Hacker News Reader

Comment by lmkg | Hacker News Reader

Without the commas, it sounds to me like the author is excited and animated about his topic, tumbling forward. I imagine him gesticulating wildly. I guess you were imagining more of a measured pronouncement, with paused emphasis to cause us to carefully consider each element. As he says, the punctuation imparts different styles.

I think the tumbling-forward interpretation works better with the list having four elements. They're meant to be a cumulative item, making his point as much by the number of different elements as by the meaning of each individual one. This style works better when it gives the impression of rushing.

View on HN

I would argue the premise of this article is invalidated if you think you can "rush" through a comma; that, in fact, argues the opposite: that the punctuation is more of a syntactic guide to the form of the sentence and is often irrelevant to the way it should be interpreted for spacing, pause, meter, and verse (<- "spacing/pause, meter and verse" there sounds great, but "spacing, pause, meter-and-verse" is lopsided; and it would be made all the worse if I had the sentence continue after "verse" in a "rushed" fasion after forcing the pause with the comma after "spacing" and "pause").

I've read that sentence a few times now both imposing and not imposing the imaginary comma in my head and it still sounds like either a mistake or a particularly weird choice of rhythm.

Without an oxford comma, a four letter list sounds like "This...that...c and d do that!" -- which in turn sounds like naming two things, then making a completely different sentence about the final two things. People don't talk that way out loud when they mean to include the first two items in the list.

I love the Oxford comma, because it leaves no room for ambiguity. I've heard anecdotes of people omitting the comma in contracts and having it come back to bite them in the butt... but who knows if there's any truth in them or not.

Except that the Oxford comma does leave room for ambiguity, c.f. http://en.wikipedia.org/wiki/Oxford_comma#Creating_ambiguity

Frankly, with writing in natural languages, hard-and-fast rules like, "Oxford Comma are always unambiguous," and, "Always avoid passives," tend to make your writing weaker, because there are still some cases in which Oxford Commas are ambiguous and passives are the right tool for the job. What makes great writing, whether it be creative writing or contract writing or what-have-you, is knowing when to apply such rules and when to disregard them.

From Wikipedia: "Without a serial comma, the above dedication would read... a phrase ambiguous only if the reader accepts the interpretation..."

In this case the serial comma isn't creating any ambiguity, the bad writing is (and the serial comma shifts the ambiguity to the second/third item being listed). When they suggest a better way to write the example sentence, it uses the Oxford comma both times (though I'd argue that the second sentence isn't a solution).

I agree that nothing is ever absolute (hah), but the Oxford comma takes care of ambiguities most of the time, and merely shifts the ambiguity around when appositive phrases are thrown into the mix. And reducing ambiguity should be the goal of every writer.

Those two suggestions use the Oxford comma because they are suggestions for how remove ambiguity while keeping the Oxford comma.

The thing about just admitting you can rewrite a sentence to be un/less ambiguous while keeping the comma is that it simultaneously admits that you can rewrite a sentence to be un/less ambiguous without the comma... which puts us back at aesthetic arguments.

There's also the semicolon.

The situations where it does create ambiguity are much fewer than without the comma, and the example given in the link is trivially fixed by altering the sequence of words. It's an unusual example, whereas omitting the Oxford Comma routinely makes the last two items sound like one melded item.

My favourite example of this is "We partied with the strippers, Stalin and Hitler".

That example is horrible.

> Without a serial comma, the above dedication would read: To my mother, Ayn Rand and God, a phrase ambiguous only if the reader accepts the interpretation my mother, who is both Ayn Rand and God.

That's exactly the same grammatical interpretation (i.e. the comma indicates an appositive) that was suggested to be ambiguous with the Oxford comma present.

The point is no one will even consider the interpretation where his mother is god; thus the sentence is unambiguous - he is dedicating to his mother, to Ayn Rand, and to god.

The ambiguity in the example is in whether his mother is Ayn Rand, not whether she's god.

>Without an oxford comma, a four letter list sounds like "This...that...c and d do that!"

It sounds more like: bit, bat, bot and bet do that to my ears. E.g the list flows in a continuous motion, and the "and" just denotes the last item ("and finally"), something like x:xs in Haskell but in reverse! "this, that, c : d do that"

>which in turn sounds like naming two things, then making a completely different sentence about the final two things.

This sounds like a projection, not from the sound of the phrase, but from your a priori idea about what the oxford comma would do there (e.g that it would evenly split the items of the list).

I think the programmers among us are infected by the use of punctuation marks as part of the syntax. Many programming languages just hijack punctuations marks: C's field.selector is plain silly when you think about it and about the fact that the colon was available; and now something less natural than the period must be used to separate statements. As for Haskell, I find that the ML family is IMO quite remarkable for its tradition of terrible syntax choices (again, why x:xs when x,xs was an option? And that convention of putting commas at the beginning of a line!?).

It seems to me that arguing about what a comma means when put in this or that place is symptomatic of this programmer mindset. Natural languages are ambiguous, redundant and sometimes inconsistent by non-design. And so is their punctuation. Nobody understand punctuation, especially programmers.

Just to clarify, in Haskell:

- [x,xs] is the list defined by exactly the element x followed by the element xs;

- x:xs is the list defined by the element x followed by the list xs.

And to add to this, the unspoken Hungarian notation rules have xs typed as a list.

So the 1st instance: [x,xs] looks unprofessional and n00b-ish.

It's the 2nd instance that's idiomatic.

The pace of the sentence is accelerating, launching into the remainder of the sentence.

Momentum - 3 syllables

Syncopation - 4 syllables

Rhythm and pattern - 5 syllables

That's true, but that's also why it doesn't work. Think about the stress patterns that actually constitute rhythm:

  momENtum, SYNcoPATion, RHYthm and PATtern  
  ___--___  ---__---___  ---___ ___ ---____

Nonomatapoeia, I fear.

I 'hear' only one stress in each.

moMENtum [beat] syncoPATion [beat] rhythm-and-PATtern [no beat so continues straight on]

And all three patterns, including the beat, take the same time, with the remainder of the sentence continuing at the same speed as PATtern.

https://en.wiktionary.org/wiki/rhythm: /ˈɹɪ.ð(ə)m/

https://en.wiktionary.org/wiki/pattern: /ˈpat(ə)n/ or /ˈpædəɹn/

I do have an English (UK) accent if that helps. But of course it's not unusual to stress different syllables when reading poetry or prose for effect.

Reply on news.ycombinator.com