Transferring meaning is what communication is for. The words are a means to an end. I have a thought, I wish to convey it to you. I construct a sentence which, to me, evokes this thought, and transfer it to you. Assuming reasonable fidelity in the method of transfer (normally ok, but questionable in, say, a noisy room) you then decode the message. There is no guarantee that what you decode is the original thought, especially when humor and sarcasm are taken into account. Back in the mid '80s, people on usenet realized that sarcasm was skewing the interpretation of their messages, so they developed a way to indicate that one was not being serious: the emoticon :). This was smart and simple.
Communication works between computers. If I share a .cpp file with someone, his compiler is guaranteed to interpret the code in the same way that mine did. There is no ambiguity: all enclosing braces are where they should be, all keywords are spelled correctly and used appropriately. Any order of operations is specified in the current language canon, which is defined by those creating the compilers. Since, for the vast majority of languages, the programmers agree, one can be sure of accurate communication.
Humans do not communicate in this way. We do not use logical structures, our grammar is often ambiguous, and there is most certainly no canon. At most, linguists track language by a democratic system, and normally accept usage that comprises over, say, 50% of people who use a given word, phrase, or construction. No thought is given to what would be the clearest. The system only works as well as it does (sometimes not very well at all) because people often are good at guessing the context of something. I saw something like the following sentence recently:
"Several Rx's are known to have side effects."
How is a person to know whether this refers to several prescriptions, or whether the words following "Rx's" are something that belong to a prescription? We know by context---'several' clues us in.
Our system of punctuation is supposed to provide a logical structure with which to communicate. Sentences would be very difficult to parse if one were to omit tokens such as the period, the space, or the comma. And as the bestseller points out, the statement "He eats, shoots, and leaves." would actually change meaning were the commas removed. This system works quite well, and makes it so that our vocabulary can still consist of relatively few words, some doing double duty with the help of punctuation ('shoots' being both a verb meaning "to fire a gun at" and a noun meaning "tall thin flora").
English grammatical syntax would benefit from some changes/additions. Changes should give clarity and reduce ambiguity in written communicative interactions. Following are some that I would propose. None of these have to do with spelling changes, or anything like that, since the words are more or less unimportant---you just have to choose some. If, in java, the keyword 'delete' were replaced by 'remove' or 'kill', this wouldn't matter so much from a standpoint of clarity. For spelling, we do have a canon, thanks to the fact that dictionaries all decided to agree on common spellings.
- Apostrophes used only for possessives and some contractions
Hardly anybody would pluralize something like 'can' with "can's". However, with so many acronyms and abbreviations floating around, it has become a sad custom that some words are pluralized with an apostrophe. Words like "Rx's", "ATM's", and so on. Sometimes the meaning is just not clear, like if someone were to say "It's the teacher's", does that mean that 'it' belongs to the teacher, or that it is the group of teachers coming up the road? I understand that some words just look plain weird without this, such as if I were going to say that all instances of the letter 'a' were showing up wrong, I might say "the a's are messed up". In this case it works, but we ought to have a rule. I propose the hyphen here: "the a-s are messed up."
- Quotations left as-is
It is the very stupid and arbitrary rule, adopted by most authors, that quotes coming at the ends of sentences have periods inside the quote, and commas if in the middle of a sentence. The following is "correct":
Orwell writes, "he came into the room," and then proceeds to describe the room.
The full sentence may well be "He came into the room." in which case having the comma there is just a lie. Why not retain fidelity. The following is unambiguous, and, moreover, actually truthful:
Orwell writes "He came into the room." and then proceeds to describe the room.
Nothing is lost by this, but the sentence now appears as it actually is. A text search of the line will work, and the reader knows that the sentence is short and terse.
- Quotes around words used as objects
Throughout this article I've used many words not as they normally are, but as objects, or nouns. If I say "I run.", run is being used as a verb, but if I'm talking about 'run', then it's a noun, an object that's being discussed. TeX has a similar facility for this when indicating a mathematical object:
"Angle $AOB$ is 30 degrees." Here I'm marking off a piece of mathematics, an angle in this case. One realizes immediately that we don't mean a word spelled "AOB", we mean an mathematical object. Philosophers, who commonly talk about words and their meanings already commonly do this, but it seems a good thing regardless when discussing linguistics. After all, wonderful sentences like "What his definition of 'is' is." really only make grammatical sense with the quotes in. - All commas left in itemized lists
Many people think it quite slick to avoid having to type one extra pesky comma when writing an itemized list. For example, they write "The planets Mars, Jupiter and Saturn cannot support life." Many times, however, an item in a list has an 'and' in it. E.g. "The gunmakers Glock, Smith and Wessen and Colt lobbied for the bill" is quite weird, since one of the items is "Smith and Wessen". If we were all to always include all commas, this list would be perfectly understandable. Moreover, there would never be any question as to when the list was ending: a comma followed by 'and' or 'or' would mean the list was done.
I don't imagine that these changes will fix completely our communication. There probably won't be a day when we compile our essays and send them off to people. But these seem like reasonable steps towards a more logical manner of writing.