Conversational Implicature, via H.P. Grice's "Logic and Conversation"

Just warning you: This here's some kooky college philosophy, straight outta the classroom. Long-winded syntax and all. The kind of B.S. where you break down something super-simple and at the end half the class is going, "Wait, what's the point of this again?" If that's not your cup of tea, might ought to skip to the next thread in the section, 'cause this will bore your face off. Not that I don't want to make it accessible to everyone, of course. And this can be useful in identifying and fixing miscommunication.

So in the process of dissecting language, one of the first topics I've been reading on is conversation. Namely, as H.P. Grice calls it, "conversational implicature." But before I explain that, I have to explain Grice's conversational maxims (rules). Now, Grice is basically saying about these maxims that everyone subconsciously tries to follow them and expects them to be followed when they converse with somebody. So just go into it with that in mind.

There's four basic rules:

• Quantity: Give as much information as is required of you. Do not give too much or too little.
• Quality: Do not say something that you believe to be false, or something for which you lack sufficient evidence.
• Relation: Be relevant.
• Manner: Be unambiguous; be brief; be orderly; do not be obscure.

Now, anybody paying enough attention can probably see that people fail at this stuff all the time. For instance, quantity. Say Tom asks Harry, "Did you feed the cat while I was out?" and Harry says, "Yeah, I sure did. I fed 'im the, y'know. The cat food. The food of the cat. He ate it and he is full. Of food." Kinda long-winded, isn't it? What do you think about Harry's answer? Do you think he's lying? It's probably because he could've just said "Yes" and been done with it. There's yer quantity maxim. Either way, for conversation to go smoothly, usually people are sticking to these rules--or trying to.

Now here's the fun part. What's conversational implicature? It's basically bending the rules. You say one thing and end up conveying something totally different. You do it all the time, and probably people don't misunderstand you too much. We're pretty damn good at it. But if you don't know what it is and how it's defined you might not even recognize you're doing it. So when you say something and somebody gives you a funny look, you don't even know how to explain yourself. So let's fix that.

I'll start with an example. You're at a friend's house, and he's cooked you both some dinner. Say he throws in a dinky little store-bought dessert. The dinner was terrible. When you're through eating he asks you this:

"Did you like the meal?"

Not wanting to offend him (too badly), you say:

"The dessert was delicious."

What is even. The conversational maxims just had a heart attack. Let's break this down. He asked you how the meal was. You told him about the dessert. You didn't even mention the meal. That's quantity, first off, 'cause you didn't give the expected information; and it could be relation, too, since he wasn't even asking about the dessert. But unless the guy is a little slow he will probably get your meaning. You didn't say anything about the meal, which means you probably didn't like it. Life goes on, conversation goes on. Well, if he doesn't throw your ass out for insulting his cooking, that is.

So what's the point of breaking this down? It's so you can explain yourself better when you say something and people give you looks. Let's take a more volatile example. Your friend just got a job. Kind of a crappy job. He's been complaining about it, but at least it's money. His girlfriend's been out of town. She's back today and she calls you to check up on him, since he's at work.

"How's Harry doing?"
"Got a job, finally."
"Oh, good! Does he like it?"
"Eh, hasn't killed anyone yet."
"...He... He wouldn't do that, right?"

Hold up now. You were just joking. It's funny because it's absurd, right? He wouldn't do that, which makes it silly to say. But she didn't get that part, for whatever reason. Maybe now she thinks he's got a mean streak she didn't know about. Might be a little out-there, but you see how this can cause issues. Either way, most of the time it works, and it's neat to break it down. Metaphor works like this too: We're well aware that "All the world's a stage, all the men and women merely players" is not trying to tell us that planet Earth is a literal wooden stage (breaking the quality maxim: don't say #$%& you know ain't true), but it's implying that the two have some things in common.

So, discussion time! Any questions? Things need clarifying? Want to hear more about or expand on Grice's idea?

Personally, I think Grice is focusing a little too much on the speaker in conversation; sometimes the speaker thinks s/he's obeying the maxims, but the listener interprets what s/he says in a way that breaks them. So I think there should be two sets of maxims: One for the speaker and one for the listener. So, here might be some listener maxims.

• Do not expect more or less information than would be logical to give.
• Allow for the possibility that, if the speaker's assertion makes sense, either it is true or the speaker believes with reason that it is true.
• Try to infer the relevance of the speaker's point.
• If unable to fulfill one of the above, ask for clarification; do not make assumptions.

Just a prototype; obviously it's up for discussion.

 

I agree that your prototype seems to solve it in a logical sense. What Grice seems to be aiming for is absolute efficiency in getting points across without regard to how one might react to what you wish to say. It leaves intuition out of the equation and favors a more literal approach to everything that you say, and would have conversations work in a machine-line way, without nuance and without intuition or manipulation. Conversations convey more than literal ideas on an enjoyable playing field, and there are many situations in which I would rather not be ruled by notions of efficiency over subtlety. Conveying or trying to invoke emotion is not covered by the rules that you listed. Is conversation the term that I am mixing up?

 

This is the point in introducing conversational implicature. The model makes no effort to constrain conversation or its mechanics; only capture them for study, in a sense. Where the conversational maxims offer a base of understanding on which to build, con-imp, as I like to shorten it, accounts for the use of aforementioned subtlety in conversation: how it works, why it works, for what it's used, etc. And, a good point, it allows for the expression of concepts for which there are few or no words. It also allows the speaker to preserve the impact of an emotion or expression, where the literal meaning would fail to do so. Metaphor is often employed to this end.

 

Oh god, I remember doing Grice's Maxims in English, and also Labov's rules of engagement as we called it.

The theory I believe was put forward many decades ago, in was more akin to the phatic conversations people would have in the quintessential conversation of the 50s and the normal colleagues and neighbours talking to each other.

I believe it gets the point across, but the maxims rely on the implication that people will act a certain way when talking to anyone. It doesn't count for factors such as how well you know the person, how often you talk, the personality of the talkers, etc, since they can all have effects on how we speak to one another.

 

It shouldn't have to cover those things. The rules of chess do not account for popular or effective strategies, or how well the players know each other's play styles. Grice's maxims are, again, a foundation upon which to be built; or, put another way, a context in which we may examine intent, familiarity, frequency, and so on. Context is a large part of con-imp and why it works (or sometimes fails, as in my example above with Harry's new job). It's not about how we speak to each other, per se. It's about, all things being equal, what we expect from each other when we converse, and how we fill in the blanks when we don't get quite what we expect.

 

But a conversation is rarely equal. You have taciturn speakers and prolific speakers, both can show different forms of being the more dominant or submissive party depending on the context of the situation.

I don't know how to make a conversation equal between all involved without it either being forcibly assigned, or just a natural occurrence.

 

Again, what you're describing is unique to the conversationalists. To incorporate that into Grice's maxims would make them unable to be applied to every situation. Context is added as needed and adjusts the exact specifications of each conversation. What you're describing is, once more, the Chess players' strategies, not the rules to which they must adhere (or appear to be adhering) even to make those strategies work.

Granted, the Chess analogy is not entirely fitting, as people can break or flout the rules of conversation. Hence con-imp. But if those maxims did not exist subconsciously in our heads for us to say, "This statement, taken literally, does not fulfill one of the maxims; thus there must be a different message implied," then we would not be able to express these subtleties to begin with.

And it is a natural occurrence. It's describing a phenomenon to which we hardly pay any attention. The point in examining it further is not to say something is wrong with it, but to better understand it and to be able to explain and resolve mis-communications when they do happen. It's taking a bit of the mystery out of it and helping us to do with purpose what we usually do just because we're used to it.

 

Maybe I did not delve into Grice's Maxims as much as I thought I had. I thought it was completely conversationalist based and not applied elsewhere. And I also didn't know that such an maxim was so readily accepted in modern speech. Still I don't know enough about it to really say whether it is still relevant.