WEBVTT 00:00:00.000 --> 00:00:02.000 I want you to picture the keyboard in front of 00:00:02.000 --> 00:00:03.980 you right now. Specifically, just look at the 00:00:03.980 --> 00:00:05.740 number five key. Oh, yeah. Right in the middle 00:00:05.740 --> 00:00:08.759 there. Exactly. Resting right there, just above 00:00:08.759 --> 00:00:12.039 that digit, is this symbol that you probably 00:00:12.039 --> 00:00:15.460 see, I don't know, every single day. Oh, without 00:00:15.460 --> 00:00:17.839 a doubt. Right. You use it in work emails. You 00:00:17.839 --> 00:00:21.199 rely on it to understand retail sales, your phone's 00:00:21.199 --> 00:00:23.699 battery life, sports statistics. I'm talking, 00:00:23.719 --> 00:00:25.980 of course, about the percent sign. It is just 00:00:25.980 --> 00:00:28.780 a... entirely ubiquitous. I mean, it's one of 00:00:28.780 --> 00:00:31.320 those rare symbols that effortlessly crosses 00:00:31.320 --> 00:00:34.079 the boundary between complex mathematics and 00:00:34.079 --> 00:00:36.119 just, you know, everyday casual conversation. 00:00:36.600 --> 00:00:38.880 But, and this is the crazy part, that mundane, 00:00:39.200 --> 00:00:41.259 everyday fraction of a hundred symbol, it is 00:00:41.259 --> 00:00:44.100 hiding a completely secret history. Oh, it really 00:00:44.100 --> 00:00:46.000 is. It's got a double, maybe even triple life. 00:00:46.119 --> 00:00:48.939 Yeah. So today, we're going to take you on a 00:00:48.939 --> 00:00:51.679 deep dive into the surprisingly chaotic life 00:00:51.679 --> 00:00:54.119 of the percent sign because we're talking about, 00:00:54.119 --> 00:00:57.399 like, lazy medieval scribes. You're a lazy stranger. 00:00:58.000 --> 00:01:00.799 Pierce international turf wars over typography 00:01:00.799 --> 00:01:03.960 and completely alternate identities in the underground 00:01:03.960 --> 00:01:07.060 worlds of computer science and linguistics. It's 00:01:07.060 --> 00:01:10.159 wild. It goes so far beyond just math. It does. 00:01:10.280 --> 00:01:12.400 And I promise you, by the time we finish this 00:01:12.400 --> 00:01:14.500 deep dive today, you are never going to look 00:01:14.500 --> 00:01:16.959 at your keyboard the same way again. And as a 00:01:16.959 --> 00:01:19.959 bonus, you're going to have the absolute ultimate 00:01:19.959 --> 00:01:22.510 trivia. in your back pocket for your next, you 00:01:22.510 --> 00:01:25.230 know, boring budget meeting. Always good to have 00:01:25.230 --> 00:01:27.810 that ready to go. Right. But before we get into 00:01:27.810 --> 00:01:29.590 the scribes and the hackers, we should probably 00:01:29.590 --> 00:01:33.329 set the baseline. What exactly is the core function 00:01:33.329 --> 00:01:35.950 of the percent sign? Well, at its most fundamental 00:01:35.950 --> 00:01:39.109 level, it's simply a symbol used to indicate 00:01:39.109 --> 00:01:42.950 a percentage, which, you know, is a number or 00:01:42.950 --> 00:01:45.409 a ratio expressed as a fraction of 100. Right. 00:01:45.569 --> 00:01:49.409 Out of 100. Exactly. But it actually has these 00:01:49.409 --> 00:01:51.489 mathematical siblings that operate on the exact 00:01:51.489 --> 00:01:54.310 same visual logic. Wait, really? Siblings? Yeah. 00:01:54.489 --> 00:01:56.250 So, for example, there's the per mille sign. 00:01:56.530 --> 00:01:58.450 It looks almost exactly like a percent sign, 00:01:58.730 --> 00:02:00.890 but it has two little zeros at the bottom right 00:02:00.890 --> 00:02:03.090 instead of one. Oh, I think I've seen that. Yeah. 00:02:03.189 --> 00:02:05.629 It indicates a fraction per thousand. And then 00:02:05.629 --> 00:02:07.530 if you want to go even further, there's the per 00:02:07.530 --> 00:02:10.189 myriad. Per myriad? Yeah. Or sometimes it's called 00:02:10.189 --> 00:02:12.750 a basis point. And that one has three zeros at 00:02:12.750 --> 00:02:15.389 the bottom. which indicates parts per 10 ,000. 00:02:15.550 --> 00:02:18.349 OK, let's unpack this. Because to really understand 00:02:18.349 --> 00:02:21.750 how the standard percent sign became this global 00:02:21.750 --> 00:02:24.469 heavyweight, we actually have to travel back 00:02:24.469 --> 00:02:27.270 to a time before it even existed. Right, way 00:02:27.270 --> 00:02:30.370 back. And this is where the story gets incredibly 00:02:30.370 --> 00:02:34.759 human. Because this symbol wasn't mathematically 00:02:34.759 --> 00:02:37.879 engineered in some pristine laboratory by scholars 00:02:37.879 --> 00:02:40.159 trying to perfect a universal language. Not at 00:02:40.159 --> 00:02:42.919 all. It was basically a happy accident born out 00:02:42.919 --> 00:02:46.400 of human laziness. Well, laziness or, you know, 00:02:46.560 --> 00:02:48.020 perhaps we should call it extreme efficiency. 00:02:48.219 --> 00:02:50.400 Fair enough. Because if you look at historical 00:02:50.400 --> 00:02:54.770 texts prior to the year 1425, There is absolutely 00:02:54.770 --> 00:02:58.110 no known evidence of a special dedicated symbol 00:02:58.110 --> 00:03:00.090 being used for percentages. So what did they 00:03:00.090 --> 00:03:02.590 do? When people needed to express a fraction 00:03:02.590 --> 00:03:05.030 of 100, they just wrote it out. They used the 00:03:05.030 --> 00:03:07.909 Italian term pretento, which literally translates 00:03:07.909 --> 00:03:11.969 to for 100. Just the full words. Yeah. Or if 00:03:11.969 --> 00:03:14.330 they were in a rush, they would use abbreviations. 00:03:14.349 --> 00:03:17.909 So you'd see things written out as per 100, P100, 00:03:18.090 --> 00:03:20.909 or eccento. Which, I mean, that... Completely 00:03:20.909 --> 00:03:22.490 makes sense if you think about the technology 00:03:22.490 --> 00:03:25.270 of the time. You are writing with a quill. Parchment 00:03:25.270 --> 00:03:28.569 and vellum are incredibly expensive. And every 00:03:28.569 --> 00:03:31.169 drop of ink matters. You want to save space and 00:03:31.169 --> 00:03:33.930 time wherever you can. Precisely. And we can 00:03:33.930 --> 00:03:36.270 actually track the visual evolution of this shorthand, 00:03:36.310 --> 00:03:38.750 which is so cool. There's this specific anonymous 00:03:38.750 --> 00:03:42.830 arithmetic text from 1339. OK. And in this text, 00:03:42.990 --> 00:03:46.460 you see the letter P. written with its descender, 00:03:46.800 --> 00:03:48.620 you know, the long tail of the pea that hangs 00:03:48.620 --> 00:03:50.639 down below the line. Yeah, yeah. It's crossed 00:03:50.639 --> 00:03:53.280 by a horizontal or diagonal stroke. Wait, but 00:03:53.280 --> 00:03:55.680 a pea with a line through the tail doesn't look 00:03:55.680 --> 00:03:58.219 anything like the two circles in a slash that 00:03:58.219 --> 00:04:00.000 we use today. That sounds more like a generic 00:04:00.000 --> 00:04:01.719 proofreaders mark or something. Yeah, well that's 00:04:01.719 --> 00:04:03.620 because it was just a standard abbreviation. 00:04:04.259 --> 00:04:06.740 In medieval and Renaissance paleography, that 00:04:06.740 --> 00:04:08.919 cross pea conventionally stood for words like 00:04:08.919 --> 00:04:12.319 purr, pour, par, or pour. Ah, I got it. So it 00:04:12.319 --> 00:04:15.000 was a generic shorthand, not a unique math symbol. 00:04:15.520 --> 00:04:18.120 Exactly. The real breakthrough, like the moment 00:04:18.120 --> 00:04:20.459 the percent sign, as we know it actually begins 00:04:20.459 --> 00:04:23.060 to emerge, happens a little later. When was this? 00:04:23.240 --> 00:04:27.100 So in a 1425 arithmetic text, well, specifically 00:04:27.100 --> 00:04:29.120 in some pages that historians believe were added 00:04:29.120 --> 00:04:34.079 around 1435, a scribe used the abbreviation PC. 00:04:34.319 --> 00:04:37.360 Uh -huh. But they added a tiny little loop. or 00:04:37.360 --> 00:04:39.500 circle floating right next to the sea. Wait, 00:04:39.560 --> 00:04:41.220 what was the point of the little loop? Was it 00:04:41.220 --> 00:04:43.839 just like a decorative flourish? No, not at all. 00:04:43.899 --> 00:04:46.040 It actually served a highly specific grammatical 00:04:46.040 --> 00:04:49.759 purpose. That tiny loop depicted the O ending 00:04:49.759 --> 00:04:53.120 used in Italian ordinal numbers. Oh, like primo 00:04:53.120 --> 00:04:55.839 or secundo? Exactly. Think of it exactly like 00:04:55.839 --> 00:04:57.699 how we write the O at the end of the English 00:04:57.699 --> 00:05:00.540 25th. It was a way of writing the concept of 00:05:00.540 --> 00:05:03.639 percent O in this incredibly compressed, visually 00:05:03.639 --> 00:05:06.209 distinct format. Oh, wow. So we essentially have 00:05:06.209 --> 00:05:08.449 a P and a C with a little O floating around it. 00:05:08.509 --> 00:05:11.250 I mean, this is basically a 15th century emoji. 00:05:12.120 --> 00:05:14.220 Yes, that's a great way to look at it. It starts 00:05:14.220 --> 00:05:16.740 out as full words, percento, then it gets abbreviated 00:05:16.740 --> 00:05:19.120 to save time, exactly like modern texting, where 00:05:19.120 --> 00:05:22.319 be right back becomes verb. And then over centuries, 00:05:22.439 --> 00:05:25.639 it just morphs from letters into a purely visual 00:05:25.639 --> 00:05:27.699 icon that doesn't even look like text anymore. 00:05:27.920 --> 00:05:30.319 Right. What's fascinating here is that a rigid, 00:05:30.920 --> 00:05:33.379 universally recognized mathematical symbol was 00:05:33.379 --> 00:05:37.019 actually born from messy, hurried human handwriting. 00:05:37.079 --> 00:05:39.500 It really is. Because the evolution didn't stop 00:05:39.500 --> 00:05:43.579 with that Italian scribe. By the year 1650, that 00:05:43.579 --> 00:05:46.980 PC with the little loop had evolved into a horizontal 00:05:46.980 --> 00:05:50.300 fraction sign and it completely dropped the per 00:05:50.300 --> 00:05:52.680 part. Wait, really? Yeah, it just became two 00:05:52.680 --> 00:05:55.279 tiny circles separated by a horizontal line. 00:05:55.800 --> 00:05:58.199 The letters had literally melted away into geometric 00:05:58.199 --> 00:06:00.579 shapes. Wait, what about the slanted line? Because 00:06:00.579 --> 00:06:01.980 I'm looking at my keyboard right now and the 00:06:01.980 --> 00:06:04.699 line between the two circles is definitely slanted, 00:06:04.839 --> 00:06:07.720 not horizontal. Right, so that slanted line is 00:06:07.720 --> 00:06:09.920 called a solidus and it is a surprisingly recent 00:06:09.920 --> 00:06:12.769 invention. Historians like D .E. Smith back in 00:06:12.769 --> 00:06:15.889 1925 noted that the solidest form of the percent 00:06:15.889 --> 00:06:18.810 sign is definitively modern. It took hundreds 00:06:18.810 --> 00:06:21.230 of years for that horizontal line to tilt. It 00:06:21.230 --> 00:06:23.610 is so strange to think about it that way. We 00:06:23.610 --> 00:06:26.689 treat the percent sign like this ancient immutable 00:06:26.689 --> 00:06:29.769 law of math, but it's really just the sloppy 00:06:29.769 --> 00:06:32.689 handwriting of an Italian scribe that got passed 00:06:32.689 --> 00:06:35.310 down and slowly distorted over centuries. It's 00:06:35.310 --> 00:06:37.790 almost like a game of typographical telephone. 00:06:37.990 --> 00:06:41.860 That's a perfect analogy. And that organic grassroots 00:06:41.860 --> 00:06:45.720 evolution is exactly why the symbol has caused 00:06:45.720 --> 00:06:48.800 so much friction in the modern era. Oh, the turf 00:06:48.800 --> 00:06:52.040 wars. Yes, because it evolved naturally across 00:06:52.040 --> 00:06:54.629 Europe. rather than being handed down by a central 00:06:54.629 --> 00:06:57.449 authority with an instruction manual, it created 00:06:57.449 --> 00:06:59.589 this massive global headache when it came to 00:06:59.589 --> 00:07:01.370 printing it. Right, because once the printing 00:07:01.370 --> 00:07:03.829 press arrives and eventually, you know, the typewriter, 00:07:03.970 --> 00:07:05.970 things have to be standardized. Exactly. But 00:07:05.970 --> 00:07:07.769 nobody could agree on what the standard should 00:07:07.769 --> 00:07:10.910 actually be. It turns into an absolute typography 00:07:10.910 --> 00:07:13.550 turf war. It really is a turf war. We have a 00:07:13.550 --> 00:07:15.649 fractured global landscape today where different 00:07:15.649 --> 00:07:17.689 languages and even different institutional style 00:07:17.689 --> 00:07:20.550 guides within the same language completely disagree 00:07:20.550 --> 00:07:23.310 on how to physically place the symbol next to 00:07:23.310 --> 00:07:25.730 a number. I mean look at the conflict just within 00:07:25.730 --> 00:07:28.970 the English language. Oh it's intense. On one 00:07:28.970 --> 00:07:31.889 side you have the classic English style guides 00:07:31.889 --> 00:07:34.769 like the Chicago Manual of Style and the APA 00:07:34.769 --> 00:07:38.269 Manual. They explicitly prescribe writing the 00:07:38.269 --> 00:07:40.670 percent sign following the number with absolutely 00:07:40.670 --> 00:07:43.670 no space between them. So 50, then immediately 00:07:43.670 --> 00:07:45.910 the percent sign. Right. But on the opposing 00:07:45.910 --> 00:07:48.110 side, you have the international heavyweights, 00:07:48.449 --> 00:07:50.550 the International System of Units, the SI, and 00:07:50.550 --> 00:07:54.470 the ISO 310 standard. They strongly prescribe 00:07:54.470 --> 00:07:56.769 placing a space between the number and the percent 00:07:56.769 --> 00:08:00.230 sign. So 50. space percent. But why? I mean, 00:08:00.230 --> 00:08:02.470 if everyone in the U .S. and the U .K. naturally 00:08:02.470 --> 00:08:04.790 writes it without a space, why are the international 00:08:04.790 --> 00:08:06.730 standards pushing for a gap? Well, because the 00:08:06.730 --> 00:08:08.930 international standards treat the percent sign 00:08:08.930 --> 00:08:11.350 in line with the general practice of using a 00:08:11.350 --> 00:08:13.850 non -breaking space between any numerical value 00:08:13.850 --> 00:08:16.149 and its corresponding unit of measurement. Oh, 00:08:16.149 --> 00:08:18.449 I see. Yeah. Just like you would write 50 kilometers 00:08:18.449 --> 00:08:20.089 with a space between the 50 and the kilometers, 00:08:20.350 --> 00:08:22.990 they argue you must write 50 percent with the 00:08:22.990 --> 00:08:26.050 space. OK, that makes sense. And they specifically 00:08:26.050 --> 00:08:30.060 mandate a non -breaking space, which is basically 00:08:30.060 --> 00:08:34.080 a piece of typographical glue. It tells the computer, 00:08:34.440 --> 00:08:37.279 never put the 50 at the end of one line of text 00:08:37.279 --> 00:08:39.360 and the percent sign at the beginning of the 00:08:39.360 --> 00:08:41.700 next line. It keeps them visually bound together. 00:08:42.000 --> 00:08:43.720 OK, I see the logic there. They want to treat 00:08:43.720 --> 00:08:46.299 it like a scientific unit. But the rules get 00:08:46.299 --> 00:08:48.460 even more complicated when you look at actual 00:08:48.460 --> 00:08:51.639 text versus tables, right? Oh, yeah. If you are 00:08:51.639 --> 00:08:53.879 writing for a newspaper, for instance, a lot 00:08:53.879 --> 00:08:56.080 of style guides say you shouldn't even use the 00:08:56.080 --> 00:08:58.580 symbol in running text. You should spell it out 00:08:58.580 --> 00:09:01.960 entirely as the word percent and strictly reserve 00:09:01.960 --> 00:09:04.580 the actual symbol for data tables and places 00:09:04.580 --> 00:09:07.000 where physical space is severely restricted. 00:09:07.340 --> 00:09:09.919 True. And if you think English is tangled, the 00:09:09.919 --> 00:09:12.360 rules in other languages show just how localized 00:09:12.360 --> 00:09:14.980 this symbol has become. Oh, do tell. Take Czech 00:09:14.980 --> 00:09:17.480 and Slovak, for instance. In Slovak, you always 00:09:17.480 --> 00:09:19.519 use a non -breaking space before the percent 00:09:19.519 --> 00:09:23.259 sign. But in Czech, the spacing depends entirely 00:09:23.259 --> 00:09:26.419 on the grammar of the sentence. Yes, if the number 00:09:26.419 --> 00:09:29.720 is used as a noun, you put a space. But if it's 00:09:29.720 --> 00:09:32.759 used as an adjective, like saying a 50 % increase, 00:09:33.600 --> 00:09:37.299 Czech drops the space completely. That is incredibly 00:09:37.299 --> 00:09:39.820 specific. The spacing changes based on the part 00:09:39.820 --> 00:09:41.700 of speech. It does. And then you have Finnish, 00:09:41.820 --> 00:09:43.840 where they always put a space. But if they need 00:09:43.840 --> 00:09:46.500 to attach a case suffix, they actually glue it 00:09:46.500 --> 00:09:49.330 to the percent sign. using a colon. Wait, they 00:09:49.330 --> 00:09:52.009 glue it with a colon? Yeah, so an increase of 00:09:52.009 --> 00:09:54.750 50 % visually looks like 50 space percent sign, 00:09:55.149 --> 00:09:58.590 colon, n space, cosfu. Looks like a math equation 00:09:58.590 --> 00:10:02.070 mixed with a typo. That is wild. It is. And the 00:10:02.070 --> 00:10:04.009 placement of the symbol changes based on geography, 00:10:04.350 --> 00:10:07.070 too. In Turkish, and some other Turkic languages, 00:10:07.549 --> 00:10:09.730 the percent sign actually precedes the number, 00:10:10.029 --> 00:10:12.809 not follows it with no space. And in Persian 00:10:12.809 --> 00:10:14.549 texts, you might see the percent sign place either 00:10:14.549 --> 00:10:17.519 before or after the number. I imagine directionality 00:10:17.519 --> 00:10:19.740 also plays a massive role there. Like Arabic 00:10:19.740 --> 00:10:21.759 is written from right to left, so the percent 00:10:21.759 --> 00:10:23.679 sign naturally follows the number which places 00:10:23.679 --> 00:10:26.500 it to the left of the digits. But Hebrew, which 00:10:26.500 --> 00:10:29.019 is also written from right to left, actually 00:10:29.019 --> 00:10:31.720 writes its numbers from left to right. So the 00:10:31.720 --> 00:10:33.759 percent sign stays on the right side of the number 00:10:33.759 --> 00:10:36.440 exactly like it does in English. It just highlights 00:10:36.440 --> 00:10:39.019 how incredibly adaptable the symbol has to be 00:10:39.019 --> 00:10:41.340 to fit into completely different writing systems. 00:10:41.919 --> 00:10:43.919 But what really stands out in all these typography 00:10:43.919 --> 00:10:46.440 rules is the human rebellion against them. Yes. 00:10:46.679 --> 00:10:49.419 Oh, I love this part. If you are someone who 00:10:49.419 --> 00:10:52.299 learns by watching how humans actually behave 00:10:52.299 --> 00:10:54.580 rather than what rule books say, the percent 00:10:54.580 --> 00:10:56.960 sign is the perfect case study. Look at the Dutch. 00:10:57.220 --> 00:10:59.250 The Dutch are a great example. Their official 00:10:59.250 --> 00:11:01.909 national rule says you must place a space between 00:11:01.909 --> 00:11:04.850 the number and the sign. But most everyday people 00:11:04.850 --> 00:11:06.990 completely ignore the rule and leave the space 00:11:06.990 --> 00:11:09.610 out. Exactly. The same phenomenon happens in 00:11:09.610 --> 00:11:13.110 North American Spanish. The real Academia Española, 00:11:13.629 --> 00:11:15.690 the royal institution that oversees the Spanish 00:11:15.690 --> 00:11:18.169 language, says the percent sign should be spaced. 00:11:18.230 --> 00:11:21.830 Right. But in Mexico and the U .S., several style 00:11:21.830 --> 00:11:24.909 guides and institutions just ignore Majid. They 00:11:24.909 --> 00:11:26.909 follow local common usage and write it with no 00:11:26.909 --> 00:11:30.830 space. So aren't these rigid typographic institutions 00:11:30.830 --> 00:11:34.110 just fighting a losing battle against common 00:11:34.110 --> 00:11:36.870 sense? I mean, if everyday people hate the space 00:11:36.870 --> 00:11:39.610 so much that they refuse to use it, why do the 00:11:39.610 --> 00:11:42.009 rule makers keep pushing it? Well, it's the classic 00:11:42.009 --> 00:11:45.110 tension between prescriptivism, you know, how 00:11:45.110 --> 00:11:47.250 things should be, and descriptivism, how things 00:11:47.250 --> 00:11:50.190 actually are. Typography is fundamentally driven 00:11:50.190 --> 00:11:54.330 by utility. The rule makers, like the ISO, prioritize 00:11:54.330 --> 00:11:57.639 logical consistency. To them, treating the percent 00:11:57.639 --> 00:11:59.960 sign exactly like a unit of measurement maintains 00:11:59.960 --> 00:12:02.100 a neat, ordered system. Makes sense for them. 00:12:02.440 --> 00:12:04.679 But human beings prioritize visual speed and 00:12:04.679 --> 00:12:07.340 aesthetic convenience. A space before a percent 00:12:07.340 --> 00:12:09.460 sign often just looks wrong to the modern eye. 00:12:09.580 --> 00:12:12.299 It feels disconnected from the number it modifies. 00:12:12.539 --> 00:12:15.009 So people simply discard the rule. human convenience 00:12:15.009 --> 00:12:17.250 will always override institutional rules. We 00:12:17.250 --> 00:12:19.370 vote with our keyboards. Exactly. And speaking 00:12:19.370 --> 00:12:21.549 of keyboards, this tension over where to place 00:12:21.549 --> 00:12:23.809 the symbol was ultimately hard -coded into our 00:12:23.809 --> 00:12:25.889 lives when the typewriter forced the percent 00:12:25.889 --> 00:12:28.730 sign into a permanent physical key. And once 00:12:28.730 --> 00:12:30.629 it was trapped on that keyboard, a completely 00:12:30.629 --> 00:12:32.750 different group of people decided to hijack it. 00:12:33.090 --> 00:12:35.450 Right. Early computer programmers looked down 00:12:35.450 --> 00:12:38.710 at their keyboards and saw a perfectly good, 00:12:39.029 --> 00:12:41.409 highly recognizable character sitting right there 00:12:41.409 --> 00:12:44.240 on the number five key. just waiting to be used. 00:12:44.799 --> 00:12:47.500 Exactly. And because a percent sign is mathematically 00:12:47.500 --> 00:12:49.759 meaningless to a computer's core binary logic, 00:12:50.500 --> 00:12:53.480 it was essentially a blank canvas. Programmers 00:12:53.480 --> 00:12:55.779 realized they could repurpose it for operations 00:12:55.779 --> 00:12:58.620 that have absolutely nothing to do with fractions 00:12:58.620 --> 00:13:01.769 of 100. And they gave it some amazing nicknames 00:13:01.769 --> 00:13:04.230 along the way. Like, in hacker jargon, the percent 00:13:04.230 --> 00:13:06.970 sign is sometimes called grapes, which, if you 00:13:06.970 --> 00:13:08.990 look at the symbol, perfectly makes sense. Two 00:13:08.990 --> 00:13:11.350 little round grapes hanging on a vine. That's 00:13:11.350 --> 00:13:13.509 very visual, yeah. And in the programming language, 00:13:13.769 --> 00:13:17.289 intercal, they humorously refer to it as 007. 00:13:17.409 --> 00:13:20.009 Very fitting for a symbol with a secret identity. 00:13:20.509 --> 00:13:23.149 Its roles in computing are vast, but let's look 00:13:23.149 --> 00:13:26.330 at how it mechanically functions. In math and 00:13:26.330 --> 00:13:28.659 logic, In programming languages derived from 00:13:28.659 --> 00:13:31.740 C and B, the percent sign is used for the modulo 00:13:31.740 --> 00:13:34.039 operation. Which is basically a fancy way of 00:13:34.039 --> 00:13:37.000 saying find the remainder, right? Exactly. If 00:13:37.000 --> 00:13:40.240 you divide 10 by 3, the answer is 3 with a remainder 00:13:40.240 --> 00:13:43.539 of 1. So if you write 10 percent sign 3 in code, 00:13:44.039 --> 00:13:46.720 the computer calculates the division, but only 00:13:46.720 --> 00:13:49.679 hands you back that remainder, the 1. OK, got 00:13:49.679 --> 00:13:51.360 it. It's a completely different mathematical 00:13:51.360 --> 00:13:53.580 function than a percentage. Here's where it gets 00:13:53.580 --> 00:13:56.379 really interesting. Because the percent sign 00:13:56.379 --> 00:13:59.440 in computing isn't just doing math, it acts more 00:13:59.440 --> 00:14:02.220 like a digital traffic cop. And its behavior 00:14:02.220 --> 00:14:04.960 totally changes depending on what neighborhood 00:14:04.960 --> 00:14:07.000 of the computer it's standing in. Oh, that is 00:14:07.000 --> 00:14:09.879 a great way to conceptualize it. Let's look at 00:14:09.879 --> 00:14:12.440 the web, for example. OK. If you look at a URI, 00:14:12.779 --> 00:14:16.500 a web address, you will often see a percent sign 00:14:16.500 --> 00:14:19.480 followed by a two -digit hexadecimal number. 00:14:20.019 --> 00:14:22.629 This is called percent encoding. The Internet's 00:14:22.629 --> 00:14:24.350 underlying architecture doesn't allow certain 00:14:24.350 --> 00:14:27.190 characters like spaces in a web address. Right, 00:14:27.190 --> 00:14:30.250 you never see a space in a URL. Exactly. So the 00:14:30.250 --> 00:14:32.190 percent sign acts as a signal to the browser. 00:14:32.269 --> 00:14:35.210 It says translate what comes next. A space becomes 00:14:35.210 --> 00:14:38.029 percent sign 2 -0. It's a translation tool. Okay, 00:14:38.029 --> 00:14:40.149 so in the web neighborhood, it's a translator. 00:14:40.750 --> 00:14:43.409 But if we go over to the database neighborhood, 00:14:43.649 --> 00:14:47.370 specifically SQL databases, it becomes a wild 00:14:47.370 --> 00:14:50.149 card detective. Yes. If a database administrator 00:14:50.149 --> 00:14:54.710 searches Lisa percent, the percent sign tells 00:14:54.710 --> 00:14:57.090 the computer to fetch every single record where 00:14:57.090 --> 00:14:59.330 the first name is Lisa, no matter what the last 00:14:59.330 --> 00:15:01.950 name is. It's the traffic cop waving all the 00:15:01.950 --> 00:15:04.450 cars through saying, yep, you start with Lisa. 00:15:04.529 --> 00:15:07.429 Come on in. Lisa Smith, Lisa Jones, all of you. 00:15:07.669 --> 00:15:09.929 That's exactly how it works. It also functions 00:15:09.929 --> 00:15:12.850 as a visual tag, or a sigil. The sigil. Yeah, 00:15:12.889 --> 00:15:15.049 like in a language like Perl, programmers use 00:15:15.049 --> 00:15:17.169 a percent sign in front of a variable name to 00:15:17.169 --> 00:15:19.710 indicate that the data is a hash, which is basically 00:15:19.710 --> 00:15:22.429 a dictionary of key value pairs. Okay. By putting 00:15:22.429 --> 00:15:24.649 the percent sign there, the programmer and the 00:15:24.649 --> 00:15:26.370 computer instantly know what kind of data they 00:15:26.370 --> 00:15:28.149 are looking at without having to read the whole 00:15:28.149 --> 00:15:30.549 file. It's also used for formatting templates 00:15:30.549 --> 00:15:34.379 in C, Python, and Ruby. or an ASP where a lesson 00:15:34.379 --> 00:15:36.820 sign and a percent sign act as a start tag. And 00:15:36.820 --> 00:15:38.779 even in command lines, right? Like in Windows 00:15:38.779 --> 00:15:41.279 batch files, it represents parameters or environment 00:15:41.279 --> 00:15:44.600 variables? Yep. Percent one, percent VAR one 00:15:44.600 --> 00:15:46.580 percent. You even need two of them to print a 00:15:46.580 --> 00:15:48.799 literal percent sign. But my absolute favorite 00:15:48.799 --> 00:15:51.879 role it plays is the silencer. Oh, the comments. 00:15:52.179 --> 00:15:55.820 Yes. In systems like Tex, Latex, PostScript, 00:15:56.080 --> 00:15:59.649 and Matlab, If a programmer types a percent sign, 00:16:00.029 --> 00:16:02.549 it tells the computer to completely stop reading. 00:16:02.710 --> 00:16:04.450 Right. It creates what's called a line comment. 00:16:04.830 --> 00:16:07.289 The machine reads the code, hits the percent 00:16:07.289 --> 00:16:09.870 sign, and immediately ignores everything else 00:16:09.870 --> 00:16:12.169 on that line. It's the traffic cop throwing up 00:16:12.169 --> 00:16:14.529 a barricade. Stop right here, machine. The rest 00:16:14.529 --> 00:16:16.990 of this line is for human eyes only. Basically, 00:16:17.090 --> 00:16:20.129 yeah. Programmers use it to leave private notes 00:16:20.129 --> 00:16:22.710 or instructions for themselves right in the middle 00:16:22.710 --> 00:16:25.649 of the code, completely hidden from the computer's 00:16:25.649 --> 00:16:29.159 execution. I mean, from a 15th century shorthand 00:16:29.159 --> 00:16:32.059 for an Italian word to a symbol that literally 00:16:32.059 --> 00:16:34.419 commands a computer to ignore data, that is an 00:16:34.419 --> 00:16:37.419 incredible evolution. It really is. But the computer 00:16:37.419 --> 00:16:39.120 scientists weren't the only ones to borrow the 00:16:39.120 --> 00:16:41.480 percent sign. Linguistics took it, too. They 00:16:41.480 --> 00:16:44.120 did. And they used it for a remarkably different 00:16:44.120 --> 00:16:47.399 purpose. I mean, computer science drafted the 00:16:47.399 --> 00:16:50.080 percent sign to communicate with rigid, perfectly 00:16:50.080 --> 00:16:53.590 logical machines. But linguistics... borrowed 00:16:53.590 --> 00:16:56.529 that exact same symbol to describe the beautiful, 00:16:56.870 --> 00:16:59.350 messy, contradictory way that humans actually 00:16:59.350 --> 00:17:02.570 speak. Exactly. In linguistics, if you place 00:17:02.570 --> 00:17:04.930 a percent sign at the very start of an example 00:17:04.930 --> 00:17:07.329 sentence, you are signaling that the sentence 00:17:07.329 --> 00:17:10.630 is judged well -formed, meaning grammatically 00:17:10.630 --> 00:17:13.650 correct by some speakers, but entirely ill -formed 00:17:13.650 --> 00:17:16.579 by others. And this usually happens due to dialect 00:17:16.579 --> 00:17:19.099 differences or even individual idiolectics. Remind 00:17:19.099 --> 00:17:22.099 me, an idiolect is? An idiolect is the unique 00:17:22.099 --> 00:17:24.680 fingerprint -like way a single specific person 00:17:24.680 --> 00:17:27.490 uses language. So what sounds perfectly natural 00:17:27.490 --> 00:17:29.309 to you might sound completely wrong to someone 00:17:29.309 --> 00:17:31.369 from a different region or even just a different 00:17:31.369 --> 00:17:33.569 neighborhood. So in linguistics, the percent 00:17:33.569 --> 00:17:36.410 sign is essentially the symbol for we agree to 00:17:36.410 --> 00:17:38.990 disagree. Essentially, yes. Which is so different 00:17:38.990 --> 00:17:41.109 from the other linguistic markup symbols out 00:17:41.109 --> 00:17:43.410 there. Like if a linguist puts an asterisk in 00:17:43.410 --> 00:17:45.910 front of a string of words, that means it is 00:17:45.910 --> 00:17:49.130 completely ill -formed. Just totally ungrammatical 00:17:49.130 --> 00:17:52.230 nonsense across the board. A question mark means 00:17:52.230 --> 00:17:55.700 the well -formedness is unclear. And my favorite, 00:17:55.859 --> 00:17:58.859 the number sign, the hashtag, marks a string 00:17:58.859 --> 00:18:01.640 that is syntactically fine, but semantically 00:18:01.640 --> 00:18:04.160 nonsensical, like saying, the angry carpet ate 00:18:04.160 --> 00:18:07.339 the sunset. Perfect grammar, absolutely no logical 00:18:07.339 --> 00:18:11.160 meaning. But the percent sign sits in that ambiguous 00:18:11.160 --> 00:18:13.500 middle ground. If we connect this to the bigger 00:18:13.500 --> 00:18:16.559 picture, it is quite profound. How so? Well, 00:18:16.599 --> 00:18:19.140 the percent sign represents relativity. In mathematics, 00:18:19.220 --> 00:18:23.049 it always means relative to 100. But in linguistics, 00:18:23.089 --> 00:18:25.250 it has evolved to mean relative to the speaker. 00:18:25.369 --> 00:18:28.089 It has become the ultimate symbol of human context. 00:18:28.329 --> 00:18:31.029 The ultimate symbol of context. Wow. And you 00:18:31.029 --> 00:18:33.450 can see how deeply embedded it has become in 00:18:33.450 --> 00:18:36.190 human context globally just by looking at how 00:18:36.190 --> 00:18:38.630 Unicode handles it. Oh, absolutely. Unicode is 00:18:38.630 --> 00:18:40.529 the international standard for encoding text 00:18:40.529 --> 00:18:42.950 on computers, ensuring that a symbol typed in 00:18:42.950 --> 00:18:45.269 Tokyo appears the exact same way in Toronto. 00:18:45.529 --> 00:18:48.130 And they don't just have 1 % sign. The standard 00:18:48.130 --> 00:18:52.460 is U plus 0, 0. But they also have a full width 00:18:52.460 --> 00:18:55.299 percent sign and a small percent sign. They even 00:18:55.299 --> 00:18:58.180 have a square Japanese percento character. What 00:18:58.180 --> 00:19:00.099 does that look like? It takes the English word 00:19:00.099 --> 00:19:03.220 percent. writes it in katakana, and squishes 00:19:03.220 --> 00:19:06.539 it into one single square character to fit vertical 00:19:06.539 --> 00:19:09.200 text layouts. Oh, that's brilliant. And let's 00:19:09.200 --> 00:19:11.880 not forget the Arabic percent sign. Because Eastern 00:19:11.880 --> 00:19:14.200 Arabic numerals use a dot to represent a zero, 00:19:14.440 --> 00:19:16.920 their percent sign adapts perfectly to that reality. 00:19:17.099 --> 00:19:19.539 It replaces the two little circles of our standard 00:19:19.539 --> 00:19:22.299 symbol with square dots set on edge to match 00:19:22.299 --> 00:19:25.259 their numbers. The symbol physically adapts to 00:19:25.259 --> 00:19:27.500 whatever culture it finds itself in. It truly 00:19:27.500 --> 00:19:30.279 is a typographical chameleon. It really is. We've 00:19:30.279 --> 00:19:32.700 covered some serious ground today. Let's trace 00:19:32.700 --> 00:19:34.880 the path of the journey you just went on. We 00:19:34.880 --> 00:19:38.259 started with a lazy 15th century Italian scribe 00:19:38.259 --> 00:19:41.480 rushing to write percento to save expensive parchment, 00:19:41.839 --> 00:19:44.579 throwing a tiny little cursive O loop onto his 00:19:44.579 --> 00:19:47.740 Pisa. We watched that messy shorthand morph into 00:19:47.740 --> 00:19:50.420 a rigid mathematical symbol that sparked international 00:19:50.420 --> 00:19:53.400 typography wars between Chicago style guides 00:19:53.400 --> 00:19:56.819 and European standard bearers. Then we saw it 00:19:56.819 --> 00:19:59.420 locked onto the typewriter. only to be hijacked 00:19:59.420 --> 00:20:02.099 by early computer programmers. It became the 00:20:02.099 --> 00:20:04.779 grapes of hacker jargon, acting as a digital 00:20:04.779 --> 00:20:07.380 traffic cop, translating web addresses, waving 00:20:07.380 --> 00:20:10.299 through SQL databases, and throwing up barricades 00:20:10.299 --> 00:20:13.049 to silence lines of MATLAB code. Before finally 00:20:13.049 --> 00:20:15.349 being claimed by linguists as a marker for the 00:20:15.349 --> 00:20:18.609 messy relative nature of human dialects, I told 00:20:18.609 --> 00:20:20.529 you at the beginning of this deep dive that you'd 00:20:20.529 --> 00:20:22.450 never look at your keyboard the same way again. 00:20:22.829 --> 00:20:24.349 The next time you reach up to hit the number 00:20:24.349 --> 00:20:27.210 five key to write 100%, I want you to realize 00:20:27.210 --> 00:20:29.930 that you are invoking centuries of human history, 00:20:30.450 --> 00:20:33.069 typographical rebellions, and overlapping digital 00:20:33.069 --> 00:20:36.009 dimensions, all in a single keystroke. It is 00:20:36.009 --> 00:20:38.089 a massive amount of historical weight for one 00:20:38.089 --> 00:20:40.490 little key to carry. So what does this all mean? 00:20:40.799 --> 00:20:43.599 We know that this mighty, globally recognized 00:20:43.599 --> 00:20:46.140 fundamental mathematical symbol evolved purely 00:20:46.140 --> 00:20:49.359 from an Italian scribe's hasty handwritten abbreviation. 00:20:49.779 --> 00:20:52.920 It makes you wonder, what messy internet -era 00:20:52.920 --> 00:20:55.559 text message shorthand or emoji are we using 00:20:55.559 --> 00:20:58.220 on our phones right now, today, that will somehow 00:20:58.220 --> 00:21:00.819 become a rigid, fundamental mathematical or coding 00:21:00.819 --> 00:21:03.700 symbol 600 years in the future? Will the laughing, 00:21:03.859 --> 00:21:06.339 crying face be the symbol for a new kind of quantum 00:21:06.339 --> 00:21:08.400 variable? We'll leave you to mull that one over 00:21:08.400 --> 00:21:09.839 the next time you look down at your keyboard. 00:21:10.089 --> 00:21:12.190 Keep questioning the everyday things around you 00:21:12.190 --> 00:21:13.009 and keep exploring.