pragmatics: communicating preferences & politeness

[gusl]

How can we achieve direct, efficient communication?

Communicating about preferences is often problematic, because language does not provide us with reliable ways of communicating fine degrees of preference.

Imagine for instance: the sound of your neighbour's TV is annoying. This is a small annoyance that you'd rather not have to deal with, and you're not sure if they are watching it anyway... and even if they are, maybe they are indifferent about watching TV or doing something else. But maybe they do want to watch it, so you would like find out whether this is the case. You could say:

"would you mind turning down the TV"?

Just the fact that you bothered to mention it is evidence to them that it is a big enough deal... because there is a convention whereby you don't complain about small annoyances: you're only supposed to complain if it's a big enough deal. Even if you explain to them that it's actually a small annoyance, they may not believe you, because a standard politeness strategy is to downplay how big a deal something is.

This is likely the consequence of an important principle of politeness (which took me a while to grasp): don't force people to make difficult choices. i.e. make it clear to them that they have the option of declining without fearing the consequences. When you don't give people that option, you risk having them frame the situation as being one of your interests against their interests, and this could damage your relationship.

The problem here is that the linguistic community eventually adapts to the polite language: "would you mind doing X?" eventually acquires the meaning that "please do X" originally had. To be polite now, you need to make the sentence even longer: "would you mind doing X, if it's not too much trouble?". This is linguistic inflation, and the real losers are clarity and efficiency of communication.

Suppose you run a business, and you'd like an employee to work during the weekend, but NOT if this means they are going to hate you, i.e. you only want them to work if they don't mind it too much. How do you find out how much they mind? How do you know that they feel the freedom to say no?

Why don't we use numbers to communicate how much we want something?
Why don't we use dollars to communicate how much we want something? One answer is that this creates perverse incentives: if you customarily pay off your neighbour to turn down his TV, he now has the incentive to turn it on exactly when it's likely to annoy you.

Comments

[gwillen.livejournal.com]

Why don't we use dollars to communicate how much we want something?

Another problem is that different people value money differently, depending on how much they have. If you offer me a dollar to turn off my TV, that dollar is (to you) equal to the value of not having to listen to my TV, and you're hoping it's greater than (to me) the value of having my TV on, but since our valuations are different this doesn't actually mean (if I take the dollar) that the value to you of having the TV off is greater than the value to me of having it on. It could just mean that, although I actually desparately want to watch that show, I'm substantially poorer than you are and value the money more than you do.

[gustavolacerda.livejournal.com]

since our valuations are different this doesn't actually mean (if I take the dollar) that the value to you of having the TV off is greater than the value to me of having it on. It could just mean that, although I actually desparately want to watch that show, I'm substantially poorer than you are and value the money more than you do.

This is true. Nevertheless, the trade is still a good deal for both of us.

[gustavolacerda.livejournal.com]

that dollar is (to you) equal to the value of not having to listen to my TV

Actually, it's less. If it were equal, there would be no point in offering the deal.

[brianfey.livejournal.com]

I like how you think.

It is so tricky to figure out among us how to make things be optimal by our own standards.
There are many ways we can communicate better by standardizing certain types of communication for mutual benefit.

[smandal.livejournal.com]

I wonder if you're missing some tacit assumption having to do with human relationships. The object is not to solve this particular task, but figure out how much you can get away with at an empathic/interpersonal level in general.

[gustavolacerda.livejournal.com]

Can you rephrase your comment? As I said, I am concerned with people's failure to signal their preferences precisely, and the suboptimal utility that this causes.

[smandal.livejournal.com]

I claim that there is often a higher utility by seeing how far you can push the other person before (s)he gets mad, to see what you can get away with in the future.

[gustavolacerda.livejournal.com]

Many people are passive-aggressive, and it's not easy to tell if they're mad.

[smandal.livejournal.com]

If they're passive-aggressive, you can probably push them a certain amount before they really try to screw you. Again, takes empathy.

[triple-entendre.livejournal.com]

If I may turn that around: The object is not to solve this particular task, but figure out how you can create enough empathy in (/with) the other person so that *they* can deduce how important what you are wanting from them is and possibly make different choices (ones that are more to your preferences) because they now have this knowledge.

And they are motivated to do so because you are creating (however temporarily) a relationship. A typical person experiences positive feelings from having a relationship with other persons, especially if they feel it is a beneficial relationship.

This is a useful principle, by the way, of Aikido -- by first harmonizing with your opponent, you can achieve your objectives through actions which amplify those of your opponent. It gives you many, many more options in that moment of interaction.

[smandal.livejournal.com]

Yes, that is a more precise description of what I was trying to convey. By building a relationhip and putting the onus on the other person, you can "engineer" his or her behavior effectively.

Thanks.

[triple-entendre.livejournal.com]

I think the problem is even deeper than language.

I'm not sure I could even maintain a consistent scale *internally* by which I could accurately describe how much I mind something.

In other words, how do we as individuals even know how much we want something?

Quantitative and/or economic models can predict individual behavior only if it is possible for that individual to internally model the "market" they are interacting with. I can only speak for myself, but I lack the "working memory" to do that very well.

Do you mind?

[triple-entendre.livejournal.com]

Another paradox is that the polite language is often taken as sarcasm. I find I have to speak slowly and carefully, with a smile and eye contact, in order to have polite language understood as such.