What’s the consonant in ‘nǐ’?
Here’s a rather stark example of how learning standardized Mandarin (pǔtōnghuà = 普通话) is a non-trivial exercise in learning new phonemes for speakers of other Chinese languages. This question is from the Peking University language forums, so you might guess the person asking is a relatively* sophisticated language learner:
普通话的“你”的声母到底是舌尖的n，还是舌面的gn？我以前一直以为是舌面的gni哦，结果听到一位同事读边音li，他竟然说他也能分清楚li和ni。还说北京话的”你”就是“ni”而不是“gni”。一般情况下，我分不清n和l，但能分清gn和l，所以我一直以为普通话的“你”是gni，因为我一听鼻音就认为声母肯定是gn
In pǔtōnghuà is the initial consonant of “nǐ” actually an apical (tongue-tip) sound ‘n’ or a laminal (tongue-blade) sound ‘gn’**? I used to always think it was gni, but then I heard a coworker read it as a lateral, “li”. Surprisingly, he said he could distinguish “li” and “ni”. He also said Beijing dialect pronounces it “ni” and not “gni”. Usually, I can’t distinguish “n” and “l”, so I always thought that 你/nǐ in pǔtōnghuà was “gni” because as soon as I heard the nasal I thought it was “gn.”
——–
*Relatively sophisticated, but it’s hard to be sure how sophisticated. For example, I’m correcting the use of “gn” to “ng” to indicate ŋ. But heck, maybe I’m not sophisticated enough myself and there’s something about “gn” that I should know…
**Big correction: I misunderstood the “gn” notation and got corrected by Chrix in the comments below. The whole thing makes a lot more sense when you realize the writer’s “gn” stands for ɲ (IPA) like the ñ in mañana (Spanish) or the gn in gnocchi (Italian). It makes me question my own ears: is it possible that northerners say ɲihao? Does anyone have a good recording of ɲihao?
That’s an interesting workaround for people who can’t distinguish /l/ and /n/.
Are you sure they mean [ŋ]? (Googling for 舌面音 seeemed to suggest mostly palatal consonants) Because the spelling is reminiscent of how [ɲ] is written in many Romance languages except for Spanish, which uses .
I checked Ladefoged and Maddieson, and various laminal configurations can go with either velar or palatal constrictions. I don’t know why they insist to use terms relating to tongue to describe consonants, they should rather use the passive place of articulation, but maybe that’s just because I’m used to that more….
Also, I think it would make sense because a close front vowel like [i] has a palatal configuration so to speak, and in many languages, like Japanese for example, they would trigger palatalisation of the preceding consonant anyways. So the Japanese particle に would actually be pronounced [ɲi]…
Chrix: I’ll bet you’re right. So ɲ is the Spanish ñ? I assume that’s what you wrote but it doesn’t show up in my browser. The description matches, like you say, and then it’d be “gn” like “gnocchi”, right?
Gotta say, I’d have a hard time differentiating between ñi hao and nihao.
I’m going to make a note of this in the post too.
Oops, I had copy and pasted it I thought, but yeah, the tilde-n.
I mean “laminal” can just be ambiguous, and I personally think that the passive articulator are more important than the active one in describing a consonant…
I do hear sometimes ñi hao, with the Spanish ñ, but I always took it to be a pronunciation quirk, or even a way to speak more “cute”. Didn’t know it was a Northern thing.
I seem to hear it more in 您 than in 你. In any case it is never as clear cut as the Spanish ñ. I think it is somewhere in the middle between our ñ and our n, if that makes sense.
Hi, I’d just like to say that just as Crix has pointed out above, the final ‘yi’ has some kind of palatal quality, hence the y in writing down the pinyin if the syllable ‘yi’ stands alone.
So both [ni] and [ɲi] are correct. If you look at a word like 音乐, it also should begin with a [yi]. In class we’ve learnt there always should be a little friction in pronouncing the i-finals.
Anyway, I can only refer to my professor’s booklet, which had a lot drilling practises and a brief overview of the theory we saw in class regarding Chinese’s sound systemn.
ɲ = ȵ = ñ
I’ve also heard ñi hao. ŋ (ng) seemed a stretch to me when I first read that. ȵ/ɲ/ñ makes much more sense.
There are dialects of Chinese that distinguish [n] and [ɲ] (often written [ȵ] by Chinese linguists), though I am not aware of any in which the difference in phonemic. It is of course phonemic in many languages of the world, including Spanish.
Among the dialects are Mandarin dialects, but not Beijing or Putonghua.* For example, Jinanhua is described as having [ȵ] in such words as 泥，娘，牛. The palatal nasal is clearly conditioned by the presence of a high front vowel.
Suzhou, a Wu dialect, is also described as having [ȵ] in words like 泥，女，严.
English speakers generally cannot easily hear the distinction between [ȵ] and [n] before [i], but such a distinction is possible and real. It doesn’t surprise me to hear the the poster of the original question was able to better distinguish [n] and [l] by pronouncing the former as
[ȵ], and thus keeping it farther away (in terms of both production and perception) from [l].
* I have no doubt that there is some small degree of allophonic palatalization in Beijing Mandarin, but it is must not be to the same extent as in those dialects described as having a palatal nasal. I certainly can’t hear it in Beijing speech.
In Wu it’s phonemic and 牛 is said /ɲiɔ/ in most dialects.
That whole ɲ vs ȵ thing has been discussed at length at Annals of Wu. Long story short, ɲ is right and ȵ is wrong but enough people got used to ȵ that it’s still widely used. There are a dozen similar ones.
if it’s phonemic in Wu, can you give some minimal pairs?
You just want the letters or do you want words that go with them? In most Wu dialects they distinguish between n/ɲ/ŋ and a linguistically self-aware speaker would tell you they’re different sounds (“letters” doesn’t seem right here).
Though looking back, at least in Shanghai it’s a lot like the difference between zh and j in Mandarin where j (and here ɲ) is always followed by /i/ or /y/.
a minimal pair by definition would involve words that are differentiated by two different sounds, thus establishing phonemehood for both of them. I’d just like to know some so I can reference them when necessary.
Most interested in words with the final /i/.
“Minimal pair” can also refer to the phonemes themselves as well, at least in my experience.
For words, 努 nu˨˧ and 我 ŋu˨˧ exist, though there is no [ɲu]. In Shanghainese at least, there are no pairs between ɲ and either ŋ or n that exist that don’t have an extra [i] or [y] following the initial nasal, however there are endings -in and -iɲ where the distinction is made.
Again it’s the same as j (or x/ɕ or q/ʨ) in Mandarin. They are only ever followed by [i] or [y]. I don’t know if that would disqualify /ɲ/ as phonemic, (unlike Sanskrit which can do [ʂɑ] [ʃɑ] and [ɕɑ] without the [i] that’s required in pinyin), at least in the initial position, though somehow that doesn’t seem right.
Would be interested to see a source for this, because that’s not how we were taught when talking courses in phonetics and phonology. (See for instance http://en.wikipedia.org/wiki/Minimal_pair)
Also let me modify my question: I guess my assumption is that while it is quite normal for a language to have n/ɲ/ŋ as distinct phonemes, the expectation would be that the contrast would be neutralised before /i/, that’s why I was specifically asking for a minimal pair involving /i/ as a coda.
A couple of interesting points have come up in this discussion — I’d like to throw my two cents in …
1) Technically, IPA [ɲ] and [ȵ] represent two different sounds, the first a true palatal and the second a slightly fronter alveo-palatal. I don’t think there are any languages that distinguish the two phonemically, so if one isn’t too concerned about phonetic precision, it doesn’t really matter which one you use. For various historical reasons, linguists practicing in China tend to favor [ȵ] (whether they are describing varieties of Chinese or non-Chinese languages), which is rarely used by Western linguists. I’m not sure if there have been any precise phonetics experiments to identify the exact place of articulation of this sound in Northern Wu.
2) In my understanding, I don’t think Yellowcard’s statement about the pinyin syllable ‘yi’ is correct. The standard description of the pronunciation of this syllable is [i], not [ji]; that is to say, more like the beginning of English ‘ear’ than like ‘year’. The pinyin base form is just ‘i’; the addition of ‘y’ when no other initial is present is merely a spelling rule, designed to make the syllable look more natural to Westerners and to prevent ambiguities of syllable boundaries when writing compound words, for example like ‘dayi’, which would end up as ‘dai’ without the ‘y’.
That’s not to say that some Mandarin speakers don’t have a slight palatal glide in front of this vowel, but such a pronunciation is not considered standard and is definitely not phonemically distinctive.
3) A minimal pair always proves phonemic distinctiveness, but the lack of minimal pairs does not necessarily prove that you’ve got allophones. For example, there are no minimal pairs contrasting English [h] and [ŋ]. The former occurs only at the beginning of syllables, the latter only at the end, but they are never regarded as allophones of one phoneme.
In addition to complementary distribution, most linguists would argue that phonetic similarity is required for allophony. That’s not a completely objective criterion, but in the case of [ȵ] and [n], we definitely have phonetic similarity. In the case of [h] and [ŋ], we don’t.
One can also appeal to the psychological reality of phonemes, but this is difficult to ascertain objectively. A statement that Wu speakers think of [ȵ] and [n] as being distinct speaks to this criterion, but how can it be proven?
Finally, it’s also well known that in some cases there is no way to definitively prove whether two sounds are phonemes or not — variant analyses are equally possible. In the 1940s and 1950s a great number of papers on the phonemic status of Mandarin j- q- x- ([tɕ- tɕʰ- ɕ-]̱) were published. Some argued they were allophones of the velar g- k- x-; some that they were allophones of the dentals z- c- s-; some that they were allophones of the retroflexes zh- ch- sh-; and others that they were separate phonemes. Eventually it was recognized that no definitive answer was possible.
My personal proclivity would be treat [ȵ] and [n] as allophones.
4) Chrix, as you know, both passive and active articulators need to be stated to precisely indicated place of articulation. Western phoneticians have the habit of just using the name of the passive articulator (dental, alveolar, palatal, velar) when it can be assumed that the most natural tongue position is touching it; Chinese phoneticians do the opposite, using the active articulator when it can be assumed that it is touching the most natural location at the top of the mouth. I think either way is okay. Retroflexes are a problem either way — you really need to name both articulators to specify a retroflex articulation (or use the term “retroflex”, which is outside of both systems.) The commentor who Syz translated was using normal Chinese practice in his terminology.