[Back to Main Page]

Language Zoo

"L1 conversational fluency does not come from the use of a lexicon of difficult words, nor from simply the most common words of the language, but from a repertoire made of the most common words."
          -Michael Lewis, Implementing the Lexical Approach, (1997)

abit absolutely actually afraid agree a little allowed to a lot
any chance appreciate at least by the way can I can't quite can you concern
consider could have could I could you decide deserve disagree doubt
else enjoy even if exactly extremely fact fairly far too
favor find out free got grateful great guess hardly
have to admit hold on hope impossible instead interest in touch just
kind know late lets like may may have mean
much must must have need now over part perhaps
personally plan point possible pretty problem promise question
quite rather1 rather2 realize really regret right say
seems seem to should have slightly so sorry sort of success
suggest sure tend to thank think try wait want
waste welcome will have will you wonder won't you worry would have
would I would you wrong you know you mind      


This "zoo" of language consists of utterances taken from TV show dialogues. I've selected the utterances that would have a fairly clear conventionalized meaning to a native speaker. I've also selected several purely for their amusement value.

These utterances provide a starting point for writing realistic dialogues. They should be useful to English teachers or writers trying to writing dialogues or translators trying to find the right way to express an idea.

English teachers can use these sentences to play a sort of guessing game with their students in which the students recreate possible contexts and unravel the potential meaning and illocutionary force of the utterances, an activity that can ultimately be extended into a creative activity in which students create their own dialogues around the utterance.

The book "Variations on a Theme" (Maley and Duff, 1978) provides a precedent for this sort of guessing game. The aim of the material in this book is to "imagine that we have broken in on other people's conversations and that we are interested in finding out about what they have been talking about or are still discussing." (10) The book seeks to develop "the ability to interpret fragments of speech, to listen for clues in what is said, and develop sensitivity to what is not said." (9) These goals are equally applicable to the sentences found in the zoo.

I put this zoo together to show what can be done easily with some simple concordancing tools. All the tools are available at this site. The utterances were extracted out of the "tvcorpus" with the "grepit" Perl program and formatted into HTML tables with the "grepprep" Perl program and then edited so that only the interesting utterances remained.

After creating this little language zoo I happened to find Michael Lewis's book "Implementing the Lexical Approach" (1997) in the Chiangmai University bookstore. I was surprised to find that he addresses explicitly what I hoped to achieve with this little zoo. He points out that many common words (e.g. mind, way, think, point, and have) "carry little meaning in themselves":

"Such words hardly have an existence independent of the multi-word phrases and expressions in which they occur. These phrases are mini-idioms, where the meaning of the phrase is not transparent from the component words, so, unless the teacher specifically draws them to the learners' attention they may not be noticed." Lewis (1997,24)

It's a common observation that the dialogues of plays or movies aren't genuine. They're not the way that people actually speak. Purists will provide you with complicated markups with length of pause and tone, speech overlap, along with all the hems, haws, you knows, and other fillers.

I find that anything other than normal punctuation and the use of commas or series of three dots "..." to represent pauses or tonal units is distracting to the student. The speakers in a real dialogue usually have fairly narrowly defined shared context that's often hard for native speakers to unravel much less language students. The speakers also usually also have a full face-to-face visual context working for them also. For these reasons strict conversation analysis seems to be a recipe for distaster in the classroom even if you take the advice of Brown and Yule (1983) in their book on teaching the spoken language and attempt to make what you do with the dialogue interesting rather than make the dialogue itself interesting.

Anyway, what's wrong with a little entertainment? Or making dialogue as realistic and useful for the student as possible, but also having the intention to amuse. I was surprised to find out that the textbook with the most entertaining listening texts that I ever used, Longman's "High Impact," was written by the prominent SLA (Second Language Acquisition) theoretician Rod Ellis!

If a dialogue is engaging or entertaining or just makes the student reflect a little bit on what has been said, the language in the dialogue has a better chance of being absorbed into the students daily life.

Since they lack the surrounding dialogue or context that they arose in, these utterances are incomplete. Each of these utterances has some range of potential meanings when it is uttered. Each has a pragmatic or illocutionary force to them, but since they are part of ongoing dialogues or series of conversational moves, a complete description of their illocutionary force would require the other utterances that surround them. Writing dialogues around the utterances similar to the ones found in "Variations on a Theme" is one way this zoo can be extended. Another way to extend it would be to create the "lexical approach" type of worksheets recommended by Michael Lewis (ch6).

The next step for this zoo as far as computers are concerned is to create a better language function list. In order to do this we'll have to grammatically parse the utterances, reduce them to a logical form, and then relate them in discourse using the discourse representation theory of Lascarides and Asher, the discourse representation trees of Frank Shilder, and linear logic. Lolli is a linear logic programming language written in ML. I want to make a Lolli interpreter embedded in Perl but nobody seems to publish papers with explicit algorithms in them except Frank Schilder. It's going to take a while.

Activity: Take sentences or snippets of dialogue from TV shows that all contain a word or phrase that you're focusing on that day in class and present it to the class completely out of context. Students use their knowledge of English and their imagination to construct a situation that matches what was said.

For each utterance try to guess the situation, the meaning of the sentence that includes the word, what the word could contributes to the meaning of the sentence, and the speech act or language function accomplished by the sentence.

This activity comes in two flavors "serious" and "silly". In the "serious" version students try to describe the actual situation, the meaning of the sentence that includes the word. The student with the most reasonable description wins.

In the "silly" version students create humorous, outlandish, or just plain absurd situations and the craziest explanation wins.

Another idea is to take one of the words, cut out all the sentences for that word, and have students or pairs of students sort them into piles with a similar meaning (i.e. the same sense.)

(Note: Some of the conversational snippets are prefixed with the name of the speaker in capital letters.)

Bibliography [
Back to the top] 1