I want to / two fish

« previous post | next post »

In the comments to "slip(per)" (7/22/14), we have had a very lively discussion on whether or not people would pronounce these two sentences differently in Mandarin:

wǒ yào tuōxié
"I want slippers."

wǒ yào tuō xié
"I want to take off my shoes."

Of the more than two dozen individuals having native fluency in Mandarin whom I surveyed, those who said that they pronounced these sentences exactly alike tended to be professional language teachers responsible for elementary and intermediate pedagogy, beginning and intermediate students, and  humanists who are highly literate in characters, while those who stated that they pronounce these two sentences differently tended to be linguists and individuals who are fluent in Mandarin but illiterate or minimally literate in characters.

I presented the views of some in the latter group in these comments to the original post:

Statement of the problem and my own approach

San Duanmu and Jiahong Yuan

Perry Link and Tom Bartlett

Brendan O'Kane, Julie Wei, and Gloria Bien

Mark Swofford and Haitao Tang

Andy Lee (here and here)

Christoph Harbsmeier

Yanyan Sui (whose dissertation dealt directly with this sort of problem)

Other comments to the thread touched on related issues.

I believe that those who maintain that they pronounce wǒ yào tuōxié and wǒ yào tuō xié exactly alike have been unduly influenced by the lexically prescribed pronunciation of the constituent morphosyllables in isolation, while those who recognize that the two sentences are pronounced differently are taking the words that make up the sentences in the overall context of their grammatical roles.

I was trying to think of comparable sentences in English, and yesterday I hit upon this pair:

I want to fish.

I want two fish.

If we pronounce the syllables of these two sentences as they are presented in a dictionary, they will be perfectly homophonous.  But no one who possesses native fluency in English will pronounce these two sentences alike in actual conversation.

As Mark Liberman explains:

The usual American way to pronounce "I want to fish" would involve a completely reduced "to" in "want to", so that the vowel becomes schwa and the consonant is just a (perhaps nasal) flap, e.g.

[wɒ̃ɾə] or [wɒ̃nə]

rather than [wɒ̃ʔtuw]

A slightly less reduced version would be something like


And, from the other side of the Big Pond, John Wells phonetically annotates the two sentences thus:

1. I want two fish (not just one)                 aɪ wɒnt ˈtuː fɪʃ

2. I want to fish (I’m an angler!)                aɪ wɒnt tə ˈfɪʃ

                                                                           aɪ wɒntə

                                                                           aɪ wɒnə

I submit that speakers of Chinese languages make the same kinds of adjustments in real life.  They are not like robots or automatons who woodenly pronounce syllables according to their stipulated sounds as found in a dictionary.  Rather, they make changes and adjustments, add emphases and stresses and pauses, and so forth, just as people do in all other languages.  Our language does not consist solely of lexical items.  A language will not function with syllables alone, but must take into account words, grammar, syntax, and various suprasegmental features of actual speech.  Just as we will not get our message across if we insist on pronouncing "I want two fish" and "I want to fish" in an identical fashion, so we will confuse people if we pronounce wǒ yào tuōxié and wǒ yào tuō xié exactly alike.


  1. Ted McClure said,

    August 3, 2014 @ 2:14 pm

    As a complete amateur thinking about how I pronounce "two/to fish", I don't see the reduction of "to" as the distinction, but rather the difference in rhythm between the sentences: "I want two fish" comes out either even in emphasis across "two fish" or with a smidge more emphasis on "two" if that's the important information = "I want two fish (rather than one or three)". In "I want to fish" the emphasis can be anywhere except "to":
    I want to fish (emphasis on "I") = I (as distinguished from someone else) want to fish.
    i WANT to fish = I want to fish (rather than not want to fish).
    i want to FISH = I want to fish (rather than do something else).
    There can be combinations, but "to" is not emphasized unless I am saying the emphatic "I. WANT. TO. FISH (and don't get in my way)."
    I'm an overeducated upper-middle-class anglo who has lived in many locations in the US (currently northern Arizona), with the slightly lazy tongue of a career US Army veteran.

  2. Martha said,

    August 3, 2014 @ 6:39 pm

    I'm certain that the reason people often mishear the chorus of "You Give Love a Bad Name" as "Shot through the heart and you're too late," rater than the actual "…you're to blame," is because of the atypically unreduced "to." "You're too late" doesn't even make sense there.

    Ted McClure, in my mind, the reduction of "to" and the difference in rhythm between the two sentences aren't two different things. The rhythm of "I want to fish" has a certain rhythm, which you've describe, which includes the reduction of "to."

  3. Victor Mair said,

    August 3, 2014 @ 6:47 pm

    @Ted McClure

    I agree with you that the rhythm and the emphasis of the two "fish" sentences are just as important as how "to" & "two" are pronounced. I stressed those same aspects in my own analysis of the two Chinese sentences. But I also agree with Martha that the pronunciation of "to" and "two" does matter in distinguishing between the spoken form of the two sentences.

  4. shangyang said,

    August 4, 2014 @ 12:57 am

    Chinese people adjust, but don't change the pronunciation. Certainly not as much as English speakers do.

    Given there are few contexts in which "I want sandals" and "I want to put my shoes off" can be confused; in the rare case that that happens you would just say, as was mentioned in the first post, 我要把鞋脱了 or 我要脱鞋子. Problem solved. No need for careful enunciation. That appears to me to be a personal quirk of yours Mr. Mair.

  5. Keith said,

    August 4, 2014 @ 2:29 am

    I think that there are two things going on in the pair of English sentences that VM proposes.

    1. "to" in "I want to fish" precedes the stressed syllable in "fish", whereas "two" in "I want two fish" is the stresses syllable; it is stressed in speech because it is the important detail that shows just how many fish I want: not one or three, I want exactly two of them.

    2. "to" and "two" are, at least in my idiolect, pronounced differently anyway, especially when preceding a word that starts with a vowel: [tə] as opposed to [ˈtuː] (becoming [ˈtuːʷ] before a vowel).

    I'm sure that older British readers of the blog will remember the Two Ronnies' sketch about "fork handles"…

  6. theslittyeye said,

    August 4, 2014 @ 3:48 am

    No, it sounds the same. 拖鞋 and 脫鞋 sounds exactly the same. Different punctuation in spoken tone is just subjective expression. You are confusing English grammar with Chinese. They sound exactly the same.

  7. Victor Mair said,

    August 4, 2014 @ 7:12 am

    @shangyang, @theslittyeye

    If you read through both the "slip(per)" post and the "I want to / two fish" posts, plus all of the abundant, detailed comments thereto, you'd realize that I'm hardly the only person (including many native speakers) who thinks the two Mandarin sentences are pronounced differently.

  8. Tingyu Liu said,

    August 4, 2014 @ 2:43 pm

    I am quite confused when I read some native speakers said they pronounce the two sentences exactly alike, and at the same time they also admit they speak them differently in daily life. This made me feel a little controversy in the same person.
    Then I start to think over how I learn these two words in elementary school, maybe that is the key to understand why the native speakers believe these two words pronounce exactly alike, because we were always asked to distinguish these two similar pronounced words that in different character forms.Here is an example in a teacher's text book of the second grade of the elementary school in China (in the fourth part):



    And we met these two words in this way (to distinguish them in shapes) for many times which made people believe they have exactly the same pronunciation although they also realize they spoke them differently in real life.

  9. Victor Mair said,

    August 4, 2014 @ 3:06 pm

    @Tingyu Liu

    Thank you for your extremely apt quotation from an elementary school teacher's textbook that speaks directly to the nub of the question in the "slip(per)" post.

  10. Elonkareon said,

    August 4, 2014 @ 10:39 pm

    I'm not sure that the "to" in my idiolect (Southern Californian + a lot of book pronunciation since I learned to read and read avidly at an early age) is reduced all the way to a schwa, at least not in all cases, but it is certainly reduced and destressed. It at the very least is still mildly rounded.

  11. Michael Watts said,

    August 5, 2014 @ 4:44 am

    If we pronounce the syllables of these two sentences as they are presented in a dictionary, they will be perfectly homophonous.

    News to me. The little-known Merriam-Webster dictionary gives three pronunciations for "to", the clitic, of which the primary is [tə]. In general, I don't think contrasting a clitic with a full word is particularly informative as to the contrast of one full word with another one. Clitics have "unstable" pronunciations because… they're clitics. In the case of a/an, we can even see unstable spelling, but what's the lesson we're supposed to draw?

  12. Victor Mair said,

    August 5, 2014 @ 6:26 am

    @Michael Watts

    I could direct you to any number of websites that refer to "to", "too", and "two" as homophones, and everyone I know would pronounce them as homophones in isolation.


    Listen to the recordings and see the pronunciation keys here:




    The pronunciations of all three are identical *in isolation*.

    It's only in grammatical context that they take on different pronunciations, which is the main point of this post.

  13. Michael Watts said,

    August 5, 2014 @ 1:49 pm

    Your chosen dictionary is actually better than merriam-webster in that it specifies that "to" is pronounced "tə when unstressed". You've made me very curious why I should believe that http://www.writingforward.com qualifies as "a dictionary", though.

    And a major point of my comment was that "to", as a clitic, can't have a pronunciation in isolation; the concept doesn't make sense. "a" and "an" are the same word; would you be as happy to state that "an" is pronounced, in isolation, without the [n]?

    I saw an English-Chinese dictionary which noted that "legitimate" might be a noun or a verb, and that the pronunciation might end with -[eɪt] or with -[ət]. I didn't conclude that some hitherto-unrecognized properties of English caused minor differences in the pronunciation from sentence to sentence; I concluded that I was looking at a low-quality dictionary. A respectable one (such as merriam-webster, cambridge, and I'm sure any number of others) will point out that the verb obligatorily ends in -[eɪt] and the noun in -[ət].

  14. Victor Mair said,

    August 5, 2014 @ 2:57 pm

    @Michael Watts

    "Your chosen dictionary is actually better than merriam-webster in that it specifies that 'to' is pronounced 'tə when unstressed'."

    I'm glad you noticed that; I thought you'd like it.

    "…would you be as happy to state that 'an' is pronounced, in isolation, without the [n]?"

    No way!

    "…the verb obligatorily ends in -[eɪt] and the noun in -[ət]."

    I wish that we had dictionaries of comparable quality for Chinese.

  15. Michael Watts said,

    August 5, 2014 @ 4:05 pm

    "…the verb [legitimate] obligatorily ends in -[eɪt] and the noun in -[ət]."

    I wish that we had dictionaries of comparable quality for Chinese.

    I'd say this is basically equivalent to noting that 长 is pronounced zhang3 when it relates to appearance, while being pronounced chang2 when it relates to length. I have issues with my chinese dictionaries, but that isn't one of them.

  16. Victor Mair said,

    August 5, 2014 @ 8:32 pm

    @Michael Watts

    You're confusing characters with words and morphemes. The phenomenon of duōyīnzì 多音字 ("polyphonic characters") or pòyīnzì 破音字 ("characters with variant readings") is well known and covered in virtually all dictionaries. What we have been talking about in this post and in the "slip(per)" post from which it sprang has to do with rhythm, emphasis, stress, and so forth, which are almost never treated in Chinese dictionaries and are known only in specialized phonological studies.

  17. Simon P said,

    August 6, 2014 @ 12:39 am

    I could see myself reading "to/two" the same way if I was emphasizing the "to". Like this:

    "I'm going fishing, honey."
    "Fishing? Why don't you just go to the store and buy a fish? It'll save you a lot of effort."
    "Because I don't want 'a fish', I want TO fish".

    … which would of course elicit the reply "I'm sure they have more than one, dear", making the thing into a hilarious double entendre joke!

  18. Michael Watts said,

    August 6, 2014 @ 12:41 am

    The distinction between "legitimate", the noun, and "legitimate", the verb, also has nothing to do with rhythm, emphasis, stress, or any other aspect of prosody; it's a simple case of two words with the same spelling and different pronunciations. As such, it would appear to be an exact analog of the distinction between the 长 of 我长得不漂亮 and the 长 of 长江.

  19. Victor Mair said,

    August 6, 2014 @ 6:56 am

    Those who are closely following the underlying issues of this post and "slip(per)" would benefit from reading "Wrecking a nice beach" and its comments.

  20. Victor Mair said,

    August 6, 2014 @ 7:08 am

    @Michael Watts

    Zhǎng ("grow") and cháng ("long") may be cognates, but 长 and 长 are just 长.

  21. cxhh said,

    August 6, 2014 @ 9:42 am


    You can also get unreduced infinitival 'to' in constructions where the verb phrase that would follow it is elided, although it doesn't get stress like it does in your examples:

    "I run a lot but I haven't had time to this week."
    "She did it even though I asked her not to."

  22. Matt_M said,

    August 6, 2014 @ 10:51 pm

    @cxhh, SimonP:

    And the preposition "to" is not always a clitic, either: in "Where are you off to?" or "Who should I give it to?" it is fully stressed and pronounced the same as "too".

  23. Simon P said,

    August 7, 2014 @ 12:59 am

    I just realized that the following sentences can be pronounced exactly equally in some American dialects:

    "I want a fish"
    "I want to fish"

  24. cxhh said,

    August 7, 2014 @ 12:47 pm

    Matt_M, But is the preposition 'to' the same word as the 'to' that introduces infinitives? Is infinitival 'to' a preposition?

RSS feed for comments on this post