Miscellaneous desk
< September 3	<< Aug \| September \| Oct >>	September 5 >

Welcome to the Wikipedia Miscellaneous Reference Desk Archives
The page you are currently viewing is a transcluded archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

September 4

Why don’t numerals look, well, numeric?

Why is it that, in most writing systems (including our own), the numerals do not look, well, numeric? Instead, they look more or less like letters that didn’t make the alphabet. This is not inevitable: look at Suzhou numerals for a system that at least ‘’tries’’ to look like numerals. In case you are wondering what provoked this question: I recently experienced semantic satiation from having spent way too much time on designing a digital clock display. — Preceding unsigned comment added by 2602:252:D14:F900:7869:BF91:694E:8EE4 (talk) 00:45, 4 September 2020 (UTC)[reply]

I'd say the numerals in our writing system do look numeric. "1" certainly does. And I think the others do as well, but I would guess there are stylizations involved that facilitate the writing process. Bus stop (talk) 01:03, 4 September 2020 (UTC)[reply]

If numerals "look like letters that didn’t make the alphabet", a natural explanation is that the digits and letters were written with the same technology, so the same sort of shapes were easy to make. --174.88.168.23 (talk) 01:26, 4 September 2020 (UTC)[reply]

They don't look enough like letters, in my opinion. I have dabbled a bit in calligraphy; it's surprisingly hard to make digits harmonize well with letters if you go far outside the roman world. This is probably akin to the problem that it's very easy for a sloppy hand to convert one digit into nearly any other (as I found when I had a temp job transcribing voter registration forms). —Tamfang (talk) 01:12, 8 September 2020 (UTC)[reply]

They did look numeric to the author of "The Arabic numerals representative and not arbitrary characters". Compare also the Hindu–Arabic numerals in various Indic scripts, which were a precursor of "our" Arabic numerals. --Lambiam 01:34, 4 September 2020 (UTC)[reply]

Agreed. Brahmi numerals needs sourcing, but indicates that they may have had roots based on letters in the Brahmi alphabet, in other words, that they may have begun as written abbreviations for numbers rather than being constructed based on the number they represent: they "appear to be arbitrary symbols in even the oldest inscriptions". I hadn't seen Suzhou numerals before: that's really interesting. From a legibility and writing-time standpoint: I think that if you do a writing system based on strokes you end up with a nine-stroke number that's gigantic and hard to draw or fit in available space unambiguously compared to a '1'. (Try drawing a character with nine strokes. It'll take up a lot of space compared to a 1 or 2 unless you're practiced.) Basically, it's a really big problem if your accountant wants you to owe ten gold pieces rather than nine because that's easier to write in a hurry!
Also, if everything is horizontal or vertical bars, you end up being unsure how many numbers you're looking at without very obvious dividers: when I see IIIIII am I looking at 2 then 4, 3 then 3 or what? And it's hard to quickly tell the difference between (say) IIIII and IIII. My guess, then, is that 1 and 2 you can represent easily with something like I and II, but by the time you get to five you're going to need to create an abstract symbol that represents it rather than have a numeral with five strokes. And conveniently, that's the point that you need to change hands, so there's an obvious logic to having a new symbol at that point.
To my eyes, Suzhou numerals feel conceptually very similar to Roman numerals, except more calligraphic: there's the same concept that at 5 you need a new symbol and start counting again, so six is 5+1 and seven is 5+1. The change in orientation in characters from 1,2,3 to 6,7,8 is a neat legibility-increasing idea the Romans didn't have, though. Blythwood (talk) 07:49, 4 September 2020 (UTC)[reply]

I expect the OP would consider binary to be sufficiently "numeric-looking"? ←Baseball Bugs ^{What's up, Doc?} carrots→ 11:00, 4 September 2020 (UTC)[reply]

Maybe! It's a topic that's interesting to me. Until the nineteenth century all numbers in print were "text figures", styled to look like letters, but lower-case letters: they have ascenders and descenders. You had to use Roman numerals if you wanted numbers of uniform height that would line up with capital letters. James Mosley, the leading historian of printing, has said that in his view the fat face typefaces of the early nineteenth century are the first to have numerals that really go well with capital letters, making them work really well for commercial advertising and posters. For a period around 1800 fonts came with "three-quarter height" figures, that are slightly smaller than capital letters (here's a modern version of one of these).
Designing numbers is still a challenge to font and publication designers, and pro digital fonts can come with a lot of varieties to ensure balance and an even colour of text on a page: text-figures, small cap, three-quarter lining (if you don't want the numbers to seem big and unbalanced on a page of text with lots of numbers on it), fully up to cap height for when you need alphanumeric codes with numbers and capitals (e.g. UK postcodes, which look like GL15 0EX). And spacing is important too, a lot of fonts come with tabular figures, which line up in columns neatly, and proportional ones, with a "natural width", that look good in body text). (Mosley pointed to Adobe Trajan as a font with numbers that in his view weren't balanced too well with the upper case-I see his point, they do look a bit narrow.) Blythwood (talk) 12:41, 4 September 2020 (UTC)[reply]

Didn't all of we commonly call "Latin numerals" come from Arabic (Western) numerals, which came from Hindu numerals, which in turn came from Brahmi? 97.82.165.112 (talk) 20:02, 6 September 2020 (UTC)[reply]

What I call Latin numerals are words like duodequinquagesimus. --Lambiam 06:55, 7 September 2020 (UTC)[reply]

Where's the upvote button? —Tamfang (talk) 01:12, 8 September 2020 (UTC)[reply]

What do we commonly call "Latin numerals"?? —Tamfang (talk) 01:12, 8 September 2020 (UTC)[reply]

Arabic numerals or Numerical digit. Graeme Bartlett (talk) 01:38, 8 September 2020 (UTC)[reply]

To make everyone happy: Isn't it true that 1, 2, 3, 4... all come from Western Arabic numerals, which came from Hindu numerals, which in turn came from Brahmi? 97.82.165.112 (talk) 17:53, 8 September 2020 (UTC)[reply]

I, II, III, IV...? 93.136.121.193 (talk) 06:20, 10 September 2020 (UTC)[reply]

For a literal case of letters that didn't make the alphabet I suggest you look at Greek numerals and Glagolitic numerals. 93.136.121.193 (talk) 06:20, 10 September 2020 (UTC)[reply]

How to get lipstick out of a wool jersey

The jersey is knitted from pure wool. Which commonly available household solvent or surfactant would effectively and cleanly remove the lipstick? Scrubbing is clearly not an option. Roger (Dodger67) (talk) 06:22, 4 September 2020 (UTC)[reply]

If there is a considerable amount, I'd start by putting the item in the freezer so that the lipstick hardens and gets brittle, then remove what you can mechanically, carefully but quickly before it gets buttery again and will easily spread. After that, it may actually help subsequent removal steps to make the lipstick softer, for example by heating it for just a few seconds in the microwave. If it is soft, make sure not to rub the stain, but only to dab it. There is an extensive WikiHow on the topic, and also advice from Cleanipedia. If rubbing alcohol is safe for (the colours of) the fabric – try this out with a Q-tip on the inside of a rim, it would be my first choice for a solvent, since it is not aggressive and does not itself leave a stain. Both web pages linked to also suggest hairspray as an alternative. If the lipstick tube is around, you can first stain an expendable piece of fabric and test which available solvents give good results. --Lambiam 08:46, 4 September 2020 (UTC)[reply]

If you have a decent professional cleaners, they may be your best option. Professional clothes cleaners may have access to cleaning materials and techniques, and have the necessary expertise, in getting out difficult stains You may want to consult them. --Jayron 32 16:24, 4 September 2020 (UTC)[reply]