Punctilious

I've been wondering why I've never learned this stuff at school proper: the more exotic members of punctuation marks, such as en and em dashes, semicolon, colon, etc. I suspect the philosophy here is like the lion cub being pushed off the cliff: we the fledgling young writers are to fend off  in the real word of writing and out of the jungle of words we would have gained the arcane secret that a cub on a rite of passage has to find out for itself.

And indeed it is best to let soak in the ocean of words to let the subtleties percolate into the deepest of your subconscious, rather than superficially reading the rules from the guidebooks (don't get me started with Strunk and White's Elements of Style, that some linguists love to hate). That said, I will attempt to write a guide of sorts, the purpose being: 1) to crystallise and organise my own knowledge, and 2) to soften the landing for a cub I'm about to push off. Here goes.

From Wiktionary:

punctuation (countable and uncountable; plural punctuations)

1. A set of symbols and marks which are used to clarify meaning in text by separating strings of words into clauses, phrases and sentences.

I will focus here on what I call the pause punctuations — those signifying reading pauses — and excluding those that, for example, denote intonations and distinguish direct and indirect speeches. A hierarchy of these, from short to long pauses, is something like this:

Hyphen (-)

Space ( )

Comma (,)

En dash (–)

Em dash  (—)

Em dash with space ( — )

Colon (:)

Semicolon (;)

Full stop (.)

Paragraph break (¶) 

Let's start from the easy ones. Space separates words, giving us words, the most basic lexical units; full stop separates sentences; paragraph break separates paragraphs. Hyphen joins words; the chief purpose is to give clarity to compound modifier. Consider:

robot gun bullet

This phrase can be parsed ambiguously, is it: a bullet for a robot gun or gun bullet that is robotic? The ambiguity evaporates with a dash of, well, a dash:

robot-gun bullet

robot gun-bullet

Some writers take this compounding utility to the extreme, which creates impact, but (and thus) it should be used sparingly. From Slate:

It's not even 10 a.m., and Mr. Farmers'-Market-Cruelty-Free-Meat-I'll-Have-a-Decaf-Thanks is having his first soda of the day.

As a side note, some words appear next to each other often enough that sometimes the hyphen is omitted. 'Manmade' and 'dillydally', for example. British English usage, though, tends to retain the hyphen.

Comma, as most already know, separates idea that follows the preceding idea directly enough that it doesn't warrant sentence separation. The key is usually to compare which will impart more clarity: with or without comma? A long sentence, for example, would warrant a comma to clearly demarcate the boundaries of ideas the writer intends to convey, while keeping the text smoothly flowing. And flow is important here: a comma is a very brief pause, not as abrupt as a full stop.

What should we do then, if the subsequent idea is quite connected to the preceding one that it is too abrupt for another sentence but not so connected to be merely separated by a comma? And the answer is staring at you—obviously you need a punctuation mark that is both a comma and a full stop, to wit, the semicolon. The recommended use is that semicolon separates clauses; in other words, if you replace the semicolon with a full stop, the resulting sentences are good enough to stand on their own as complete sentences. Why not a full stop then? This is important especially in narrative writing, suspense thriller for example, to simulate a fast-paced, adrenaline-infused narrative that takes your breath away because the sentences are deliberately long like this one; imagine the feeling of being chased, lines of thought come relentlessly, zipping past through one after another; get the idea?

To fine-tune the flow of writing further, though, you would need the longer dashes. The first usage is similar to the semicolon as above, but personally I think it is more fluid:

And the answer is staring at you—obviously you need a punctuation mark that is both a comma and a full stop, to wit, the semicolon.

As an additional bonus, you can modulate the flow a little more by adding spaces; it gives quite a different feel:

And the answer is staring at you — obviously you need a punctuation mark that is both a comma and a full stop, to wit, the semicolon.

Or substituting with an en dash:

And the answer is staring at you – obviously you need a punctuation mark that is both a comma and a full stop, to wit, the semicolon.

(For pure writing purpose, I usually regards en and em dashes as interchangeable. The difference is mostly concerning typographical aesthetics. For example, it is recommended to use en dash for number range, like 25–50, instead of a normal hyphen, but only typographers or OCD patients would notice the difference.)

The second usage is to enclose a modifier. Again, from Slate:

Hanna keeps walking through to the kitchen—25 feet away—to get food for the kids.

'25 feet away' here explains, or modifies, the preceding clause. Again, you can modulate the flow like before. Additionally, there are other ways to enclose a modifier, all coming back to how much flow you want. Compare:

Hanna keeps walking through to the kitchen, 25 feet away, to get food for the kids.

Hanna keeps walking through to the kitchen (25 feet away) to get food for the kids.

Caveat emptor: commas may not be desirable if your sentence is already choked full of them; while parentheses would give the feeling of afterthought (sometimes desirable, say, to voice out the narrator's inner thought).   

Finally, the colon. There was an antiquated use of colon to separate clauses, much like the semicolon, but it's hardly used anymore. The rule of thumb for the colon is that the subsequent idea should elaborate on the preceding one. Thus our common usage to list items: item 1, item 2, etc. Other examples that I shamelessly smooched from Wikipedia:

There was only one possible explanation: the train had never arrived. (syntactic-deductive)

Luruns could not speak: He was drunk. (appositive)

In closing, I would like to reiterate again that in writing, second only to content, fluency is very, very important. Master your pauses and it's one step closer to master the flow. Do check the individual Wikipedia pages for each punctuation if you want to learn more.