Toddler Speech and the MACAW Lab

Language is the basis for all human connection. It allows us to communicate our ideas, express our feelings, and have discussions with the people in our lives. The language skills we develop as children lay the foundation for how we communicate as adults. Typically developing children begin speaking between 12 to 18 months and start regularly incorporating verbs around the age of two. Research suggests that the number of verbs a child knows at age two is an indicator of their grammar skills at age two. But what happens when these language skills are delayed?

In this episode, Dr. Sabrina Horvath, an assistant professor of rehabilitation sciences in the College of Health Professions at MUSC shares her research in childhood language development. Horvath is the Principal Investigator for the MACAW Lab, which stands for Methods of Acquiring Concepts and Words. Horvath leads research in child language development and disorder, late talkers, and vocabulary development.

Read The Transcript

[00:00:09] Gwen Bouchie: From the Medical University of South Carolina, this is Science Never Sleeps, a show that explores the science, the people, and the stories behind the scenes of biomedical research happening at MUSC. I'm Gwen Bouchie. Have you ever stopped to think about the role language plays in our lives? From casual greetings to public speeches, language shapes our interactions, perceptions, and identities. The language skills we use as adults begin to form in our earliest years. and we carry these skills with us throughout our lifetimes. But what happens when children have delays in language development during those formative years? How can we course correct to make sure that children develop language skills that will set them up for a future of success? In this episode of Science Never Sleeps, we're joined by Dr. Sabrina Horvath, an assistant professor of rehabilitation sciences in the College of Health Professions at MUSC. Dr. Horvath is also the principal investigator for the MACAW Lab. Standing for methods of acquiring concepts and words. The MACAW Lab explores how children develop their vocabularies and broader language skills and the importance of verbs in later language development. Stay with us. Dr. Horvath, welcome to Science Never Sleeps.

[00:01:33] Sabrina Horvath, Ph.D., CCC-SLP: Thank you so much for having me. It's a real pleasure.

[00:01:35] Bouchie: So you are a speech pathologist. And I want to really jump in first. One of the things we talk about on Science Never Sleeps are researchers' journeys. How do they get where they are? And you have this really fascinating journey of starting in your early career, your early college career. just in linguistics and history, just sort of immersed in language, and then that brought you to where you are now. So I would love to hear you talk a little bit about what that journey has been like for you and how you got to where you are today.

[00:02:10] Horvath: Absolutely. Well, as you mentioned, I started my undergraduate career as actually a history major and a Spanish-German double minor. But more broadly, I've just always loved and been fascinated by languages. I grew up just outside of Washington, D.C., which is a beautifully diverse area to live, and had several friends growing up who spoke languages other than English at home. And it's the sort of place where you go to the grocery store and hear half a dozen different languages on your way to get eggs. So I've always been exposed to this idea of the richness of language and the variability of language, and the idea that language connects us to others, to our own culture, and... through language we can experience other cultures and other ideas that we might not be familiar with from our own particular experiences.

So that was my motivation for studying two languages as minors, but I very quickly found and fell in love with linguistics, which is the study of language in a scientific way. How do we put words together and make sentences, and how might that look different from one language to another? What are the rules that govern how we... can use and formulate sentences socially, structurally in the language itself. And importantly to my future work, how do children ever seem to figure this out? Because it's a very complicated system. And yet so many children seem to do it effortlessly without any sort of direct instruction. So this was kind of the beginning of my journey. And speech pathology, of course, is one very obvious pathway from linguistics and one that I was drawn to because of those clinical implications of helping children for whom this process is not unfolding as it should, for whom there are challenges. So I got actually the history and linguistics double major. The history has not served me particularly well moving forward, but I went straight into a master's program in speech language pathology. Did a little bit of research as part of that because I was interested in the how and the why and worked clinically for a couple of years before deciding that research was really where I wanted to be. So went back, got my Ph.D., did a postdoc and was very, very lucky to end up here at MUSC for a faculty position.

[00:04:46] Bouchie: And here you are. And you're working now with children, particularly toddlers. And so before we begin talking about your work today, I wanted to define a toddler. So when we talk about toddlers and the kids that you're working with, what does that mean? What is that age range? What are we looking at?

[00:05:06] Horvath: Absolutely. Most of my research is from the two to three age range. But I've run studies with children as young as about 18 months as part of my work, and as old as about seven years of age, relevant to some of my, depending on what the question is, of course.

[00:05:27] Bouchie: Right, right, depending on the research.

[00:05:29] Horvath: Yes. Yeah. With two- to three-year-olds, what we're seeing a lot of is this rapid growth in vocabulary development and the beginnings of putting words and short sentences together. So I'm interested in how this process is happening and why some children might be struggling with this. And the children who struggle with this, we typically refer to as late talkers.

[00:05:54] Bouchie: And so we do, so many of us have either children that we know in our lives who have been late talkers or perhaps ourselves we were late talkers. So this really is sort of a common thing across lots of different children, I would imagine.

[00:06:10] Horvath: Absolutely. Late talking impacts about 12 to 13 percent of kids. And late talking itself is not a diagnosis. However, children who are late talkers are at increased risk for ultimately having a diagnosis of something called developmental language disorder. Developmental language disorder is a disorder in which an individual has difficulty understanding and using language. And that challenge is not attributable to something like autism spectrum disorder. or cerebral palsy, or an intellectual disability. These challenges can manifest in many ways, such as difficulty finding the words that they want to say, being unspecific in their word choices, or perhaps having difficulty following complex directions or understanding figurative language. And as I said, it's the most common disorder you've probably never heard of. It impacts about seven percent to eight percent of the population. So this is about three times more common than autism spectrum disorder and about as common as ADHD.

[00:07:20] Bouchie: Wow, yeah.

[00:07:22] Horvath: So our goal in the McCall lab is to first, help all late talkers who could all benefit from therapy services, but second, to try to identify which of our late talkers are at greatest risk for ultimately having developmental language disorder, so that we know who needs the greatest support in those preschool years. because early intervention is critical in improving long-term outcomes.

[00:07:49] Bouchie: And as part of your work, you really have this focus on verbs. And so what is it about the verb in language that makes it such a critical piece of assessing language skill?

[00:08:03] Horvath: That is such a good question and one I love talking about. Verbs are unique because they bridge the gap between the vocabulary, words and their meanings, and the structure of language itself. So it's the verb that actually determines what a sentence is going to look like. For example, the verb sleep can only take a subject. I sleep, you sleep, the dog sleeps. Comparatively, the verb give. must take a subject and an object and an indirect object. I give you the book, for example. So the verb is really what dictates what can and needs to appear in that larger sentence structure. The reason I focus on verbs is because this is potentially a very early way of getting a sense of children's later grammatical skills and development. And that grammar piece is what children who have developmental language disorder struggle most with. Unfortunately, we can't back up our current age of diagnosis for developmental language disorder, which is about four to five, because we've traditionally been waiting for these milestones that typically developing children have to hit, where we can then compare and say the child isn't doing X, Y, and Z. Therefore, they have developmental language disorder. I think if we focus on verbs, we can predict which children are struggling, even in the earliest stages, with what will become their grammatical system.

Verbs are also interesting because learning verbs themselves requires some knowledge of grammar. So we're getting insight into their learning process, their ability to use grammar at an early age, and then when they learn verbs, they're reinforcing what they know of grammar and expanding their ability to use structure and language. So the idea is then that if we provide intervention focusing on verbs, we're supporting not only vocabulary development but some of those grammatical skills as well.

[00:10:15] Bouchie: And it reminds me a little of, if you think about the early primers, see, spot, run, those type of things are very basic sentence structure as children are learning. So it really is, they're learning from that very young age, the sentence structure of how to tie the action of something or someone together through language. And your hope is to try to assess their skill or ability to be able to do that or not do that much earlier so that they can get intervention sooner.

[00:10:50] Horvath: One of the things we do know about verb learning is that at very, very young ages, children are sensitive to these patterns in language and use those patterns to make predictions about what new words will mean. So this is a process for verb learning known as syntactic bootstrapping. Children will assume if we teach them a made-up verb that the duck is gorping the bunny means something very different from the duck and the bunny are gorping. And so this attention to early structure is really important in them being able to build their verb vocabularies. or I should say, typically developing children, being able to build their verb vocabularies, because we don't know if late talkers can use this skill or not.

[00:11:43] Bouchie: So talk a little bit about your work in terms of the specificity around typical development versus some development that falls outside of that typical range. So we would be thinking about children with autism, perhaps other disorders. Can you talk a little bit about what you're exploring in that space?

[00:12:03] Horvath: Absolutely. Most of my work is comparing typically developing children to late talkers, but I've also done some work with children on the autism spectrum as well. Autism is a unique disorder to look at when we think of language because language challenges are no longer part of the diagnosis itself. However, the majority of children on the autism spectrum do have some challenge with language. We are interested in the relationship between those characteristic features of autistic children, like the challenges they have with joint attention and social engagement, and what those implications might be for the language learning process.

[00:12:51] Bouchie: I'm sure a day in your lab or in your study environment is wildly fun because I can't imagine, you know, having these toddlers coming in and their families. So let's talk a little bit about what is a typical study visit look like for the kids who are coming in to work with you?

[00:13:11] Horvath: Well, as you said, we try to make it as fun as possible for the families and their children. We know it can be a challenge to drive into downtown Charleston, but we want to make sure that the visit is enjoyable and a positive experience for everyone. We always start with a little play to make sure that the child feels comfortable in the space and with the research team. And we also take that time to sign consent forms, which are always important, and gather any additional information that we might need from the family, such as completing a vocabulary checklist of the words that they know, or maybe providing some demographic information for us. The actual study will vary based on our particular research questions and goals, but a lot of our studies use eye tracking methodology, which is a lot of fun for children and their families. Our children sit in our special race car chair in front of an eye tracking machine where we display videos and it captures in real time the exact coordinates on the screen where the child is looking. And we can analyze this data. to get a lot of information about language skills, language processing, language knowledge, and really a far more robust data set than we would get from just pointing, for example.

[00:14:37] Bouchie: Yeah, the eye tracking is really interesting because it does give you that extra set of data. And what are we really looking for when we're looking at eye tracking?

[00:14:49} Horvath: That's a great question. One thing we can look at is accuracy. So when we run studies that have made up verbs, we can look at whether the child can, for example, find groping when we ask them to do so. Another thing we can look at is something called language processing speed. And for these, we're using familiar words. We might display a picture of a ball and a shoe and say, find the ball. And what's very interesting is the speed at which a child will look to the ball is predictive not only of their concurrent vocabulary size, but also their language growth and later language outcomes. At least if you're looking at nouns, we've been exploring this in verb vocabulary as well.

[00:15:44] Bouchie: So how is it that you can tie those things together in terms of it being predictive of a future vocabulary?

[00:15:52] Horvath: Well, these have been longitudinal studies traditionally, where the children who are faster to find the item have better language scores on standardized tests to age eight. We have worked on how to quantify processing speed when children are watching videos of verb actions, as opposed to looking at static images, because eye behaviors will change if you're looking at a video versus an image. Right now we've only done concurrent studies. We've done one with late talkers and one with autistic children. But our hope in the future is to do some longitudinal work to see if this processing of verb vocabulary also predicts later language outcomes. One of the things that we found is that late talkers and typically developing children have different types of verbs in their early verb vocabularies. So I'll explain what that means. We've looked at several different aspects of how verbs encode meaning. And one that is important across the world's languages is manner versus result verbs. Manner verbs are verbs that encode how something happens, like walk, run, jump, hop, skip. These are all manners of motion. Result verbs, by contrast, encode the end state of the event, like open or close or finish. What we've found, very interestingly, is that late talkers have a preference for result over manner meanings even from the earliest stages of development, whereas typically developing children prefer manner over result verbs.

So this is a notable difference. We've also found that irrespective of whether a child is or is not a late talker, Children who have proportionally more manner than result verbs have larger vocabularies overall. So this has opened up several interesting lines of research for us in thinking about why each group might prefer one verb type over another and what it might mean for intervention in terms of picking therapy targets. Another notable finding that we have out of our lab is actually publishing the first ever study to look at verb learning mechanisms in late talkers. In this particular study, we were considering whether it was better for late talkers and typically developing children, when learning a new verb, to hear only content nouns, the boy is daxing the balloon, the boy is daxing the balloon, or to hear a mix of content nouns and pronouns, the boy is daxing the balloon, he is daxing it. We found, somewhat to our surprise, that both typically developing children and late talkers performed better given only content nouns. But late talkers in particular performed very poorly when they were given the suboptimal content noun pronoun mix. So this is evidence for the idea that when a verb learning situation is suboptimal, late talkers are particularly negatively impacted. And further support for the idea that we need to be conscientious about how we're teaching new vocabulary items to children who are already struggling with vocabulary.

[00:19:23] Bouchie: So let's talk for a second about these made up words. You have mentioned a few, and so I just, it's so fascinating that there's these, you know, just made up words that you're working with in your lab. So what is the significance of these words and working with the kids with these particular words?

[00:19:43] Horvath: Well, our purpose in using made-up words is to make sure that the child doesn't already know what that word is. That way we can say confidently that when they demonstrate knowledge of this made-up word, it's not because they came in already knowing what daxing or gorping or ziffing is. And in the actual experiments, we're able to make subtle changes to learning environments. For example, comparing only content nouns versus content nouns and pronouns, and make conclusions from our experiment about which learning situations are optimal or suboptimal. This is one of the major goals of our research because identifying what is optimal is necessary when we think about designing efficacious interventions.

[00:20:36] Bouchie: Do you think there are implications for this research to... for... is generally the way we teach language to children?

[00:20:46] Horvath: Absolutely. And I think you're tapping into a really interesting question about our perceptions of how much teaching needs to happen in the process of language development. Interestingly, there are cultures across the globe where it's inappropriate socially to talk to a child before the child is at a stage where they can be conversational in return. And yet, children learn language beautifully in these cultures as well. What we see is that the child mind, whether it's pre-programmed for language or not, a theoretical debate for another time, is ready to embrace the rich complex system that is language to work through it, make sense of it, and then employ it themselves. In a variety of circumstances and with a variety of experiences. Our challenge, particularly, is for those children who struggle with this process and making sure that we can support them as best we are able to improve their long-term outcomes.

[00:21:56] Bouchie: With your study, you're looking at things like verbal cues, various different methods of doing this assessment. Can you tell us a little bit about what that looks like and maybe where that comes from?

[00:22:08] Horvath: Figuring out what an unfamiliar word means is actually an incredibly difficult task. And sometimes we don't think about how hard it is for children to make sense of a word that they have never heard before. There's been wonderful research over the past 60 to 70 years to identify some of the cues that children use when making sense of an unfamiliar word.

One of the cues is visual information. It's a lot easier to figure out what a word means if the referent of that word is visually available to the child in real time. Children also use social cues to figure out what words mean. For example, a parent might point to an item or direct the child's attention in some other way to bring attention to the word that they are teaching.

One cue that's incredibly important, but perhaps not so immediately obvious, is using linguistic information to figure out what that word means. And there are all sorts of linguistic cues that children use. For example, by about 18 months of age, children can reliably class words into nouns or verbs based on how they're used in a sentence. If they hear a-ziff, a-mody, they know that that's probably a noun. Whereas if they hear gorping or zebbing, they know that that's probably a verb. They're also aware by this age that nouns denote objects and verbs denote actions. So this is one way that they narrow down their hypotheses of what an unfamiliar verb might mean. So this is one way they may narrow down their hypotheses of what an unfamiliar word might mean.

When we consider verb learning specifically, typically developing children use a process known as syntactic bootstrapping. This means they use the structure of the sentence and the relationship of the nouns to the verb to make inferences about the type of meaning encoded by that verb. I'll give you an example. Typically developing children by age two know that the duck is gorping the bunny; means that the duck is doing something to a bunny. Conversely, the duck and the bunny are groping means that the duck is doing something and the bunny is doing something separately from that duck. We actually don't know if late talkers use syntactic bootstrapping at all in their process of verb learning. This is one of the studies we're running right now in the lab, and I'm very excited to see what the results are. There are different implications about the verb learning process and intervention techniques based on whether this cue is available to late talkers in learning new verbs.

[00:25:05] Bouchie: Do you have any tips for parents of toddlers or even really honestly parents at any age, probably kids sub five, I guess, based on your work and what you know, particularly as a researcher and also as a speech language pathologist?

[00:25:23] Horvath: This is a question I get asked a lot. The first recommendation I make to any parent is to read to your child. Reading is a wonderful way to expand a child's vocabulary and their grammar skills, because we use grammatical structures and vocabulary items in literature, even children's literature, that we don't use in our day-to-day conversation. For example, ball means one thing when we're playing outside and another thing altogether when reading Cinderella. Another thing that I encourage parents to do is to talk to and with and around their children. You are enriching their language environment by giving them exposure to language even if you're not talking directly to them. Narrating what you're doing or narrating what you're seeing is a wonderful way to spark their language growth.

I also get a lot of questions about language development from parents who are concerned that their child might be a late talker. It's hard to know what the milestones are. To any parent who asks me, I always recommend that they talk with their pediatrician. If you have serious concerns, it is worth it to follow up and potentially pursue an evaluation.

[00:26:47] Bouchie: Why are things like the number of verbs that a child knows and sentence structure important?

[00:26:56] Horvath: These are parts of language known as morphology and syntax. And together, morphology and syntax make grammar. They're so important because children who have developmental language disorder struggle with grammar particularly. We're interested in verbs as an early predictor of where they'll end up with, with their grammar skills, and who might struggle the most. But I've used these very technical terms, morphology and syntax, and I feel like I should define them a little bit more. Syntax is essentially the word order of a language.

For example, in English, we have a subject, verb, object word order. In other languages like Hungarian, the verb goes at the end of the sentence. Morphology are these little pieces of language that let us know who did what to whom and when it happened. They don't have any meaning in and of themselves other than conveying this information. Examples in English include the progressive ing, walking, eating, the past tense ed, walked, talked, or the plural s, shoes, socks. What we found is that the number of verbs a child has at age two predicts how many morphemes they know at age three and how complex their sentences are. We find this very suggestive of the role that verbs play in developing grammar.

[00:28:37] Bouchie: So understanding how children learn verbs and how to possibly improve that learning is key to their language success in the future.

[00:28:47] Horvath: Absolutely.

[00:28:50] Bouchie: We've been talking to Dr. Sabrina Horvath about childhood language development. Have an idea for a future episode of Science Never Sleeps? Click on the link in the show notes to share with us. Science Never Sleeps is produced by the Office of the Vice President for Research at the Medical University of South Carolina. Special thanks to the Office of Instructional Technology for support on this episode.