Linguistics questions?

[Tahnan]

All right, I feel a bit silly doing it, but Antrax did say the forum was for people with a PhD in the appropriate field to offer answers. And I don't have a PhD, yet, but I'm pretty close to ABD, so, what the heck.

Mostly I feel silly because I can't imagine what questions people might have, but: if you've got any questions that you think a linguist would be best qualified to answer, heck, go ahead, ask 'em.

[A Rhumba of Rattlesnakes]

What issssssssss a linguisssssssssst?

------------------
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

[Vinny]

First: I love linguists! Love them! I have a whole yoodle of questions regarding this field. I wished I have gotten into this field early on, but what the heck. I can't imagine doing anything with a linguist degree (Except for being interpreters and translators, which is really hard and stressful, from what I have heard).


Ok, questions:

Do you think Asianic written languages, by way of pictograms (such as Chinese and Japanese) will eventually go away? Why?

What about the difficulites in indexing catalogues and directories and such because of the lack of alphabets? Can that be solved?

My brother and I had a pretty long discussion a few months back on how languages were evolved. Mainly, we were wondering why American English differs in sound so much from the British. Is this because the first few Americans, from the pilgrimages, were from a certain niche of British population that speaks without the British accent, and thus the main language was evolved from that, without the accent?

How did the mascoline/feminine form came about in the first place? I understand that all of the romance languages (latin, spanish, french, etc), have these, so they were *before* English. Why did they do this? Why didn't English adapted the form? Is it useful in any way at all?

Was English derived from the Gaelic language? And where was the Gaelic language derived from?

That's it for now.

[This message has been edited by Vinny (edited 03-19-2004 04:12 PM).]

[RequiemEternam]

The questions:

What field of linguistics are you studying?

Are you an X-bar kind of guy? What alternatives seem plausible to X-bar theory?

I've looked at a dozen sites online trying to understand verb islands. They are all extremely dense. The only things I can understand are subjacency conditions and binding. Do you know of a good site/document that will help me?

[Tahnan]

No fair, an actual linguistics student asking the hard questions!

I'll tackle Rattlesnake's question first, since it'll be more foundational: a linguist is someone who studies language. That's vague, but linguistics covers a wide variety of topics. There are sociolinguists, who study language's effect on culture; neuro- and psycholinguists, who study the brain as a language processor. There are historical linguists, who study the development of language over time, often with an eye to finding relations between languages and language families.

What I do is "formal linguistics": treating language as a data set and trying to find the rules behind it. This comprises, among other things, phonology (the study of the sounds), syntax (the study of the structure), and semantics (the study of the meaning). That's a vast oversimplification, of course, since the borders are not distinct and there are other topics that fall under formal linguistics (morphology, the study of piecing words together, and pragmatics, the study of the meanings of sentences beyond their literal meanings).

To say nothing of people who study writing systems, people who study body language....

I'll tackle the harder ones in a bit.

[Agamemnon]

What English accent would you class as being the closest to it's England origin?
I've read books that state American is more to the true England English and I've read others that state Australian is more true. One book, which I forget the name of, states that possibly folk from the Cook Islands are nearer original English, but then again it also stated that there is a place in Yankeeland (Possibly somewhere in deepest Carolina or Somewhere south in the first 12 states) that have folk who still speak an accent very close to old England English.
?
I agree that Blighty English has been intergrated and corrupted over the years with the influx of forign intergration, but I also see it too with American English and Australian English.
I know there is too a theory that Celtic is a form of English back in the days when the Romans pushed the old Britons across to Cornwall, Wales and over to Ireland, but since this is more of a condensed dialect I believe it to be a false theory.

Can you shed any light on the subject old chum?

[Tahnan]

So, let's see, the slightly easier ones, which are Vinny's--now that I've made it somewhat clear that the questions are within the realm of linguistics but wholly outside my area of expertise. :-)

I don't see syllabic writing systems like Chinese and Japanese going away any time soon. They've survived for quite a while now; I don't know anything about cataloguing in such alphabets, but surely methods have been found. (Additionally, though this is way outside my field, I feel as if globalization of culture is on the decline, and there is a rising interest in national identity.)

--um, and even more in even more of a bit.

[Bicho the Inhaler]

Tahnan,

Here's a really easy question. What's your opinion of the grammaticality of this sentence:

I persuaded the book.

To me, the answer is pretty clear, but this isn't a joke question, since I've actually heard different claims (from linguists) about this sentence.

[Legion]

What is the root of the Basque Language? How old is it? Where did it originate? Are there versions or derivations of it anywhere other than in the Basque Country?

[Antrax]

How come Russian and Hebrew have a letter in common? (ie, same picture, same sound).
Antrax

------------------
"Look, that's why there's rules, understand? So that you think before you break 'em" - Lu-Tze, Thief of Time

[Agamemnon]

[Ningal]

Ant: Cyrillic was stolen largely from Greek, but Hebrew was used to fill in the gaps.

Edit: Or one of them, anyhows.

[This message has been edited by Ningal (edited 03-24-2004 02:01 AM).]

[Vegetable]

Do you know sandskrit?
If so, explain your theories about the similarities between sandskrit and latin.

[Noam Chomsky]

Are you familiar with Vivian Lin's work related to Coordination and Sharing at the Interfaces? In particular, I am interested in your opinion of her theory of Gapping, in which sharing structures play a central role. Do you agree that Determiner Heads must be licensed by a head higher in the tree?

[Professor Henry Higgins]

Repeat after me:

The rain in Spain falls mainly on the plain.

[Colonel Pickering]

Really, Higgins... be reasonable.

[zeek]

Did we kill Tahnan?
I'm certainly no linguist but what the hell, I'll try to tackle Bicho's question.

We want to know the "grammaticality" of this sentence:

[DrJones]

Japanese use different methods to index catalogues and dictionaries. Some of them are ordered by syllabes, using the IROHA ordering. However, that requires knowing beforehand the pronounciation of each kanji.

There are other ordering methods based on the number of strokes, position of radicals, and such.


I'm interested in learn how many steps are needed to take the ambiguity out from natural languages.

How many languages could be written as they sounded? How many could if there weren't silent letters in them?

EDIT: What's the grammar of the sentence "I am George"?

[This message has been edited by DrJones (edited 03-25-2004 06:37 AM).]

[Tahnan]

Ack! I got pulled away from the computer on the 20th, forgot to check back, and boom, explosion. All right, as much as I can handle in one sitting:

"We were wondering why American English differs in sound so much from the British": essentially, this is just a sort of standard language drift over the course of 200 years. Australian English is also somewhat different from British English (though perhaps closer than American); and, heck, Yorkshire English and London English don't quite sound the same. Consider, too, the changes from Middle English to Modern English; then consider what might have happened if there had been two separate communities of Middle English speakers who let the language drift over a few centuries. It's unlikely that they would have drifted in the same way.

The people who settled in the New World had the same accents as their fellows back home, but things drift over time--especially in America, where there would be so much contact with settlers from other countries. It's been claimed that people in the Appalacian Mountains in the Eastern US speak "the English spoken in Shakespeare's day" (that's the group Agamemnon was thinking of); that's an exaggeration, as well as I understand it, but their English has in fact changed much less than other Americans' has, primarily because of their isolation, which gave them less contact with outside influences.

As to Aga's question of which modern dialect of English is most true to its origins, it's a little outside my field, which as I said isn't really historical linguistics. I think the best answer is that everything's changed in different ways--again consider Middle English, for instance. Once English went through the Great Vowel Shift, you're not going to be left with anything that sounds like Middle English any more. For what it's worth, it's my understanding that American vowels are more diphthongized than British, but I'm not sure--and I haven't any idea what they sounded like in London in 1600. (Speaking of London in 1600: it would be a mistake to think that everyone spoke English identically then, either. Cockney is hardly a recent development. I suspect you're asking about "Received Pronunciation," though.)

"How did the masculine/feminine form came about in the first place" and "Why didn't English adopt the form": that, I genuinely don't know. I do know that masculine/feminine is hardly a universal; many African languages have far more than two or three classes of nouns. Sesotho has 18 (though 11-13 are mostly used for loanwords, and some are pluralizations of other noun classes). They don't quite break down into perfect semantic groups--Class 1 is roughly "people" (though "sister" is Class 9), but other classes of words--body parts, food, animals--range across the classes. Masculine and feminine are the same thing, essentially, except that there are far fewer classes. (And even in languages with grammatical gender, it doesn't always quite correlate to meaning. Setting aside questions of why tables are female and fish are male, you get exceptions like German Mädchen 'maiden', which is neuter.)

OK, new post.

[Vinny]

Thanks, Tahnan! The Great Vowel Shift, NeatO!

Since you're overloaded with questions right now, I won't ask you anymore ... fer nouw ...

[Tahnan]

RequiemEternam wanted to know what I'm studying. I'm studying semantics; my semantics qualifying paper was on the topic of "concealed questions," which are nouns that act as questions. In a nutshell, I set out to explain why I know what time it is can be rephrased as I know the time, and He asked me what time it is can be rephrased as He asked me the time, but I wonder what time it is can't be rephrased as *I wonder the time.

X-Bar, for those who don't know, is a theory of syntax which claims, in essence, that phrases are made up of "heads" and then the things that either modify the head or serve as a continuation of the head. "An intelligent student of linguistics", for example, is a Noun Phrase (NP); "student" is the noun that's the head of the NP, "of linguistics" is its complement--think of it as "student of linguistics" being all one unit, which is called an "N-bar"; the adjective is added to the N-bar to make another N-bar, and then the article "an" closes it off and makes the whole thing an NP.

I think there's definitely some sort of structure going on--there's a clear intuition that "student of linguistics" is a unit in a way that "intelligent student of" is not. But I don't take the bar labels literally (nor does Chomsky, these days). By and large I try to ignore as much of the syntax as possible, or at least to come up with theories of interpretation that depend on the structure but not on what you call it.

One reason for doing that, in fact, is that other theories of syntax--Categorial Grammar, for example--reflect the structure in ways other than literal drawn trees with labels. Staying neutral about the syntax helps my theories be more general, more adaptable to the work of non-X-bar people, and more likely to survive if people stop using X-bar theory.

Alas, I don't know of any website that explains verb island--just a little bit outside of what I spend most of my linguistics time thinking about.

Whew. Deep breath. On to Bicho, who asked about the grammaticality of I persuaded the book. (A quick note: when a linguist talks about "grammar" or a sentence being "grammatical", she isn't referring to a set of rules that are written down in a book and that William Safire and your English teacher will call you incorrect if you fail to follow. The grammar is the set of internalized, instinctive rules of language that native speakers know without necessarily knowing they know, and a grammatical sentence is one which a native speaker accepts as part of the language.)

My judgment is that I persuaded the book is syntactically well-formed, but semantically anomalous. In other words, it doesn't have the structural impossiblity of, say, *the persuade I booked, but there's a problem with the meaning, namely that the object of the verb "persuade" must be capable of making a decision.

The syntax, by the way, isn't quite perfect, since really "persuade" requires two objects: a person persuaded, and an action (e.g. I persuaded him to leave). But if it's obvious from context, the action can be elided.

In other words, if you and I find ourselves transported to a magical realm where written material is alive, and we're asking a book and a pamphlet to help us get home, I might ask, "Did you persuade the pamphlet to help us?" and you might reply, "No, I persuaded the book." In that context, where books can make decisions, the sentence is fine semantically (which suggests there's nothing wrong with its syntax).

(With respect to what zeek said: many linguists do use "grammar" to mean specifically the syntax, which is more or less "combining words to make a valid sentence". I tend to use "grammatical" to refer to both this and semantics, i.e. "rhetoric," which is why I separated the two out, above, rather than giving a yes or no answer.)

Long enough for one post. One more coming.

[Tahnan]

Basque. The current theory among linguists is that Basque was brought to Spain by aliens; much debate rages over whether they were from inside our own solar system ("Martians" is regaining popularity, with the recent discovery of water) or outside ("Alpha Centauri" is still the most common within this camp).

All right, not true. But it might as well be. When linguists categorize languages by their relations to one another, there's Indo-European (Greek, Latin, Sanskrit, German, Slavic languages, all their derivatives), there's Finno-Ugric (Hungarian, Finnish, etc.), there's Semitic (Arabic, Hebrew, etc. etc.). And then, um, there's Basque. Which is apparently unrelated to anything at all. I've heard it claimed that it's related to Ainu, the aboriginal language of Japan, but then, I've also heard about (though never found) a web page that argues that every language in the world is derived from Basque.

There are other languages that are "isolated," i.e. unrelated to anything else we know. (Basque's language family is usually called "Basque," on the strength of dialects of Basque that are not entirely mutually intelligable.) http://www.ethnologue.org/show_family.asp?subid=311 is a list of thirty of them, including Zuni (a native language of the Americas, apparently unrelated to other Native American language families) and Korean.

Beyond that, however, I'm not remotely qualified to speculate on its origins, and I doubt anyone really is.

Antrax's question about Hebrew and Cyrillic: Ningal is essentially correct. As far as I understand (which, again, is not very far, as I'm not an expert on writing systems), Cyrillic was specifically invented (like Korean) rather than evolving over time. As a result, much of the alphabet resembles Greek. Greek, however, has no "sh" sound (Classical Greek did not; I believe Modern Greek does not). Insofar as one of the main early uses of Cyrillic was religious (St. Cyril being sent from Greece on a mission of conversion), it's not surprising that he would borrow the "sh" symbol from another theological language, Hebrew.

Vegetable: I don't know a single word of Sanskrit, alas. It was a British colonial governor whose name escapes me who noticed the similarity between Sanskrit and Latin, and that point was more or less the birth of our understanding of "Indo-European" (and Proto-Indo-European), the ancient language that developed into the European languages (German, Latin, etc.) and the Indic languages (such as Sanskrit).

Noam: unfortunately, Vivian and I only overlapped here for a couple of years, so I didn't get much of a chance to really become familiar with her work (and, of course, it's more syntactic than most of what I do). If you want, however, set up a time to meet with me, and swing by my office (it's down a floor from you), and I'll take a look at it and see what I think.

Professor: the Rhine in Spine sties minely in the pline.

[Hitchhiker]

English has a definite order for multiple adjectives before a noun: something like number, size, age, shape, color, gender (four young black female students, two little old hunchbacked men, five large square pink cakes, etc.) It isn't taught to children in school, but native speakers do it automatically, because ordering the words another way sounds "wrong."

Does this have a name? Do other languages have something similar?

[This message has been edited by Hitchhiker (edited 03-25-2004 04:12 PM).]

[Tahnan]

Ooops, one more--sorry, DrJones!

Taking the ambiguity out of natural language is rather like trying to take the salt out of sea water: it's pervasive, and ultimately it's a lot of what makes sea water sea water. You can take the salt out, and what you're left with is, well, water, not sea water. (Or at least water with a lot of seaweed and fish poop in it.)

All natural languages have ambiguity at almost every level: words that sound the same but mean different things, strings of words that can be parsed into structures in different ways, structures that can be interpreted with different meanings.

(For instance: I went to the bank--did I go to a financial institution or the edge of a river? I like flying planes--do I enjoy being a pilot, or watching planes that are flying? Everyone didn't leave--did each person stay, or is it merely not the case that every person left?)

One can construct a new language that's free of such ambiguities, or at least one can try. My opinion is that such a language would be great for computers but not for people, whose brains are hardwired to learn certain possibilities as languages. (I believe that people raising their children as Esperanto speakers are finding that their children have already begun to change the language--making some verbs irregular, for instance. Benjamin Bergen has a paper on this topic on his web page--it's not long, and if you don't want to read the paper, at least look at the photo of him facing his arch-nemesis.

"How many languages could be written as they sounded? How many could if there weren't silent letters in them?" Oy. Well, all of them can be written as they sound, because after all we've got the International Phonetic Alphabet, which you can use to write the sounds of any language.

But if you meant, "for how many languages is the method of writing a strict representation of the pronunciation"--that is, each sound is represented by only one symbol and each symbol represents only one sound--I can't actually answer it, because there are 6,000 languages in the world and I know so few of them. Many languages, however, had no written form until European missionaries wanted to translate the Bible into their language, so many languages that bear no relation to Latin use the Latin alphabet, and use it more or less in that way. (African languages, for example, and Native American languages.) Not that the sound-letter pairing is the same across all such languages; Sesotho distinguishes different vowel sounds with accent marks, whereas Wampanoag uses "8" for the "oooo" sound. ("Moose" comes from the Wampanoag word "m8s".) The first missionary to the Wampanoag wrote the "oo" as a sort of infinity symbol, two o's conjoined; but that couldn't really be typeset very well, so "8" was substituted for it.

I am George...is a lot harder than it looks. For one thing, "I" isn't an easy word to define, since it changes by context; for another, the copular ("to be") is quite nuanced. The syntax is not terribly different from, say, "I know George"--it's a subject, verb, object. Sort of. Except for the fact that the "object" gets treated like a subject...why can't you ask about something easy like "John seemed to want to be arrested" or "Floyd broke the glass"?

[Tahnan]

...and I'll get to Hitchhiker's question tonight or tomorrow; I want to review Larson before tackling that one, to see if he says anything. (I suspect not.)

(For those who were wondering: the following was the structure, circa 1968, for Floyd broke the glass:

[kevinatilusa]

This is rather tangentially related to linguistics, but...

How do translators handle wordplay/anagrams, etc. when translating a novel from one language to another? Specifically, in the second Harry Potter book, There is a scene where an anagram of Tom Marvolo Riddle to an English phrase plays a critical role. How is such a scene translated into a picographic language like Chinese?

[This message has been edited by kevinatilusa (edited 03-25-2004 08:04 PM).]

[Lucky Wizard]

kevin, this page deals with how one translator chose to deal with wordplay when translating Terry Pratchett's book Soul Music into German.

(Side note: In a few weeks, I'll have a copy of Soul Music!)

I don't know how your specific example would be dealt with... I suppose if I were faced with this task, I'd translate the English phrase, and then anagram the result into a plausible-sounding name that's as similar to the original name as possible. But I'm not a professional translator, so this might not be what translators do. And I have no idea how it would be handled with a pictographic script.

[This message has been edited by Lucky Wizard (edited 03-25-2004 10:10 PM).]

[Hitchhiker]

I read about some difficulties one Japanese-English translator had with a conversation in a Japanese children's book. One child-character used a very formal word for "you" (like the Spanish "Usted" or German "Sie") and the other child thought he was being pretentious. So, if translated literally, you got something like:

"Are you coming over to play?" he asked her.
She thought: He said "you." What a jerk!

Which makes absolutely no sense -- so the translator got permission to do a looser interpretation:

"I presume you're coming over to play?" he asked.
She thought: He "presumes?" What a jerk!

[Bicho the Inhaler]

Originally posted by Tahnan:

---

For those who were wondering: the following was the structure, circa 1968, for Floyd broke the glass:

---

*Gasp* It's generative semantics! Run away!

Thanks for answering my question, Tahnan. I've heard people using selectional restrictions to exclude things like "I persuaded the book" from a syntactic position, but my opinion is that whatever the failings of "I persuaded the book" are, they are purely semantic. When I hear "I persuaded the book," I don't think "that sentence is ill-formed" or "that sentence is wrong"; I think "that's a funny thing to say."

I think an important point is that there is a cogent idea behind "I persuaded the book"; it's just an unusual one. The same goes for "colorless green ideas sleep furiously," which is even more problematic--semantically, of course. It's only the content of the idea that is anomalous, not the form in which it's expressed. In other words, the questionable nature of "I persuaded the book" is not linguistic. In fact, to me, the questionable nature of such utterances is hardly worthy of serious attention, but maybe that's just me.

If you care, the supposedly ungrammatical "I persuaded the book" came up in a lecture for a linguistics course.

Also, I don't mean to imply that selectional restrictions are bad (I don't know enough about them to be able to make such a claim), just that this would be a misapplication.

[RequiemEternam]

I totally buy into the idea that verbs select for phrasal categories as arguments, and that thematic roles (as a syntactic concept) are discharged, thereby keeping some of the semantics within the sphere of syntax. That "persuade" specifies an animate Patient seems perfectly logical. All optional adjuntcts are infinitival clauses which would have the Patient as an Actor within its minimal domain, were the inifinitical verb able to discharge a thematic role.

But that's just me.

[Tahnan]

The best argument I've heard against encoding semantic selectional restrictions dates back to 1968, in an article by, I believe, McCawley. One could claim that "persuade" can only take an object that's marked as [+animate]. However, in that case, "deveined" would have to take an argument that's [+shrimp] and "diagonalize" would have to take an argument that's [+matrix]. And who wants to have to mark every single lexical item for its shrimpness and matrixocity--or, really, more to the point, these aren't the sorts of semantic features we think of as basic.

A better way of encoding such things, perhaps, would be to put them into a domain restriction. For instance, "devein" can be thought of as a function that maps a shrimp S into the property of being someone who deveins S (and this combines with an individual to give the value "true" if that individual deveins the shrimp S). Then, given I deveined the VCR, the words fit together syntactically but the function fails to apply: "devein" is simply undefined when applied to a non-shrimp.

At that point you've got a presupposition failure (i.e. "for this sentence to work, the VCR would have to be a shrimp, and I know it's not"), and you can respond by saying "But the VCR isn't a shrimp!" or by thinking of metaphorical extensions...but now we're into pragmatics.

I tend to feel a little dubious, myself, of theta roles--why put the semantics into the syntax at all? This is one reason I stay away from the syntax/semantics interface.

(If this didn't make sense to you because you haven't taken a linguistics course--you being the reader, not you being Bicho and RE, who I know have taken linguistics courses--feel free to ask for clarifications of terms or arguments. That's what this thread is here for, after all.)

To answer kevinatilusa: ironically, although I clearly like language and I clearly like wordplay, I know very little about wordplay in other languages. As Hitchhiker suggests, translation is often quite complicated. (Translate the following into French: "Bonjour!" she called out.) Often the best one can do is capture the spirit of the text, as in the usted/presume example, and I suspect that's what non-alphabetic translations of books with anagrams do. But I don't know enough about wordplay in Chinese or Japanese to know what technique gets used.

(I'm now trying, and completely failing, to imagine translating The Westing Game into another language. If you've read it, you know why, and if you haven't, go read it, you'll like it. I suspect it would be even less successful than that apparently dismal attempt to translate Pratchett's humor.)

And I still owe Hitchhiker a discussion of adjectives. Man, keeping track of this is harder than Botticelli.

[DP]

[hijack]One of the best translations I have seen was a spoof Spanish dubbed version of Fawlty Towers, which was included in one of the British KYTV shows (itself a spoof on SkyTV). The sketch taken is the one where O'Reilly's men have made a mistake and eliminated the door to the dining roomn. Basil obviously speaks Spanish in the dubbed version, but Manuel speaks English. "Onde este la puerto?" "It's right here mate!" [/hijack]

disclaimer: DP no hablo Espanol, no canno mako the reversed ? or the ~ on the n, and the quote in Spanish above is most likely wrong.

[Tahnan]

Not unlike the British comedy 'Allo 'Allo!, in which English was represented as English, good French as English with a French accent, bad French as English with a bad French accent, and so forth.

[sk]

is there a list of all the sounds that are used in languages? in english, letters have many different pronounciations. are there symbols or somethign that identify the sounds- or how else do you denote which sound is which?
(this is in regard to referee's GL tongue thread, where we're going to make a language) i figure this might help.

[Dragon Phoenix]

What is the distinction between a dialect and a language?

[Lucky Wizard]

SK, regarding the last part of your question, about denoting which sound is which:

In addition to the pronunciation keys found in dictionaries, there are three phonetic alphabets for distinguishing between sounds:

SAMPA
X-SAMPA
IPA

[Tahnan]

I know I still owe Hitchhiker an answer, but I'll take care of some of the stuff I can answer off the cuff while I'm here:

Lucky Wizard's links are more or less the right answer to the symbols-for-sounds question. I've never heard of SAMPA, and it looks like it's mostly used for computational work. The standard among linguists is the International Phonetic Alphabet--the Wikipedia article also has their chart, I see. (Ladusaw and Pullum have a book that goes through the symbols, one by one, and explains what sound each represents and what the symbol is called.)

That's pretty much all of them, really. You may also find Peter Ladefoged's site helpful--it's a collection of audio and video files that accompany a book, but it's explanatory without the book.

DP: I can only cite Max Weinreich, who wrote, "A language is a dialect with an army and a navy."

That's not quite true--neither his formulation, nor the fact that that's all I can do. There are 6,000 languages in the world (give or take--depending, of course, on how you count), but there are hardly 6,000 countries.

But it does go a long way: hence "Serbian" and "Croatian," two "languages" that used to be called "Serbo-Croatian" before certain political occurrences, and which continue to be the same language. Similarly, Danish and Norwegian are essentially the same language, but have separate names because they are the national languages of separate countries. On the flip side, there's no language one could really, properly call "Chinese"--there are a large number of languages in China, of which Mandarin Chinese is one, but somehow they all fall under the term "Chinese."

The real test is "mutual intelligibility": if two people understand each other, they speak the same language, and any differences in their speech are differences in dialect.

You may have already realized that this isn't especially helpful. A monolingual Parisian and a monolingual Roman probably understand each other better than I would understand a Glaswegian. (In America, if not in London, one scene of "Trainspotting" was subtitled because the speech was so incomprehensible to Americans. Out of fairness, it was the scene in the bar, where there was a whole lot of background noise; but even so.)

Indeed, at the borders of many European countries, the people speak something that's understood, say, a hundred miles in one direction and a hundred miles in the other, though those two may not be mutually intelligible. At that point, one tends to say that if the people are in Spain, they're speaking a dialect of Spanish...at which point you're right back to Weinreich's formulation.

So, ultimately: the answer is that the distinction between a dialect and a language isn't perfectly clear, and is in some cases a little arbitrary, but generally speaking it has to do with whether two speakers understand one another.

[Noam Chomsky]

With a name like yours, do you ever see yourself as a clue and answer in the Sunday New York Times crossword puzzle? A Penny Press puzzle?

(For the edification of the other readers here, my first name, clued by my last, has been an answer word in crossword puzzles for more than twenty years. Qaqaq can't seem to create a puzzle without using me.)

[Dragon Phoenix]

Thanks. That was a very clear answer.

[Tahnan]

If you hadn't understood it, DP, it'd speak badly for the "mutually intelligible" definition. :-)

Noam: on the one hand, both LANCE and NATHAN have common letters in useful patterns, such that one might expect them to be needed in a crossword puzzle. However, I've rarely seen them as grid answers, so it doesn't seem too likely--though I believe I recall seeing my uncle's name, Robert, used to clue NATHAN (as an author, though in retrospect I wonder if the clue referred to Robert Nathan, the poet; I can't recall if it mentioned one of his books in particular).

Insofar as you're much better-known than me, though, Noam (and insofar as there are far fewer ways to clue Noam, as opposed to my first and last names, which are shared by a number of other people or objects), it seems pretty unlikely.