Language quizzes

[Qwynegold]

I'm making these language quizzes, and I'm using the YouTube channel I Love Languages as a source, because it has a lot of language samples. Here are two examples of my quizzes:

Guess the Language #1
Guess the Language #2

However, since it's YouTube I can't just copy texts I see on screen. That's a problem when there's a script I'm not familiar with. So I was wondering if someone who knows Arabic could re-type the text in https://youtu.be/XTeiW-Ps6AE, at 2:20 for me? The one that's translated as "And that cry resulted in [...] rid her of Romans".

Also, is there anyone with experience in Sumerian cuneiform, Egyptian hieroglyphics or other ancient scripts, or with any Brahmic scripts?

[Richard W]

I'm making these language quizzes, and I'm using the YouTube channel I Love Languages as a source, because it has a lot of language samples. Here are two examples of my quizzes:

Guess the Language #1
Guess the Language #2

However, since it's YouTube I can't just copy texts I see on screen. That's a problem when there's a script I'm not familiar with. So I was wondering if someone who knows Arabic could re-type the text in https://youtu.be/XTeiW-Ps6AE, at 2:20 for me? The one that's translated as "And that cry resulted in [...] rid her of Romans".

Also, is there anyone with experience in Sumerian cuneiform, Egyptian hieroglyphics or other ancient scripts, or with any Brahmic scripts?

Well, you might be able to twist my arm, but i only know the script, not the language. Fortunately, the ambiguous transliteration provides a way of proof-reading the typing. I've transcribed my Middle Egyptian, but i haven't used the Manual shaping controls - in that respect I am at the MdeC level. For Sumerian cuneiform, you will have to specify the glyph development stage. Brahmic, I can struggle through the basics - Brahmi, Devanagari, Sinhala, Burmese and Thai. In all these cases, a proof-reading mechanism would be needed. (Well, perhaps not Thai, if the text were in an easy to read font.)

Incidentally, the Arabic text suddenly become a lot clearer - it's a lot more readable at 2:29. It was hard slog to do the first line of the transliteration/transcription: " ونتج عن تلك اصرخةفتح عظيم و هوفتح عمورية" having to use Emacs character selection by name.

[Kuchigakatai]

However, since it's YouTube I can't just copy texts I see on screen. That's a problem when there's a script I'm not familiar with. So I was wondering if someone who knows Arabic could re-type the text in https://youtu.be/XTeiW-Ps6AE, at 2:20 for me? The one that's translated as "And that cry resulted in [...] rid her of Romans".

ونتج عن تلك الصرخة فتح عظيم وهو فتح عمورية الواقعة في آسيا الصُغرى، وقد كان يحكم عمورية الروم البزنطيين، وكانت هذه المرأة تتلقى التعذيب من حاكم عمورية فصرخت باسم المعتصم الخليفة العباسيّ الأمر الذي دفعه للتوجه إلى المنطقة وتخليصها من الروم

[Qwynegold]

However, since it's YouTube I can't just copy texts I see on screen. That's a problem when there's a script I'm not familiar with. So I was wondering if someone who knows Arabic could re-type the text in https://youtu.be/XTeiW-Ps6AE, at 2:20 for me? The one that's translated as "And that cry resulted in [...] rid her of Romans".

ونتج عن تلك الصرخة فتح عظيم وهو فتح عمورية الواقعة في آسيا الصُغرى، وقد كان يحكم عمورية الروم البزنطيين، وكانت هذه المرأة تتلقى التعذيب من حاكم عمورية فصرخت باسم المعتصم الخليفة العباسيّ الأمر الذي دفعه للتوجه إلى المنطقة وتخليصها من الروم

Thank you so much!

Oh, and can you please check if this transliteration is okay?

Wa nataj an telka al-sarkhah fathun athim wa howa fath amoureiah al-wakeaa fy asia al-soghra, wa was kan yahkom amoureiah al-roum al-beezantyeen, wa kanat hathehy al-mar'ah tatalaqa al-ta'theeb men hakem amoureiah fasarakhat b'esm Al-Motasem al-Khalifa al-abasy al-amr alathy dafaahou lel-tawajoh ila al-mentakah wa takhleeseha men al-room.

The original transliteration ended with a comma, but I replaced it with a period. I hope that still makes sense.

[Qwynegold]

I'm making these language quizzes, and I'm using the YouTube channel I Love Languages as a source, because it has a lot of language samples. Here are two examples of my quizzes:

Guess the Language #1
Guess the Language #2

However, since it's YouTube I can't just copy texts I see on screen. That's a problem when there's a script I'm not familiar with. So I was wondering if someone who knows Arabic could re-type the text in https://youtu.be/XTeiW-Ps6AE, at 2:20 for me? The one that's translated as "And that cry resulted in [...] rid her of Romans".

Also, is there anyone with experience in Sumerian cuneiform, Egyptian hieroglyphics or other ancient scripts, or with any Brahmic scripts?

Well, you might be able to twist my arm, but i only know the script, not the language. Fortunately, the ambiguous transliteration provides a way of proof-reading the typing. I've transcribed my Middle Egyptian, but i haven't used the Manual shaping controls - in that respect I am at the MdeC level. For Sumerian cuneiform, you will have to specify the glyph development stage. Brahmic, I can struggle through the basics - Brahmi, Devanagari, Sinhala, Burmese and Thai. In all these cases, a proof-reading mechanism would be needed. (Well, perhaps not Thai, if the text were in an easy to read font.)

Incidentally, the Arabic text suddenly become a lot clearer - it's a lot more readable at 2:29. It was hard slog to do the first line of the transliteration/transcription: " ونتج عن تلك اصرخةفتح عظيم و هوفتح عمورية" having to use Emacs character selection by name.

I'd be grateful for any language that you would transcribe. I don't know any of the languages that I'm putting in my quizzes, so I'll just have to trust that they're correct. I'll look up samples that I'd be interesting in getting transcribed and I'll post them here, maybe later today.

[Kuchigakatai]

Oh, and can you please check if this transliteration is okay?

I'd say it's okay or not okay depending on what your purpose is.

If you mean to transcribe it exactly as it is in the video, your transcription is good.

If you ask whether it's okay as a transcription of the Arabic text and the audio recording, I'd say it's not, because of a mistake in it: it has "wa was kan yahkom" where it should have "wa qad kan yahkom". This is a mistake in the video of course. A more minor mistake is that the recording begins with [wæ ˈnætæʒæ ʕan tilkæ sˤˈsˤɑrxɑ], but [ˈnætæʒæ] appears transcribed as "nataj", not "nataja", and also has "(wa was) kan" when the audio has [ˈkæ:næ].

If you ask whether it's okay as an accurate transcription of the Arabic, I'd also say it's not, because it's written in a very ambiguous and informal romanization. The transcription was likely made by a native speaker who doesn't know of unambiguous schemes. For example, "wa howa fath amoureiah al-wakeaa fy asia al-soghra" would be more accurately "wa-huwa fatḥ ʿamūriyya l-wāqiʿ fī ʾāsiyā ṣ-ṣuġrā".

The original transliteration ended with a comma, but I replaced it with a period. I hope that still makes sense.

It's alright.

[Qwynegold]

Here are some other languages. I don't expect to get all these typed out for me, but any one of these would be nice to have. I can do e.g. Hangul myself by just clicking characters on Character Map, but the following languages have scripts with huge character sets, or complicated rules regarding combining characters or ligatures, so I don't dare to attempt them myself. I should probably look for some other source with a large collection of languages, that I can just copy-paste from. Does anyone have any suggestions?

Sumerian
https://youtu.be/igLEwLkU1HY?t=29
0:29-0:46
"Enlil, king of all the lands [...] set up a stele there."

Akkadian
https://youtu.be/7yVK0qIe1t8?t=11
0:11-0:21
"Enkidu, whom I love deeply [...] day and night."

Egyptian
https://youtu.be/MS0ZnLlsi1o?t=24
0:24
"Now I have entered [...] right and truth."

Hittite
https://youtu.be/EQAJgTPwv4M?t=16
0:16-0:21
"Only you have joined [...] directed my deeds."

Hindi
https://youtu.be/8efVP1u-ENI?t=216
3:36
"She looked up [...] bread from the crow."

Persian
https://youtu.be/vdURA4OzkEo?t=150
2:30
"Captivity of [...] existence and the world."

Tigrinya
https://youtu.be/bFAS5kTFsMk?t=199
3:19
"It was fitting to [...] and is found."

Yiddish
https://youtu.be/ir3UUop_0OU?t=58
0:58
"Everyone is entitled to [...] birth or other status."
(please exclude the words "article 2" if it's in the Hebrew script, because it's not in the transcription"

Central Tibetan
https://youtu.be/F48fqNGDGDw?t=75
1:15
"Still other seed [...] a hundred times."

Telugu
https://youtu.be/SLizVfacZ7k?t=44
0:44
"The younger one [...] property between them."

Malayalam
https://youtu.be/-SdpPJZgINI?t=54
0:54
"After he had spent [...] be in need."

[Qwynegold]

Oh, and can you please check if this transliteration is okay?

I'd say it's okay or not okay depending on what your purpose is.

If you mean to transcribe it exactly as it is in the video, your transcription is good.

If you ask whether it's okay as a transcription of the Arabic text and the audio recording, I'd say it's not, because of a mistake in it: it has "wa was kan yahkom" where it should have "wa qad kan yahkom". This is a mistake in the video of course. A more minor mistake is that the recording begins with [wæ ˈnætæʒæ ʕan tilkæ sˤˈsˤɑrxɑ], but [ˈnætæʒæ] appears transcribed as "nataj", not "nataja", and also has "(wa was) kan" when the audio has [ˈkæ:næ].

Aha. The audio doesn't really matter, because I'm only copying the text. So the romanization needs to match the Arabic text.

If you ask whether it's okay as an accurate transcription of the Arabic, I'd also say it's not, because it's written in a very ambiguous and informal romanization. The transcription was likely made by a native speaker who doesn't know of unambiguous schemes. For example, "wa howa fath amoureiah al-wakeaa fy asia al-soghra" would be more accurately "wa-huwa fatḥ ʿamūriyya l-wāqiʿ fī ʾāsiyā ṣ-ṣuġrā".

The original transliteration ended with a comma, but I replaced it with a period. I hope that still makes sense.

It's alright.

Yeah, I saw that it's not one of those transcriptions that use lots diacritics. But I guess there is no way to say which one of all possible transcriptions one should use. I think most of the quiz takers will just identify the language by the Arabic text.

[Qwynegold]

Here's the result: Guess the Language #3

[Richard W]

Hindi
https://youtu.be/8efVP1u-ENI?t=216
3:36
"She looked up [...] bread from the crow."

अचानक उसकी नचर ऊपर गई । पेड पर एक कौआ बैठा हुआ था ।
उसके मुंह में रोटी का एक टुकडा था। कौवे को देखकर लोमरी के मुँह
में पानी भर आया। वह कौवे से रोटी छीनने का उपाय सोचने लगी।

[Kuchigakatai]

Aha. The audio doesn't really matter, because I'm only copying the text. So the romanization needs to match the Arabic text.

If that's so, then all you need to do is change "wa was kan yahkom" to "wa qad kan yahkom".

Persian
https://youtu.be/vdURA4OzkEo?t=150
2:30
"Captivity of [...] existence and the world."

اسارت ضحاک، پادشاه اهریمنی و بدی ها در دل دماوند شاید اشاره ای باشه به اينکه بدی و شر در ذات بشر و دنیا وجود داره، جزئی از کل ما و هستی و جهان هست

[Richard W]

Egyptian
https://youtu.be/MS0ZnLlsi1o?t=24
0:24
"Now I have entered [...] right and truth."

I'm not quite sure you want with this. In Manuel de Codage (MdeC), the text is:

i:r-zA-q:D54-k-w-A1-r N37A:t*U30-Z9:Y1-Z3
i-w aHa-a:D54 M15-A-tp:Z1 Dd-A40 h:r-ra:Z1 n:t
U35-D40 n:N37A-n:Z4:D40 DA-A-P3-pA U4:a-Sw-t:Z4A

There's a Wikimedia tool that will render this sort of stuff, WikiHiero.

For horizontal text, the hyphens represent side-by-side placement, colons vertical placement with a higher priority, and asterisks side-by-side placement with an even tighter grouping. There are now Unicode characters to command this layout, but very few fonts and rendering systems that will implement this. I've seen them working on Andrew Glass's laptop, but I don't know how much has been released. In hybrid notation
(Unicode for the glyphs, MdeC for the layout), the text is:

𓇋:𓂋𓅭𓈎:𓂻𓎡𓅱𓀀𓂋 𓈚:𓏏*𓍔𓏴:𓏛𓏪
𓇋𓅱 𓊢𓂝:𓂻 𓇇𓄿𓁶:𓏤 𓊽𓀭 𓉔:𓂋𓇳:𓏤 𓈖:𓏏
𓍚𓂡 𓈖:𓈚𓈖:𓏭:𓂡 𓍑𓄿𓊞𓅮 𓌶:𓂝𓆄𓏏:𓏮

[Qwynegold]

Thanks you Richard W and Ser for the Hindi and Persian!

Egyptian
https://youtu.be/MS0ZnLlsi1o?t=24
0:24
"Now I have entered [...] right and truth."

I'm not quite sure you want with this.

Oh, you mean the text in the video is wrong?

For horizontal text, the hyphens represent side-by-side placement, colons vertical placement with a higher priority, and asterisks side-by-side placement with an even tighter grouping. There are now Unicode characters to command this layout, but very few fonts and rendering systems that will implement this. I've seen them working on Andrew Glass's laptop, but I don't know how much has been released. In hybrid notation
(Unicode for the glyphs, MdeC for the layout), the text is:

𓇋:𓂋𓅭𓈎:𓂻𓎡𓅱𓀀𓂋 𓈚:𓏏*𓍔𓏴:𓏛𓏪
𓇋𓅱 𓊢𓂝:𓂻 𓇇𓄿𓁶:𓏤 𓊽𓀭 𓉔:𓂋𓇳:𓏤 𓈖:𓏏
𓍚𓂡 𓈖:𓈚𓈖:𓏭:𓂡 𓍑𓄿𓊞𓅮 𓌶:𓂝𓆄𓏏:𓏮

I see. This is a problem. I'll look into if it's possible to put in an image instead of text.

[Richard W]

I see. This is a problem. I'll look into if it's possible to put in an image instead of text.

If the law is the issue, one can use Wikihiero to generate HTML to use images available under the GNU Free documentation License. You don't need to grab the original license. The WikiHiero glyphs are monochrome - so were Andrew's. You'll need to tweak the image paths in the generated code HTML segment. The grey-scaling results in images that scale very well. Having seen the principle, I wrote an HTML generator of my own that's coded in bash and converts MdeC notation to HTML.

[Linguoboy]

Yiddish:

יעדער איז באַרעכטיקט אױף די אַלע רעכט און פֿרײַהײטן װאָס שטײען געשריבן אין אָט דער דעקלאַראַציע, ניט געקוקט אױף װאָסער־ניט־איז אונטערשײד, זאָל זײַן פֿאָלק, הױטפֿאַרב, לשון, רעליגיע, פּאָליטישע אָדער אַנדערע געדאַנקען, נאַציאָנאַליטעט, קולטור, פֿאַרמאָג, געבורט, אָדער װאָסער־ניט־איז אַנדערן סטאַטוס.

Source: https://unicode.org/udhr/d/udhr_ydd.html.

[bradrn]

Yiddish:

יעדער איז באַרעכטיקט אױף די אַלע רעכט און פֿרײַהײטן װאָס שטײען געשריבן אין אָט דער דעקלאַראַציע, ניט געקוקט אױף װאָסער־ניט־איז אונטערשײד, זאָל זײַן פֿאָלק, הױטפֿאַרב, לשון, רעליגיע, פּאָליטישע אָדער אַנדערע געדאַנקען, נאַציאָנאַליטעט, קולטור, פֿאַרמאָג, געבורט, אָדער װאָסער־ניט־איז אַנדערן סטאַטוס.

Source: https://unicode.org/udhr/d/udhr_ydd.html.

Well, I don’t speak Yiddish, but I can read Hebrew (though I don’t understand it), so here’s my attempt at a transcription:

Yeder iz barekhtikt oyf di ale rekht un frayheytn vos shteyen geshribn in ot der deklaratsye, nit gekukt oyf voser-nit-iz untersheyd, zol zayn folk, hoytfarb, lshun, religye, politishe oder andere gedanken, natsyonalitet, kultur, farmog, geburt, oder voser-nit-iz andern status.

[Qwynegold]

Here are my latest quizzes that I've been able to create thanks to you guys:
Guess the Language #4
Guess the Language #5

[bradrn]

Here are my latest quizzes that I've been able to create thanks to you guys:
Guess the Language #4
Guess the Language #5

I tried these and managed to get 6/10 for both of them. Obviously I’m not too good at this!

[Qwynegold]

I see. This is a problem. I'll look into if it's possible to put in an image instead of text.

If the law is the issue, one can use Wikihiero to generate HTML to use images available under the GNU Free documentation License. You don't need to grab the original license. The WikiHiero glyphs are monochrome - so were Andrew's. You'll need to tweak the image paths in the generated code HTML segment. The grey-scaling results in images that scale very well. Having seen the principle, I wrote an HTML generator of my own that's coded in bash and converts MdeC notation to HTML.

No it has to do with the quiz creator. When it comes to hieroglyphs and other uncommon scripts, I have the following two choices.

1. Make a text quiz. Images may not be inserted into text quizzes, so everything has to be actual text. So this is a problem for those scripts that are not simply linear and horizontal. Even simpler scripts have the problem that they may not show up if the viewer doesn't have the right font installed. For example, I've heard someone complain that they couldn't see Burmese text, and that's not a very obscure script.
2. Make a picture quiz. The problem with this is that all "questions" have to be images. A question may not consist of text. And even more importantly, you can't just upload any pictures. You have to choose from the images available at either Wikimedia Commons or Pixabay. And I would require both the foreign text and a romanization of it to be present on the same image. So this is not feasible.

I have thought about a third option though. On a text quiz you may upload one SVG image that goes above the whole answers row. This is basically meant for map quizzes where different indicated areas of a map change color when you guess the names of those places. But I think it would be possible to just upload a non-interactive .svg image. Then I would write something like: <see image above> as the hint for Egyptian.

[Qwynegold]

Here are my latest quizzes that I've been able to create thanks to you guys:
Guess the Language #4
Guess the Language #5

I tried these and managed to get 6/10 for both of them. Obviously I’m not too good at this!

Thanks for trying! In each quiz I have five easy and five hard questions. But I can see from the statistics that people get fewer than five correct answers on average. Someone didn't even recognize French!