The Birth of the Semicolon | Hacker News Reader

The Birth of the Semicolon

theparisreview.org

45 pointsOsiris3040 comments

If we're speaking about punctuation, there really needs to be another mark in between a period and exclamation point, something that means what you said is sincere.

Responding to a text by saying "Thanks." or "Thanks!" is so different. If I was talking to someone and they gave a compliment, I'd want to say a sincere "thanks", but there isn't a way to do that in a text that doesn't come off as weird.

thanks :)

"Thank You"?

This. Expanding words and phrases out of their acronyms, abbreviations, or shorted versions lends an air of genuineness and intention to what you are saying.

Well, even "thank you" is an abbreviation of "I thank you." :-)

Even better, “thank you {name}.”

> something that means what you said is sincere

Ideally, everything one says should be sincere, so your idea sounds weird to me. I think I understand you want more options in the communication medium of text since we can't have body language and voice tone here.

Personally, I just try to be consistent with my expressions. If my "Thanks." is always used in situations where I'm "sincere" or "deeply appreciative" or otherwise, then that's what it means from me.

There are people that express themselves with absurd amounts of exclamation points and all sorts of hearts and blushing emojis, and I understand that they haven't gone absolutely crazy for me because of that. I just keep tabs in the different expressions different people have for the same things.

The meanings of expressions are determined by those who use them.

> The meanings of expressions are determined by those who use them

communication is a team sport though. "intended meaning" is only worth anything if the recipient understands it

> Ideally, everything one says should be sincere, so your idea sounds weird to me.

This is an extremely characteristic HN/engineer comment.

People use irony in communication. It's part of the human repertoire. Saying "oh great" in response to bad news, for example. The notion that ideally there would be no irony in textual communication isn't a practical or desirable possibility. It's in the same Zuckerbergian category as beliefs in the transparency of communication or the sacred integrity of having a single online identity.

The thing that really kills the "sincerity mark" is that there's no reason why it can't be abused and lose its meaning, as has happened with the word "literally".

> The thing that really kills the "sincerity mark" is that there's no reason why it can't be abused and lose its meaning, as has happened with the word "literally".

Yeah, I was thinking that too. I forgot to mention.

Thanks, sincerely.

We just use emoji for that now :sunglasses:

„‚cultural rebirth’ after the gloomy Middle Ages”

Not the point of the article, but that slanderous characterization of the Middle Ages for some reason refuses to die even centuries following the fanaticism of the Reformation and the Enlightenment. It is well understood that the Middle Ages were a remarkeable period, culturally speaking. The Renaissance frankly grew out of the Middle Ages.

Interesting that the semicolon was born out of a need for a longer pause in spoken language, rather than as a means to connect two independent clauses as is taught today

Or as an unnecessary end of statement character to clutter up your code with.

> Or as an unnecessary end of statement character to clutter up your code with.

To the benefit of compiler writers, in detriment of everyone else.

I've come in to fix bugs caused by ASI many a time. Having an explicit character to denote the end of a statement is much more valuable than people pretend.

And don't get me started on meaningful whitespace vs brackets.

> Having an explicit character to denote the end of a statement is much more valuable than people pretend.

No, I'm not talking about automatic semicolon insertion. I'm talking about _ not having semicolons at all _.

Many languages are like that – in fact, almost all languages which are not descendants of the C branch.

I think you quoted the wrong part of the message there, but I understand.

Still, without a line terminator you either have to use newlines, which sometimes will help make cleaner code and sometimes will make a mess of temporary variables and such. Or you have ambiguity, which is why I bring up the the anecdotes about fixing ASI errors.

Which makes it interesting that the Go creators, being heavily into C before, decided to ditch the end of statement semicolon.

Virtually all punctuation marks started as prosody hints, indicating intonation and pauses, suggesting how the text is to be read aloud. In fact you can argue that this is still the case: it was only later that we invented grammatical structure rules for them, and even those are applied only in certain contexts like professional writing. In casual writing, we tend fall back to using punctuation as prosody markers, for example: "Worst. Episode. Ever."

Yes, reading a transcript without punctuation can give one a visceral sense of what drove its introduction.

If there are two independent clauses, why not keep it 2 sentences? I personally use it to connect two separate thoughts that have a related subject.

"Independent clause" doesn't mean "unrelated thought," it's grammar terminology for a phrase that makes sense as a simple sentence: https://en.wikipedia.org/wiki/Independent_clause

Also that the article talked about the colon as having a longer pause than a semicolon (different to how it is used today).

Did they not have the ellipsis back then?

> The semicolon was born in Venice in 1494

Again! Ten years ago I met my (now) wife in Venezia (Venice), and since then I've become really interested in Venezia's history, inventions, etc.

It's incredible how many ideas, inventions, trends, important people are inextricably linked with Venezia between the 14th and 17th century. It was never a huge city, but it has been deeply influential, in a way similar to today's concept of "alpha city" [0], but not just limited to economic importance.

[0]: https://en.wikipedia.org/wiki/Global_city

The typography on the scanned 1494 page is beautiful. I also remember seeing the ancient Roman grave stones with the very nice letters, but I admire how they developed the types for printing.

Aldus Manutius and his workshop assistants had an incredible eye for design. Not just the typography but the layouts as well tend to be more elegant (to my eyes) than what we have today. One of my favorite examples: https://commons.wikimedia.org/wiki/Category:Hypnerotomachia_...

Is there a reason why programming languages use semicolons instead of dots as terminators?

I guess because benefits of 'dot notation' vastly overpass the ones of eol-with-dot.

I always figured its position on the querty keyboard had something to do with it, but would love to know the actual history.

A lot of programming languages are descendants of ALGOL [1], where the ; is the separator, not a terminator (you don't have to use it on the last statement of a block of code).

Pascal (being based on algol) is the same, and the final "end." statement of the program, which is the main program, has a period/full stop. See end of this sample: [2] A begin / end block is one statement so they end like so "end;" unless it's the last one in a procedure, in which case you don't need the semi-colon.

C was similar, except they made ; a terminator instead of separator, but you don't have to put it at the end of being/end blocks. C just uses { } to make it clear. But then, in C++ I always forget to put ; at the end of a class definition. I get why it's there, but it looks more like a function definition to me which doesn't need a semicolon at the end.

Many languages are also from C or at least the Algol family tree (but usually C), like C++, Java, Javascript, etc.

More about ALGOL: https://en.wikipedia.org/wiki/ALGOL

[1] Algol sample: http://sandbox.mc.edu/~bennet/cs404/doc/primes_alg.html

[2] Pascal sample: http://courses.washington.edu/css448/zander/Code/array.p

some languages (eg Erlang) do use dots as terminators. Dots are harder to see and verify.

Imagine: A = something+"some string.".

Erlang and prolog use dots in this way.

The author's description of the various typefaces and their semicolons is just exquisite. What a fucking writer.

anyone knows what "ufq" means in the latin page?

It's "usq", with a long S: https://en.wikipedia.org/wiki/Long_s - so it's an abbrevation for "usque".

It's "usq"; that's a medial S (ſ) rather than an f. It's short for "usque", meaning "up to" or "as far as".

A rough translation of the sentence would be "The experience teaches you as far as it can; the author can't say beyond what they've been given to speak."

“usque” (see <http://www.perseus.tufts.edu/hopper/text?doc=Perseus%3Atext%...), because the second letter is a long s (“ſ”), not an “f”, and “q;” is the abbreviation for “que” as mentioned in the text:

That semicolonish mark at the end of the fourth line from the bottom isn’t a semicolon, it’s an abbreviation for que, Latin for “and.”

Reply on news.ycombinator.com