Mining professional forums to supplement a Business English course: some examples for IT

In the previous post  I wrote about some difficulties I’ve been facing in tailoring Business English courses to the needs of my students, who are IT professionals. As I mentioned there, the main problem is that it’s difficult to pinpoint what language is used in workplace communication in IT but under-represented in BE coursebooks.

To address that issue, I started using http://www.webcorp.org.uk/, a web concordancing tool to retrieve concordance lines from stackoverflow.com, the main forum for IT specialists. Therefore all examples here can only be directly used with IT groups, but there are professional forums for almost every profession out there, so I hope that this might be useful in other contexts too. (If you clicked the link and the tool is not loading, just check it out a bit later: it is down sometimes, but when it stops working it generally goes back online in a few hours).

I’ve been using the forum for three weeks now and lessons that use this material seem to be giving my students exactly the language they need. Students especially at lower levels find this very motivating. It’s also been a lot of fun for me, because I enjoy exploring the language so much.

So far I’ve used the tool and the forum to

• follow up on lexical and grammatical areas presented in the coursebook by exploring how it’s used on the forum
• focus on grammar structures the students tend to confuse (e.g. should, must, have to and be supposed to) by supplying them with lots of examples from the forum and asking them to figure out the difference in meaning
• provide the group with grammar drills contrasting two or more structures
• look for discussion topics that would appeal to my students

Here are some examples.

Vocabulary area >>>>> specific vocabulary, patterns and contexts

Here is an example of an area we looked at with a pre-intermediate group. With the coursebook we’d looked at the topic of change: we started off by looking at some simple verbs like fall and rise, matching them with graphs and discussing the changes the students have noticed in the amount of traffic, prices and so on in the past ten years.

After that I elicited some modifiers (e.g. considerably and slightly, which had come up in on-the-spot feedback) and focused the group at the  word ‘dramatic’. We looked at concordance lines from stackoverflow and noticed the expressions containing the word ‘dramatic’ in the first four lines (a dramatic speed difference between … and; might be causing the dramatic increase in time; a dramatic increase in server response time).

 

 

 

 

 

 

 

 

The students underlined the expressions in the remaining lines (I monitored encouraging them to look both to the left and to the right of the word) and then we organized these expressions into a table (initially working as a whole group, and then the students completed their tables individually):

Finally, the students

• tried to reproduce their tables from memory – when they were stuck, I referred them back to the worksheet with lines
• wrote two or three examples from their own professional experience using target expressions and then shared those expressions in a mingling activity without looking at their notes
• for homework, wrote a few more examples and emailed them to me.

Another search (e.g. ‘cause [a|an] [*] increase‘) retrieves other adjectives commonly used in this context.

The big advantage I see here is that all grammar and vocabulary associated with the situation comes up. A similar search on an accounting forum (http://www.accountantforums.com/) might well return the same verbs used with ‘dramatic’, but there’s no question that the last column will be different. I think that that is the vocabulary this language needs to be practiced on (to me this looks much more natural and useful than describing graphs, a task this language is practiced on in quite a few BE coursebooks – unless the students do need to describe graphs).

Chunks containing a structure >>>>> Patterns and contexts

We tried the same procedure with B1 and B2 groups with the chunk ‘it’s supposed to’. Again, it is clear from the concordance lines that there are distinct recurring patterns.

We ended up with the following table:

 

 

 

Again, after the students underlined expressions in the concordance lines, organized them into a table and reproduced the table, they shared a few examples from their work experience – this time talking about situations when something wasn’t working the way it was supposed to and sharing what this led to and how they overcame the problem.

Grammar presentation: topics and texts

Because webcorp allows pattern matching, it can be easily used to find forum threads that contain a lot of examples of a specific grammar point. As on forums each page is centred around just one question, it might be possible to choose a general enough question, centre the lesson around that topic and use the forum thread for grammar presentation and/or grammar exercises.

For instance, this search [was|were|am|are|is|be|being|been] [*]ed returned several pages that have more than 10 instances of the passive with regular verbs (the question is too technical to centre a lesson around, but so I’d use this page for follow-up exercises).

On stackoverflow it’s also possible to search for questions that received a lot of replies and are thus clearly appealing to wide audience: http://stackoverflow.com/search?q=answers%3A100-1000 returns all topics with 100 to 1000 replies.

There are quite a few questions there that are general enough and can be used as a warmer or a lesson topic and they might naturally call for certain grammar, which will be evident from the replies.

For instance, here’s a topic in which people share ‘the toughest bugs they’ve ever found and fixed’.  This particular page features dozens of instances of ‘would’ used for past repeated action:

Grammar drills 

I’ve also used concordance lines with an elementary group, simply to give them some practice in using correct auxiliaries (we focussed on ‘are’ and ‘do’). The group told me that they go to the forum and read it using google translate, and they were clearly motivated because they understood quite a few lines. Also, the task exposed the students to some very relevant vocabulary. In the next few weeks I hope to get the students to produce such exercises for themselves and do them autonomously.

 

If you teach in an IT company and would like to use or adapt some of these materials, here is a Microsoft Office document with these concordance lines and tables.

Some tips

I’ve found it useful to

• look for the typical patterns containing target language before the lesson, using the ‘show collocations’ option at the bottom of the page and then running a few more searches to see how typical these patterns are
• don’t forget to delete the http:// part of the url when you specify the site to search; use the word filter if the word/structure you’re looking for comes up in someone’s signature on every page they comment on – this will exclude those pages (e.g. in the following search, the pages containing the word ‘loud’ will be filtered out).
  
• retrieve concordance lines a day or two before the lesson as webcorp is sometimes down for a few hours
• edit the lines: make sure that all typical patterns are represented with enough examples; proofread for mistakes in target language
• use pattern matching and keep a record of the queries.  Here are some sample queries:
  [have|has] [[*]ed|known|done|seen|been] – to retrieve examples of present perfect
  [was|were|am|are|is|be|being|been] [*]ed – passive with regular verbs
  [am|is|are|was|were] [*]ing – continuous tenses
  

Some issues

Ideally what I’d like to be able to do is use in my setting the approach suggested by Michael Handford in his book The Language of Business Meetings. He analyzed 64 business English meetings from Cambridge and Nottingham Spoken Business English Corpus by

• identifying and categorizing potentially important linguistic features, that is, words and clusters that are more frequent in business meetings than in other contexts (e.g. we’ve, customer, if, which; and then; you know, sort of, the rest of it)
• understand the specific meaning and use of those items in their contexts (he finds that chunks have varying meaning, depending on who’s using them and the context, e.g. that in external meetings  you have to rarely expresses any sense of obligation, ‘with speakers instead offering non-face-threatening and sometimes general advice, as in Mm. what you have to do is get the paper and twist it round)
• infer what discursive practices they realise and what goals they serve (e.g. and then is used for linking; sort of, for softening)
• interpret what communicative activities are enacted through these linguistic features (page 39)

Right now I’ve got little idea how to find language that features more prominently in IT related communication and therefore need to be focussed on as a supplement to working with the coursebook, and the task does seem daunting. Running searches and looking for recurring structures on specific pages is a lot of fun, but it’s unlikely to produce comprehensive results; I’m going to look into WordSmith Tools but I don’t expect it will be easy to figure out how to use it with a huge forum, which is not a neatly cleaned up corpus.

Also, while some chunks clearly tend to occur in patterns, others don’t. This makes it even more time-consuming to look for language that can be explored in this way.

Over to you. It would be great to hear some other ideas how to use webcorp and forums in teaching. I’d be also greatful for pointers to other useful tools, as well as for suggestions what other chunks and lexical areas to explore and how to approach this in a more systematic way.