Using ICT to Analysis Language

Analysis and English teaching at first sight not the most creative of combinations! Yet you can see statistical approaches everywhere.

            You can literally ‘see’ statistics in Wordle, perhaps the most visually accessible way to view analysis of text by frequency.

            And you can spot statistical methods in collapsing or alphabetizing a text, now widely used to re-present or deconstruct a text.

            Statistic inform the modern dictionary. No longer do lexicographers chew pencils and stare into space to come up with word definitions. Instead they use powerful computers to analyse huge collections of text (a ‘corpus’) and then read the resultant start to gauge a word’s range of meanings, its collocation and its dominant patterns in the language.

            This approach has led to a whole new area of study – Corpus Linguistics. Again, using computers, researchers count frequencies, identify patterns and sort text according to keywords. New discoveries and insights emerge all the time.

            There was a time that these methods were the exclusive preserve of academics with access to the sophisticated technology required. However, there are now many free websites that offer you an entry into this brave new world. But for small-scale experiments we can use Word and Excel, and free programs such as AntConc, to achieve modest result of our own.

            So, what can statistical analysis offer the serving English teacher? Are there any special advantages for students? Can we use it to enhance the reading of poetry? What about creative writing?

            This chapter will explore what Corpus approaches and concordance software can offer the English classroom – without pursuing the crude numbers involved, or the pseudo-science and jargon associated with corpus linguistics, interesting though it is. By contrast, we are looking for critical insights, creative spin-offs, way of bringing to the surface inspirational patterns and revelations that have a direct bearing on the classroom.

Chapter Outline

1.      Wordle  

A piece of text which has been rearranged into a visual pattern of words

'Here's a wordle of the G20 communique, with support, financial and agreed (surely greed?) especially prominent.'

A wordle is a visual depiction of the words contained in a piece of text, as exemplified within the citation above. Generated by a web-based tool of the same name, a wordle is created by manipulating the words of an input text and arranging them into a kind of graphic. The more frequent a particular word was within the source text, the bigger it's displayed in the wordle. Font and color variation, as well as adding visual appeal, may also give weight to particular words, which are positioned vertically as well as horizontally.

Wordles, also sometimes referred to as word clouds or text clouds, have recently been popularly used to visualize the topical content of political speeches. For an example, check out this recent article in the New York Times, which features a wordle of President Obama's inaugural speech, and note the emphasis on words such as America, new, nation and every.

In the educational domain, teachers have been quick to pick up on the pedagogic value of wordles, which are particularly useful in language teaching as a text analysis tool. They can also be used to elicit speaking and creative writing (see Further reading, below).

If you want to have a go at creating your own wordles, check out www.wordle.net.

2.      Using a corpus to investigate language

A corpus is a collection of texts of written (or spoken) language presented in electronic form. It provides the evidence of how language is used in real situations, from which lexicographers can write accurate and meaningful dictionary entries. The Oxford English Corpus is at the heart of dictionary-making in Oxford in the 21^stcentury and ensures that we can track and record the very latest developments in language today. By analysing the corpus and using special software, we can see words in context and find out how new words and senses are emerging, as well as spotting other trends in usage, spelling, world English, and so on.

Using the corpus enables lexicographers to examine a word in detail by looking at all the different contexts in which it occurs. Below is a typical way of viewing the results of a search of the corpus, using a display format called KWIC (or ‘key word in context’?)

The Oxford English Corpus gives us the fullest, most accurate picture of the language today. It represents all types of English, from literary novels and specialist journals to everyday newspapers and magazines, and even the language of blogs, emails, and Internet message boards. And, as English is a global language, used by an estimated one third of the world’s population, the Oxford English Corpus contains language from all parts of the world – not only from the UK and the United States but also from Ireland, Australia, New Zealand, the Caribbean, Canada, India, Singapore, and South Africa. It is the largest English corpus of its type: the most representative slice of the English language available.

Try this: Using a corpus to enrich a critical reading of a poem

A wonderful side-product of an online corpus is its ability to enrich our understanding by giving us a quick insight into the associations carried by specific words

           

3.      Activity outline

This technique involves the following stages:

·         A close reading of the text ( preferably in electronic form on screen, or the best of all, a combination of on screen and paper formats)

·         Identification of phrases that seem to catch the eye (Target Phrases) that appear to conceal information; or in any other way signal interest (this process is a much gut instinct as any else and needs practice and experiment – some leads turn out to be dead-ends, others reap rich harvests of association and meaning). Mark each one for examination.

·         Analyzed : students type the target phrases into the online corpus and review the result, asking themselves these sorts of questions:

-          What contexts are associated with the phrase?

-          What other words are friendly with the phrase? – near collocations

-          Are there any inseparable friends? A best friend?

-          Do you spot any patterns? ( write them down)

-          Is there a reason for the pattern? Can you propose a theory? What can be seen in the words?

-          Invent a theory to cover what you can see ( your hypothesis)

·         Go back to the original text and apply the result to it

·         Quantify (if possible) the enrichment of reading and understanding that this process has given. This is best done through whole class discussion.

·         Ask students to return to the hypothesis and test it further using the corpus

·         Modify the hypothesis in the light of what is found

·         Has that led to further enrichment of the reading?

For all the above, the process is experimental – there may be some wonderful insight waiting

; There may be nothing obvious.

However, without testing the ideas using corpus, it’s impossible to know. The key thing is to try it out. Applying these ideas to a study of ‘Heaven-Haven’ by Gerard Manley Hopkins, we might proceed as follows:

Read the poem and mark the phrase for analysis:

Heaven-Haven

A nun takes the veil

            I have desired to go

                        Where springs not fail,

            To fields where flies no sharp and sided hail

                        And a few lilies plow

            And I have asked to be

                        Where no storms come,

            Where the green swell is in the havens dumb

                        And out of the swing of the sea

Are the atmospheric effects observed in this poem borne out by other association? A corpus may be able to tell us. So the experiments might follow this order:

1.      ‘swell’

2.      ‘ a few lilies’

3.      ‘ I have asked’

4.      ‘I have desired’

5.      ‘ out of the swing’

So that’s the routine – identify target phrases and then look them up experimentally in an online corpus.

A free online copy of The British National Corpus is housed at Brigham Young University. You will find it at http://bit.ly/te_17 (or you can call it up easily by typing BNC BYU into Google). There’s a search window and you simply type your phrases into the box and click on Search.

                                                           

A result, in upper case letters will appear in the right hand pane, showing you how many times that term was found. Click on it to see the concordance lines. These will be arranged with the target phrase underlined and in bold.

You can copy and paste the result from the search into a word processor to print out classroom resources of your own. Of course, if students have direct access to the system they can read and interpret the result online. However, with a very large corpus there will be material that could be considered unsuitable for some school use. This is because the text-collection will be spread as wide as possible, including adult fiction. With younger pupils it may be wise to adopt a cautious approach and check the searches you ask them to make in advance.

4.      Using an offline corpus

The online corpus approach means working with a preexisting set of reference texts, and although you can limit the choices to a certain extent, you don’t have control over the composition of the corpus.

It is quite possible to obtain concordance software that will allow you to investigate your own text collection offline. The very best examples cost money. However, an extremely useful free concordance is available for download. It is called AntConc (http://bit.ly/te_26). You can install it on your own computer so you don’t have to be online to use it. The first task is to assemble the texts you want to investigate. These have to be in text format (.txt.). Luckily there are many sources of these texts – Project Gutenberg offers all its vast collection in the text format, and it is easy to build a collection of classic texts this way. If you were about to teach the Gothic or Pastoral, or you wanted to find quick cross references for Shakespeare or Dickens, you could assemble the collection works in a special folder for use with AntConc.

When you launch the program, you first load these texts. Searches are lightning fast, even if you have 30 or 40 novels loaded. Impressive! The output comes in the form of statistically sorted concordance lines, and you can export the result of your search as a text file for import into a word processor.

The program will give you a word list with frequency counts, will let you see collocation patterns, and will even allow  you to compare one text with another one to identify ‘key words’ that have special prominence in each text. AntConc is a powerful tool for creating really exciting new classroom resources, and will repay the effort required to learn its features.

5.      Working with concordance lines in a word processor

If you’ve retrieved raw concordance lines from a corpus, what you’ll have is a series of lines, each containing the target phrase somewhere in the centre of the text. It would be nice to get control over the lines – display the target word (s) in the centre of the page and have the option of blanking out the target word (s) to make exercises and introductory lesson materials.

Here’s a typical sample from a corpus, retrieved the phrase ‘have asked to’.

United in typically forthright fashion. ‘We have asked to be kept informed of alan shearer’s situation in

The deputy prime minister. The union is believed to have asked to see Mr. Prescott this evening.

See what I mean?

This is fine, but unwieldy. I would be best to put this into table format, each row having three cell and the middle cell containing the target phrase. The techniques for making this resource using word can be found on the continuum site.

Try this: Use a corpus to investigate word order rules

A corpus can very quickly provide hundreds of word-patterns to enable you to create active lesson where the rules of the game are deduced by pupils rather than given by teacher.

In this activities, students simply investigate color words:

·         What sort of words team up with color words? (the collocates)

·         What order do they go in?

·         When there are several adjectives working together, what position do dimension or ‘size’ words take (big, little,huge,tiny etc)?

·         What are the rules?

·         What happens to the meaning if the normal order is changed?

For instance, your students may find instance of this word order:

Little green eyes

(Possible rule: color word usually precedes the modified noun: dimensions word comes first in the order)

Then what is the effect of breaking the rules?

Green little eyes

Insult? A sense of the sinister? The word ‘little’ is doing the job of belittling of diminishing status. A bit, like saying of someone he’s a funny little man when the person in question may in fact be physically big?

Some collocations are dominant. Both little green and little red have overwhelming associations. Little green collocates with ‘men’-and the alien connection seems to be sprinkled through many of its other contexts.

Try using a corpus to give you a list of collocates for a color word. For instance, for blue. This gives you a harvest of such expressions as ‘poison bottle blue’ and ‘kingfisher blue’-some juicy metaphorical language to discuss, play with and use as models for creative writing.

Try this: odd-man-out

This is an activity designed for students who have already had some experience using a corpus.

It requires students to look for and establish the dominant pattern of use of a specified phrase so that they can then identify where the usage is being subverted or played with.

Almost always these examples give an instant display of stylistic skill: they will embody techniques that can be easily copied, learnt and appreciated.

6.      What are the dominant pattern of use for ‘state of the art’?

The concordance lines over whelming show these collocations-computers, technology, kitchens, system. But there are some clear exceptions

Example of odd-man, playful usage 

The presumptions is that it has to be so good-that every extra ounce expands what is traditionally referred to as sheer gaming power. This however is not the case. Centering through the box’s capabilities with the state-of-the-art alien massacre halo and the snowboarding dude-a-rama of amped, i and the men grew slowly deflated.

I thought i was going to the land of oz ; joe said sadly after an of Halo.

This is nust oz from auf wiedresehen, pet.’

Notes, the context is still computer/high tech

The French facilities were often a state rather than state-of-the-art. During black’s time Metz, the centerpiece of the training ground was a red ash pitch to the ones dotted around school and housing in Scotland’s cities.

Sports facilities often associate with ‘state-of-the-art’- here the play on words is between the phrase ‘in a state’ and ‘state-of-the-art.’

            Notice the well-worked word play, the matched compound word (state-of-the-art and beer-guzzling) the alternative pairs (‘hunky heroes’, ‘chalk and cheese’ and ‘charismatic as a carthorse’), the similes (icepick and carthorse). All of these effects can be copied and experimented with effectively.

This is quite a distance from the usual haunts of the phrase. Here it is describing a chat-show format.

            Notice how the journalist has played with another aspect of the target phrase-the hyphens-to make up two long hyphenated compound adjectives.

By change, the database throws up a definitions. Notice how little the orgins of phrase affect its current uses.

My own attempt at state-of-the-art dorset-‘clarionet, you’m playin’ too vaarst.

Here the author uses the phrase ironically-there is nothing in that use of dorset dialect that suggests modern and up-to-date: the reverse is true.

You could set the class this task:

            Can you think of an ironic use of ‘state-of-the-art’ where you mean clapped out and ald-fashioned?for example, you might say ‘my dad has a state-of the-art-computer-it runs window 95 but struggles with X,fine with elastic band, but not a chance of broadband’.

7.      Follow-through

Where the process throws up an interesting grammatical pattern, you can capitalize on it by weaving it into your teaching.

            Avoid asking children to look up individual words - always choose phrases. Because of the size of corpus database, single words will deliver tens of thousands of result - completely overwhelming and counter-productive.

            How do you find useful target phrases? Well they seem to occur without too much trouble in every other sentence! For instance, I used the phrase ‘throw up’ - and it immediately strikes one: ‘What’s the metaphor behind the phrase? What other meanings are carried by it? Nausea? Discovery? Physical action?

            This phrase provides good raw material for a poem. You could set this task: write a poem that has to incorporate the phrase in every section.

            You can read the poem I wrote (not a particularly good one, I hasten to add) as a model, based on five main strands of meaning as defined by a corpus search.

                        Throw Up

                                I throw up my hands

                                In despair

                                                In shock

                                                In horror

                                In ultimate defeated

                                My life conspires to throw up

                                Surprises, problems, difficulties

                                                And

                                Unexpected shattering conclusions

                                Living every day throws up

                                Faults in character,

                                                Embarrassing blemishes

                                                Half-healed scars . . .

                                No good concealing them.

                                They rise, dark, half-hidden whales,

                                To the surface of my eyes

                                From the deeps of long-gone wars

                                Like ghosts of sunken battleship.

                                And no matter that I try

                                To throw up a palisade

                                The outside world still lobs

                                A sudden grenade

                                Into the foxhole

                                Wrecking everything I’ve made.

                                It makes me want

                                                It makes me want

                                                                To throw up

                                                                                                                                                                Chris Warren

8.      Preposition dance

This investigation explores the way prepositions work with verbs to give us in English a wide spectrum of meaning and nuances, of implications and subtleties. The activity has the added benefit of cementing in students’ minds how prepositions work.

            The idea is simple

·         Make a long list of preposition as a class brainstorm.

·         Choose a common verb.

·         Match it with each preposition.

·         Investigate each variant in a corpus.

·         Each variant will carry several meanings. Make a list of the meanings and the common collocations for each variant.

For instance, preposition such as ‘in’ ‘out’ ‘up’ ‘down’ ‘to’ ‘by’ ‘off’ ‘on’ ‘through’ can be matched with common verb such as ‘eat’ ‘sit’ or ‘run’ to produce a preposition dance.

eat in eat out eat up eat down eat to eat by eat off eat on eat through

sit in sit out sit up sit down sit to sit by sit off sit on sit through

run in run out run up run down run to run by run off run on run through

            Students can ask themselves these sort of question:

·         What contexts are associated with the phrase?

·         What are its collocations?

·         Are there any words that always occur with the phrase?

·         Do you spot any pattern? (Write them down!)

·         Is there a reason for the pattern? Can you purpose a theory? What can be seen in the words? Invent a theory to cover what you can see (your hypothesis)

·         Where appropriate, return to the hypothesis and test it further using the corpus (e.g. try the same preposition with another verb to see if it has the same effect)

·         Modify the hypothesis in the light of what is found