Careful Speech

[Seirios]

(Also, numbers are deceptively complex in English; There's different ways to read the same number depending on whether it's an amount or a year. For example, how does one read 1998? If it's a year: "nineteen ninety-eight", if it's an amount: "one-thousand nine-hundred ninety-eight". For a Chinese speaker, it's even more complex, because in Chinese, numbers are separated every 4 digits instead of every 3 digits.)

Chinese speakers differentiate numbers as used in years and just regular numbers too. "1998" would be "one nine nine eight year" (一九九八年) or "one thousand nine hundred ninety eight" (一千九百九十八). It gets fun when there are zeros involved, though, as any zero between 2 non-zero digits is said, but any zero afterwards is not. Thus "1,050" is read as "one thousand zero five"( 一千零五) while "1,005" is "one thousand zero zero five" (一千零零五). As years, those would be "one zero five zero" and "one zero zero five" respectively (一零五零 and 一零零五).

English has the added complexity that in certain cases, numbers in the thousands can be stated as an amount of hundreds ("I have fifteen hundred dollars")

The convention I've known of is different from that you described. "One thousand zero five"（一千零五）means 1005 for me and literally everybody I've known of. "One thousand zero zero five" is something I haven't heard/seen of, and to express 1050 we would say "one thousand zero five ten/one thousand zero fifty"（一千零五十）, "fifty" being expressed as "five ten" in Chinese. That is, any numbers of zeros between two nonzero digits are compressed into only one "zero", and there doesn't exist any special rule about zeros at the end; you can say we then only read nonzero digits together with their position/location.

204,1203,3500,0320 is "[[two hundred zero four] ten-thousand, one thousand two hundred zero three] hundred-million, [three thousand five hundred] ten-thousand, zero three hundred two ten" （两百零四万一千两百零三亿三千五百万零三百二十）, brackets being only aid I added to help Western users better understand the grouping system of numbers in Chinese, and commas only to match the commas used with the Arabic number. The number separated in Western tradition is 204,120,335,000,320.

BTW I don't think the 4-digit comma-split practice is taught in China. When numbers are split by comma I always saw them done in the Western, 3-digit-a-comma way, which (used in Chinese is a useless distraction) has actually annoyed me a lot >.>

[Rui]

Huh interesting. I've never heard someone say e.g. 五十 in a number larger than 100. The 十 is not said. Even in "150" it's just 一百五 (vs. 105 is 一百零五). Maybe I have to listen harder haha.

[Seirios]

Huh interesting. I've never heard someone say e.g. 五十 in a number larger than 100. The 十 is not said. Even in "150" it's just 一百五 (vs. 105 is 一百零五). Maybe I have to listen harder haha.

Well on that we are the same...一百五十 (one hundred five) is used, but we also use this form a lot (colloquially? in some cases?) when said alone or when the number is not modifying anything after it as a quantifier. e.g. -"How much is it?" -"two hundred three"（“多少钱？”“两百三”）, and "a bill of 63,000"（六万三的账单）(here the "63,000" stands before "bill" in Mandarin but does not quantify it; I mean the phrase does not mean something like "63,000 bills").

Actually it seems like we can omit the magnitude of the last nonzero number whenever no zero that need to be said is in between. e.g. 15000, 16380. and also 106000 as its full form can be "ten ten-thousand six thousand" （十万六千）(if you choose to use its alternative full(-er) form "ten ten-thousand zero six thousand"（十万零六千）, however, you cannot omit), but not 100060 as its full form is "ten ten-hundred zero six ten"（十万零六十）. That is, if any zero is before a nonzero number in any 4-digit grouping of a number (i.e. if any zero is said), then this omission does not apply.

In non-formal contexts I feel like this is overwhelmingly preferred to the full form and I feel like there is an articulation/stress on the number if the full form is used, but I'm not sure. I don't ever expect (highly) formal passage to have any such omission though (e.g. legal contracts).

I also feel like, in some (highly?) special cases probably, we might omit the zero even when it stands on the first of a 4-digit grouping, but I'm not sure.

By the way, I'm sure that the shorter expressions without zero are found in ancient and archaic texts. In ancient Chinese no zero is needed (a "有", "also/and", may be added), but I think this omission of magnitude marker is absent from it.

[Qwynegold]

Or twenty-first instead of twenty-oneth, although that's not so surprising in a way

She had generalized it another way. She also asked why it was 11th, and not 11st; It's 1st, 21st, 31st, etc, after all.

Or yeah, the look of mild despair in my kid students' eyes when they've assumed that a million means 10,000 and I tell them otherwise.

I didn't know this, but it makes sense; They want to recreate the Chinese system in English, and the first step is to have a number for 1,0000 to multiply from.

I think you misunderstood what I was saying about the commas there: I mean technically I can't speak for China, but here they put the commas in the same place as the usual Western system, so they write 10,000 or 100,000 instead of 1,0000 or 10,0000 (although as often they write it like １万 or １０万, using a combination of characters and numerals). Incidentally, when I type that in the Japanese IME it gives the options of 100,000, 100000, or １０万, but not 10,0000. I don't know why they do this, but it means that the commas for them are meaningless decorations.

I've heard that when speaking English, many Japanese people need to write down large numbers to be able to translate them from Japanese to English, or vice versa.

[finlay]

Huh interesting. I've never heard someone say e.g. 五十 in a number larger than 100. The 十 is not said. Even in "150" it's just 一百五 (vs. 105 is 一百零五). Maybe I have to listen harder haha.

This sounds awfully complicated - like in japanese they're 百五十 and 百五... you don't say 一百 and this business of inserting spoken zeroes is strange.

Rendered in characters, I guess that longer one would be like
二百四兆千二百三億三千五百万三百二十
But I've never seen such a long number written that way - they're more likely to write it out in Arabic numerals, write it in a place-value version of kanji numerals, or do a combination of numbers and kanji, viz.
204兆1203億3500万0320

[svld]

The more formal way I "say" 204,120,335,000,320 in Chinese is 兩百零四兆一千兩百零三億三千五百萬零三百二十 (or 二百零四兆一千二百零三億三千五百萬零三百二十).
Informal ways varies, it can even be something like 二零四兆一二零三億三五零零萬零三二零 or 二零四兆一二零三億三千五百萬零三百二.

[Yng]

In Arabic, careful speech (e.g. to emphasise a word which has been misheard, or whatever) can sometimes involve introducing features from the literary language - pronouncing /ʔ/ as /q/ (ʔāl > qāl), for example, or restoring vowels (mʕallim > muʕallim). In Persian, careful speech generally seems to involve pronouncing words almost exactly as they are spelt, including producing all written <ʕ ʔ h> (as /ʔ ʔ h/ - this surprised me, because I didn't think written ʔ was anything more than a device for marking hiatus, but), as well as using the written versions of various morphemes with contracted forms. For example, the 3rd person singular suffix <-ad> is almost universally produced as /e/ in Tehrani Persian, whilst the copular clitic <-ast> is /e/ or /s[t]/ depending on position; in 'careful speech', though, both can be pronounced in literary style. Likewise, the common word ending -e (found in lots of Persian words as well as as a reflex of Arabic -ah) is pronounced as /a/ in rapid Tehrani speech before a clitic or suffix, accompanied by loss of the clitic's initial vowel (xunevâda-m 'my family') but in literary Persian is followed by a hiatus (خانواده‌ام xânevâde-am), and this may be restored in careful speech.

[zompist]

A lot of this discussion makes me wonder if 'careful speech' is really 'spelling pronunciation'. Most people, unlike linguists, take the written form of the language as primary.

Anyone have any examples from unwritten or morphemographic systems?

[Rui]

A lot of this discussion makes me wonder if 'careful speech' is really 'spelling pronunciation'. Most people, unlike linguists, take the written form of the language as primary.

Anyone have any examples from unwritten or morphemographic systems?

Mandarin 3rd tone, though again it could be because of some educational bias.

It's traditionally taught as being a falling-rising tone, when phonetically it's more like a low falling tone when before 1, 2, and 4 tone, and merges with rising (2nd) tone before another 3rd tone. When emphasizing syllables (e.g. a recitation or pronouncing a syllable in isolation), speakers tend to actually do a falling-rising tone. In my experience.

[zompist]

My understanding is that you get the 213 pronunciation at the end of a phrasal group, including the end of a sentence. So it makes sense that isolated words would get 213.

But it may also be learned behavior. There must be some bound third-tone morphemes that always occur before another third tone, and thus are always heard as second tone? I wonder how people say them in isolation, and how they learn what the underlying tone is...

[gach]

A lot of this discussion makes me wonder if 'careful speech' is really 'spelling pronunciation'.

I'd say we need to distinguish between "careful speech" as spelling pronunciation and as tending towards the literary or some other prestige register. In descriptions of many minority languages you can often read how the dialect of one community is regarded as more "correct" than the others by the speakers of the language themselves.

In Finnish much of careful speech is just replacing colloquial features, including inflectional patterns, with literary ones. For example, you'll see the restoration of the possessive suffixes for attributive possession in more "careful" Finnish where in ordinary colloquial varieties only the SG2 suffix is present: coll. mun talo ("my house"), sun talos ("your house") > lit./"careful" minun taloni, sinun talosi.

But careful speech can also introduce hypercorrection and result in features that are not present in the written variant of the language and can't count as spelling pronunciation. In Finnish many Vi diphthongs in non-initial syllables have lost the i in the dialects and colloquial varieties. In particular this is the case for verbs that have the derivational affix -OttA- which invariably has a monophthong o/ö in its first syllable in colloquial speech. In the literary norm, however, there are both verbs where this affix has been fixed into the monophthongal form -OttA- (irrottaa, korottaa) and others that have been fixed into the diphthongal form -OittA- (aloittaa, kirjoittaa). This is a quirk of the development of the literary language and means that very often speakers don't know intuitively which is the correct and approved variant of a particular verb. The result is hypercorrection where "too many" verbs get injected with the missing i. A common example would be hypercorrected irroittaa ("detach") as opposed to both colloquial and literary irrottaa.

[Terra]

A lot of this discussion makes me wonder if 'careful speech' is really 'spelling pronunciation'. Most people, unlike linguists, take the written form of the language as primary.

Anyone have any examples from unwritten or morphemographic systems?

Well, there is the opposite case: eggcorns.