Hacker News Reader

Comment by notdrunkatall | Hacker News Reader

Buffalo buffalo buffalo buffalo buffalo buffalo buffalo buffalo.

http://en.wikipedia.org/wiki/Buffalo_buffalo_Buffalo_buffalo...

I can't resist mentioning the Lion-Eating Poet in the Stone Den poem I came across recently. The poet plays with the many tones and many variants of the sound 'sh' in Chinese:

  « Shī Shì shí shī shǐ »
  Shíshì shīshì Shī Shì, shì shī, shì shí shí shī.
  Shì shíshí shì shì shì shī.
  Shí shí, shì shí shī shì shì.
  Shì shí, shì Shī Shì shì shì.
  Shì shì shì shí shī, shì shǐ shì, shǐ shì shí shī shìshì.
  Shì shí shì shí shī shī, shì shíshì.
  Shíshì shī, Shì shǐ shì shì shíshì.
  Shíshì shì, Shì shǐ shì shí shì shí shī.
  Shí shí, shǐ shí shì shí shī shī, shí shí shí shī shī.
  Shì shì shì shì.

Translation:

  « Lion-Eating Poet in the Stone Den »
  In a stone den was a poet called Shi, who was a lion addict, and had resolved to eat ten lions.
  He often went to the market to look for lions.
  At ten o'clock, ten lions had just arrived at the market.
  At that time, Shi had just arrived at the market.
  He saw those ten lions, and using his trusty arrows, caused the ten lions to die.
  He brought the corpses of the ten lions to the stone den.
  The stone den was damp. He asked his servants to wipe it.
  After the stone den was wiped, he tried to eat those ten lions.
  When he ate, he realized that these ten lions were in fact ten stone lion corpses.
  Try to explain this matter.

http://en.wikipedia.org/wiki/Lion-Eating_Poet_in_the_Stone_D...

Note that this poem isn't in any "real" Chinese language. Basically, it's written in Classical Chinese, but pronounced as if it were Mandarin.

Old Chinese (the spoken language on which the written Classical form was based) had a much more complex phonology, which was simplified in Mandarin, resulting in many Old Chinese words becoming homophones or near homophones. To reduce the ambiguity, Mandarin uses many more compound words than Classical Chinese.

For example, a poet is called shīrén in Mandarin (literally, "poet person") while this poem just uses shī.

Chinese... you're drunk go home.

(On a more serious note that's pretty cool.)

When written in chinese, is there any more indication that it's not just the same word forty or so times than in the romanised version?

Check the wiki link, it has the characters instead of the pinyin. To answer your question yes the characters are very different.

The singularity is coming: the day google translate can correctly make that translation.

Didn't someone show that every single sequence of words buffalo* (buffalo, buffalo buffalo, buffalo buffalo buffalo, ...) is valid?

Here's how it works:

"buffalo" can function as a verb (bully) and a plural noun (bison).

So we can write "Bison bully bison." as "Buffalo buffalo buffalo."

I'm going to tag my groups of bison with numbers.

> Bison(1) bully bison(0).

> Buffalo(1) buffalo buffalo(0). (3 words)

Now, we can qualify bison(1), by stating that they are bullied by another group of bison, which I'll call bison(2). We can rewrite "who are bullied by bison(2)" as "bison(2) bully" (similar to rewriting "food that is liked by me" as "food I like") so we get:

> Bison(1) who are bullied by bison(2) bully, bison(0).

> Buffalo(1) buffalo(2) buffalo buffalo buffalo(0). (5 words)

We can modify bison(2) in the same way we modified bison(1), by stating that they are bullied by yet another group of bison, called bison(3):

> Bison(1) bison(2) bison(3) bully bully bully bison(0).

> Buffalo(1) buffalo(2) buffalo(3) buffalo buffalo buffalo buffalo(0). (7 words)

And so on...

> Buffalo(1) buffalo(2) buffalo(3) buffalo(4) buffalo buffalo buffalo buffalo buffalo(0). (9 words)

> Buffalo(1) buffalo(2) buffalo(3) buffalo(4) buffalo(5) buffalo buffalo buffalo buffalo buffalo buffalo(0). (11 words)

So that gives us sentences of lengths 3, 5, 7, 9, 11, ... words. We can add a word to each sentence by just adding the qualifier "Buffalo" (meaning "from the city of Buffalo") before buffalo(0). This gives us sentences of lengths 4, 6, 8, ...:

> Bison(1) bully bison(0) from the city of Buffalo.

> Buffalo(1) buffalo Buffalo buffalo(0). (4 words)

> Buffalo(1) buffalo(2) buffalo buffalo Buffalo buffalo(0). (6 words)

> Buffalo(1) buffalo(2) buffalo(3) buffalo buffalo buffalo Buffalo buffalo(0). (8 words)

So now we just need the sentences of lengths 1 and 2, which we can do by using the imperative. I.e. we give the general instruction "Bully." or the more specific instruction "Bully bison.":

> Buffalo. (1 word)

> Buffalo buffalo. (2 words)

The issue remains: can we call a sentence "valid" if it is entirely incomprehensible?

The sentence is nonsensical, because it refers to the same group three times - Bison from Buffalo. If bison from Buffalo bully other bison from Buffalo, it's tautological to again repeat the point that they bully bison from Buffalo.

You'd need more specific demographics to make the sentence valid, methinks, and when that happens, you're adding in a new word.

You're making the classic nerd error of assuming that natural languages should be as precisely defined as mathematics or computer languages. You're going to tell me that there's no such thing as deceleration next!

(To be fair, "Buffalo buffalo buffalo buffalo buffalo." isn't exactly a "natural" sentence.)

And in any case, the sentence is still syntactically valid, which is all I'm really concerned about here.

I'm not operating from logic, but from a 'natural' perspective. I've no problems with the syntax ("French citizens German citizens bully, bully Belgian citizens") though it's quite an awkward sentence in that format and any editor would require a rewrite, but it's just not something that someone would say if all the demographics were the same ("French citizens French citizens bully, bully French citizens") - ie, it's nonsense for that reason.

While the sentence is technically ambiguous, we say ambiguous things all the time and still manage to understand one another. This is the domain of the study of Pragmatics.

In natural communication, we START by assuming that the speaker had a meaningful intention, and then attempt to infer that meaning from context and assumptions about shared knowledge. I can easily think of contexts in which your "French citizens" sentence make sense.

I don't have the source for this, but I'm almost positive you are correct. I believe I remember an inductive proof that demonstrated this.

http://www.ci.buffalo.mn.us/Admin/Citycode/1004.htm

> Subd. 14. C. Notwithstanding the language contained in Subdivision 6 of this section, all “potentially dangerous animals,” as defined in this Ordinance, which are outside of the owner’s residence, must be kept on a suitable leash, or in an enclosure which restricts the animal’s ability to egress from the owner’s property.

Buffalo buffalo are not buffaloed by fences, but it might buffalo a Buffalo buffalo to see another Buffalo buffalo on a leash.

Based on a simple grammar I wrote in Python there are 21 valid interpretations of this sentence: https://github.com/taliesinb/chartreuse

Reply on news.ycombinator.com