Designing conversations
By looking at speech ‘beyond the level of the sentence’, discourse analysis reveals some of the conventions that underlie it. Speakers use conversational structures to engage listeners, create cohesion, and facilitate comprehension. As we talk, we subtly adjust the lexis and structures in accordance with roles, status, and context. There’s a lot going on under the surface of your average chat.
Most conversations have a beginning, middle, and an end (with lots of subtle movements in between). We signpost to each other where we are in the process through appropriate phrases, from, What’s up?, to I’d better be getting back to the office, for example. We indicate turn-taking through pitch, pauses, and discourse markers (well, anyway, right then, ok, etc).
Meanwhile, speaker roles determine why this exchange would be expected between a doctor and patient, but not between a commuter and ticket collector at Waterloo Station.
Thank you so much.
Not at all.
Conversational structures can frame relationship roles and status, as Deborah Tannen has argued for some time. (Framing roles inappropriately can send very odd signals.)
After recording 400 podcasts and 200 mobile lessons (or learning objects) it is very clear to me that each medium is a new form of discourse with different rules of engagement.
Engagement is a good word here because without it there is very little learning. There are things that work and things that don’t, and content designers definitely work in the R&D department of new media. Clive Shepherd has done a tremendous job in laying the groundwork for this, with his 60 Minute Masters, while people like Tom Kuhlmann , Cammy Bean and others also explore the subject.
I believe discourse analysis offers helpful insights for the new media instructional designer, by helping to make explicit the structures, strategies, and cohesive devices that work for them. There’s no space here to go into every aspect of this but let’s take a look at roles and relationships as they might be framed in an audio-based learning object.
The first question we ask inevitably concerns the whom: Whose problem am I solving? and How do we frame the relationsip? We tend to open lessons with propositions or calls to action to indicate a direct message for the listener and signal their involvement. It’s important that we project a sense of authority on the subject but without framing it in a negative way. (Getting the human element right makes a big difference. Contrast the ChinesePod approach with that of a traditional textbook. I think it’s the parasocial dimension that explains ChinesePod’s stunning success.)
There are many other dimensions and many other questions. One project we’re working on is a set of 90 second learning objects for Nokia cellphones. This might sound incredibly short but by boiling the medium down to its essentials and describing the elements of that discourse format, I beleive we are finding ways to make them effective. Questions like How do we organize/structure the conversations? Is there cohesion? How do we convey the movement of the conversation? etc, are all truly relevant when you only have 90 seconds to make something happen. Here is an example of a prototype lesson that targets young Chinese managers on the subject of time management. (It comes with a transcript, plus translation, etc, on the learner’s handset.)
More on this nascent subject later. In the meantime, you thoughts are requested.
Ken Carroll
January 24th, 2008 at 2:11 am
[…] Read the rest of this great post here […]
January 24th, 2008 at 9:34 am
Hey there, Ken! I agree that engaging your students (or your audience) is a vital part of what makes any learning endeavor successful. At the heart of it, it’s not unlike sales. You need to sell each lesson and have the user “buy in,” to “engage,” in order for it to be effective. At first, as was the case with ChinesePod, people bought in lesson by lesson. After a time, though, people buy into the concept, into the brand, and you win people over by virtue of the process, as opposed to the individual content of each lesson.
Thinking about how that content is delivered, how ANY content is delivered in a Web 2.0 world, is illustrated even here in this post. You deliver your points with great clarity and authority, but it’s also a dynamic discourse. There are links embedded throughout for people to further explore whatever concepts or people might interest them.
I’m curious, though… obviously you have the ability to track how many people are clicking on these links, how long they stay on the site, etc. How much does that data influence how you write these posts? How quickly are the Web 2.0 elite adapting to their audience?
January 25th, 2008 at 5:54 am
Ken,
It stuck me as I listening to your sample lesson that you have become an expert at condensing discourse rather than capturing natural discourse. I could be wrong but the key would be to ask yourself if and to what degree you create a script before you do a recording or after the recording has been finished.
If you are creating condensed discourse (my own terminology) then this has implications both for learning and for teaching.It occured to me that condensed discourse is also a feature of Chinesepod which, to pick up on an earlier thread, calls into question what non-linearity actually means if the discourse is scripted ahead of time and follows set “house” rules which you have put into place over time.
Believe me I understand the issues involved in capturing natural discourse and translating it into teaching materials. Students often find real discourse boring and abtuse and are usually better served by the condensed, manufactured version.
Anyway I find the intersection of condensed manufactured conversation and the rules that we can uncover from actual observations of speaking “in the wild” to be of interest. I’m curious if you believe the second can really inform the first so better lessons can be built for the student.
January 25th, 2008 at 7:10 am
My favorite teachers have always been the ones who engage with their students on a “real” level. That is, have a conversation WITH us, rather than talk TO us.
Attempting to create conversation without the face-to-face interaction is way-challenging. You may have done a good audience analysis up-front, but you’re still lacking a clear view of Tannen’s conversational structures.
Some writers/instructional designers do this exceptionally well — but is it just a matter of finding the right voice to draw in the learner?
January 25th, 2008 at 10:35 pm
Frank,
I measure responsiveness on the blog in terms of comments, not the numnber of them, in terms of what can be learned from them. This is kind of what guodes it. I don’t think I’ve ever written a post without learning something hugely important from the users’ comments. The posts are usually based on some insight that I want to discuss with people like you to see if they’re in any way sensible or totally off the wall. Users will let you know soon enough! A popst is neve compete wihtout some input and that’s why I do it.
Michael,
Very pointed question. I don’t think we can say that there is only one ‘natural’ conversation format. We all engage in code switching and other discourse stategies throughout our day (from conversing with a child one minute, to answering the phone to talk to a colleague,the next, for example). Switching like this often requires abrupt changes in conversational strategy, but it’s all natural and all good.
Meanwhile the medium (partly) determines the message. If I want to connect with readers in book format, then I adopt the most effective discourse starategies for that medium. But those are not the same stategies I’d use for smoke signals or a slide show. And they’re certainly not the same as those I’d need for a 90 second mobile lesson, not least because it’s entirely ‘natural’ to address different constituencies in different ways.
So, ‘authentic’ language scenarios are great when the learner is in an authentic scenario, but they do not transpose wll into communications media. As designers I beleive we have to design for the medium, but also for the audience, and especially for the environment in which the content will be consumed. No user needs to hear small talk, grunting or ellipsis in this format, though these are all things that occur in face to face conversation.
Scripting is another huge question that you touched on. We do not script these lessons. It is essential that the exchange between Sheryl and I (in this case) happend spontaneously, so we prep it but we do not script it. (We write the transcripts and choose the vocabulary items after the actualy recording - if you want to know a trade secret, in order to make the conversation really happen.)
Ken Carroll
January 27th, 2008 at 7:15 pm
Interesting comments, as usual. I wanted to add that part of what’s going on “under the surface” is culturally based too. One brief example, when I was in Shanghai a couple of weeks ago, one of my interviewees came into the office and I asked if he would like something to drink. He said no, and then grabbed the cups, filled them up from the cooler, and then gave one to me and took the other for himself. Linguistically I handled the situation in Chinese, and I even knew that people say no when they really mean yes. But I was still amazed at how it all happened in front of me so quickly. So it wasn’t just a matter of chinese language, discourse, or even turn taking. The cultural behavior seemed to move the whole thing along. (I might add the very next day someone asked me if I would like some coffee. I said no, and two minutes later he showed up with a cup of coffee for me anyway. Same situation from the opposite side.)
January 29th, 2008 at 8:44 pm
Designing for the 90-second mobile medium is a challenge. It’s also really interesting from a discourse analysis perspective (and many other perspectives!), and we’re only just beginning to scratch the surface.
There’s a lot going on. There are issues of how discourse markers, grammar, turn-taking affect everything differently in this medium - from defining speakers’ roles in relation to the listener (different from radio or conversation) to providing coherence and cohesion. Compare, for example, the importance discourse strategies play in conversational maintenance in normal conversation - a strategy we have to stay away from in this medium because the effect does not benefit the listener/learner - it distracts and there is no time.
Another interesting aspect from an instructional design point of view is the role the questions play in guiding the listener/learner to the learning objectives in the message. These, I think, provide the framework necessary for a mini learning object, which is not a conversation.
January 30th, 2008 at 7:38 am
Thanks for clarifying things for me Sheryl. I hear you saying that nominally a podcast isn’t giving students a model for a “normal” two-person conversation (And yes Ken, I agree there is not such a thing as “normal” but we need to start somewhere and students are looking for models).
If moreover the discourse dynamics of a pod-cast are unique to that medium what are students learning from a podcast and what should we be asking them to attend to?
Are we teaching students how to create good English podcasts (where the performance bar is quite high thanks to Ken’s artful acomplishments)? Or are we suggesting that they can learn grammar, or vocabulary or “conversation” by listening to these very informationally dense examples of somewhat staged discourse? Yikes, what is a 90-second podcast good for educationally speaking?
January 30th, 2008 at 7:15 pm
Really good question, Michael: what is a 90-second podcast good for educationally speaking?
I don’t believe anyone is under the illusion that anyone could learn to communicate in English entirely through this medium. At this stage, I see it as supplementary. For example, one of our “channels” focuses specifically (and in an insightful manner) on high-frequency phrases for the workplace that you would not normally find in a textbook. The point is that the learner can use the time spent commuting to work learning about (yes, about) a new chunk of language which, because it’s high-frequency, they will probably hear that day, or week, or month, in context. Which could mean they actually acquire that chunk. So yes, they can learn vocabulary this way, and this, I believe, is what we’re asking them to attend to.
Arguably, though, the only way to learn to communicate in a language is through first-hand experience of authentic communication in that language. And in the particular situation you wish, ultimately, to become communicatively proficient in. From a communication skills point of view, watching TV news would only benefit you if your goal were to learn to understand the news or become a newsreader; the discourse dynamics of this medium will not help in conversation, and I’m sure that in most (if not all) languages, many of the phrases and vocabulary are also not likely to be used in conversation. The discourse dynamics of every medium are different, and that includes the recreation of ‘authentic’ scenarios found in textbooks or on audio products for learners. I think that the discourse dynamics of ‘conversation’ in a textbook are also very specific to that medium.
January 31st, 2008 at 2:57 am
Sheryl,
Things for responding and kudos on the podcast. I have a couple of comments on what you said. First, my guess is that the podcast that I heard runs at a clip much faster than a normal two-person conversational dialog and faster than what I usually record for intermediate students. I didn’t do a WPM count so I can’t be sure.
If so, given the volume of info (vocab, grammar) that you covered I’d be very skeptical that a student would pick up what you wanted/intended him to pick up given that nothing is highlighted during this recording. Students need “our” help to pick out the trees from the forest especially when they only get one or two shots at the material.
In other words there is no guarantee that they will attend to what you want them to attend to and, if what they are hearing is new, (the idea behind instruction in the first place) there is a good chance that what you are targeting will be ignored or mis-heard.
Note: All my comments are not nearly as valid when repeated listening takes place. However I’m under the impression that this medium is not particularly designed for repeated listenings nor is it priced to encourage repeated listening. In fact, just the opposite. It’s for students on the go in crowded, noisy, multi-tasking environments.
Next, I would question that you offer high-frequency phases that are not found in textbooks. I would think the opposite would be true. The very nature of high-frequency phrases makes me believe that they ARE the ones covered by textbooks.
Of course then we come to the issue of what constitutes a high-frequency phrase. Whose data are you using? I would be really impressed if you were doing this “research” on your own but if it is public data then I would be more inclined to believe that other material writers are using these same high-frequency phrases in their work.
In other words, this all doesn’t add up to me. Heck, I believe in the intuition and art of good teachers but I am a bit put off when art and intuition is clothed in academic language like 99% of this industry indulges in!
I don’t mean this to be critical, I’m just saying it as I see it.
February 3rd, 2008 at 6:13 pm
Thanks for your feedback, Michael - responding to your comments definitely makes me look at what we’re doing more critically, and rising to the challenge will, I’m sure, inform work we do in the future.
With regards to repeated opportunities to listen - the users on this platform can listen as many times as they would like to after they have paid only once, i.e. paying for a month of podcasts means unlimited access to all those podcasts.
There is a text guide in the form of a transcript on their handsets, with a glossary of useful phrases and, on some channels, main points highlighted. This, I believe, helps to highlight what they could ‘attend to’. I say ‘could’ for a reason.
I think there’s a danger in suggesting that ‘we’ should or need to tell the learners what to attend to - this is too much in line with the prescriptive nature of so much of the material out there. We can make suggestions, give guidelines, highlight certain things, but we should not underestimate the learner’s own awareness of what they need or want. In fact, I think the painful gaps in their communication abilities might show very clearly what they need (and want) to learn.
As for high frequency phrases in coursebooks - yes, there is, of course, lots of useful and high frequency language in the books. However, I think that certain areas are often overdone - for example, something like ‘expressing an opinion’ has a long list of possible ways to do this in most coursebook series when research (corpus linguistics) shows that most proficient speakers of English rarely go beyond the ‘I think…’ phrase. Learning all the other possible ways to do this means a very bad return on time investment.
I think (!) that the coursebooks’ extrapolation of a fairly limited number of functions means that some really important and useful language is ignored. Before learners embarck on expanding their list of ‘expressing an opinion’ phrases, there are other actual gaps in their language that would make a big difference to their communication skills. Things like ’scroll down the page’ when talking to a colleague about any site or document on a computer screen. This is the sort of language I haven’t seen addressed in the coursebooks, and the sort of language that provides fluency and sharpens the ability to communicate. In my experience, it’s exactly the sort of awareness of collocation that most learners lack, even those with great swathes of vocabulary.
Choosing which language to use involves listening to language in a variety of media related to the workplace, actual use of the language by proficient speakers in the workplace (and what the learners of English need), and - yes - intuition that comes from being a language teacher for very many years and seeing people struggle with communication in the real world. The experience of being a language learner whose communication needs were barely addressed by the coursebooks is also invaluable.
February 11th, 2008 at 11:49 pm
Sheryl,
Thanks for your reply. I enjoyed reading it and it clarified a great deal.
I agree with you 110% when you say:
I think (!) that the coursebooks’ extrapolation of a fairly limited number of functions means that some really important and useful language is ignored.
I would also agree that course books are mostly behind the times when it comes to a lot of technology and even idiomatic language. However, I can understand why they ignore the stuff- coursebooks mostly have long lifetimes of 3-5 years between changes.
I would think in this context that technology related vocabulary (and business buzzwords) would become fertile ground for your broadcasts.
I was coming from a bit different perspective when I questioned your basis for talking about the relative frequency of word use. I have seen statistical studies that try to compare lexical items between and among both Chinese and English textbooks. I find such studies quite interesting and very informative to someone like myself who likes to use such information to make informed decisions. I am perhaps unduly sensitive to these issues.
Thanks again for being so informative and continued good luck!
February 11th, 2008 at 11:50 pm
Oops, between and among English textbooks and between and among Chinese textbooks.