On learning new words while reading, with Dr Holly Joseph

Holly Joseph is a developmental cognitive psychologist at the University of Reading who specialises in reading development and difficulties. During this podcast she chats to Sue about a piece of work looking at how children learn new words they encounter during reading, and how some children find this more difficult than others.

You can follow Holly on Twitter here.

The paper discussed in this episode is:
Joseph, H., & Nation, K. (2018). Examining incidental word learning during reading in children: The role of context. Journal of Experimental Child Psychology, 166, 190-211.

Transcript

Sue: 0:09

[Podcast jingle][ringtone] Hello? Oh, it is recording. I see the little figure. Okay, great. I will do my little spiel and then I'll introduce you. Nice. Okay. Here I go. Hi, this is Sue from the Salvesen Mindroom Research Centre at the University of Edinburgh. And we are recording another podcast episode, just trying to, u m, take advantage maybe if the i nterest that there is at the moment in child development and w ellbeing and so on, and see if we can add to that conversation in a useful way. And today I'm talking to Holly Joseph, who is very dear old friend of mine, as well as being an associate professor of Language, Education, and Literacy development at the Institute of Education at the University of Reading. Hello Holly!

Holly: 0:53

Hello! Very long title I have there!

Sue: 0:58

Yeah, well, we've got to give you all your credit for all of your hard work that we've done together, absolutely! Um, so Holly we are going to talk about your paper that was looking at incidental word learning during reading in children. So this is, I guess, how children learn new words while they're reading?

Holly: 1:19

Exactly yeah!

Sue: 1:19

Um, perfect! Glad I understood that correctly. It's a good start. So tell me, what did you find when you were doing this study? What did you discover?

Holly: 1:29

Well, um, very simply we discovered that children do learn new words, um, as they're reading. So this is words... They're not trying to learn them on purpose. It's not deliberate. It's just, as they're reading, do they kind of pick up information about, um, the meaning of the word? So children can do this, they can learn, um, about both the spelling and the meaning, um, of words. Um, but I guess the most interesting finding was that children differ quite a lot in how efficient they are in doing this and how efficient they are in learning these new word meanings. And so I guess, if you're thinking about educational relevance or practical implications, I guess this shows that children well, what we found was that children who we know were already not so good at reading comprehension in general, those children, um, so-called poor comprehenders, were slower and less successful at learning these new word meanings.

Sue: 2:31

Um, so it's going to have a kind of, um, uh, uh, a snowball and negative kind of, uh, cycle of effect, right?

Holly: 2:40

Excatly.

Sue: 2:40

In terms of acquiring new words as well.

Holly: 2:43

Exactly, exactly yeah. Sorry, go on.

Sue: 2:46

No, go on.

Holly: 2:47

I was just going to say the sort of, we call them, Matthew, um, effect, so the idea of the rich getting richer and the poor getting poorer. So...

Sue: 2:55

Right.

Holly: 2:55

Um, if you're already good at reading, then you kind of get more out of reading, learn more vocabulary, and then if you've got better vocabulary, it's har-... it's easier for you to understand text, and that kind of goes on and on.

Sue: 3:07

Right...

Holly: 3:07

And then the other side, if you're poor reading, it's harder for you to learn new words and then it's harder for you to read new text. So over time, the difference between those two groups of children increases.

Sue: 3:21

I see, yeah that's, that's important. Um, so, so why... So it's interesting, you started off by saying that, you know, you discovered that children learn new words while reading, right? So that's, it's one of those findings in science where I think people probably listening want to be like"uh, duh!"[Laughs] Why did you, can you sort of explain a bit about what we knew already and, and what you know, why, what motivated you to, to be asking this question in the first place?

Holly: 3:51

Yeah. I mean, I should say that we didn't discover that! We knew[laughs], we already knew that children learn new words, um, through reading. And in fact, from around age nine onwards, most of the new words we learn, we learned through reading them in context, and rather than to deliberately trying to learn them or being taught them in a classroom, for example. Um, and I guess we were interested in the process of that. So I dunno, it might be useful to think about an example. So let's say the word"benign", the first time you encounter the word"benign", you might encounter it in a medical context. So something to do with a benign tumor, right? So that first context that you encounter it- it in will give you a little bit information about the meaning of the word"tumor", but it's not the whole story. So what you need is then to, um, encounter"benign" again, and then again, in various contexts, and over time you might hear about, you might read about a benign leader or benign weather, and over time you start to extract the kind of core meaning of that word as something like"harmless". And that means, you know, your, your knowledge of that word increases, and you're able to then use it yourself in productive language, so in your, in your speaking, in your writing. So we knew, we already knew that children did this and that they did it a lot. So around 2 to 3000 words a year are learned this way from around age nine onward.

Sue: 5:17

Wow! That's amazing!

Holly: 5:17

Isn't it! Um, um, but what we don't know is, is how that happens, the kind of process by which that happens. Um, and so we were interested in: okay, what happened the first time you see a new word in context, and then the second time, and then the third time and so on and so on. And what can we see in terms of the kind of development of understanding of that word, um, meaning? So we wanted to look at the process. We already knew it happened, but we didn't know how it happened.

Sue: 5:44

Right...

Holly: 5:44

And we also didn't know what conditions might be the most helpful for word learning, um, in this way to occur and which children might find it particularly hard.

Sue: 5:54

Okay. This is great. So, so how... So that's really helpful context. And now I'm curious about how you went about measuring that. So you have to give them... How did you know that they didn't already know the word, for example, did you give the made up words or, you know, wh- what was the process?

Holly: 6:11

Yeah, I mean, we've done a few studies like this. So in this study we've used real words. There's lots of, um, arguments sort of for and against using real versus, um, made up words. So obviously if you use made up words the children will never have encountered those before...

Sue: 6:28

Yeah.

Holly: 6:28

Um, but if you use... The ethical issues around, um, you know, having children read many times these made up words, so we use very rare, um, um, real words, which were all actually past tense, so things like"confabulated" and"languished". And these are children in years 5 and 6. So they were kind of 9 to 11 years old. Um, and what we wanted to do was kind of recreate what happens in the real world, and people kind of just read independently and encounter words that they don't know, recreate that within a kind of controlled experiment. So we created these kind of sentence frames, or contexts, and we had, um, six, I think it was, um, the target rare words, and each of them had 10 contexts. So you sort of had, the first time you encountered it would give you some information about its meaning and the second time, third time, et cetera. Um, and so there were 60 sentences in total across these six, um, these six words. Um, and so children read these 60 sentences over the course of two days. Um, and we used an eye tracker to track their eye movements as they read. Um, and the reason we did this was to try and get at the process that I was talking about. So we know from eye movement research that the longer you look at a word or a phrase, the harder it is for you to process. Um, so, um, so we thought we might see this interesting pattern over the course of these 10 different presentations of each word, that wh-, the first time children saw it they'd take a long time to read it because it's unfamiliar and difficult to understand. And then the second time they'd spend a little bit less time and so on and so on. Um, and in this way we could kind of map children's learning trajectories of these individual words and find out some things interesting. Um, and that's how we made the conclusion that poor comprehenders were less efficient at, um, learning these new words, because their reduction in reading times was much less dramatic than the children with better reading comprehension skills.

Sue: 8:40

Right, right... So did those, um, poor comprehenders... Did- did everyone start out reading sort of slowly to the, to the same, you know, were they all equally slow the first time they read"confabulated" and then it was just that the confident readers got, got super fast by the time it was the 10th time, or was there a difference right at the very beginning?

Holly: 9:04

I-, I should say actually, I've sort of slightly been confusing, that we didn't actually have a group of poor comprehenders and good comprehenders.

Sue: 9:10

Right, right, right yeah....

Holly: 9:10

We just had children with varying reading comprehension.

Sue: 9:13

Sure, sure...

Holly: 9:13

So not everybody was the same, but everybody was within a margin of kind of second or millisecond.

Sue: 9:19

Yeah, yeah.

Holly: 9:19

Um, and everybody very, very consistently reduced um, their reading times. So if you look at it on a graph, you basically just see this relatively steep, um, slope down for the, um, for everybody. Um, but if you kind of show it in the graph of the kind of worst compounders versus the best comprehenders. You can see a much less steep slope for the poor, poorer compounders.

Sue: 9:45

Yeah... And so is this something that maybe you didn't look at but I'm curious about: so you, you talked, when you were given the"benign" example at the beginning, obviously the other words that"benign" is paired with are a big part of revealing the meaning of the word"benign", right, and therefore, I guess, contributing to this faster reading. So, you know,"benign tumor","benign leader","begign whatever" et cetera...

Holly: 10:05

Yeah...

Sue: 10:06

So did you look at patterns of how they were looking at the word around the kind of target word, the new word, you know, was there some, maybe looking back and forth or anything like that or, or is reading just not really subject to those kinds of effects? I'm not... I don't know it well enough.

Holly: 10:23

You absolutely can do that. We didn't do that in the study. We have subsequently done another study where we had kind of clue words, which we control for various things. And so then you can look at how long they look at the clue word and how long they look at the target word. Um, I mean, the difficulty with eye tracking is that you have to control- If you've got tons of words, you've got to control them for length and frequency and various other things...

Sue: 10:46

Yeah, yeah...

Holly: 10:46

Because you don't want the longer reading times to be influenced by one of those variables that we know have huge impact on reading times. So it's very, very hard to have the multiple, um, regions of interest, uh, while making sure that they're controlled. Um, so yeah, I d-, I don't know if they were looking, I mean, I guess we made the assumption that if they're looking less time at the word, they're looking at the rest of the sentence more, although that's not necessarily the case, so yeah. In this study, we can't really speak to that, unfortunately.

Sue: 11:24

Yeah, yeah, yeah... I love this about reading research though, because you and I shared an office when we were doing our PhDs right, and I remember, you know, having to make up impossible sentences and unlikely sentences, right. Then you'd have, you'd have"the man played the guitar", that would be a, a normal sentence. And then you'd have, I don't know,"the dog played the guitar", that'd be like an unlikely sentence. And then, you know,"the house played the guitar", which would be an impossible sentence, but it was... it's so fun! Right? Do you enjoy that part of your job?

Holly: 11:57

Yeah, I love that part! I- I actually really love, um, also trying to find words that are the same length in the same frequency in the same age of acquisition and the same concreteness and the same, all the things you have to think about, that fit into a sentence. I mean, it does result in some really weird sentences. So you I'm sure, you know, the famous one it's about, um, sort of grammatical processing not vocabulary, about"the horse raced past the barn f ell", do you know that one?

Sue: 12:25

No!

Holly: 12:25

So this is incredibly hard for people to understand. Um, but it does actually make sense.

Sue: 12:32

" The horse raced... past... the barn"... Uh, I don't understand it...[laughs] could you...[laughs]

Holly: 12:38

That's probably because you're a good comprehender and you have[laughs] I mean, so people, when they see that they assume that the verb"raced" is, um, an active verb the horse is doing the racing, but actually in this sentence, it's passive. So if you sort of, you know, uh, made it longer, you could say"the horse that was raced past the barn", brackets"by the jockey","fell".

Sue: 13:06

Ooh! Ooh! Right, right, right! So the horse is falling, not the barn![laughs].

Holly: 13:11

Yes! No, the barn is not falling.

Sue: 13:13

Yeah, cause I just, I just thought"the horse raced past" comma,"the barn fell", right?

Holly: 13:18

Oh okay! That's interesting!

Sue: 13:18

That was the best attempt I could make at making sense of it. Like if it was two, two separate events... Anyway! Yeah, yeah.

Holly: 13:27

I guess that's an interesting distinction between oral language and written language, right? Because you couldn't have that in written language without a comma or a semicolon I don't think, I don't know what the punctuation would be. Whereas in, in, in spoken language, you can kind of insert your own punctuation to try and help you make sense of it.

Sue: 13:42

Yeah. I think my brain is desperately putting in punctuation to help me understand![Laughs] Um, oh this is great! And so you touched at the beginning on the kind of implications of this, you know, in the, the kind of"rich gets richer, poor gets poorer" effect here. So I guess the question for, you know, maybe parents, obviously at home, homeschooling and, um, teachers, you know, um, thinking about their practice when, whenever that, um, is permitted to return to normal again, what can we do about that? Is there any way that we can sort of reverse these effects or take account of them in the way that we, um, encourage children with their reading and support their reading?

Holly: 14:24

Yeah. I mean, that's a good question because in some ways the findings from this study are a little bit negative because on the basis of this study, there's nothing specific that I could recommend. Um, I mean, what I think it... I mean, we know that the reading comprehension is important right, and schools are, you know, really promoting this, and quite rightly, we all know as parents that we should really be encouraging our children to read. And, and if you have any young children you should be reading story books to them... We all know that. Um, and I wouldn't for a second question the benefits of independent reading, but I guess for teachers and parents potentially it's just worth remembering that, that the reading is likely to benefit some children more than others. And those who already struggle may need additional support in order to gain the same advantage. I mean, one obvious way off the top of my head, not based on research to support a child who we know already has reading comprehension difficulties is to pre-teach vocabulary or, or at least, you know, sort of be aware that some of the words that they are going to encounter are going to be difficult for them. But, you know, we're talking about independent reading, which is what I am talking about, you know, how, how, how do you do that? It, it requires a lot of intervention from a third party. So, so it's tricky. Um...

Sue: 15:51

Yeah, but I think, I mean, I think you're right. Just being cognizant of the difficulties is a really important part of the puzzle, right? And just being, you know, patient and understanding with people who, um, whose reading progresses maybe going to be that bit slower, you know?

Holly: 16:09

Yeah, yeah...

Sue: 16:10

I guess as well... I'm always interested in the debate about sort of the role of things like audiobooks right?

Holly: 16:17

Mmhh...

Sue: 16:17

Off the top of my head, not in any way an expert in this area, it seems to me that perhaps acquiring new vocabulary for someone who's struggling with the mechanics of reading might work better...

Holly: 16:31

Yes...

Sue: 16:31

You know, it might be easier when it's an audiobook, right. Because you can then take the time to dwell on the new meanings of the words that you're encountering without the hurdle of, of kind of interpreting what's on the page. Does that...?

Holly: 16:44

I mean, I think, I think audiobooks are a fantastic resource for children with a dyslexic profile, because if what's hard for you is decoding,...

Sue: 16:53

Yeah!

Holly: 16:53

... is efficiently recognizing, and accurately reading words, then A: you're probably not going to really like reading, um, and B: you're gonna sort of missread words or not be able to read words, and that's going to have, um, a detrimental effect on your understanding. Although if you have dyslexia by definition, you don't have a problem with understanding language, but rather with as you say the mechanics of reading. With a poor comprehender it's kind of different, right? Because a poor comprehender is somebody who has, um, accurate and fluent word reading. So they don't have difficulty kind of, you know, um, reading words quickly and correctly. But, um, their difficulty is in understanding language, and we know now it's not just written language, it's language, so poor comprehenders tend to have poor vocabulary, and they tend to be poor at comprehending language. And we see that, that often becomes very apparent in late primary school, because it's harder to see somebody who hasn't got a sort of more significant language um... impairment. It's harder to see, you know, when they're very young. Um, so... So for those children, arguably audiobooks are not, they might help a little bit because you've got these sort of added clues o f prosody and, you know, using the voice to, um, you know, to emphasize meaning, et cetera. In, in theory, at least audiobooks would be thought to be more helpful for a child with a sort of dyslexic rather than poor comprehender profile. It is really hard to, to, to help these children, um, because it's really hard to improve vocabulary and high level reading comprehension skills. Um, I have got one little happy aspect to this research that was not actually...

Sue: 18:48

Yeah give it! Give us the happy aspect!

Holly: 18:48

Um, it's that we ran a similar study with children who speak English as an additional language. And so those children tend to look a bit like poor comprehenders in English. So they tend to have lower vocabulary than monolingual children, and they, um, tend to struggle with reading comprehension, but be fine at word reading. So they are accurate and fluent. For these children who on paper kind of look like poor comprehenders, but they're not really poor comprehenders, right? They're just really relatively new to English. and they've had less exposure to the English language. These children seem to be particularly good, particularly efficient at learning new words through reading in a very similar design. So comparing them with non poor comprehenders monolingual children, their reduction over the course of the exposure was much steeper.

Sue: 19:39

Wow!

Holly: 19:39

So it's not just, if you've not got so much vocabulary and reading comprehension skills, you're kind of, you know... I can't think of a word that's not a swear word! You're in trouble![laughs] It's not that! Um, so, so that's another reason why we have to be... We ha-... It is complicated. And as a teacher, practitioner, parent, you know, it's not a kind of"if this, then this", but rather different groups of children to different degree might be able to benefit more or less from independent reading.

Sue: 20:13

Uh, this is such juicy stuff. I feel like we're going to have to do a second follow up podcast cause I've got so many more questions I want to ask you about the category of poor comprehenders, but I'm going to force myself to stop for now and just end by asking you whether you have any, um, pearls of wisdom for the students and early career researchers and so on, who might be listening, who are, you know, um, perhaps, uh, dealing with derailed research plans, uncertain futures and so on. So what, what would you like to say to them?

Holly: 20:46

Um, so I mean I haven't listened to other podcasts in the series, which I probably should have done because probably everyone else said this as well. But, um, you know, academia can be a really strange world. And, and during this crisis I've seen lots of tweets from academics, talking about how productive they're being during lockdown, and what an opportunity this is to catch up with writing and grant applications. Um, but at the same time, there were others who were struggling to get any work done due to caring for children or other people or for other reasons. And so I kind of, I knew you were gonna ask me this and I thought"what would I say to myself", you know, 10, 15 years ago. And I think I would say: try to ignore what other people are doing and find a path that works for you. So you can have a successful academic career without working more than 40 hours a week. You don't have to work evenings and weekends. You can prioritize family and friends and free time over work, and it's okay to do that. Of course, it's also fine for people who genuinely want to work more than 60 hours a week. And you know, their first love is their research, and they prioritize that over everything. Of course, that's fine too. Um, but you don't have to be like that. Um, and I think it's, it can be really hard to remember that particularly if you were at a sort of very, you know, high achieving institution where people are incredibly successful, and it doesn't look... When you look around it doesn't look like anybody's going home at 5:30 to, you know, make dinner for their family or, you know, nobody's taking an entire weekend off to go skiing or, or whatever, um, um, whatever they like to do in their free time. And, and, and I think that if more people, you know, thought about why they got into it in the first place, thought about their passion for research, not for publishing and bringing in money. Cause I mean, who starts in academia because they've got a passion for publishing and bringing in money, you know. If more of us sort of remembered that then maybe academia would be a little bit less scary and a bit more welcoming to people just starting off, um, in this weird world we live in.[laughs]

Sue: 23:01

Uh, that is not just a pearl of wisdom. I am going to say that that's a diamond of wisdom. That is really, really excellent, um, beautifully put point.

Holly: 23:10

I mean, if more people did what you have done Sue, and said... What is it-, what is it your"radical no" to these extra requests?

Sue: 23:17

Yeah, a year of saying" No". Yeah. Well, I also broke my own rules a few times, Holly. So I don't know if we should hold me up on this stuff...[Laughs]

Holly: 23:26

But I mean, just, just, just trying, I mean, I'm sure it had an impact on both your work, but also on the other parts of your life, which are important...

Sue: 23:35

Yeah, yeah!

Holly: 23:35

And we need to remind ourselves of that.

Sue: 23:39

Yeah, absolutely. Absolutely. And it's been good practice. I think I've got better at saying"No", which is a skill that it is worth developing, um, cause it's not easy to do.

Holly: 23:47

Yeah, yeah... Because it takes... yeah, no, exactly. It takes confidence and that's hard when you're just starting out, and you look around you, and everybody seems to be staying there. So yeah... Say"No", that's my pearl of wisdom! Say no![Laughs]

Sue: 24:04

Well, thank you very much, Holly, for your time. And this is a real cracker and anyone who's listening, you'll be able to find out more about Holly's work, and of the work specifically we talked about today by following the links that we'll put on the podcast page, which is at ed.ac.uk/salvesen-research. Thank you very much, Holly. Bye!

Holly: 24:28

Bye! Thank you![ringtone]

Sue: 24:29

Okay we did it! I thought that went quite smoothly![Podcast jingle]