Wikipedia:Featured article candidates/Chinese classifier/archive1


 * The following is an archived discussion of a featured article nomination. Please do not modify it. Subsequent comments should be made on the article's talk page or in Wikipedia talk:Featured article candidates. No further edits should be made to this page.

The article was not promoted by SandyGeorgia 02:05, 25 July 2009.

Chinese classifier

 * Nominator(s): r ʨ anaɢ talk/contribs 04:08, 30 June 2009 (UTC)

I am nominating this for featured article because it is thoroughly researched and has been through several rounds of copyediting (GA, PR, and repeated copyedits from me over about a month and a half). I feel it's a comprehensive treatment of this subject that should be both an informative introduction for people unfamiliar with it, and a well-researched analysis for people who already have a background in linguistics and/or Chinese language. r ʨ anaɢ talk/contribs 04:08, 30 June 2009 (UTC)


 * Kwami's comments
 * I'd link classifier (linguistics) in the introduction.
 * Linked; also linked measure word (linguistics). r ʨ anaɢ talk/contribs 21:38, 1 July 2009 (UTC)


 * "Different nouns often require different classifiers": Reword. Maybe s.t. like "nouns require specific classifiers (classifiers are lexicalized)" would make it clear that it isn't just random or speaker choice. I think much of this is lexicalized: other than 个 being generic, and apart from poetry, AFAIK each noun has its "correct" classifier, which means that it is not being assigned semantically, even if semantics or analogy is the historical explanation for the assignment. ("different dialects often using different classifiers to count the same item" makes it clear that speakers aren't deciding which classifier to use while speaking, but that the assignments are already set.)
 * While I personally believe they are lexicalized&mdash;at least for the most common words&mdash;just like you do (at the very least, they are certainly taught that way, as classifier-noun pairs) I'm not sure we could call it a fact yet. While most of the papers I'm familiar with seem to like the prototype theory and argue that classifier-noun pairings are often lexicalized/arbitrary, I don't think it's 100% accepted yet, and even some of these papers think there are still some semantic factors floating around (for example, the Zhang 2007 paper); it also gets messy, of course, trying to draw a distinction between classifier use synchronically (probably lexicalized) vs. where the classifiers came from in recent history (probably arose through analogy to prototypes, a bit more semantic). Mainly I was just trying to keep the lede simple (ie, what a person just starting to learn Chinese would know-- that books go with ben, pants go with tiao, etc.). I will try to think of other possible rewordings to make this clear. r ʨ anaɢ talk/contribs 21:38, 1 July 2009 (UTC)'
 * After looking at this more, my impression is that the word "require" does make it clear that this isn't random or speaker choice. I can't really think of any better way to put it that still flows well; maybe something like "each noun can only be used with certain classifiers (based on inherent properties of the noun or on learned/lexicalized associations), and cannot be used with other classifiers." I'm hoping to avoid using intimidating jargon like "lexicalized" and "association", though, in the lede, since it might make the article less accessible to lay readers, so that parenthetical might not be so good to include. r ʨ anaɢ talk/contribs 05:03, 3 July 2009 (UTC)


 * "they are just motivated by analogy": there's no "just" about it. This is an ongoing and productive avenue of research. Might want to link prototype theory.
 * Removed "just" and linked. r ʨ anaɢ talk/contribs 21:38, 1 July 2009 (UTC)


 * "the mass-classifier 盒 (hé, "box") may be used with anything that fits into boxes": Reword. You mean 'anything that's in a box', or maybe the quantity of a box, don't you? You wouldn't use "1盒cigarette" for just one cigarette, which is how I could read this. Would "1盒cigarette" mean only 'a pack of cigarettes', or could it be used to mean 20 cigarettes that aren't actually in a pack?
 * Ah, yes, you're right. Changed to "may be used with anything in a box". As for the other question, I'm pretty sure it would only mean a box of cigarettes, not "a box's worth of cigarettes" (for that I would use "1盒的cigarette", lit. "1 box COMP cigarette" or "1 box's cigarettes"); I don't think it has that second meaning, although I'd have to consult a native speaker to be sure. r ʨ anaɢ talk/contribs 21:38, 1 July 2009 (UTC) update: I checked with a friend from Shanghai and she agreed with this. r ʨ anaɢ</b> talk/contribs 03:58, 3 July 2009 (UTC)


 * "where the Chinese system originally came from": "originally" is redundant.
 * Good catch, removed. <b class="Unicode">r ʨ anaɢ</b> talk/contribs 21:38, 1 July 2009 (UTC)


 * "and slowly had their meanings bleached away": Reword as active voice.
 * Do you mean like "and their meanings were slowly bleached away"? That is still passive but less passive (I guess "had their meanings bleached away" is doubly passive, not sure what the official description of it is); I don't see any way to reword this as truly passive, since there's no agent. Certainly couldn't say "people bleached their meanings away", and I don't know if 'bleached' can be used intransitively ("their meanings bleached away")... I've changed it to "their meanings were gradually bleached away" for now, does that work? <b class="Unicode">r ʨ anaɢ</b> talk/contribs 21:38, 1 July 2009 (UTC)


 * usage: you might want to use 'cattle' in your English examples, in place of perhaps in addition to 'cat', because that requires a classifier in English ('cattle' is not directly countable in English, and has no singular ("neat" being obsolete), much like Chinese nouns): a head of cattle, this head of cattle, three head of cattle, these three head of cattle, three head of black cattle, three head. (Yes, 'this head of cattle' is not common in English, but it does occur: I can sell you this head of cattle; At five years of age this head of cattle is worth perhaps $40.) Also: do you say 'many head of cattle' in Chinese, or just 'many cattle'?
 * Ah, wow, I could have sworn I mentioned cattle in there somewhere (AFAIK "head of cattle" the most famous, if not the only, example of a count-classifier in mod. English), but I can't find it. I will try to add mention of it tonight, at the very least in a note (something along the lines of "English doesn't really have classifiers, but there is this one"). As for 'many'.... just "many cattle" is probably the best. "Many头cattle" sounds like a students' error to me, although again I'd have to check with a native speaker to see if it's entirely impossible or not. <b class="Unicode">r ʨ anaɢ</b> talk/contribs 21:38, 1 July 2009 (UTC) Update I asked a couple friends and there was disagreement, but the general consensus I've taken away from it is that "many-CL-N" may not be technically incorrect but it's never, or almost never, said. (It seems complicated...sometimes I miss the comforting arms of prescriptive grammar and ignorance...) <b class="Unicode">r ʨ anaɢ</b> talk/contribs 03:58, 3 July 2009 (UTC)
 * I added a brief note on "head of cattle". A lot of the sources I've read mention it, but I can't find it right now (and most of them are scans of things, rather than nice PDFs, so they're not searchable), but I think the one source there is sufficient. Personally, I don't agree with the view that "head of cattle" is an example of a count-classifier in English ('cattle' seems more like a mass noun for me, and 'head of cattle' just like 'grain of salt'...note that we don't say "five head of cows"), but the example is cited enough that it's worth mentioning. As for using it in the "usage" explanation... I actually specifically avoided it, partly because it's unusual English (either it's a count-classifier, which makes it different from the norm, or it's a mass-classifier, where I was trying more to give an example of a count-classifier), and partly because I wanted to use an example with a real countable noun to highlight the major difference between English and Chinese--that Chinese uses classifiers for counting discrete objects, but English leaves that space empty. I'm worried that if I used "head of cattle" as the example, then readers might not realize that with other English nouns, a classifier-less phrase ("five cats") would have a classifier added when translated into Chinese. <b class="Unicode">r ʨ anaɢ</b> talk/contribs 04:23, 3 July 2009 (UTC)


 * other uses: you might want to clarify that 片 is not the classifier for clouds.
 * Added a brief clarification, let me know if you think it's enough. <b class="Unicode">r ʨ anaɢ</b> talk/contribs 21:38, 1 July 2009 (UTC)


 * I haven't read this far, but if you don't explain where 个 came from, that would be of interest to me as a reader.
 * The article doesn't currently mention it, but there is a ton of literature on it and it would be easy to add (off the top of my head, out of the current references, Erbaugh, Ahrens, and Wang all have sections on it; Erbaugh even has sort of a timeline of all the different things 个 has meant over the years). Appropriate places to add a little bit on it would be the Neutralization section (if I put it there, it would probably have to go in a footnote, unless I do a bit of reorganizing) or the Origins/History section. <b class="Unicode">r ʨ anaɢ</b> talk/contribs 21:38, 1 July 2009 (UTC)

Okay, enough for now. I'll try to take up the rest of the article later. kwami (talk) 20:35, 1 July 2009 (UTC)
 * Thanks for the comments; I'll leave responses above (I know the FAC guidelines say not to, but I find it easier to keep track of which responses go to which comments that way; if you don't want your points to be split up, let me know and I'll move my responses down below). <b class="Unicode">r ʨ anaɢ</b> talk/contribs 21:38, 1 July 2009 (UTC)


 * Well, with "Different nouns often require different classifiers", the problem I see is that the meaning, that for each noun there is a specific classifier, is rather opaque. Yes, it is logically required by the following text, but I think it would be better to simply say it to begin with.


 * "their meanings slowly bleached away" sounds fine to me.


 * Good point on 'cattle'. You could maybe have the 'cat' example to show the difference, then the 'cattle' example in parentheses to make the construction intuitive. I think it's important to tie material in with things the reader already knows, as far as reasonably possible.


 * 'Head of cattle' has nothing to do with mass nouns. You can say either 'five cattle' or 'five head of cattle', 'five sheep' or 'five head of sheep'. The 'head' shows that you're treating the animals as livestock rather than generic animals. AFAIK, 'five head of deer' or 'five head of quail' would only be used if they were being hunted, farmed, considered for wildlife management, etc. But then you get "1000 head of cattle and to 20 head of horses", "30 head of horses and mares", with a clear plural. kwami (talk) 06:56, 3 July 2009 (UTC)


 * About "different nouns often require different classifiers"...not sure if I'm understanding you, are you basically saying that it might be better for the introduction to have a bit of a "white lie" for simplicity's sake (ie, suggesting that every noun has one and only one classifier that it absolutely must use) and then further along in the article clarifying the issues of variability, usage, etc?
 * As for the cattle example...I suppose I could put a cattle sentence directly below each cat sentence, to give two examples of each structure. That would fit the most nicely into a table, and I could add a note to the table along the lines of "notice how in English, 'three cats' doesn't need a classifier but 'three head of cattle' does, whereas in Chinese they both need classifiers". <b class="Unicode">r ʨ anaɢ</b> talk/contribs 07:09, 3 July 2009 (UTC)
 * Ok, I made a mock-up example here. The top example is both cat & cow sentences in one table, as I had suggested above; as it turns out, once I put it together I think it's hard to read. Below it is the examples in separate tables, which I like a little better. (Another problem, though, is that I think the cow examples sound very awkward in English...can you think of any times where you can use an adjective in "X head of cattle"? My adjective examples don't translate well, but I figured since there were adjective examples for cats there should be adjective examples for cows as well.) <b class="Unicode">r ʨ anaɢ</b> talk/contribs 15:11, 3 July 2009 (UTC)


 * "head" is uncountable: 5 head of cattle, not 5 *heads of cattle.


 * No, I don't think we should lie in the intro. I just don't think the word "different" in that sentence is very clear. "Nouns are associated with particular classifiers", maybe?


 * Color: As is, the color doesn't add much to the chart. First, I think the Chinese words should be color coded to match the schema. Second, IMO the CL should be the brightest color. NUM and DEM could share the second-brightest color. It's the adjective IMO that should be left black.


 * Arrangement: I think the simpler examples should come first: num-cl N and dem-cl N. Then the one w/o the noun. Only then introduce the adj. Also, we might want to add san zhi hei-de for "three black ones", "five head of big ones". kwami (talk) 19:25, 3 July 2009 (UTC)


 * Your suggestions for arrangement of the table are good, and I've reorganized it and also redone the colors: here is what it looks like now. I tried to use a bright color, red, for the CL, and dull colors for all the others (I ended up putting color on the ADJ because when left black it stood out, whereas we probably just want CL to be standing out...so I made it brown--I tried orange, but it was too light to read easily). I just want to make sure it doesn't cause accessibility problems for people with screen settings different than mine...if there is a chance that it would cause accessibility problems, maybe it would be better just to leave the whole thing black?
 * As for the lede, "Nouns are associated with particular classifiers" sounds pretty good...I just also want to make sure it's clear that noun X is also associated with different classifiers than noun Y is, since I think that is one of the main ideas. I'll have to try to think of a good way to express both of these things at once without getting overly complicated in the lede. <b class="Unicode">r ʨ anaɢ</b> talk/contribs 20:18, 3 July 2009 (UTC)
 * Ok, I've added the new tables to the article, I think they're ready. As for the sentence in the lede...I realized that what I was worried about isn't a big deal, since the following sentence clarifies the point I'm trying to make. So here are the possible rewordings I have in mind; they're pretty similar to what you suggested:
 * "Each noun is associated with particular classifiers." (even shorter than what's there now, but the rest of the paragraph explains it)
 * "Each classifier is associated with a particular kind of noun." (different emphasis... probably more accurate than the other sentence, but might also be more confusing for readers)
 * Any preference? <b class="Unicode">r ʨ anaɢ</b> talk/contribs 20:39, 4 July 2009 (UTC)


 * Yes, the first. The second is linguistically dubious: nouns don't likely come in "kinds", but in semantic networks. But nouns are associated w particular CL, AFAIK either lexically or though association with semantically similar nouns. kwami (talk) 21:32, 6 July 2009 (UTC)


 * Also, regarding your question about 个...after starting to read Wang a little more closely, there is easily enough on this topic for a good paragraph about the origins/history of 个. The question is just where to put it--in the Neutralization section, or the "classifiers themselves" subsection of the History section. I am leaning towards the second, partly because it would flow better (it will probably be easiest to talk about the historical development of a particular classifier after we've already discussed the historical development of classifiers in general), and partly because that section is already pretty small and could use some beefing up. <b class="Unicode">r ʨ anaɢ</b> talk/contribs 23:14, 4 July 2009 (UTC) Update Added it, in a new subsection. <b class="Unicode">r ʨ anaɢ</b> talk/contribs 05:21, 5 July 2009 (UTC)

Types :

While you say 书本 has a "plural" sense, your examples both involve "all". Is it perhaps exhaustive rather than simply plural? Chinese obviously doesn't normally use number: what makes these constructions different?

Since some of the most salient mass classifiers in English are "loaf/slice/piece of bread/cheese", etc., it might be instructive to give a Chinese equivalent.

I'm a little concerned that so much attention is given to mass-CL, which are not of much interest to an English speaker, compared to the amount of time on count-CL. Also, the section on verbal CL could be and maybe should be expanded.

What kind of "event" does 場 count? I assume that it's extremely general, like 个 for nouns. Are there also more narrowly applicable verbal classifiers, or is Chinese rather semantically impoverished in this area? (E.g., are there different ways of counting human activities that might actually take place in an arena, as opposed to weather or calendrical events? Would 場 be used for "2 solar eclipses"? Is it the only CL that could be so used?) If spoken Chinese uses two dozen noun classifiers, how many verbal classifiers does it use? And come to think of it, how many of those two dozen are count-CL? An English speaker wouldn't think twice of learning the Chinese for a "cup" of tea or "slice" of pie, but would consider 只, 头, etc. to be a challenge, and it would be considerate to be explicit about the extent of the challenge.

I haven't read the rest of the article, but do you cover how much semantic play is involved? For example, in medieval Japan, the 'wing' classifier for birds was also used for rabbits, though I don't know how seriously. (People will laugh at this today.) The supposed motivation was that their ears resembled wings, but some suspect that it may have been a way to justify violating Buddhist proscriptions against eating meat. kwami (talk) 07:30, 23 July 2009 (UTC)


 * About 书本...I think you're right. I don't know much about the formal differences between plural and exhaustive, but AFAIK these forms are closer to exhaustive&mdash;it doesn't necessarily describe every item in the set, but describes them all in a general fashion (for example, 书本很多 would literally mean "the books are very many", ie 'there are a lot of books'&mdash;on the shelf, or whatever...slightly awkward example, I think 车辆 'the cars' is more common). I could modify that sentence to say "to convey a plural or exhaustive sense); I just can't find anything on WP to link "exhaustive" to.
 * "Exhaustive" isn't a linguistic term. I would just describe the sense rather than trying to find a (probably unhelpful) label for it.


 * About loaves of bread/slices of cheese...I'll do some brainstorming and then add one. Interestingly, the best examples (the ones you named) are precisely the foods most people in China don't eat ;). Perhaps the best example is "pizza": 一张比萨 one-CL-pizza has a count-classifier and refers to a whole pizza, whereas 一块比萨 one piece pizza has a mass-classifier and refers to a single piece of pizza. It doesn't sound too awkward in Chinese, and it's still a phrase that all readers can recognized.
 * Yes, that occurred to be when I wrote it! The pizza example would be good. "Bread" is an odd word this way (where you need a measure word for the basic unit), and I can't think of another example like it. I know: for pizza you could give the literal translation "one pie of pizza" (or "one pizza pie") as well as the idiomatic "one pizza".


 * As for attention to mass-CL...actually, my concern has always been the opposite. Beyond the "types" section, the article focuses almost entirely on count-classifiers to the exclusion of mass-classifiers; there simply has not been much research on mass-classifiers, since count-classifiers are more interesting (especially back when people still believed studying count-classifiers was going to give us insight into how categorization works in the human brain...although that has not really been the case). For example, the "relation to nouns" section, where prototype theory and neutralization and usage variation is discussed, is pretty much all about count-classifiers (since mass-classifiers can be used with pretty much anything, there's nothing interesting to say about 'which nouns they pattern with'); same goes for the "purpose" section, since mass-classifiers are nearly universal cross-linguistically, whereas count-classifiers are somewhat special.
 * Okay. I just haven't read enough of the article.


 * As for verbal classifiers....it's pretty much the same issue, there's not much written on them. Li & Thompson don't even acknowledge them as their own category (they essentially give them two sentences, and describe them as "another type of measure word is one that denotes an instance of occurrence of an event"), and most papers on classifiers tend to have a footnote somewhere near the beginning basically saying "there are some verbal classifiers out there, but I'm only gonna talk about nominal classifiers here"...for example, Zhang 2007 has "Chinese classifiers are not limited to nominal ones (mingliang ci) but also include verb classifiers (dongliangci), measurement units (danwei liangci), and so on. In this study, I am only concerned with nominal classifiers." Personally, I'm not even really convinced yet that verbal classifiers deserve their own category, because they seem similar to nominal classifiers to me; for example, one of Li & Thompson's six examples is 那场火没人死 (that-CL-fire not.have people die "Nobody died in the fire")&mdash;seems to me like it's just another nominal classifier, where the classified noun happens to be an event. The only verbal classifiers that I'm really confident about are all the ones that roughly mean "times" (次 ci, 遍 bian, 回 hui, etc.), which is also the only example I included in the section; those are the only ones that seem clearly "verbal" to me.
 * Yes, I was a bit dubious about the distinction myself. I wonder if we could make this more overt?


 * As for your question on 場, it is basically just a classifier for events in general (perhaps I should say that instead of "general classifier for events", to avoid sounding like I'm comparing it to the super-special 个/個). For example, Li & Thompson's examples are 那场球很紧 (that CL ballgame was very tense), 张那场火没人死 (no one died in that CL fire), and 昨天有一场电影 (yesterday there was a CL movie); I could include these examples in the article if you think it would help.
 * As you said, it seems 球, 火, and 电影 are just nouns, that 'event' nouns take a separate classifier than long thin nouns or small animal nouns, not that they're verbal. I can't see creating a special subsection for them. The "times" counters, okay. That would seem to be a distinct category.


 * Questions about other verbal classifiers.... well, I believe there are pretty specific and pretty general verbal classifiers, like there are for nouns. For example, the ones I listed above (次、遍、回) all roughly mean "times" and can be used pretty generally ("I did X however-many times"), but there is also 躺 specifically for trips/journeys, so you can say "I went to Beijing one 次" or "I went to Beijing one 趟"; the meanings sound a tiny bit different to me but not in any way that's translatable. (Also, with ones like 躺 there is the same problem I mentioned above, that I'm not totally convinced it's a verbal classifier, rather than just being the nominal classifier for trips/journeys.)
 * Worth mentioning.


 * As for the number of verbal classifiers, I have not yet found any data on this (probably because not until fairly recently did anyone bother trying to consider them as anything separate from nominal classifiers, and even then the boundary between 'verbal' and 'nominal' classifier is very unclear). I also have been unable to get my hands on any of the Chinese-language "classifier dictionaries" that are mentioned in the article. My impression is that the system of verbal classifiers, while not necessarily "impoverished", is certainly much smaller than that for nominal classifiers; while they're similar in that both systems have a small subset of classifiers that do 99% of the work in real-life, for nominal classifiers there are hundreds of extra/rare classifiers beyond that core set, whereas for verbal classifiers there seem to be very few. Likewise, while the so-called "core" set of nominal classifiers consists of over 20 common classifiers, anything that could be called a "core" set of verbal classifiers would probably be more like 5.
 * Worth mentioning too.


 * About semantic play.... the "Purpose" and "Variation in usage" sections both briefly mention ways in which classifiers can be used for stylistic purposes, etc. I don't know of any specific examples as interesting as the birds/rabbits one you mention in Japanese, but there is definitely a good deal of similar stylistic use in Chinese (especially if you listen to enough pop music...). <b class="Unicode">r ʨ anaɢ</b> talk/contribs 18:43, 23 July 2009 (UTC)
 * Haven't gotten to those sections yet. That's what I find interesting. kwami (talk) 23:46, 23 July 2009 (UTC)

Comment I started the article because I thought it would be of interest - I'm somewhat of an armchair linguist. Before I finished the introduction, though, I got bogged down. I'm actually familiar with classifiers because I learned to speak a tiny bit of Thai a while ago, but I didn't even recognize that this was basically the same thing I was already familiar with until quite a ways through the introduction. I would try to make the intro even less technical because if I didn't really get the picture--as a well-educated person who actually has used classifiers in the past (though never learning their name, or maybe just forgetting what they were called)--I don't think a lot of other people are going to get it either. Maybe it would help to frontload the intro with stuff from the second paragraph so that readers understand that one word is used for things that are similar to each other etc and get a bit of a practical understanding of what a classifier is before (or immediately after) confusing most readers with reference to bound morphemes. Calliopejen1 (talk) 21:06, 1 July 2009 (UTC)
 * Calliopejen
 * I'll take a look at it and see if I might be able to reorganize things. You're right, there's a difficult balance to strike&mdash;for the sake of comprehensiveness and stuff I feel obligated to give a basic linguistic description of what they are (how they're bound morphemes, yada yada), but on the other hand I also want to start off with a simple description for lay readers and gradually work up to the more complicated stuff. In my experience the first sentence of a lede often has the more formal/technical definition of a topic and then the rest of the lede goes on to explain it in simpler English, which is what I tried to do here, but I'll try and see what I can do to simplify things more. <b class="Unicode">r ʨ anaɢ</b> talk/contribs 21:45, 1 July 2009 (UTC)


 * Comments - sources look okay, links checked out with the link checker tool. Ealdgyth - Talk 14:02, 4 July 2009 (UTC)


 * Comment. Done; thanks. Images need alt text as per WP:ALT. Eubulides (talk) 08:16, 13 July 2009 (UTC)
 * Added. Thanks for catching that. <b class="Unicode">r ʨ anaɢ</b> talk/contribs 10:52, 13 July 2009 (UTC)

Support I reviewed this article for GA; it was a good article then, and has improved since. I believe it meets the FA criteria, and is an impressive and useful explication of a complicated topic (I now use it as my go-to reference on the subject in my studies of Mandarin, as I have not found a better, more concise discussion of the topic). Ricardiana (talk) 23:11, 13 July 2009 (UTC)

Comments
 * Explain convention of underlining classifiers. Maybe use a coloured line so as to distinguish the "underline" from an actual "character line".
 * link Mary Erbaugh, Lianqing Wang (maybe redlink)
 * "There may be specific patterns behind which classifier-noun pairs may be "neutralized" to use the general classifier, and which may not." Grammar problem.
 * 一层楼 -> ceng is not underlined; same for 条鱼 tiáo yú
 * "*"three muds" is ungrammatical" why is there a star?
 * "Researchers" is used 4 times in the article and is a bit vague. Replace each occurrence with a more specific word: grammarian, linguist, linguistic historian, etc.

92.134.30.204 (talk) 12:14, 23 July 2009 (UTC)


 * replies from <b class="Unicode">r ʨ anaɢ</b> talk/contribs 
 * I think underlining is most conventional in the academic literature; as far as I know, there's no way to make an underline that is a different color than the character (although if anyone knows how to do so, I'm open to suggestions). A brief conversation was held at Talk:Chinese classifier about the underlying, and we weren't able to settle on any good alternatives. (If underlining is the only way to go, I guess I could always add a note, like you suggested, explaining that the classifiers are underlined). The best alternative I've thought of so far is putting classifiers themselves in a different color (although we'd have to worry about accessibility for colorblind readers or different screens).
 * Added a link for Erbaugh in the text; I don't think one is necessary for Wang, because I don't think she meets WP:PROF and I doubt there will be an article about her anywhere in the near future (and the thing that's cited in this article is only her PhD dissertation); the only reason her name is used in the article is for avoiding weasel words, more or less. Erbaugh, though, is relatively prolific, and could have an article written about her.
 * I don't see anything grammatically wrong with "There may be specific patterns behind which classifier-noun pairs may be "neutralized" to use the general classifier, and which may not"&mdash;although one can always argue that there are stylistic problems. How about a rewording to Classifier neutralization may follow specific patterns&mdash;certain kinds of classifier-noun pairs may neutralize where others do not.
 * Added underlines, thanks.
 * Asterisks * are conventional in all linguistic literature to indicate that an example sentence is ungrammatical.
 * Replaced "researchers" with other terms.
 * Thanks for the comments, <b class="Unicode">r ʨ anaɢ</b> talk/contribs 18:03, 23 July 2009 (UTC)
 * The above discussion is preserved as an archive. Please do not modify it. No further edits should be made to this page.