The art of conversation

« previous post | next post »

The Onion News Network, 7/20/2010:


Scott Adams, "Conversation", 7/20/2010:

Imagine an advanced alien life form that materializes on Earth in the middle of a popular dance club. The alien has a cloak of invisibility and observes the humans dancing. He is here to watch and learn. My question is this: Would the alien ever learn to distinguish good dancers from poor dancers?

Now suppose the alien leaves the club and finds a bar that is open late. He observes a lot of what we call "conversation" happening. The alien's universal interpreter device allows him to understand the content of the conversations. My question is this: Would the alien ever learn to distinguish a good conversationalist from a poor one?

I started thinking about this after reading that people with Asperger syndrome have trouble understanding the subtleties of human social interaction. That skill doesn't come as a package deal with general intelligence. The advanced alien can't figure out who the good conversationalists are, nor can the fellow with Asperger syndrome even if he has an otherwise exceptional IQ.

The sad reality behind the Onion's joke is well covered in the Wikipedia article on "feral children". And Simon Baron-Cohen's theory of autism as "mind blindness", which implies difficulties in conversational skill, is discussed (and questioned) here. Note that Scott Adams takes a different view of the situation:

You might think that everyone on earth understands what a conversation is and how to engage in one. My observation is that no more than a quarter of the population has that understanding. I was solidly in the conversationally clueless camp until I took the Dale Carnegie course, in which one small part of the learning dealt with the mechanics of conversation. It was a life-changing bit of knowledge.

Prior to the Dale Carnegie course I believed that conversation was a process by which I could demonstrate my cleverness, complain about what was bugging me, and argue with people in order to teach them how dumb they were. To me, listening was the same thing as being bored.  I figured it was the other person's responsibility to find some entertainment in the conversation. That wasn't my job. Yes, I was that asshole. But I didn't know it. The good news is that once I learned the rules of conversation, I was socially reborn. It turns out that active listening is more fun than talking, although sometimes you need to guide the conversation toward common interests.

Three-quarters of the people reading this post just thought "Uh-oh. I didn't know conversation had rules."

There's less linguistic (or other serious) research on "conversational rules" in this sense than there might be, given how interested the general public is in it. Most of such work that I know about is motivated by the engineering goals of human-computer interaction, where for so far, the general choice seems to be between too little logic and too much. A recent Partially Clips strip on the "too much" side of the balance sheet:



13 Comments

  1. Doctor Science said,

    July 24, 2010 @ 9:33 am

    Scott Adams has a point, but he is missing the obvious context.

    I believed that conversation was a process by which I could demonstrate my cleverness, complain about what was bugging me, and argue with people in order to teach them how dumb they were.

    This is a conversational style, all right — the one usually called "Men Are From Mars", though as Deborah Cameron points out it's really the way a high-status person talks to inferiors. If 3/4 of the people Adams observes are talking this way, maybe they're all trying to talk like superiors. Everyone is trying to be above average.

    It's "talk for success" (like "dress for success", only in conversation): when high-status people are self-absorbed assholes, one way to look successful is to be a self-absorbed asshole. As Cameron also discusses, this is crucially involved in gender dynamics. Women are supposed to be "good listeners", and so a man who's a good listener is immediately a loser under subtractive masculinity .

    [(myl) Adams' discussion is surely oversimplified — a better way to think about it, as you suggest, would be that the situations and goals of conversational participants generate a lot of different "rules" — and his estimates of proportions are, let's say, subjective at best.

    But your characterization of the conversational role of status is also incomplete, I think. In many situations, the higher-status participant will generally talk less and listen more: a job interview, a sales pitch, an oral exam, a doctor-patient exchange, and so on.

    In any case, the thing that I liked about Adams' post was the idea that many people could benefit from lessons in how to carry out (certain kinds of) conversations — and not just people tagged with a clinical label like "Asperger's".

    Someone that I know had a fourth-grade teacher who gave her class a lesson in how to start a conversation with a stranger, and how to use that conversation to turn the stranger into an acquaintance if not a friend. The kids had to try the ideas out in practice on other kids in their grade, as a sort of homework assignment, and then in class they discussed the details of their experiences and the lessons to be learned from their successes and failures. My impression is that for many of the students, this experience affected them in the way that Adams describes the effects on him of the Dale Carnegie program, which I gather does some of the same sort of thing for adults.

    We give people a lot of guidance and practice in writing, where it's also true that there are many different situations, goals, and appropriate methods. Less often, we give lessons in public speaking. It's much rarer to give explicit instruction in conversation. Maybe it shouldn't be.]

  2. Mr Fnortner said,

    July 24, 2010 @ 11:32 am

    If status was the X axis and call it expositive speech was the Y axis, then I suspect a humped curve such that the pretenders and poseurs–those with some status but still aggressively apiring–would have the highest Y values. I wonder if this has been/could be tested.

  3. Allison said,

    July 24, 2010 @ 12:31 pm

    Could an alien tell which conversation was going well? If the alien has the same ability to interpret human emotional states that humans do – of course! The good conversationalist has a partner who is engaged, smiling, laughing, paying close attention – those without "mind-blindness" would be able to tell as well.

    What Scott Adams is talking about is not strictly speaking linguistic knowledge. His brain collected all the information he needed to know he was a jerk – he just didn't bother paying attention to it.

    And what he learned weren't "rules" anymore than style guides teach you "rules". The rules are what guides whether or not a particular conversation choice will lead to a particular social response (highly reliant on the full package that is his conversation partner, the context and their social interaction) – Dale Carnegie is teaching you the equivalent of not to use passives. Except maybe this style training might actually modify your social responses in the way it claims it can.

    We should be as skeptical about "The Loss Art of Conversation" stories as we are about "This New Technology will make Children Stupid" stories.

  4. TS said,

    July 24, 2010 @ 2:07 pm

    "My question is this: Would the alien ever learn to distinguish good dancers from poor dancers?"

    Yes. The alien would probably take a result-oriented approach, similar to the way a human would study mating rituals in animals. So you observe who gets laid afterwards, and then work backwards to analyze what physical features and dance movements were associated with that. The human would reject any idea that there might be some other, artistic, value associated with the ritual.

  5. Ben Hemmens said,

    July 24, 2010 @ 4:55 pm

    So you observe who gets laid afterwards,

    How would the alien know getting laid was the point?

  6. Trond Engen said,

    July 24, 2010 @ 6:06 pm

    If status was the X axis and call it expositive speech was the Y axis, then I suspect a humped curve such that the pretenders and poseurs–those with some status but still aggressively apiring–would have the highest Y values. I wonder if this has been/could be tested.

    Has it been tested? This is what sociolinguistics is about! You're not far from describing the classical sociolinguistic curve as laid out by e.g. Labov's findings in NYC pronunciations of /r/:

    For the purpose of sociolinguistic research speakers are categorized in classes on socio-economic criteria. In Labov's study there were six classes: middle and lower middle class, higher, middle and lower working class and lower class. He found that as the level of formality increases the percentage of prestige markers will also increase for speakers of all classes, but the internal ranking between the classes will stay the same. With one interesting exception: in the most formal situations the "lower middle class", the second highest, will catch up with and eventually surpass the "higher middle class". This is presumed to be a near-universal since it's found in very different languages and cultures. (As I recall there are some interesting exceptions, but they seem to be due to competing prestige varieties or social upheavals.)

    Caveats:
    1. I'm no linguist, so I suspect that almost everyone around here will have a better understanding of this.
    2. I haven't read Labov. I'm mainly paraphrasing Robert McColl Millar's revised edition of Trask's Historical Linguistics.

  7. Trond Engen said,

    July 24, 2010 @ 6:32 pm

    I forgot to add a link to the word curve.

  8. the other Mark P said,

    July 25, 2010 @ 1:41 am

    Yes. The alien would probably take a result-oriented approach, similar to the way a human would study mating rituals in animals. So you observe who gets laid afterwards,

    I'm guessing you are male. ;-)

    The people selected by the alien would be those most into casual sex, not the best dancers. Good looks might also factor into it. With girls, alcohol consumption would correlate better than dancing ability.

    Many people dance for the sheet enjoyment of it. They, unsurprisingly, tend to be the best dancers.

    But why need an alien? A person from Tibet would have no way of distinguishing good from merely OK, lacking enough cultural references.

    I can't even say how well dressed my own kids are, by the standards of their peers!

  9. ?! said,

    July 25, 2010 @ 9:00 am

    That Scott Adams note is very interesting. I suspect I am a bit like him (pre-Dale Carnegie). I might spend a few days teaching my five-year-old boy how to converse.

  10. Janice Huth Byer said,

    July 25, 2010 @ 4:40 pm

    Corroborating Mark's insights into the conversational speech habits of higher status individuals is an observation made by the Bulgarian-born, German-educated, British Nobel laureate, Elias Canetti (1905-1994):

    The perk guarded most dearly by those in power is not the right to speak but the right to be silent.

  11. Rubrick said,

    July 25, 2010 @ 4:51 pm

    An alien should be able to at least tell which conversations went well by observing who left together. (The same might apply to the dancing…)

  12. KevinM said,

    July 27, 2010 @ 3:47 pm

    "The opposite of talking isn't listening. The opposite of talking is waiting." Fran Lebowitz

  13. ruidh said,

    July 27, 2010 @ 10:21 pm

    What is a "good" conversation? What is a "bad" one? What is a "good"/"bad" dancer. I don't think two humans could agree on such a value judgment, much less an alien. Perhaps to an alien, a "good" conversation is one where the words are most true. Perhaps a "good" conversation is one with long pauses which would seem awkward to us.

RSS feed for comments on this post