"Becoming a Nation of Word Learners"

On-screen: [Becoming a Nation of Word Learners, EST Forum, Paul Deane, Educational Testing Service, October 21, 2015]

On-screen: [Goal: Becoming a nation of word learners.
Question: How do we get there? Margaret McKeown, University of Pittsburgh, mckeown@pitt.edu; Paul Deane, Educational Testing Service, PDeane@ETS.org]

People in this video:

Margaret McKeown, University of Pittsburgh, mckeown@pitt.edu
Paul Deane, Educational Testing Service, PDeane@ETS.org

Margaret McKeown: Our topic today is vocabulary and we're promoting a goal of becoming a nation of word learners. And we're thinking here mainly about kids in school and getting them on a good word-learning trajectory.

On-screen:  [A Collaborative Conference to Expand and Reimagine the Nature of Vocabulary Assessment; Supported by a grant from the Education Research Conference Program: American Educational Research Association]

As Joanna mentioned, today's talk, as well as the goal of becoming a nation of word learners, is based on a conference that we were awarded a grant for from the American Educational Research Association, AERA, and we're very indebted to them. The title, yes, we imagined ourselves reimagining vocabulary assessment. It was designed as a working conference leading to consensus about what's known in the area and to set some future directions for work. And we kind of examined three broad themes defining the construct of vocabulary knowledge, thinking about the design of new assessments, and then what the impact of those assessments might be on education practice.

On-screen: [Participants from educational research, assessment, and linguistics: Camille Blachowicz, National Louis University, Michael Coyne, University of Connecticut, Amy Crosson, University of Pittsburgh, Susan Flinspach, University of California-Santa Cruz, Elfrieda Hiebert, The Text Project Inc., Michael Kieffer, New York University, Joshua Lawrence, University of California-Irvine, Margaret McKeown, University of Pittsburgh, William Nagy, Seattle Pacific University, P. David Pearson, University of California, Berkeley, Judith Scott, University of California-Santa Cruz, Jack Vevea, University of California-Merced, Ben Bergen, University of California-San Diego, Kevyn Collins-Thompson, University of Michigan, Bill Croft, University of New Mexico Merced, Paul Deane, Educational Testing Service, Bonnie Dorr, Florida Institute for Human and Machine Cognition, Gwen Frishkoff, Georgia State University, Bob Krovetz, Lexical Research, Rene Lawless, Educational Testing Service, Charles Perfetti, University of Pittsburgh, Jenny Sullivan, Seneca College]

I know you can't read this, but these are the participants, so you might be able to see it on your handout. It gives you an idea of where everybody came from. We had a set of education researchers who have been working in vocabulary. That's the top folks. And then we had folks from assessment, and psychology and linguistics because those groups of people don't usually come together and we thought that would be a productive way to have a conference.

On-screen: [Energizing and Collegial!]

This gives you some sense of the atmosphere of the conference. This is participant Bill Nagy. I think of him as the academic in his suit with his Einstein hair on a bike. And as you may or may not be able to tell, this was on the ETS campus, which was where the conference was held. That's the best place in the world to have a conference. We had a wonderful time thinking, discussing and eating—as any of you who have been to ETS know, they feed you very well—and so it was a clearly joyful and thoughtful conference.

On-screen: [Becoming a Nation of Word Learners; Vocabulary underlies academic achievement; Words, reading, good life opportunities]

But we're here today to discuss becoming a nation of word learners. This title grew out of the conference, but it was mostly due to participant Judy Scott who came up with that. It's now also the title of a book that's under contract with Guilford Press based on the conference. We think becoming a nation of word learners is a worthy goal because vocabulary underlies academic achievement. More specifically, knowing a lot of words and knowing about words leads to being a good reader, and having good reading ability leads to good life opportunities. It opens the door to better opportunities in further education, in career choices and in civic involvement. And it's said that many states use fourth grade reading scores to predict how many prison cells they'll need ten years later. And while that turns out not precisely to be the case, there's a lot of truth underlying that. For example, 60 percent of inmates are illiterate and 85 percent of juvenile offenders have struggled with reading.

To consider how we might reach our goal of becoming a nation of word learners, we're going to address three questions today. What is the universe of word knowledge worth knowing? What conceptions of vocabulary knowledge should underlie assessments? And what recent work in psychology and linguistics can inform these conceptions? And I'll provide an educator's response to the question and Paul will provide a linguist's response.

So the universe, the idea of a universe of word knowledge, by calling it a universe we mean to draw attention to the fact that we don't just mean how many words can you provide definitions for. We know that skilled readers have abundant high quality understandings of words relative to less skilled readers, but this means not only that they know more words, but they know more about those words. They have a lot of general knowledge about language. They know how words work, morphology predominantly, and how words work together in context.

A big obstacle we have is that vocabulary, the conception of vocabulary instruction in schools is not likely to lead us to our goal of becoming a nation of word learners. Both instruction and assessment of vocabulary in schools is focused on definitions of words. So you may say what's wrong with definitions? A definition provides a handy generalization of what a word means. But a lot depends, as far as how valuable that is, how useful to a learner that is, depends on how that definition came to be represented in memory. A definition, a learner's knowledge of a word, a definition might be developed from abundant experiences with a word and deducing a word's nuances and features. That kind of definition would have a strong foundation, like an iceberg. But definitions can also be like an oil slick that's just a surface form and have no connection to the word's nuances or how it's used. And typically the oil slick version is what results when students are given lists of words to look up in a dictionary or given simple definitions to study. And the drill in school usually is kids are given lists of words to look up in a dictionary, and then they're tested on those definitions, and that constitutes vocabulary study.

So if the universe of vocabulary knowledge worth knowing is knowing words and knowing about words, what do we do about that? How do we start? How do we support students in getting access to that universe? So the Oxford English Dictionary has over 170,000 entries, and I think Paul is going to give you even a greater number when he talks, so we can't teach all those words in school. However if we focus on some key words, we might be able to make some progress. Most vocabulary researchers work with a segment of the language characterized as academic vocabulary or tier 2 words. These are words that typify academic text and they're found frequently across domains, so they appear commonly enough in texts, mostly, that some investment in instructing them might well pay off.

On-screen:  [What is the universe of word meanings worth knowing?  Here are some key words: analyze, fluctuate, inherent, consistent, duration]

Here are some examples of the kinds of words I have in mind. You'll notice they are general domain words. They don't apply just to one specific discipline. They're also abstract, and that's a typical characteristic of academic words, and this makes these kinds of words harder for students to learn on their own, independently, from reading or from conversation, something that most students can do pretty readily with concrete, very, very high-frequency words in the language.

So what do we do with these words if we're going to introduce those in school? Well, for a while now there's been a consensus that the kind of vocabulary instruction that would lead to comprehension enhancement has a couple of features. One is you need multiple encounters with the words you're teaching over a variety of contexts, and you need to give students opportunities for active processing, doing something more with the words than just receiving information in the form of definitions or even contexts.

So we can start by introducing words effectively, and here's one way that some colleagues and I developed in a recent vocabulary program that we implemented. We'd provide a rich context like, "Teresa was disappointed when Heather was her lab partner because she didn't know her very well. But it turned out that she and Heather were very compatible." And then provide a student-friendly definition for that, for the word, when people or things are compatible, they work well together and get along with each other. But then the even more important part is to task students with explaining how that word makes the context meaningful. So we asked them, "How does the meaning of compatible fit the context?" And we're looking for students to say something like, "Well, the girls got along really well when they worked together in science."

On-screen:  [What should we do: Practice words.  Focus on interactions around words!]

Then we have to get students to practice with the word. Engaging students in language around and about the word is critical, because what will boost students' academic achievement is enhanced ability to use their language, not isolated bits of information about individual words.

So what does engaging students look like? I'm going to give you a couple simple examples. These are real ones. So a teacher; it's a fourth grade classroom. The teacher asks the student, "Can you use eavesdrop and gregarious in a sentence?" Now these words had been introduced to the students. They'd been working with them. And one student responded, "If somebody reports you for talking, they're eavesdropping and you're gregarious." I think she pretty much nailed those words. Seventh grade classroom, students were given a writing prompt and they had to use six of the words that they had been studying. The writing prompt was, "If I had a trillion dollars." One student says, "If I had a trillion dollars," this should be italicized, but you can probably catch on to which are the target words, "it would be very complex to come up with an idea on what to spend it on. I would adopt a diverse group of kids from Africa, Asia, Europe and India. I would also buy two domestic hedgehogs. I would send $1,000 all over the world, which would be global," and my favorite, "I would make an invention that integrates pizza and swimming. I will reside in Hawaii." Now that might not hang together as a paragraph so well, but the use of the word is really pretty good and that's the point, get students to use their words.

So how are those active and different from the really standard, "Use your word in a sentence?" So when we first started doing vocabulary interventions, we'd say, "Okay, write a sentence using the word cozy," and the kids would come up with, "I am cozy." Thank you very much. Yes, we knew that wasn't going to work, so we had to change it up a little bit. We came up with a format of what we called sentence starters. "My cat looks cozy when." So in that example a student has to kind of imagine a scenario of cat coziness and then describe it, so that gets a lot more thinking going. Similarly if you ask a student to put two words into one sentence, like the eavesdrop-gregarious example, it changes the thinking process. They then have to come up with a situation that will accommodate both of those words. Or if you ask them to put a lot of the words, like six words under one topic for a writing prompt, it usually produces more thoughtfulness. It prompts a different kind of way of looking at using the words.

So if we're going to become a nation of word learners, we also need to work on how vocabulary knowledge is assessed because that really can drive the instruction. And I think that depends, that begins with thinking about what is it that we want to learn about the students' understanding of a word from an assessment? Historically vocabulary assessments, both standardized and those that teachers use in classrooms daily, were based on definitions like this one, distraught, and you choose the synonym for it.

On-screen:  [Consider….Distraught: 1. happy, 2. depressed, 3. upset, 4. hungry]

But think about whether you'd have more confidence that a student understood the word distraught if they could answer this or respond to an item like this, "Mrs. Thomas was distraught when she looked at her garden. How did her garden look? How does that fit with being distraught? So in this context we take a garden, which usually looking at a garden, it's something pleasant, but here we brought the idea of being distraught into garden-looking, so the student has to accommodate that into figuring out what was going on with Mrs. Thomas's garden. Maybe it was that rabbit. An item like this requires a different set of thinking processes than just picking out a synonym, but it's a much better representation of how we have to use words when we're reading and understanding text. We don't just bring meanings to mind as we're reading, but we have to integrate words within the context to come up with an understanding of the situation.

So as far as conceptions that should underlie assessments, we want assessments that capture words in use, like the Mrs. Thomas example, and words in relationship, assessments that capture word relationships. One prominent kind of relationship we want to capture, because we want to promote students' learning and understanding it, is morphological relationships. So first of all, derivational morphology. For example, if students learn the word satisfy, can they then understand the sentence, "Pam was happy with her purchase, but she became dissatisfied?" And then there's also lexical morphology, working with Latin roots. So if students learn the word diminish and the Latin root min that means small, can they then explain or have some access to a meaning for the sentence, "The girls talked about the minutia of daily life?"

Another kind of relationship we want to capture in assessment is the semantic relationship. What are the consequences of a word in context? Can students deal with that? One way to do that might be to provide a scenario, "We love to eat ice cream," and then challenge students to explain the consequences of different words in that scenario. Like, what's the difference between your mother monitoring how much ice cream you eat and your mother regulating how much ice cream you eat? Now this kind of task can be used either as an assessment item or as an instructional activity.

I'll leave the question of recent work that informs conceptions of vocabulary knowledge mostly to Paul, but I do want to mention one topic. Evidence continues to accumulate that interactive talk has a long-term impact on students' vocabulary knowledge and associated skills, reading comprehension, etcetera. Interactive talk most simply can be defined as talking, listening and responding. It's often characterized as including, as requiring making inferences and calling on sophisticated vocabulary. And one particular finding is recent work that provides strong evidence involving preschool students, so interactive talk in preschool, and this interactive talk was defined as requiring inferences and using sophisticated vocabulary. When preschool classrooms had that kind of talk it contributed to stronger vocabulary and reading for those students through fourth grade, so that's a big boost. So some basic ingredients for becoming a nation of word learners is talk to your kids, make sure that classroom talk is robust and gets kids to think about, talk about, and use their words, and this applies preschool through high school.

So some quick, easy ways to motivate that kind of talk in classrooms, just finding opportunities to keep attention on new vocabulary. So, for example, a teacher might start the day just using a word. "Today is Tuesday. Let's get inspired." Or to ask students a question that uses one of their words. "Was anyone reluctant to get out of bed this morning," and have a little bit of a discussion. Or teachers can use words, take the vocabulary and incorporate that into other kinds of lessons. So let's say the class was just read a story and the teacher asks, "Do you see any of the characters in that story as manipulative?" to get kids to consider that.

Or have some fun. What we did in a couple of the classrooms we worked with was post some pictures and challenge students to come up with captions for the pictures that used their words. Anybody want to take a try? Anybody got a caption for that? That's my own fat cat, by the way. No? Okay. [Inaudible comment from audience.] Yeah, exactly, yeah. All right, how about, "Have you heard the latest trend is taking your cat on vacation?" Or what I think he's really saying, my definitive answer is, "You're not going anywhere."

But there's also a need to get kids to take interactions with language beyond the classroom because there just is not enough classroom time to provide students all the experiences they need to develop a rich facility with language, especially for students who don't have that kind of rich language environment outside of the classroom, and we've done this in several studies, and we have data that it works. Just challenge students to find their vocabulary outside of school.

On-screen:  [Beyond the Classroom: Challenge students to find their words outside of school: in books, websites, newspapers, billboards, bus ads, radio, television, video games, magazines, menus, music, etc.]

Tell them they can find it anywhere, in books, in websites, billboards, radio. They can hear it or read it, magazines, menus. And then they bring evidence of that into school, post a big chart on the classroom wall, tally points for bringing in those words, make it public and motivating and you can get some kids into vocabulary. And finally, my final ingredient for becoming a nation of word learners is put down the cell phone and read, and this applies preschool through senility. And now I'm going to turn the floor over to Paul.

Paul D. Deane, Principal Research Scientist, ETS:  I'm going to go ahead and follow up on our first presentation by taking a dive into some of the linguistic issues that are connected with all the concepts we've been discussing.

On-screen:  [Becoming a Nation of Word Learners, ETS Forum, Paul Deane, Educational Testing Service, October 21, 2015]

So let me start by reviewing my take on some of the ideas that we've already been talking about from the previous talk. So one of the things that's really important is there's a lot of words, and there's even more word meanings and teachers can't teach all of that. And so everything has to be leveraged to make people more sophisticated word learners or they're not going to learn the words, because there's just a limited number of words that all the teachers in twelve years of school can actually try to teach students and teach anything else.

Something else that we really have emphasized is the fastest way to learn new words is to hear or use them in conversation, or failing that, to read them and do a lot of reading. Ultimately there's no substitute for rich and varied experiences with words. And the whole goal of becoming a nation of word learners must be to create an environment in society and in the classroom where a rich vocabulary is just part of life. If it isn't, it's going to be very hard, and that's why students who are from disadvantaged backgrounds can be challenged with acquiring vocabulary to the extent that their life does not already expose them to a rich vocabulary and does not encourage them to learn vocabulary as they go along, it's easier and easier to fall behind.

So let's now consider the issue of our first question, what's the universe of word meanings worth knowing? So the important thing to realize here is that this is something of a chicken-and-egg problem. In vocabulary the rich get richer. Students who come in with large vocabularies learn more words. Students who start behind keep falling behind. And so the major challenge educationally is how to help the students who do not come in with those advantages, how to create an environment both in instruction and assessment which encourages the right attitude and approach towards becoming a word learner.

So let's consider this chicken-and-egg problem from a very nuts-and-bolts perspective, how big a problem is it?

On-screen: [Graph: Just How Big is the Problem?  Vocabulary Size by Age.  Y Axis describes 'Vocabulary Size' from 0 to 200,000.  X Axis describes Core Working Vocabulary, Total Recognition Vocabulary, Total Word Meanings based on Age 7 and Age 16.  Age 7 has approximately 2,000 in the Core Working Vocabulary, approximately 15,000 in the Total Recognition Vocabulary, approximately 35,000 in the Total Word Meanings.  Age 16 has approximately 45,000 in the Core Working Vocabulary, approximately 85,000 in the Total Recognition Vocabulary and approximately 182,000 in Total Word Meanings.]

Well, this graph will give you a sense. I've got three bar charts here. The first one involves what is called core working vocabulary, which is simply words you have to memorize. You can't know them by knowing some other word. And in seventh grade, that's pretty small, several thousand words. And the data here, by the way, comes from a study that Jenny Sullivan, one of our conference presenters did, where she tried to estimate the size of vocabularies actually known by students at different grade levels. And so according to her estimates between the age seven and sixteen you're going from a vocabulary of several thousand words to well over 45,000 words that are part of core vocabulary. Those are words that you not only recognize, but you can reuse without relating them to any other word.

Total recognition vocabulary is receptive but reading vocabulary, that has to grow even more from around 20,000 to around 80,000 words in that same timespan according to Jenny Sullivan's study. And if you actually pay attention to the specific word meanings rather than to the words, there's an even bigger problem. You're going from around some 30,000-some thousand words at age seven to, word meanings, to well over 180,000 word meanings that sixteen-year-olds actually know. Many of those are quite, are differences that may seem fairly small, but are pretty important. If you consider a word like asylum, if you're talking about an institution or you're talking about what the people in Europe are currently doing, those are very different meanings of the word. If you use a word like alien, well, again that has different meanings in different contexts, and some of those meanings may be different uses of the same abstract meaning, but people have to learn all of those different uses. So there's a lot going on under the hood that involves a massive, massive growth in vocabulary during the school years.

So what is behind this massive growth? What are some of the things that account for the difference? Where do the big growths happen? Well, one of them we've already talked about. It's morphology. What that is, is being able to put together a word from its parts.

On-screen:  [Graph: Word Families Matter!  Morphological Types – Number of Meanings Known.  The Y Axis describes the 'Number of Word Meanings' form 0 to 80,000.  The X Axis describes Grades 2 through 11.  It is charted by 'Derived (Music>Musical)', 'Compound (Firefly<Fire+fly'), 'Inflected (Walk>Walked)', 'Root (Friend, Employ') and 'Idiom ("The blues, "Sick as a dog", "Pull no punches")'.  Grade 2 shows approximately 200 Idioms, approximately 5,000 Compounds, approximately 7,000 Roots and approximately 10,000 Inflected.  Grade 5 shows approximately 3,000 Idioms, approximately 10,000 Compounds, approximately 18,000 Roots and approximately 25,000 Inflected.  Grade 8 shows approximately 7,000 Idioms, approximately 23,000 Compounds, approximately 30,000 Roots and approximately 38,000 Inflected.  Grade 11 shows approximately 9,000 Idioms, approximately 37,000 Compounds, approximately 37,000 Roots and approximately 40,000 Inflected.  There are 2 other 'Derived' on the chart with approximately 64,000 and approximately 75,000 numbers of word meanings.]

And so here is another chart from Jenny Sullivan's dissertation work that really highlights one of the places where growth is really happening. You see that line with the red dots? Those are derived words, such as musical from music. There's other kinds of morphological relationships that don't grow so fast. So being able to form compounds like firefly from fire plus fly grows pretty linearly. Knowing inflections, well, that more or less grows at the same speed as the base words you know them from, which are the root words. And of course there's idioms and phrases that you have to learn and all of those are kind of growing linearly.

But being able to guess the meanings of derived words, there's a huge jump between second grade and eleventh grade in the number of derived words people know. And so it becomes; so knowing that students are able to hit a word like tautological and have any idea that that's related to tautology, and knowing that tautology has something, because ology is in it, that kind of knowledge, if someone has good language skill, happens automatically and tacitly and doesn't require a lot of effort. If you have someone who is struggling with reading and is having trouble even sounding out words with multiple syllables, those kinds of derived words become pretty hopeless.

On-screen:  [Metaphorical Extensions Matter!  Number of Multiple Types of Word Meanings Known.  The Y Axis has 'Number of Word Meanings Shown' from 0 to 90,000.  The X Axis has 'Grades 2 through 11'.  The items that are graphed are Homonyms (Yard, trunk, trip), Conversions (Bat, duck, match) and Metaphorical Extensions (Mouse, head, bug).   Grade 2 shows approximately 0 in Homonyms, approximately 5,000 Conversions, approximately 8,000 Metaphorical Extensions.  Grade 5 shows approximately 3,000 Homonyms, approximately 7,000 Conversions, approximately 28,000 Metaphorical Extensions.  Grade 8 shows approximately 8,000 Homonyms, approximately 12,000 Conversions, approximately 70,000 Metaphorical Extensions.  Grade 11 shows approximately 8,000 Homonyms, approximately 15,000 Conversions, approximately 80,000 Metaphorical Extensions.]

Another huge issue is multiple meanings of the same word, which in Jenny Sullivan's dissertation she called metaphorical extensions, though that actually includes a number of other things. So if you consider things like talking about a mouse, whether it's this kind or what's scurrying across the floor, whether it's the head of a trail or head of a person, whether you're talking about; there's a whole series of cases where words have related meanings. That's where a lot of the growth in vocabulary is happening in the school years according to Jenny Sullivan's work, whereas other kind of things like conversions, a cook cooks, well, that's easy, but there's not a lot of growth in that. And knowing let's say a bank is a financial institution or the edge of a river, another kind of multiple meaning, that doesn't grow that much either. Those things are kind of growing linearly at best, but learning new meanings of old words is a huge, huge factor in the increase in people's vocabulary knowledge during the school years.

So there's a number of things that are going on that we can put together to call, these are generative mechanisms. These are ways of leveraging what you already know to know even more. And to the extent that people develop the ability to do that kind of leveraging, much of which linguistically is happening without any conscious awareness, then the better word learners they are. So the problem from a vocabulary instruction and assessment point of view is you want to find out if people have this ability and you want to actually develop that ability in them. And a lot of the things that we are hearing about recommendations in the classroom are things that will help create the necessary skills that will transfer to learning new words, which ultimately means being able to deal with these many different related forms of a word and many different related meanings of the same word.

So one moral we can draw from this basic linguistic information is that effective word learners know how to make educated guesses about what words mean, and this is a skill that in young kids in an oral context is pretty much there. They learn a lot of words without needing any instruction. And to the extent you can create a rich context where people can learn those words without a lot of instruction, so much the better. But you've got students who are behind, who don't have that rich context. You've got to do something for them. A large part of it is building up that rich background knowledge and context, but there may be other things that you can do. We'll talk about that a bit.

Now too I'd like to drive some of these ideas home by moving to some vocabulary you would need to deal with in order to handle complex academic texts. So in particular, giving a bit of a bow to some things in the news, I'm going to take a look at things having to do with immigrants.

On-screen:  [Middle School Student Judgments about Words Statistically Related to Citizenship.

Know the word but don't think its's related

Know the word and think it's related

  1. Percentage of Students Indicating They Know the Word




Exile (Root Word) 97% A; 28% B

Status (Root Word) 98% A; 45% B

Identity (Derived Word) 100% A; 69%B

Borders (Root Word) 99% A; 80% B

Migrant (Derived Word) 98% A; 83% B

Immigrant (Derived Word) 98% A; 90% B



Refugee (Derived Word) 91% A; 48% B

Nationality (Derived Word) 97% A; 71% B

Illegal (Derived Word) 97% A; 79% B

Citizen (Root Word) 97% A; 89% B

Emigrants (Derived Word) 93% A; 82% B



Aliens (Extended Meaning) 83% A; 51% B

Undocumented (Extended and Derived) 90% A; 70% B

Naturalization (Extended and Derived) 90% A; 63% B

Foreigners (Derived Word) 90% A; 91% B

Multicultural (Derived Word) 86% A; 80% B




Ethnicity (Derived Word) 73% A; 73% B



Deportation (Derived Word) 86% A; 70% B


Asylum (Extended Meanings) 63% A; 30% B

Expatriate (Extended and Derived) 65% A; 50% B














  1. Percentage Judging Word to be Related to Citizenship


So this is vocabulary you might see in the newspaper or in a text about immigrants, but this chart is actually based upon a study I did as part of a grant a couple of years ago where we asked students to make judgments. We gave them words and we said, "Tell us if this word is related to, not related to, or I don't know the word at all, with respect to a topic." So we gave them a topic of citizens, of being a citizen, and we said, "Which of these words are related, not related, or if you're not sure you can just say I don't know the word?"

And so this chart is basically, on one dimension it's the percentage of kids who said they knew the word. That is they made a judgment either related or not related. And the other dimension is the percentage of those students who made a judgment who said it was related. So what you can see is that in this middle school population there's a cluster of words in the upper right-hand quadrant where the students know they're related to the topic and they know that, and they know the words. There are other words that they, that those who know them know they're related, though only a few such as ethnicity. There are other words such as expatriates or asylum that most people don't know and those who do don't think it has anything to do with the subject. And then there are others that people think they know, but don't think they're related.

Now what's interesting is we think about the different kinds of generative mechanisms that are at play here, that the words that are in white are root words, and in this kind of academic vocabulary, there aren't a lot of root words. Almost everything is either a derived word, an extended meaning from a known word, or both extended and derived. So if you look at a word like undocumented, that's a derived word. You have to know it's related to document, and you'd have to know the prefix, you have to know un means not, you have to know what ed does. But even if you know all those things, if you say, "Oh, undocumented means you don't have documents. So it might be you'd say, and this is where say the classroom exercises that we were hearing about in the previous talk would sort of come in, "Use the word undocumented in a sentence." With no background, what is the kid going to do? "Sally came to school undocumented," because in fact you have to know that this is referring to a specific kind of document, citizenship or other kinds of legal documents of that sort. That's what the meaning is actually referring to.

A whole series of derived words that don't have that extra wrinkle, so words like identity, nationality, migrant, illegal, immigrant, emigrant—notice the word family there—foreigners, multicultural, deportation, ethnicity, refugee, all those words, if you have some knowledge of the morphology of the language, anyone with a core vocabulary, even if you haven't encountered this before, you'd make a pretty good guess. But then we also have extended meanings. So if you look at words like asylum or aliens, well, these are kind of on the hard category. A lot of kids in this middle school study didn't know these words, and those that did didn't think they were related to the topic. Why? Well, because they knew another meaning and that meaning blocked their ability to learn or recognize that this word was related to the topic.

And so one of the issues that you run up against, and I've seen studies that kind of highlight this, is that you can have struggling readers who assume that a word means the meaning it already has, and the context of that meaning doesn't make any sense at all. And so one of the issues that's really important is developing an awareness of multiple meanings and a willingness to be flexible and try to figure out what a word means in a context, and that ability is really important.

Okay, so this is a kind of a picture of the nature of the problem. So let us now consider how this might link in to issues of assessment and the relationship of assessment to instruction.

On-screen:  [Part II.  What conceptions of Vocabulary Should Underlie Assessments? What Recent Work Can Inform these Conceptions?]

So one of the things that is going to come into this is related to work I've been doing on other topics of vocabulary, which is thinking about the fact that what you assess has a big impact on what people teach. And so if vocabulary is done as it was say in the SAT when I was growing up, with a lot of analogies and other kinds of hard things, that there was a certain strategy that made sense, memorize lots of definitions. And what was the effect? Well, one, the effect was not to get people to go out and read a lot of books, which is actually the good way to improve your vocabulary.

On-screen:  [A Starting Point…Shouldn't assessment support effective instruction?]

So we need to think about how assessment can support effective instruction. What kinds of assessments do you want for what purposes and how can you structure them to make them more useful?

So for a starting point I'd like to look at a diagram that comes out of some research that's been done on effective vocabulary instruction to get some sense of the richness of what we ought to be thinking about when we're trying to assess vocabulary.

On-screen:  [Chart with circles and boxes stemming off of other circles and boxes.  Middle circle says 'Multi-Faceted, Comprehensive Vocabulary Instruction Program'.  There are 4 boxes stemming off of this that say 'Provide Rich and Varied Language Experiences', 'Teaching Individual Words', ' Teaching Word Learning Strategies', 'Fostering Word Consciousness'.  Then, there are 9 circles stemming off of these boxes.  Stemming off of the box labeled 'Provide Rich and Varied Language Experiences' there are 3 circles that say 'Character Traits', 'Vocabulary Graphics', 'Read-Alouds Independent Reading/Writing'.  Stemming off of the box labeled 'Teaching Individual Words' there are 2 circles that say 'High Frequency Words' and 'Academic Words'.  Stemming off of the box labeled 'Teaching Word Learning Strategies' there are 2 circles that say 'Context Clues' and 'Morphology'.  Stemming off of the box labeled 'Fostering Word Consciousness' there are 2 circles that say 'Word Consciousness Lessons' and 'Word Play'.]

So this is work that comes out of a number of people, including Bauman, Manyak(?), Blacktivus(?), Graves and a number of others, and what they're emphasizing is that there's a lot of things that go into effective vocabulary instruction. It isn't just teaching individual words. It isn't just teaching word-learning strategies. It's also providing a rich and varied language experience that makes that word meaning and use actually something that matters and fostering a word consciousness, a metacognitive awareness of vocabulary, a willingness to think about vocabulary, a willingness to reason, make inferences about vocabulary. That kind of rich experience, including that metacognitive awareness, is really, really important to becoming a word learner.

So in the vocabulary conference we tried to develop a definition of what we think the construct is that should be assessed and taught when we talk about vocabulary, and it's not just definitions.  So, yeah, it involves specific knowledge of words, but that specific knowledge is really rich. It doesn't just involve definitions. It involves understanding of how words are used in context. It involves understanding of its relationship to other words. It involves understanding of multiple meanings and relationships to one another. It involves knowing the morphological relationships in words. There's a whole range of rich knowledge that knowing a word really stands in for, and all of that ties into knowledge about concepts in the world. So there's a very strong link between knowing a lot of words about baseball and knowing a lot about baseball, for example. So there's a whole series of things where the more people know about a lot of different subjects and have vocabulary that ties it, how to use that vocabulary to write and read and talk about it, the more likely they are to develop this rich specific knowledge.

Tacit knowledge of generative patterns, that's something I emphasized earlier in the talk. Being able to guess at extended meanings and derived words is really important, and that depends upon a really rich unconscious base of knowledge about a lot of words and their relationships to one another. That can support conscious strategies for effective word learning, which can help provide those other things are there. And ultimately when you're using words, you're not just using them to know words, you're using them for other purposes as part of rich literacy activities. And that rich metalinguistic awareness, when it's integrated with other literacy skills, creates a much higher likelihood that people will really learn words. That's where the example about distraught, about the garden, comes in because understanding distraught is knowing how to make inferences about people's social reactions, and if you can't make those inferences when you hear the word, it doesn't matter if you can recite a synonym.

So some of the implications that this has, if we want assessment to support instruction there are things we should be thinking about. Does the assessment reflect the richness of the underlying construct? If you just have one type of question dealing with one kind of knowledge of a word, such as synonyms, and that's all you ever test, you're encouraging and narrowing what people's conception of vocabulary is. You should also worry about measuring students' word learning ability. It's not just how good is a student at knowing words they already know. You care about whether they have the disposition to pick up word meanings when they're exposed to them. If they don't have that disposition, all of the instruction in the world is probably not going to help them that much because there's too many words out there.

Ultimately a teacher needs to know patterns of strength or weakness, know where an intervention is actually going to help. So you might want to know, for example, if you have a lot of ELL students, who need help with English morphology because their languages don't have a lot of morphology, or they're from a language that isn't romance and, therefore, a lot of cognates they're not going to know from their own language. And ultimately, and I think we saw a lot of examples of these in the first talk, we should have assessments that model effective classroom practices. So these are some of the implications that the people in the conference more or less came up to a consensus. What I'd like to do is draw on some of the research from the conference presentations to give you a sense of some of the kinds of things that might move in the directions we're talking about.

On-screen:  [Example:  Probing Many Different Aspects of Vocabulary Knowledge (VASE Project, Scott et al., 2008).  There is a list labeled 'Fill in the Small Circle beside the Correct Answer (Only one answer).  The first question on the list is 'How well do you know the word official?'  The answers can be 'I'm sure I know what this word means', 'I think I know what this word means', 'I am not sure what this word means' or 'I don't know what this word means'.  The second question on the list is 'I think that these words are related to official':  The answers can be 'Slowly, inefficient, dragging on', 'Suggesting, guidance, informative', 'Jewelry, ornamental, wearing', or 'Rightful, endorsed, permission'.  The third question is 'Which word do you think is a correct word connected in meaning to official?:  The answers can be 'Reofficial', 'Officialen', 'Unofficial' or 'Officialest'.  The fourth question on the list is 'Use the sentence that you think makes the most sense': The answers can be 'There is an official website for the new movie', 'Any outfit looks fancy when you put on a fine official', 'The teacher wanted to official the student about going to college', or 'Jessica is so official that it takes her hours to do a simple job.'  The fifth question on the list is 'I think the word official is closest in meaning to':  The answers can be 'Giving good advice to someone', 'Unable to finish something in authority', 'Approved by someone in authority' or 'A large decorative pin or clasp'.  A sentence in between the questions says 'Some words have one part of speech, but some have more than one.  Fill in the circle beside the correct answer or answers.'  The sixth and last question on the list is 'I think the word official may be':  The answers can be 'A noun', 'A verb', 'An adjective', or 'An adverb'.]

So this is a set of items from a test of vocabulary depth developed by Judith Scott. And so the critical thing is that while the individual items are somewhat traditional looking, the range of things they're testing is not at all traditional to do it all for one word. So notice the first question deals with whether or not you even know the word. Then, do you understand something about the word's meaning? Then can you; let's see, sorry. Okay, so the first one is do you know the word? The second one is dealing with word associations. The third one is dealing with morphology, with knowing which is a correct derived word. The fifth one is dealing with using the word in a sentence context. Then you have connecting it to a definition, and finally some explicit knowledge about part of speech. So the key thing in an assessment like this is that it's going to give you a much richer picture of the words that it tests. It does so at a cost. It's testing fewer words, but it's modeling a much more typical range of things that somebody actually has to know in order to know a word.

Now what I put up here is a screen from an experimental tool that Gwen Frishkoff and Kevyn Collins-Thompson have developed to assess students' word, ability to learn new words. Now the way this tool works, a word is presented, and in this particular case it's the word operose, which they chose because nobody knows it, and so any guesses people make will be pure word learning ability. It will have nothing to do with whether or not they know the word. But of course in a more realistic context, you'd pick more realistic words because a lot of students won't know those more realistic words. But in this context what the experimental study wanted to know, given contexts and multiple chances and feedback, how quickly do people zero in on a particular form of the word, a particular meaning of the word.

So they're given a context, they're given the word, asked for a guess. They make a guess. There's an automated system that looks at their guess and says how close is it to the word that we're testing? And they're told, "Yes, you got it," or, "You're sort of close," or, "No, you're way off," and they're given a new context and another chance to guess at the word, and they do that multiple times until they get it. And so the really interesting thing is that this kind of tool gives you a sense of how quickly students can focus in on what a word means from contexts, and it gives you a sense of what contexts are actually helpful and useful in enabling students to guess the word, and they've been developing technology to try to classify contexts to support this kind of practice and learning.

On-screen:  [Patterns of Strength and Weakness-(RISE, Sabatini, Bruce, O'Reilly, RISE)]

So another thing that is worth thinking about is how to provide a richer assessment of vocabulary knowledge that links into things like reading abilities. So some of my colleagues, John Sabatini and Tenaha and O'Reilly and others, have developed an assessment that's essentially upon a skills test for reading, but most of the items are really vocabulary items. So is that a word? You're given it so it might not be; it might be a real word, it might be just nonsense syllables. Do you know if it's a word?

Can you give some kind of synonym information? Can you use it in a sentence? Can you figure out which word should be used in a context in a passage and so forth? So there's a series of vocabulary-related tasks that are gradually upping the ante and require richer and richer vocabulary knowledge. And the critical thing with this is that it provides you a profile that can be then used to identify where students are running into weaknesses.

On-screen:  [This is a page from a textbook.  The title is 'Floating Eggs'.  Here is the text:  Each spring a new life cycle begins as frogs lay their eggs.  Frog eggs are called spawn.  Most spawn is found in calm, shallow water.  It sticks together in clumps of up to 4000 eggs. Only about half of the eggs will hatch.  Many will be eaten by turtles, fish, and other animals.  Even if an egg does hatch, the tadpole may be eaten before it becomes an adult.  Frogs lay a lot of eggs to make sure that at least some of their young will survive and grow into adults.  The next page has questions:  Mr. Johnson: When you define something, you should always start by saying what kind of thing it is.  What kind of thing is spawn a word for?  Marie:  A new life cycle?  That's the first think the passage talks about.  Jaime:  A kind of food?  Turtles and fish eat them.  Sarah:  What about eggs?  It does say that frog eggs are called spawn.  The directions at the bottom of the page say: Directions: Which student has the best idea of what spawn is?  Select the number of the correct answer (1, 2, or 3) and drag it into the box.]

This last example, it comes from the work we've been doing on CBAL™. This is from a fourth-fifth grade assessment on the topic of frog life cycles, and it's part of a much richer, larger scenario where students read a book about frog life cycles and walk through a whole series of steps, including pre-reading, making inference about vocabulary, doing summary and paraphrase, and ultimately sharing what they've learned with other people. And in this particular case the thing I wanted to focus on is that we've tried to present a vocabulary item in a way that contextualizes it as why would you care about doing this at all?

So anybody could be given a test item where you're asked, "What's the right meaning of this thing in context?" But here we've modeled a conversation about a text, and we're asking students to assess the context clues that are pointed to by the different students and pick which context clue actually gives the best information about the word. So you can imagine that if you have assessments that sort of model the relationship between what you're assessing, what are people going to do with the knowledge where they're in a classroom context, that that's giving a certain link back so that if someone then takes the assessment and says, "I want to do something to teach the kids these words," they're more likely to be modeling effective classroom practices.

So to sum up everything that we've talked about so far, vocabulary knowledge is critical, but there are too many words to teach one by one. So it's really important to build up the kind of classroom and context and social environment in which word learning is natural, and that's pretty much the sum of both of our presentations. And so ultimately the best goal for vocabulary instruction assessment is becoming a nation of word learners, which means a lot of different things. It means really encouraging rich interactive conversation and talk with rich vocabularies in the classroom, really encouraging people to read, assessing vocabulary in ways that really focus on the richness of the vocabulary knowledge you have, that encourage people to understand why they would care about vocabulary in ways that really link back to this rich understanding of the world that a rich vocabulary knowledge ultimately gives you.

End of Video