[goto_self]

There's a common argument that the purpose of language is communication, that mistakes don't matter if the desired information is communicated, and that prescriptive rules for punctuation and grammar are merely pedantry.

However, that argument is self-defeating. The purpose of language is obviously communication, yes. Mistakes might not hinder communication in some cases, sure. The rules of grammar have changed throughout history and are somewhat arbitrary, granted.

However, as the ideas one wishes to communicate become more complex, abandoning the greater set of grammatical tools makes ones sentences more difficult to read. I see this all the time in professional environments of all places. I frequently have to reread communications because someone was too lazy to proofread.

And as for language changing through time? It does. But that's not carte blanche for making whatever changes one wishes. Language change has to come about by collective agreement, not by some cowboy who doesn't like grammatical rules.

[jkingsbery]

Agreed, a lot of grammarians are pedantic for sport. To bring this around, since this is hacker news: the way I think about it is as error correcting codes. For a given sentence, with a certain amount of errors, others can still understand it, but beyond a threshold, you can't (without thinking hard) . The Rules give us that margin for when we make mistakes.

I think this gels with your point about professional environments - as ideas get more complex, that margin for error gets slimmer.

[kmill]

I try to operate under the principle of taking care to be clear with what I say or write and to take care to understand what I hear or read. (It's the human version of internet protocols or combinational logic levels.)

This is a very strong position if you care about trying to transfer thoughts from one person to another, but there are two problems. The first is that the people you are interacting with might not share the same protocol, which can be frustrating (even with just slight variations in levels of effort), and the other is that not all communication is about transferring thoughts.

Normative statements about how communication ought to occur, like whether one needs to follow prescriptive rules, are akin to the way engineers debate an API design. In the end, it is about setting a boundary that determines who has to do what work. The more clear you are (and clarity can take quite a lot of effort!) means less work for the recipient (which, if it is something you think they must know, then it might be worth it). The more effort you put into understanding someone means less work for the speaker/writer (which, if it is something you want to know, then it might be worth it). Conflict arises when there are different expectations, here like anywhere else -- for example, two housemates with different expectations for cleanliness.

Many people believe at some level that if they say something, then the utterance has exactly the meaning they intended it to have. This is despite the fact that symbols have no inherent meaning, but rather they are constructed to trigger something in the mind of the recipient. Such people might actively resist answering any questions for clarification since they believe they were clear enough and don't need to do any more work. They might think the question comes from a place of pedantry since they know what they meant.

There are some people who don't care to be clear and don't care to figure out what you are trying to say. Some people even add a bit of noise into their interpreter if it's something they know they don't want to understand, for whatever reason that might be.

When one person corrects or questions the speech of another, this is sort of a side channel, communicating either "I want you to be clearer" or "I am taking care to understand you." Unfortunately, there is not a good way to distinguish between the two without knowing the person well enough. Lots of talking is just to maintain a relationship of some kind, and highly precise language can be inappropriate since sometimes all someone wants is to be heard, even if not 100% understood.

It's nice when two people know the channel model well enough, so to speak, to be able to have high-bandwidth low-effort interactions.

[goto_self]

The high-bandwidth, low-effort interactions you're talking about might only work with quick, two-way exchanges, as in a spoken conversation. In that regard, I definitely have a weird English variant among friends. I enjoy the broken English of so many memes.

There's an additional layer of language where you can switch to a different variant of the rules as part of communication. Speaking in another register defines the tone of the interaction.

But I'd still like to defend standard English as the default even in informal groups, because the standard makes it readily accessible to many people and ensures as much permanence as one can get from language.

All that said, I have one more flip-flopping disclaimer. I think some of the rules agreed upon before are really dumb. Putting punctuation in quotation marks if the punctuation is not from the quote is monstrous. I will not follow that rule in any context.