It would be fascinating to analyze this in verbal communication. I've thought about this in the past, specifically in how idiomatic expressions, "inside jokes" and the like are disseminated through social groups. It always seems that certain individuals in a group are responsible for setting the tone and style of humor and colloquial expression, while the rest end up mimicing the leaders.
In a more practical sense, simply being able to recognize these signals would be very valuable in meetings, especially sales meetings involving pitching a group when it is unclear who the actual decision makers are.
The article mostly talks about structural words. The real breakthrough is that they provide a lot of clues into our very being. We can predict lying, age, gender, and power using them. For those interested in this it would be worth checking out The Secret Life of Pronouns[1]
Your second paragraph talks about being able to notice these things in meetings. Penbrooke's research[1] shows that our brains ignore structural words. So it's pretty impossible to take advantage of this particular research.
However, I think there are a few things you can do.
1. You're first paragraph about "inside jokes" is spot on. When I was in college it was pretty common to to get what were called "Dorm Floor" shirts. The front had some slogan or name on them and the back had quotes from your dorm floor. When we went to create ours there were three of us who dominated the floor quotes. So they made a limit for how many quotes a person could have. Most of these quotes were essentially inside jokes.
If you want to know where you stand in a group it's pretty easy to test this using inside jokes. Can you start something? If they respect you they will pick up your inside joke.
2.When I was in college, I took a lot of discussion based classes related to political theory. I noticed my self picking up the mannerisms of a kid in my class. I then realized that others were too. This was a huge eye opening experience for me. I realized that mannerism and the way we speak is organic and if we can record it's evolution it tells us something about how power and society works.
The down side of this is that in a meeting you don't see the birth of the mannerism so you don't know who the decision makers are. You also can't mentally process the structural words, so I don't think there's much advantage there.
I have noticed the copying of mannerisms as well in college. I didn't know about all this research.
But then I decided to play along and started to pick professors and copy their mannerisms and verbal style on purpose. So on week I would talk and act like a math professor, then like the physics one. I didn't do it in an overt, derogatory, or offensive, but in a subtle way. Doubt anyone even noticed. It was interesting that it was easier to copy from those that I respected and looked up to. some that I didn't care about just came harder to do.
Very interesting. At first blush (based on the title and first bit of the article) I wanted to offer an alternative explanation, something like:
1. People make sense of things through language. (You can get this sentiment in sound-bite form from any number of linguists or language philosophers, e.g. Wittgenstein's "The limits of my language mean the limits of my world.")
2. People with more linguistic resources are better able to make sense of things, and thus tend to rise to the top of the pecking order. (N.B.: Language is more than just words, so this isn't simply an issue of vocabulary.)
3. People with fewer linguistic resources are drawn to the expressions of people with more linguistic resources because those expressions give them new ways to make sense of things. They're like design patterns for thought.
4. Thus, what's being measured isn't power or influence, at least not directly.
But of course none of this is relevant to the article, which draws its conclusions from the degree of periphrasis in a discussion, rather than the substance of what is said. So, again, very interesting.
Yeah, that could be clearer. What initially struck me, before I'd finished reading the article, was the impending sense of post hoc ergo propter hoc. So my response, at that point, was to invert the relationship between copying and pecking order: a) "pecking order is an effect of copying" vs. b) "copying is an effect of pecking order."
If b) is the case then it makes sense to frame the issue as "The way people copy each other's linguistic style reveals their pecking order." But if a) is the case then the article's conclusion is tautological. Of course you're right that that doesn't stop us from taking it as a direct measure of power or influence, since they're still strongly correlated. I was just thinking of it more as a measure of linguistic reserves [1].
[1] http://goo.gl/NOUwo (Sorry to use a link shortener, but it's a long Google Books URL.)
I still think this is the major effect that makes it harder for adults to learn a foreign language. Brain chemistry changes a little between child and adult but your resistance to social inferiority signals changes a lot.
If you can reach a zen state of submission to the ocean of expression around you and follow the currents, you could learn like a child again. That's dang hard to do.
I've actually observed this many times, having lived for quite some time in America and now quite some time in India - not just with linguistic style but also with accent (when the people talking to each other cross geographic boundaries).
So over the years, I've internally made a (somewhat silly) point to be aware of this and not try and change the way I speak based on whom I'm speaking too.
James Pennebaker wrote about this in "The Secret Life of Pronouns". It was a good book released late last year.
He also does twitter stream analysis using his research and you can find that here --> http://analyzewords.com
People might dismiss it as tripe, but after you've read some of his assertions and look over your own emails you'll probably believe it makes a bit of sense.
I thought Pennebaker found an absolute correlation between status and pronoun usage (e.g. higher-status uses "we" more than lower-status), where this work seems to be about relative patterns between individuals.
He didn't really have a definite pattern for higher status except they used more articles, however lower status used more "I"s.
Yet later in the book he also mentioned mirrored language and who reflects who. He also had some sort of web program that measured the difference between the two. I think that correlates with the goals of this study.
And linguistic style reflects how a person thinks (so the theory goes). So when you mimic someone's linguistic style, you may find yourself succumbing to their way of thinking.
The theory continues that different programming languages reflect different linguistic styles and thus your choice of programming language can affect the way you think.
As an example, I do a large amount of text processing work using sed. Interestingly (imo), sed is a language designed by someone who had a degree in Psychology.
As a result, to some extent my way of thinking would probably differ from someone who uses, e.g., Perl as their preferred language for processing text.
Or vice-versa, i.e., one's thinking style affects one's choice of programming language. I believe this is why some programming languages just never clicked with me. My brain definitely resisted Perl...at first....
No doubt. But would you agree that the more you use someone else's language, the more your thinking adapts to their way of thinking, as evidenced in how they designed the language?
When I talk to people who have language problems (I was a TA at a US university, so I meet a diverse set of language skill levels in my students), I notice that I sometimes copy the amount of broken-ness in their language (involuntarily). This is not in a condescending way, because I can sense they become comfortable at once. To understand this in a different way, try to imagine one person speaking broken English and the other person just completely keeping up his fast fluent pace in all his replies. That would seem pretty rude.
What I wanted to get at is this, pecking order might not be the only reason to copy verbal styles. Empathy might be one more reason. Does this study control for this?
There was a post on this a while back, the idea was that copying helps you communicate because you are using the same style as them. So you are instinctively trying to match so you understand each other better.
"There are a number of techniques that are supposed to be beneficial in building rapport such as: matching your body language (i.e., posture, gesture, etc.); maintaining eye contact; and matching breathing rhythm."
You are simply reducing the power differential between you and the less articulate individual since they perceived you stand superior on the speaking dimension to begin with. You are being humble and inviting. This seems to be an implementation of the principle described in the article, and consistent with the implied hierarchy when one imitates another's speech patterns.
I do interviews over the phone. Often I find myself subtly copying the local dialect of the interviewee. I don't really do this on purpose, perhaps it's some kind of an automated mechanism to make myself more understandable or bring myself on the same level with the interviewee.
I am Spanish and my English is good but not perfect. Any way when I find myself talking English close to people who´s English is worst than mine, I change my pronunciation to one much less accurate (almost Spanish pronunciation of English words).
I just don´t feel confortable pronouncing properly (well kind off properly) in that situation, I just hide my ability. (maybe to avoid the mythical Spanish envy?).
It's also a lot easier to understand. I am constantly asking my German colleagues to pronounce that word 'as if they were imitating my accent', so I can get some idea of the spelling, and not just a smudge of weird consonants.
I cannot actually speak German, but isn't German relation of pronunciation to spelling much clearer than in English (for which it's notoriously intractable, see the famous "ghoti" joke)?
Spanish here. I remember that back in high school or uni people would make fun of you if you tried to make your pronunciation sound a bit more English.
I think that's changing nowadays. People are more exposed to English via TV series and movies and stuff.
With regard to your point, whenever I'm London and find myself having a conversation with someone with a cockney accent. After a few minutes, I start to mimic their accent. I can't help it. I love the accent.
I found the following somewhat funny given that this article discusses pecking orders and linguistic choices:
This was submitted 2 weeks ago [1], with the title "Computer Scientists Create Algorithm That Measures Human Pecking Order", which is appended to the main title for the article. (reached 48 points)
pg has resubmitted this 2 weeks later, using the subtitle of the article, "The way people copy each other's linguistic style reveals their pecking order." (so far 96 points)
Now I'm not nitpicking about credit, I don't care. My question is to what degree is this submission getting more traction from pg's status, and to what degree is it getting more traction due to the linguistic choice of a more specific, better(?) title?
I'd say pg, at least here on HN, is higher up in the pecking order than the person who submitted the article first (llambda).
On the other hand, the subtitle is certainly more descriptive, since computer scientists have been making algorithms to measure human pecking orders for a while, with large companies seeking to build brand confidence by giving away their products to "influential individuals", etc.
This is different than the effect described in the article. The original article contrasts their new work with previous approaches that focus on structural/network effects: who's talking to whom, etc.
I guess what spiked my interest is that your power and/or reputation still affect the linguistic choice of whether to respond at all.
The earlier submission appears to be from Christmas Eve, when the entire world was on a weekend and most English-speaking countries were either on a major holiday or preparing for its imminent arrival. This submission was late on New Years Day in the USA, at a time when much of the world was already halfway through Monday, January 2. (I believe this was submitted at noon on Monday Sydney time, for example.)
That's just one of many alternative hypotheses to explain the difference in traffic. Obviously, there's also a raft of "competitive" hypotheses: Articles on HN don't sink or swim in a vacuum, they have to compete with the other articles on the site at any given moment.
I clicked both links because i didn't know what "pecking order" is. I don't think i ever notice who's submitting what. Regardless, it should be easy to test your hypothesis if we have data about the submissions from karmillionaires compared to the rest of the herd.
While it is always nice to have harder proof than is usual in the humanities and to a lesser degree social sciences, I feel mildly irritated by the coverage which presents this as a "discovery" by clever tech people. These facts are pretty basic in sociolinguistics, the study of discourse, psychology of language etc., as amply acknowledged in the paper itself.
I'd be curious to know how much of this applies to non-native English speakers.
I guess their native language can shape the way they build sentences and influence the way they communicate in different way. Or, maybe, they just attempt to correlate more with native English speakers.
It's interesting, I would have thought something different. I'd have thought that people who are more capable of adapting to the communication style of their peers had a subtle advantage (though that may not reveal pecking order).
I've always thought of a person that can hold a high level, technical conversation one moment, then adjust their vocabulary and speaking style to communicate with some neighborhood children the next, then adopt a pidgin to communicate better when dealing with a non-native speaking friend, then switch to a local drawl when getting their car fixed by the local mechanic, to have a special sort of power w/r to communicating effectively with their audience.
People who don't seem capable of adjusting their patterns based on audience seem more..."brittle" somehow.
It's interesting that linguistic co-ordination is being used to determine "pecking order". Maybe so, but I observe things a bit differently.
I grew up in a part of the country that was mostly rural, toward the midwest/southeast part of the US. I was raised believing that "y'all" was a label for referencing a group of people.
I no longer live in this part of the country, but return from time to time for family. Re-connecting with family and friends, my wife will notice that I start to sound like everyone with whom we're surrounded (she kindly refers to this as my "inner hick".)
If this is true (I assume it is), it is completely sub-conscious. If there is anything I'm consciously doing, it is attempting to bridge any perceived gap in communication.
In terms of pecking order, it may be the case that I'm conforming to my surroundings. I tend to think of it more as "when in Rome".
I don't care for the value judgment implied by the term "pecking order", as it denotes that someone is in control or holds more influence. Communication carries an underlying intent, a reason for the communication to occur in the first place. Linguistic analysis can measure syntax and output, but it doesn't measure intent. And intent is important to understanding the value of communication.
When the Google guys came up with PageRank, the assumption was that the intent of including a link was to point to a more credible source. As we now know, assuming the intent of a website to link to another as a means of assessing credibility is flawed.
ah, that explains why both my manager, his manager and the team's group manager keep making the "... is is that" mistake. I'm a poor ESL guy, and my grammar parsing brain paths get fizzed whenever I hear them speak like that.
43 comments
[ 3.8 ms ] story [ 113 ms ] threadIn a more practical sense, simply being able to recognize these signals would be very valuable in meetings, especially sales meetings involving pitching a group when it is unclear who the actual decision makers are.
Your second paragraph talks about being able to notice these things in meetings. Penbrooke's research[1] shows that our brains ignore structural words. So it's pretty impossible to take advantage of this particular research.
However, I think there are a few things you can do.
1. You're first paragraph about "inside jokes" is spot on. When I was in college it was pretty common to to get what were called "Dorm Floor" shirts. The front had some slogan or name on them and the back had quotes from your dorm floor. When we went to create ours there were three of us who dominated the floor quotes. So they made a limit for how many quotes a person could have. Most of these quotes were essentially inside jokes.
If you want to know where you stand in a group it's pretty easy to test this using inside jokes. Can you start something? If they respect you they will pick up your inside joke.
2.When I was in college, I took a lot of discussion based classes related to political theory. I noticed my self picking up the mannerisms of a kid in my class. I then realized that others were too. This was a huge eye opening experience for me. I realized that mannerism and the way we speak is organic and if we can record it's evolution it tells us something about how power and society works.
The down side of this is that in a meeting you don't see the birth of the mannerism so you don't know who the decision makers are. You also can't mentally process the structural words, so I don't think there's much advantage there.
[1]http://secretlifeofpronouns.com/
But then I decided to play along and started to pick professors and copy their mannerisms and verbal style on purpose. So on week I would talk and act like a math professor, then like the physics one. I didn't do it in an overt, derogatory, or offensive, but in a subtle way. Doubt anyone even noticed. It was interesting that it was easier to copy from those that I respected and looked up to. some that I didn't care about just came harder to do.
1. People make sense of things through language. (You can get this sentiment in sound-bite form from any number of linguists or language philosophers, e.g. Wittgenstein's "The limits of my language mean the limits of my world.")
2. People with more linguistic resources are better able to make sense of things, and thus tend to rise to the top of the pecking order. (N.B.: Language is more than just words, so this isn't simply an issue of vocabulary.)
3. People with fewer linguistic resources are drawn to the expressions of people with more linguistic resources because those expressions give them new ways to make sense of things. They're like design patterns for thought.
4. Thus, what's being measured isn't power or influence, at least not directly.
But of course none of this is relevant to the article, which draws its conclusions from the degree of periphrasis in a discussion, rather than the substance of what is said. So, again, very interesting.
I'm with you until this because it doesn't seem to draw from points 1-3.
If b) is the case then it makes sense to frame the issue as "The way people copy each other's linguistic style reveals their pecking order." But if a) is the case then the article's conclusion is tautological. Of course you're right that that doesn't stop us from taking it as a direct measure of power or influence, since they're still strongly correlated. I was just thinking of it more as a measure of linguistic reserves [1].
[1] http://goo.gl/NOUwo (Sorry to use a link shortener, but it's a long Google Books URL.)
HN shortens automatically...
If you can reach a zen state of submission to the ocean of expression around you and follow the currents, you could learn like a child again. That's dang hard to do.
So over the years, I've internally made a (somewhat silly) point to be aware of this and not try and change the way I speak based on whom I'm speaking too.
He also does twitter stream analysis using his research and you can find that here --> http://analyzewords.com
People might dismiss it as tripe, but after you've read some of his assertions and look over your own emails you'll probably believe it makes a bit of sense.
Yet later in the book he also mentioned mirrored language and who reflects who. He also had some sort of web program that measured the difference between the two. I think that correlates with the goals of this study.
The theory continues that different programming languages reflect different linguistic styles and thus your choice of programming language can affect the way you think.
As an example, I do a large amount of text processing work using sed. Interestingly (imo), sed is a language designed by someone who had a degree in Psychology.
As a result, to some extent my way of thinking would probably differ from someone who uses, e.g., Perl as their preferred language for processing text.
What I wanted to get at is this, pecking order might not be the only reason to copy verbal styles. Empathy might be one more reason. Does this study control for this?
"There are a number of techniques that are supposed to be beneficial in building rapport such as: matching your body language (i.e., posture, gesture, etc.); maintaining eye contact; and matching breathing rhythm."
source:
http://en.wikipedia.org/wiki/Rapport
I think that's changing nowadays. People are more exposed to English via TV series and movies and stuff.
With regard to your point, whenever I'm London and find myself having a conversation with someone with a cockney accent. After a few minutes, I start to mimic their accent. I can't help it. I love the accent.
When Kevin Rudd was the Australian Prime Minister, every Labor MP sprinkled their conversations with "can I just say" and asked themselves questions.
Now it's Julia Gillard and we're being treated to being spoken to like a class of special-needs 5 year olds.
This was submitted 2 weeks ago [1], with the title "Computer Scientists Create Algorithm That Measures Human Pecking Order", which is appended to the main title for the article. (reached 48 points)
pg has resubmitted this 2 weeks later, using the subtitle of the article, "The way people copy each other's linguistic style reveals their pecking order." (so far 96 points)
Now I'm not nitpicking about credit, I don't care. My question is to what degree is this submission getting more traction from pg's status, and to what degree is it getting more traction due to the linguistic choice of a more specific, better(?) title?
I'd say pg, at least here on HN, is higher up in the pecking order than the person who submitted the article first (llambda).
On the other hand, the subtitle is certainly more descriptive, since computer scientists have been making algorithms to measure human pecking orders for a while, with large companies seeking to build brand confidence by giving away their products to "influential individuals", etc.
This is different than the effect described in the article. The original article contrasts their new work with previous approaches that focus on structural/network effects: who's talking to whom, etc.
I guess what spiked my interest is that your power and/or reputation still affect the linguistic choice of whether to respond at all.
[1] http://news.ycombinator.com/item?id=3385517
That's just one of many alternative hypotheses to explain the difference in traffic. Obviously, there's also a raft of "competitive" hypotheses: Articles on HN don't sink or swim in a vacuum, they have to compete with the other articles on the site at any given moment.
Perhaps it reveals some subtle "pecking order" or perhaps it simply reveals I submitted it at a bad time. :)
I've always thought of a person that can hold a high level, technical conversation one moment, then adjust their vocabulary and speaking style to communicate with some neighborhood children the next, then adopt a pidgin to communicate better when dealing with a non-native speaking friend, then switch to a local drawl when getting their car fixed by the local mechanic, to have a special sort of power w/r to communicating effectively with their audience.
People who don't seem capable of adjusting their patterns based on audience seem more..."brittle" somehow.
I grew up in a part of the country that was mostly rural, toward the midwest/southeast part of the US. I was raised believing that "y'all" was a label for referencing a group of people.
I no longer live in this part of the country, but return from time to time for family. Re-connecting with family and friends, my wife will notice that I start to sound like everyone with whom we're surrounded (she kindly refers to this as my "inner hick".)
If this is true (I assume it is), it is completely sub-conscious. If there is anything I'm consciously doing, it is attempting to bridge any perceived gap in communication.
In terms of pecking order, it may be the case that I'm conforming to my surroundings. I tend to think of it more as "when in Rome".
I don't care for the value judgment implied by the term "pecking order", as it denotes that someone is in control or holds more influence. Communication carries an underlying intent, a reason for the communication to occur in the first place. Linguistic analysis can measure syntax and output, but it doesn't measure intent. And intent is important to understanding the value of communication.
When the Google guys came up with PageRank, the assumption was that the intent of including a link was to point to a more credible source. As we now know, assuming the intent of a website to link to another as a means of assessing credibility is flawed.