't think it's independent of input and the number of examples but but it does look like you know those certain kinds of examples are surprisingly sticky even with a lot of input that they non-native speakers will will continue to make them you know even when they're fairly proficient even when they're going to college saying in the United States what general advice would you give to to an adult who who really wants to learn a language is that should they focus on one kind of input or or output or what would be your your general advice based on what you know about the way that we learn languages so I think yeah input is important and one thing we know plays a big role is and and we it's more useful to know the words and phrases that are used most often so focus on those first right because you're gonna use those most and that will help you then understand more so I think I was talking to some second language instructors here and they were saying that the way it's often taught is by topics so if you're teaching about traveling you might learn the word for windshield wiper but you don't you're never gonna need the word for windshield wiper um it's so low frequency and you can always point if you have made it changed and so it seems like more naturalistic input where the natural frequencies appear in there with their natural preferences I think would be important and just one final question because obviously you've dedicated most of your life to two languages and and understanding languages what why do you think it's important language well you know I think language is endlessly fascinated a fascinating I'm interested in how people think and and what makes humans special and and really how language is this incredibly complex skill you know we're able to formulate sentences without without reflecting at all and we're able to comprehend sentences on the order of milliseconds so it's a stunningly amazing ability that everyone can do you know except if severely impaired people rare populations some autistic children never learn to speak but aside from them most people learn a language so how do we do this it's just this puzzle to me that I find a mostly fascinating um well adult Goldberg thank you very much for your time oh thank you guys soon really [Music]

Channel: Canguro English

Views: 32,335

Keywords: learning english, learn english, english teacher, english grammar, grammar, esl, adele goldberg, input vs output, Krashen's input hypothesis, cognitive scientist, language acquisition, how to learn vocabulary, what is the best way to learn vocabulary, how to learn english, comprehensible input, second language acquisition theory, acquire vs learn a language, canguro english, kangaroo english, canguru english

Id: oRX5EWdm7Co

Length: 22min 44sec (1364 seconds)

Published: Wed Sep 18 2019