Need help with Chinese non-words

[Lauranen]

Hi all,

 

I am trying to create Chinese non-words (e.g. words that can be pronounced in Chinese but that don't mean anything). A non-word in English would be for example "snarp", "vonk" or "tornaduce".

I know it is challenging as most syllables used in Mandarin Chinese have a meaning in all 4 tones. But this has been done before for certain psycholinguistic experiments.

Ideally I would have 3 syllables long non-words that don't resemble an existing word in Mandarin Chinese too much (i.e. if a Chinese speaker hears this word, they would perceive it as nonsense).

There are two ways to go about this: either trying to find existing syllables that don't happen to be used in one of the tones or finding syllables that don't exist but could be pronounced, i.e. in pinyin "chian" or "rian" don't seem to exist according to this table:

 

https://www.yoyochinese.com/chinese-learning-tools/Mandarin-Chinese-pronunciation-lesson/pinyin-chart-table

 

Could anyone think of any non-words in Chinese, or point me to the right direction with this? 

Any help with this is highly appreciated!

[Lu]

Have you talked to other linguists about this? Last year I took part in an experiment where the words I heard were Chinese-but-not-Chinese, so very similar to what you need. In other words, the wheel has already been invented :-)

 

Syllables like 'rian' or 'chian' are non-words, but they're not Chinese. They're impossible syllables in Mandarin, much as srspi is an impossible syllable in English. So I think that won't work. You could consider using two-syllable words that happen to not exist in that combination. Three-syllable words would work as well, but three-syllable words are more rare in Chinese, so it would be a bit less suitable, I think.

[Xiao Kui]

Quote

You could consider using two-syllable words that happen to not exist in that combination.

I think this is a useful workaround, since so many words are compound.  I know on several occasions I have heard 2 syllable  Chinese words incorrectly, and when I went to look up the "word" discovered it didn't exist, so this might work.  good luck!

here are some examples: gelun, banou, kaiwo, binsan  (for gelun I just chopped the end off 哥伦比亚 - maybe some I listed are really words, but they are not in my dictionary - hopefully they are not in the Chinese urban dictionary either lol - if so please pardon my naivety!)

[Luxi]

The one that immediately comes to mind is the very recent monosyllabic "duang", it even has a simplified and full form version. 

https://en.wikipedia.org/wiki/Duang

 

If you were able to lift the pronunciation requirement and start with the characters, you could use many. many more examples of invented characters. The artist Xu Bing has thousands:

https://teachartwiki.wikispaces.com/Book+from+the+Sky--Xu+Bing

You could be creative and attach a Mandarin pronunciation to any of those!

 

One example is the amusingly complicated character "biang" (illustration from this Wikipedia article). It does have a meaning, biang biang is a particular type of noodles, but you won't find it in dictionaries and it doesn't exist in Unicode. It must have been invented after the Kangxi dictionary was compiled, as it doesn't figure there.

 

[somethingfunny]

2 hours ago, Lauranen said:

find existing syllables that don't happen to be used in one of the tones

 

There should be lots of these.  Two that immediately spring to mind are 'gei' which is only used in the third tone and 're' which is only used in the third and fourth tone.

 

There should be a table somewhere showing all initial/final/tone combinations and from this you could work out which ones aren't used.

[lips]

pinyin is not Chinese.  It's just a way of romanizing Chinese character pronunciation.  If you really want “non-Chinese” as you described, then use the suggestions in previous posts that are based on Chinese characters , or even writing 别字, i.e. creating (non-existing) characters using standard Chinese strokes.

[Lu]

Depending on the experiment the OP wants to do, it's possible that she needs the pronunciation rather than a character. To have people listen and react to hearing certain 'words', for example.

[lips]

If that's the case just speak some random sound (or perfectly good English, or French, or Swahili, etc.) and tell people it's Chinese.  Whether the sound can be written in pinyin is irrelevant.

[zander1]

I imagine it would be a linguistics experiment designed for native Chinese speakers? This sort of experiment is extremely common in linguistics, for instance I remember studying the 'This is a Wug' test (https://en.wikipedia.org/wiki/Jean_Berko_Gleason#Children.27s_learning_of_English_morphology.E2.80.8D.E2.80.94.E2.80.8Cthe_Wug_Test)

[Lu]

1 minute ago, lips said:

If that's the case just speak some random sound (or perfectly good English, or French, or Swahili, etc.) and tell people it's Chinese.  Whether the sound can be written in pinyin is irrelevant.

Depending on the participants and the nature of the experiment, that might not be what the OP needs. There's a difference between non-words that still sound like valid Chinese (or English, or Dutch, or Swahili) words and words/non-words that sound like they are from another language. 'Tornaduce' is clearly an English non-word, while 'prakkersmaap' or 'diangbou' are not. 'Diangbou' could be a Chinese non-word, but clearly not Dutch. Etcetera.

[zander1]

1 minute ago, Lu said:

There's a difference between non-words that still sound like valid Chinese (or English, or Dutch, or Swahili) words and words/non-words that sound like they are from another language

 

I think the term you're looking for is pseudoword: https://en.wikipedia.org/wiki/Pseudoword

[lips]

3 hours ago, Lauranen said:

Ideally I would have 3 syllables long non-words that don't resemble an existing word in Mandarin Chinese too much (i.e. if a Chinese speaker hears this word, they would perceive it as nonsense)

"sounds Chinese“ doesn't seem to be a requirement.

 

 

[zander1]

Just now, lips said:

too much

 

'too much' implies it has to resemble it at least in some sense?

[lips]

42 minutes ago, zander1 said:

I think the term you're looking for is pseudoword: https://en.wikipedia.org/wiki/Pseudoword

If the OP really wants pseudoword, then according to the definition in wikipedia it has to sound or look like Chinese.  The OP should clarify if he/she wants it to sound like or look like Chinese.

[Lu]

I think maybe this is a good point to pause the discussion a bit and wait for the OP, who can give us more information on what exactly she needs and why.

[Lauranen]

Hi all,

 

Apologies for my vague question, I didn't mean to leave you speculating on all the different ways non-words could be created! What I'm looking for is:

1. The phonological form rather than the visual representation of Chinese non-words (they should sound like Chinese, they don't need to look like Chinese, the non-words would be played to the participants, not shown).

2. The non-words should be phonotactically legal in Mandarin Chinese (i.e. they can be pronounced and they sound like they could be Chinese, see the examples provided by Lu: "'Tornaduce' is clearly an English non-word, while 'prakkersmaap' or 'diangbou' are not. 'Diangbou' could be a Chinese non-word, but clearly not Dutch. Etcetera.") 

3. The non-words should not resemble an existing word. When a participant hears the word, they should not be able to twist their auditory perception to the closest sounds that would have a meaning. (I said "resemble an existing word in Mandarin Chinese too much" as I imagine they would necessarily resemble existing words to some extent as they are meant to sound like Chinese).

 

This is for an exiperiment for native Chinese speakers.

 

I think I would have to go for the "existing phonemes, non-existent tone -combination" approach, as somethingfunny suggested. The problem is that as I'm trying to create at least bisyllabic non-words (still preferably trisyllabic), combining syllables with non-existent tones might still make them sound like some existing words, as the tone might change depending on its phonological context, for example: When a 3rd tone (such as “yě”) is followed by another 3rd tone in a group, the first 3rd tone changes to a 2nd tone (such as “yé”). But I suppose this could be easily avoided by having the created syllable combinations tested by a native Chinese speaker..  

 

Thank you for your input so far, it's been helpful and more than I expected!

[dwq]

Not words, but paragraph of text:

 

Chinese lorum ipsum

http://www.richyli.com/tool/loremipsum/

 

Not sure if it uses some algorithm other than just stringing random characters together.

[Lu]

If you're using bisyllabic (or trisyllabic) words, you don't have to make them from non-existing syllables, you can just combine existing syllables in non-existing combinations. For example, fābiǎo is a perfectly normal word, while biǎofā doesn't seem to exist.

 

You could also come up with new syllables (as you already planned), but the effect is a little different: non-words like fòupu or diǎngbóu sound like words in a language works similarly to Chinese, while something like biǎofā sounds like a Chinese word that the listener just happens not to know.

 

I think if you could find a linguist, especially a native Chinese speaking linguist, who has worked on these things, you could have a very interesting and fruitful conversation. (Actually, is this something @OneEye would know about? Perhaps it's not his field though.)

[somethingfunny]

OK, nouns should be pretty easy:

 

re1gei1zi5

 

That should confuse the hell out of some people.

 

 

[Lauranen]

Yes, I will consult a Chinese speaking linguist later on, just trying to get as much of this done/figured out before that.

I think combining existing syllables in a non-existent way would be risky, because ultimately the non-words should not elicit any kind of meaning in the listeners, which biǎo and fā might do as they seem to have meanings independently, even if they were combined in a non-existent way (in the same way that "boardblack" wouldn't mean anything in English but the compounds "board" and "black" would still activate the semantic representations of these items).

 

You've been very helpful!