Avoid teaching sematically-related words together

Work Teaching English in Taiwan
1

Often, textbooks will present you with a bunch of new vocabulary from a specific lexical set, for example “clothes”. What I’ve usually done in the past is to break up the vocab into small groups and teach them over a few lessons leading up to the one in which students encounter the vocab set in their books. In this way, the book page serves as useful review rather than overwhelming new language.

However, I’ve recently found out that there may be even better ways to teach vocab: by completely avoiding teaching semantically-related items in the same lesson. A number of studies have found that teaching vocabulary from the same lexical set at the same time results in interference and confusion, making the vocabulary harder to learn. What is often better is teaching a group of words which have some kind of story, thematic, or discourse relationship to each other. Off the top of my head, one example group could be “ugly,” “sister,” “ballroom,” “prince,” “glass,” “slipper,” and “pumpkin!”

Here’s one of the studies I mentioned:
www1.harenet.ne.jp/~waring/papers/Sys2_97.html

Waring gives some practical suggestions for implementation of the findings; using this principle in the context of standard textbooks. Obviously it isn’t always possible to separate lexical sets in this way, however, and other remedies such as ensuring rich narrative or other discourse contexts may well reduce the interference effect.

2

I’ve also read (in a textbook about how to teach vocabulary) that learning similar words at the same time is more difficult- the authors suggested that vocabulary should not be taught in semantic groups - ie., don’t teach the kids “zoo animals” all in one go - and that words similar in form shouldn’t be taught together either - ie., that teaching ‘to get married’ ‘marriage’ ‘married’ ‘to marry’ is not helpful.
I have found that this advice is true in my own learning of languages - as a beginner in Japanese I got ‘inu’ and ‘neko’ (dog and cat) mixed up, and the words for ‘yesterday’ and’tomorrow’.
The problem is textbooks always present the vocabulary in semantic groups. The kids’ textbooks at Kojen for example taught all vocabulary in groups. As a teacher, I especially hated the clothing chapter - shirt, skirt, shorts, socks and shoes - where one can definitely see that the vocab. is confusing the students.
In my Chinese class the teacher always ‘helped’ us learn the vocabulary by presenting words related to the word on the list. I felt it was useless, because I could never remember which of the very similar words to use in which context.

3

[quote=“bababa”]I’ve also read (in a textbook about how to teach vocabulary) that learning similar words at the same time is more difficult- the authors suggested that vocabulary should not be taught in semantic groups - ie., don’t teach the kids “zoo animals” all in one go - and that words similar in form shouldn’t be taught together either - ie., that teaching ‘to get married’ ‘marriage’ ‘married’ ‘to marry’ is not helpful.
I have found that this advice is true in my own learning of languages - as a beginner in Japanese I got ‘inu’ and ‘neko’ (dog and cat) mixed up, and the words for ‘yesterday’ and’tomorrow’.
The problem is textbooks always present the vocabulary in semantic groups. The kids’ textbooks at Kojen for example taught all vocabulary in groups. As a teacher, I especially hated the clothing chapter - shirt, skirt, shorts, socks and shoes - where one can definitely see that the vocab. is confusing the students.
In my Chinese class the teacher always ‘helped’ us learn the vocabulary by presenting words related to the word on the list. I felt it was useless, because I could never remember which of the very similar words to use in which context.[/quote]I think that’s exactly the same thing that the research shows. And I agree that clothing is one of the worst lexical sets in this regard. What I’ve always done to minimise the confusion is pre-teach such vocab, 3 words at a time, over several lessons, usually starting with TPR and progressing to more output-based stuff. But it seems from the research that even this isn’t ideal. I find it very interesting that in the part of Waring’s study where he taught six words – three from the same lexical set and three from other sets – learners tended to confuse the three from the same set with each other, but not with the words from other sets. Waring used made-up words, but I guess a real-life example would be something like teaching “shirt,” “skirt,” “shoes,” “tennis,” “frog,” and “park,” at the same time. Students might tend to produce “shoes” in place of “skirt,” but not in place of “tennis,” “frog,” or “park”.

I think it’s also a bad idea to teach synonyms or antonyms together, for the same reason. Work on synonyms, antonyms, lexical sets, and other “semantic elaboration” is fine once learners already have a basic feel for the particular words they’re working on, but not as a first step.

4

What level of student are you referring to here?

With my supposed-to-be-advanced adults recently I’ve been doing the opposite. At that level they’re supposed to be able to figure out from the context whether a given word is a noun, verb, adjective, etc. I’m not actually teaching them new words, but I am expecting them to figure out whether to use ‘to get married’, ‘marriage’, ‘married’, ‘to marry’, etc.

It may be confusing, but they’re not going to get far if they can’t infer meaning.

5

[quote=“tmwc”]What level of student are you referring to here?

With my supposed-to-be-advanced adults recently I’ve been doing the opposite. At that level they’re supposed to be able to figure out from the context whether a given word is a noun, verb, adjective, etc. I’m not actually teaching them new words, but I am expecting them to figure out whether to use ‘to get married’, ‘marriage’, ‘married’, ‘to marry’, etc.

It may be confusing, but they’re not going to get far if they can’t infer meaning.[/quote]I don’t see any problem with what you describe. I think that these semantic elaboration activities are good as a second stage in vocabulary learning, to deepen and broaden lexical knowledge.

I don’t think that’s the kind of vocabulary set that was referred to. If you were actually teaching completely new words, and you wanted to teach “married,” “going out,” “hitched,” “divorced,” “split up,” and “separated” in the same lesson, maybe that would create some confusion.

Did you read the article? It’s very interesting.

6

I didn’t read the article but I know what you are talking about.

I am writing a book called “This is an instruction book” for low level adults designed not only to conextualize all the vocab but to require that my students actually be able to “do” things in order to move ahead through the book. It is complicated but fascinating to me and over diner I started boring my wife with the details of it when she got that glazed over look again. I said “It’s not just me you know? There is this guy at forumosa who is fascinated by scientific proofs of all the theories that get batted around. Today he said that it was a mistake to sort words into catagories before the meaning of each word has been learned individually. The mind naturally seeks out that sort of orginaization but if you provide it too soon the units get confused. For example I know that I learned colours that way. All together with translations of each word and even now, years later, I still get confused and think that lu4se4 means green and lan2se4 means blue.” My wife thought that was pretty funny.

7

My school has a great way to deal with this built into our specialized curriculum: thematic monthly themes so that you can fit all the words together and tell stories, but not overlap them semantically. The themes are aimed at kids, so they get into learning a set of vocabulary that enables them to talk about X kid-friendly and relevant topic, but without realizing it we are nefariously teaching them key words that will make them learn how to read more serious things when they get older. They seem to be making better progress than ones in other schools, so I have to conclude it works.

8

[quote=“bob”]For example I know that I learned colours that way. All together with translations of each word and even now, years later, I still get confused and think that lu4se4 means green and lan2se4 means blue." My wife thought that was pretty funny.[/quote]I find that lots of kids confuse twelve/twenty and kangaroo/giraffe, maybe for the same reason.

[quote=“superemma”]My school has a great way to deal with this built into our specialized curriculum: thematic monthly themes so that you can fit all the words together and tell stories, but not overlap them semantically. The themes are aimed at kids, so they get into learning a set of vocabulary that enables them to talk about X kid-friendly and relevant topic, but without realizing it we are nefariously teaching them key words that will make them learn how to read more serious things when they get older. They seem to be making better progress than ones in other schools, so I have to conclude it works.[/quote]Do you mean that, for example, in one story you’d teach a sport, an animal, and an item of clothing? That would seem to work well. Actually, even putting items from the same lexical set together in a story might not be so bad. It seems to be when studying contextless items from the same set together that problems occur. Of course, contextualised learning is a very good thing overall. But I’d say that supplementing it with some contextless vocabulary study, such as individuals using flashcards, is an efficient way of learning. It’s difficult to ensure sufficient repetitions of a wide range of vocabulary given natural contexts alone.

9

I teach vocabulary through incidence. I have a list of the words the students come across in their textbooks (one textbook of language arts with context to arts, sciences, culture, etc. and one textbook of world history or science depending on the semester they are in, plus a young adult novel) and add words that also come up in class… one of the first words that usually comes up, thanks to there always being one socially inept kid no matter where I teach, is “antagonize”. Since it’s a transitive verb, it’s easy to teach them how to use it correctly. It also works as a behavior deterrent for the kid at whom it’s aimed.

The problem with teaching a semantical group of words is that it’s believed to work since it’s “within a context”. The only thing is, for most lessons in books the context is an artificial one. How often do we go shopping and talk about all of the clothes that we want to buy? How often do we go to a restaurant and ask about every dish on the menu? However, if you were to take many lower-level English (and other languages are often as guilty) books as the gospel truth, you’d believe that people walk up to everyone at their school and introduce themselves and ask for their phone numbers, after knowing someone for one chapter they tell you everything about every member of their families, that they do 20 different things every day, they go to 15 different places 15 different ways (and talk about them all with their classmates), etc.

Granted this is done so when the students get into those situations, they can at least make conversation with one of the sentence patterns they had to practice and tweak the vocabulary to match their situation, but it is more meaningful if it is presented in a realistic situation rather than one big gob of vocabulary “within a context”.

This is also why almost all of the Chinese vocabulary I’ve learned has been outside of a textbook - because they are words I learn in a real-life context rather than an artificial one in a book.

10

[quote=“ImaniOU”]The problem with teaching a semantical group of words is that it’s believed to work since it’s “within a context”. The only thing is, for most lessons in books the context is an artificial one. How often do we go shopping and talk about the clothes that we want to buy? How often do we go to a restaurant and ask about every dish on the menu? However, if you were to take many lower-level English (and other languages are often as guilty) books as the gospel truth, you’d believe that people walk up to everyone at their school and introduce themselves and ask for their phone numbers, after knowing someone for one chapter they tell you everything about every member of their families, that they do 20 different things every day, they go to 15 different places 15 different ways (and talk about them all with their classmates), etc.[/quote]I hate those fake texbook dialogues too. And my worst bugbear is 2-dimensional “characters” who in fact are so characterless that even I can’t remember their names!

[quote=“ImaniOU”]This is also why almost all of the Chinese vocabulary I’ve learned has been outside of a textbook - because they are words I learn in a real-life context rather than an artificial one in a book.[/quote]I know what you mean, but there’s good research evidence for the benefit of some decontextualised vocabulary learning, such as self-study with flashcards. And my own experience backs that up. I’ve learned nearly 1000 words in the last six weeks or so (mainly 2-character combinations so obviously a lot of the same characters are used in several words). This study is already having a beneficial effect on my communication in Chinese. I can’t produce all the words at will. But I can produce some of them in real communication, and I often recognise others in speech and writing.

11

This has made me realize why I hated teaching the lower level textbooks at my school. There are the usual suspect of characters, but it doesn’t help that #1 they have no real personalities (which is better than Joy’s which goes so far as to even give them blood types and horoscopes and really pathetic dialogue with strict sentence patterns) and #2 they are drawn very differently by different artists. I like to call the characters in the last section of the 2nd level book, “the football-heads” (as in American football).

12

Despite the evidence that grouping semantically-related words together makes them more difficult to learn, I’ve found that there’s a lot of resistance to changing this method of textbook organizing; especially in Taiwan. Parents, teachers, schools and publishers all feel a certain safety with grouping by theme. The fact that learning English in Taiwan has to conform to the constraints of “The Business of Selling English” makes it almost unavoidable in fact.
The constant demand by parents to know “What did you learn this week?” finds comfort in the simple answer of “colors”, rather than “Frog, car, rain and walk.”
Teachers also find it makes keeping track of what their students have learned, and ensuring adequate review, much easier when thematic groupings are used. Even when presented with material that is not grouped this way, I’ve seen many teachers reorganize it so that it is.
And given this preference of parents and teachers, the schools and publishers simply follow the money.

I did a lot of research this Fall on Kindergarten programs in North America and found a lot of similarities in how the best programs organized their themes. Instead of having a whole week dedicated to colors for example, only one color would be chosen for a week. This then would serve as a way to group together several otherwise non-semantically related words. Ex: Red: apple, firetruck, sock, and bird.
Other common weekly themes that worked well this way where:
Shape for this week
Letter for this week
Season for this week
etc…

Luckily in Taiwan, most English programs do not consist of simply one textbook, but also phonics books, alphabet books, songs, etc…
So, by the time students get to the “Zoo Animals” unit, they have usually encountered at least half the words previously, such as elephant, hippo and monkey, as alphabet nouns. It’s certainly a good idea to examine a list of new words though, and determine how many of them are review vs. new to determine how you will group them for teaching.

I personally like using the method Joesax advised of laying out flashcards with several different theme-groups represented, and having students use them as cues to create their own sentences. Much more interesting than simply showing all clothes. Am constantly surprised and amused by the different ways my students choose to make stories with the different flashcards I lay out.

13

[quote=“Ktownboy”]I personally like using the method Joesax advised of laying out flashcards with several different theme-groups represented, and having students use them as cues to create their own sentences. Much more interesting than simply showing all clothes. Am constantly surprised and amused by the different ways my students choose to make stories with the different flashcards I lay out.[/quote]What I had in mind when I mentioned flashcards was review by individual students in their own time. I don’t think I’ve tried doing what you describe, but it sounds interesting. Do you mean that you use a random selection of vocabulary – one colour, one animal, one food, etc., and get students to generate little stories based on that? I think that could work well. Setting artificial limiting parameters like this can really get the creative juices flowing – sometimes more so than when starting with a completely “blank page”!

[quote=“Ktownboy”]Teachers also find it makes keeping track of what their students have learned, and ensuring adequate review, much easier when thematic groupings are used. Even when presented with material that is not grouped this way, I’ve seen many teachers reorganize it so that it is.[/quote]Right. To be fair, it is difficult to devise systems which ensure adequate review of all new vocabulary, and thematic grouping does make the task easier. Ideally, students would have their own individual vocabulary review systems going, but that tends to work better with motivated adults than with kids. I guess a way to overcome this, provided a school had the resources, would be to set up a computer review program such as Supermemo on multiple PCs. Kids would have their own login IDs, and the program would optimise vocab review for each one.

Still, at a basic level it’s easy enough to look through a course ahead of time and plan a new vocab and review schedule. And unplanned incidental new vocabulary can be noted down and also reviewed in subsequent lessons.

I think that your point is useful in a wider sense too: as teachers we should try to avoid practices which seem to make life easier for ourselves but are not optimal in terms of students’ learning.

[quote=“Ktownboy”]I did a lot of research this Fall on Kindergarten programs in North America and found a lot of similarities in how the best programs organized their themes. Instead of having a whole week dedicated to colors for example, only one color would be chosen for a week. This then would serve as a way to group together several otherwise non-semantically related words. Ex: Red: apple, firetruck, sock, and bird.
Other common weekly themes that worked well this way where:
Shape for this week
Letter for this week
Season for this week
etc…[/quote]I’ve always liked Sesame Street (at least the version for native English-speaking kids), and that follows much the same principle.

[quote=“Ktownboy”]Luckily in Taiwan, most English programs do not consist of simply one textbook, but also phonics books, alphabet books, songs, etc…
So, by the time students get to the “Zoo Animals” unit, they have usually encountered at least half the words previously, such as elephant, hippo and monkey, as alphabet nouns. It’s certainly a good idea to examine a list of new words though, and determine how many of them are review vs. new to determine how you will group them for teaching.[/quote]That’s true. On a little bit of a tangent, though, the presentation of new words in phonics books is a pet peeve of mine. A common practice seems to be to present a bunch of new and relatively obscure words, then immediately abstract phonemic principles from them. My experience tells me that it’s much better for kids to really know a word (that entails having encountered it multiple times in a variety of contexts) before they use it as a basis for phonemic abstraction. And it would make sense to choose useful, fairly common words for this purpose wherever possible, rather than obscure or ambiguous ones.