Hacker News Reader

Comment by Cushman | Hacker News Reader

I don't know if it's completely true that English has less character variation than Arabic— as you say, in English there is a choice of upper or lower-case, with occasional changes in meaning. In Arabic, the form of a letter is completely determined by the letter it follows. It's purely a display difference, not a separate character set. There are the diacritics to think about, but outside of the Quran they are simply ignored.

That's still a big problem for a universal language of the internet though, since written Arabic is highly non-phonetic.

View on HN

A couple of corrections I feel I have to make as an Arab.

> There are the diacritics to think about, but outside of the Quran they are simply ignored.

They are not ignored - when they are present, attention is paid to them. I think what you mean is that Arabic speakers, knowing what the diacritics are, do not bother writing them down. That is not because they ignore them, it is because we have paid such close attention to them when learning Arabic that we no longer need to be reminded of them.

> That's still a big problem for a universal language of the internet though, since written Arabic is highly non-phonetic.

Arabic is highly phonetic, and if being phonetic was a criterion for being universal language of the internet, English should be disqualified immediately.

On arrival in England at the age of 10, I had no idea how English people knew how to pronounce their words. Now I know that non-Arab speakers may think the same way about Arabic because we do not write down the diacritics by default... but it is easy to buy books that have these diacritics written, and thus to crack the code.

But English seemed designed to trap foreigners into mispronunciations, to the great amusement of my classmates. (Traveling to America after college, it was mostly place names that tripped me up.)

Yeah, I'm sorry, I was highly unclear. What you said is what I meant :) I was saying that the Arabic character set is actually simpler, but vocalized Arabic becomes harder again.

American place names are a constant source of confusion and amusement even among Americans, largely because many of them are adapted from American Indian words. You might be a perfectly normal English-speaking American, but if you've never been to the state of Washington before you won't know how to pronounce "Puyallup" or "Sequim" just by reading them.

Someone more familiar with Arabic may need to correct me, however I was under the impression that the shape of the letters is not at all affected by the letters it follows, but by their position in the word.

Arabic letters look different depending on whether they are in the initial, medial, final position in a word (as well as having a 4th form when they appear in isolation). However, there are patterns of similarity so it's not as difficult as having to learn 4 completely random shapes for each letter.

Not quite. There are several letters that don't connect to the letter that follows them, so a letter that follows them will appear in the 'initial' form even though it's not at a word boundary.

That would be me :) No, the preceding letter matters. Some letters (د،ذ،ر،ز،ؤ) have no medial form-- they take the terminal form instead, and the following letter takes the initial form.

There are patterns (17 by my count?) among which letters differ only minorly, so it's not as bad as learning four forms for each letter, but there are some additional gotchas too. For example "ل-ا" ("laa") is always written as a single character "lamalif": لا. (Bonus knowledge: that's also the word for "no". You can see it at the beginning of the Shahada: ...لا إله إلا الله <- "There is no god but God..." etc.)

There are some differences in the way the way the characters are laid out in English too, unless you use a monospace font.

Those differences are not to my knowledge as a native English speaker used to convey any sort of meaning, and as demonstrated by the existence of monospaced fonts have no particular importance. They are merely an artifact of the glyph geometry.

They don't convey any sort of meaning in Arabic either - letters just change their shapes depending on where they occur in a word. Cursive English handwriting does the same thing, to a lesser degree, and for the same reasons.

Like where? In cursive, at best what changes is that some ending curse goes up or down a bit more. The existence of cursive computer fonts shows that even in cursive writing, the shape of letters can be the same across a text. Minor changes are caused by the speed of writing the cursive.

On the contrary, and as I understand it, in Arabic there are rules on how to change the shape of letters in certain contexts. In Arabic it's part of the writing system, in Western European languages it's not.

I give you Typographic Ligature: http://en.wikipedia.org/wiki/Typographic_ligature

Exactly, ligatures don't have any grammatical meaning. They're merely that - 'typographic'.

And the Arabic letters don't have any grammatical meaning. They are also merely typographic.

I think we are saying the same thing here.

Reply on news.ycombinator.com