Linguistic Miscellany Thread

[Raphael]

Do I get this right that in current French, the vowel in "oui" is often something a bit like /ai/?

[bradrn]

Do I get this right that in current French, the vowel in "oui" is often something a bit like /ai/?

As I understand it, oui and ouais are considered two different words, analogous to English ‘yes’ vs ‘yeah’.

[Raphael]

Do I get this right that in current French, the vowel in "oui" is often something a bit like /ai/?

As I understand it, oui and ouais are considered two different words, analogous to English ‘yes’ vs ‘yeah’.

Ah, thank you, that explains it!

[Starbeam]

To my ears, Arapaho /o/ has only ever sounded like [ʌ]. Historically, it is a straight-forward reflex of PA /a/, so probably the question of how /a/-y versus how /o/-y it is, is a matter for the philosophers.

That works for me, thanks for the help. I wonder if /e/ is properly marked as well

[WeepingElf]

A few hours ago, I met the phrase hen to pan (italics in original) in a German-language essay discussing physics and philosophy. First I read it as English, wondering what frying chicken was doing there, but I soon realized that it's Greek, 'one is all'.

[Travis B.]

A few hours ago, I met the phrase hen to pan (italics in original) in a German-language essay discussing physics and philosophy. First I read it as English, wondering what frying chicken was doing there, but I soon realized that it's Greek, 'one is all'.

Ah, one of those gratuitously untranslated phrases in academic texts...

[zompist]

A few hours ago, I met the phrase hen to pan (italics in original) in a German-language essay discussing physics and philosophy. First I read it as English, wondering what frying chicken was doing there, but I soon realized that it's Greek, 'one is all'.

Ah, one of those gratuitously untranslated phrases in academic texts...

People need to be more pedantic! This confusion would have been averted if they wrote ἓν τὸ πᾶν.)

[Travis B.]

A few hours ago, I met the phrase hen to pan (italics in original) in a German-language essay discussing physics and philosophy. First I read it as English, wondering what frying chicken was doing there, but I soon realized that it's Greek, 'one is all'.

Ah, one of those gratuitously untranslated phrases in academic texts...

People need to be more pedantic! This confusion would have been averted if they wrote ἓν τὸ πᾶν.)

I was thinking that too!

[Raphael]

A few hours ago, I met the phrase hen to pan (italics in original) in a German-language essay discussing physics and philosophy. First I read it as English, wondering what frying chicken was doing there, but I soon realized that it's Greek, 'one is all'.

Yeah, I think I would have thought of frying chicken at first, too.

[Travis B.]

A few hours ago, I met the phrase hen to pan (italics in original) in a German-language essay discussing physics and philosophy. First I read it as English, wondering what frying chicken was doing there, but I soon realized that it's Greek, 'one is all'.

Yeah, I think I would have thought of frying chicken at first, too.

To me as a native English-speaker that would seem like nonsense or at least something a non-native English-speaker (a German?) would write.

[WeepingElf]

A few hours ago, I met the phrase hen to pan (italics in original) in a German-language essay discussing physics and philosophy. First I read it as English, wondering what frying chicken was doing there, but I soon realized that it's Greek, 'one is all'.

Ah, one of those gratuitously untranslated phrases in academic texts...

People need to be more pedantic! This confusion would have been averted if they wrote ἓν τὸ πᾶν.)

Yes, that would have avoided that ambiguity by making it obvious that it is Greek, not English. Indeed, I wondered only for about a second or so, as the translation into German followed which made things clear. Still, a funny coincidence.

[Richard W]

In examinations on arithmetic, one is frequently asked to give an answer to so many decimal places. I was watching an English-language video on the orderly extraction of square roots digit by digit, and it seemed to imply that when extracting the square root of 4.7 to one decimal place, the answer is 2.1. Now, as the square root of 4.7 is 2.167.., shouldn't the answer be 2.2? (To 4 significant figures, the square root of 4.7 is 2.168.) How does understanding vary amongst English speakers, and what happens in other languages?

Now, I am young enough not to have been examined on this extraction method, and it is just possible that there is a different handling of curtailing numerical answers for this particular problem. Now, a quicker method of extracting square roots may be Newton-Raphson (iteratively replace approximation with the mean if it and the square divided by the approximation), but I don't recall any rigorous rules for stopping conditions. (Newton-Raphson consistently overestimates the square root.) There's also a piecewise linear approximation for the length of the hypotenuse, accurate to 3%, which is good enough for many purposes.

[Lērisama]

In examinations on arithmetic, one is frequently asked to give an answer to so many decimal places. I was watching an English-language video on the orderly extraction of square roots digit by digit, and it seemed to imply that when extracting the square root of 4.7 to one decimal place, the answer is 2.1. Now, as the square root of 4.7 is 2.167.., shouldn't the answer be 2.2? (To 4 significant figures, the square root of 4.7 is 2.168.) How does understanding vary amongst English speakers, and what happens in other languages?

At least in maths exams in the UK, that would be marked as wrong. It's not really something that comes up much outside that context, so I only have what was drilled into me by maths teachers¹.

¹ Work it out to 1 or 2 more dp than you need, then round at the very end

[hwhatting]

At least in maths exams in the UK, that would be marked as wrong. It's not really something that comes up much outside that context, so I only have what was drilled into me by maths teachers¹.

¹ Work it out to 1 or 2 more dp than you need, then round at the very end

Just for clarity, 2.2 would be the correct answer in UK exams?

[Richard W]

At least in maths exams in the UK, that would be marked as wrong. It's not really something that comes up much outside that context, so I only have what was drilled into me by maths teachers¹.

¹ Work it out to 1 or 2 more dp than you need, then round at the very end

Just for clarity, 2.2 would be the correct answer in UK exams?

We would say so. However, it seems that extracting square roots (at least by this method) is no longer examined in the UK or the US, so I've raised the question. I've heard that it is still examined in India.

I had thought there might be rounding issues with the Newton-Raphson method of square-rooting, apparently also known as 'divide and average', e.g. with getting approximations like 3.15000000007, but for square rooting it is fairly quick to check that the rounded answer brackets the exact answer, which one should do unless one is currently well-drilled in calculating square roots.

[Travis B.]

At least in maths exams in the UK, that would be marked as wrong. It's not really something that comes up much outside that context, so I only have what was drilled into me by maths teachers¹.

¹ Work it out to 1 or 2 more dp than you need, then round at the very end

Just for clarity, 2.2 would be the correct answer in UK exams?

We would say so. However, it seems that extracting square roots (at least by this method) is no longer examined in the UK or the US, so I've raised the question. I've heard that it is still examined in India.

I had thought there might be rounding issues with the Newton-Raphson method of square-rooting, apparently also known as 'divide and average', e.g. with getting approximations like 3.15000000007, but for square rooting it is fairly quick to check that the rounded answer brackets the exact answer, which one should do unless one is currently well-drilled in calculating square roots.

When I studied math in school, including in both high school and college, I was never taught to calculate square roots even though I was taught other numerical methods such as Taylor and Maclaurin series; this was something I had to learn on my own when I later implemented fixed-point and floating-point routines for my own purpose.

[Lērisama]

At least in maths exams in the UK, that would be marked as wrong. It's not really something that comes up much outside that context, so I only have what was drilled into me by maths teachers¹.

¹ Work it out to 1 or 2 more dp than you need, then round at the very end

Just for clarity, 2.2 would be the correct answer in UK exams?

As Richard W said, yes

[hwhatting]

I vaguely remember learning an interpolation method for square roots back in school, but in real life I either know the square root by heart (for the numbers in the multiplication table) or I use a calculator / an Excel formula.

[Richard W]

I had thought there might be rounding issues with the Newton-Raphson method of square-rooting, apparently also known as 'divide and average', e.g. with getting approximations like 3.15000000007, ...

For evaluating the limit of a convergent sequence, giving the limit to 1 decimal place can be a non-halting computation. However, giving the limit to 1 or more decimal places is not. In an example like this, for a 'defined' provably convergent sequence (sorry, I don't know the correct brain-frying term), one can get an approximation accurate to 0.001. Then there are three possibilities:

• The limit is clearly closer to 3.1 than to 3.2 - return 3.1
• The limit is clearly closer to 3.2 than to 3.1 - return 3.2
• The limit is closer to 3.15 than to 3.14 or 3.16 - return 3.15. This answer is to two decimal places, but is allowed by the relaxed instruction.

I haven't optimised the process. Also, it's conceivable that my argument may fall foul of the axiom of choice.

[WeepingElf]

A few hours ago, I found the diminutive Säundchen of the English loanword Sound in a German-language music forum. Fun fact is that the form is perfectly regular in German: the suffix -chen triggers umlaut, /au/ umlauts to /oy/, and the spelling is regular (except the initial voiceless /s/, which doesn't occur in that position in native German words and there is therefore no convention how to spell it).