WEBVTT 00:00:00.000 --> 00:00:03.960 Picture the scenario. You are sitting at your 00:00:03.960 --> 00:00:06.339 desk drafting a critical piece of communication. 00:00:06.540 --> 00:00:08.419 Right. Maybe it's a strategy memo for your executive 00:00:08.419 --> 00:00:10.939 team. Yeah, exactly. Or a technical proposal 00:00:10.939 --> 00:00:14.019 for a new infrastructure client. Or even a white 00:00:14.019 --> 00:00:16.760 paper intended for publication. Your fingers 00:00:16.760 --> 00:00:19.199 are moving rapidly. You're navigating the complexities 00:00:19.199 --> 00:00:22.579 of your argument. And then... you hit a localized 00:00:22.579 --> 00:00:25.519 speed bump. You pause. You pause. Your pinky 00:00:25.519 --> 00:00:28.460 hovers over the shift key, you stare in the screen, 00:00:28.519 --> 00:00:31.239 and you confront a very specific orthographic 00:00:31.239 --> 00:00:34.740 dilemma, one that has quietly divided technologists, 00:00:34.939 --> 00:00:37.840 linguists, and engineers for, well, the better 00:00:37.840 --> 00:00:39.920 part of two decades. And you ask yourself that 00:00:39.920 --> 00:00:42.380 one question. Right. You ask, do I capitalize 00:00:42.380 --> 00:00:45.200 the IA in internet? I mean, it is a moment of 00:00:45.200 --> 00:00:47.719 hesitation that almost everyone who writes about 00:00:47.719 --> 00:00:50.240 technology has experienced at some point. And 00:00:50.240 --> 00:00:53.719 it warrants examination because it's not merely 00:00:53.719 --> 00:00:56.079 a question of arbitrary formatting. No, not at 00:00:56.079 --> 00:00:59.039 all. That single keystroke represents a profound 00:00:59.039 --> 00:01:01.689 philosophical debate. When we decide whether 00:01:01.689 --> 00:01:04.950 to use a capital I or a lowercase i, we are actually 00:01:04.950 --> 00:01:07.890 participating in an ongoing negotiation about 00:01:07.890 --> 00:01:09.909 the ontological status of the network itself. 00:01:10.209 --> 00:01:12.829 And that is precisely what we are unpacking in 00:01:12.829 --> 00:01:16.609 this deep dive. Today we are analyzing the historical, 00:01:17.049 --> 00:01:19.390 technical, and linguistic timeline of the word 00:01:19.390 --> 00:01:21.810 internet. drawing from a really comprehensive 00:01:21.810 --> 00:01:24.890 collection of historical records, technical RFCs, 00:01:24.989 --> 00:01:27.370 and major linguistic studies. Exactly. Our mission 00:01:27.370 --> 00:01:29.650 today is to trace the evolution of this term 00:01:29.650 --> 00:01:33.430 from its origins in 1970s DARPA research, all 00:01:33.430 --> 00:01:36.390 the way to the pivotal style guide wars of 2016. 00:01:36.949 --> 00:01:39.370 We are looking past the simple grammar rules 00:01:39.370 --> 00:01:41.750 to understand what the capitalization or the 00:01:41.750 --> 00:01:45.290 lack thereof reveals about human agency, technological 00:01:45.290 --> 00:01:47.870 ubiquity, and the shifting power dynamics between 00:01:47.870 --> 00:01:49.810 the architects of the digital realm and the people 00:01:49.810 --> 00:01:52.469 who use it. The premise here is that orthography 00:01:52.469 --> 00:01:55.409 is not static. It is a direct reflection of societal 00:01:55.409 --> 00:01:57.790 integration. A capital letter is a linguistic 00:01:57.790 --> 00:02:00.730 marker of distinction. It denotes novelty, specificity, 00:02:00.750 --> 00:02:02.829 and often reverence. Reverence is a great word 00:02:02.829 --> 00:02:05.530 for it. Right. So when a society collectively 00:02:05.530 --> 00:02:08.680 decides to alter the capitalization of a foundational 00:02:08.680 --> 00:02:12.240 technology. They are signaling a paradigm shift 00:02:12.240 --> 00:02:14.960 in how they conceptualize that technology's role 00:02:14.960 --> 00:02:17.979 within their daily lives. To quantify just how 00:02:17.979 --> 00:02:20.219 fractured that conceptualization had become, 00:02:20.599 --> 00:02:23.240 let's look at a critical piece of data from 2016. 00:02:23.659 --> 00:02:26.520 The Oxford English Dictionary conducted a massive 00:02:26.520 --> 00:02:29.080 corpus analysis to determine how the word internet 00:02:29.080 --> 00:02:31.280 was actually being written in the wild. And they 00:02:31.280 --> 00:02:33.419 didn't just rely on a small sample size. Oh, 00:02:33.419 --> 00:02:36.599 no, they ingested an astonishing We're talking 00:02:36.599 --> 00:02:40.639 2 .5 billion printed and online sources. 2 .5 00:02:40.639 --> 00:02:43.129 billion. Yeah. This corpus encompassed academic 00:02:43.129 --> 00:02:45.370 journals, international news media, corporate 00:02:45.370 --> 00:02:48.189 literature, and broader web publications. And 00:02:48.189 --> 00:02:50.330 the analysis revealed that the word was capitalized 00:02:50.330 --> 00:02:53.389 in exactly 54 % of the cases. Which is fascinating. 00:02:53.530 --> 00:02:55.770 From a linguistic and sociological standpoint, 00:02:56.189 --> 00:02:59.030 a 54 % usage rate is incredibly revealing. It 00:02:59.030 --> 00:03:01.490 represents a statistical deadlock. A proofreader's 00:03:01.490 --> 00:03:04.030 nightmare. Absolute nightmare. In the context 00:03:04.030 --> 00:03:07.930 of a 2 .5 billion word corpus, a near 50 -50 00:03:07.930 --> 00:03:10.610 split indicates that the English -speaking world 00:03:10.669 --> 00:03:12.750 was in the absolute center of a transitional 00:03:12.750 --> 00:03:15.289 phase. There was no consensus. The historical 00:03:15.289 --> 00:03:17.969 standard of capitalizing the term had lost its 00:03:17.969 --> 00:03:20.469 absolute authority. Exactly. But the movement 00:03:20.469 --> 00:03:23.250 toward lower casing it had not yet achieved dominance. 00:03:23.969 --> 00:03:26.449 It reflects a society caught between two distinct 00:03:26.449 --> 00:03:29.770 paradigms. You have one side viewing the digital 00:03:29.770 --> 00:03:32.629 network as a unique monumental infrastructure 00:03:32.629 --> 00:03:35.830 and the other side treating it as a generic ambient 00:03:35.830 --> 00:03:39.599 utility. But the OED data gets even more granular 00:03:39.599 --> 00:03:42.240 than that. It reveals a sharp geographic divergence. 00:03:42.860 --> 00:03:44.840 The analysis showed that the capitalized internet 00:03:44.840 --> 00:03:46.860 maintained a much stronger hold in the United 00:03:46.860 --> 00:03:49.080 States, whereas the lowercase internet had already 00:03:49.080 --> 00:03:51.360 become the predominant styling in the United 00:03:51.360 --> 00:03:53.259 Kingdom. The transatlantic divide? Yeah, the 00:03:53.259 --> 00:03:55.919 transatlantic divide. Which is fascinating, right? 00:03:55.960 --> 00:03:58.719 We are talking about a globally distributed borderless 00:03:58.719 --> 00:04:01.960 protocol suite. Why would the geographic location 00:04:01.960 --> 00:04:04.340 of the writer dictate the orthographic treatment 00:04:04.340 --> 00:04:07.280 of the network? Well... The geographic divergence 00:04:07.280 --> 00:04:10.139 likely stems from the differing proximity to 00:04:10.139 --> 00:04:13.360 the technology's origins. That and the distinct 00:04:13.360 --> 00:04:15.840 cultures of tech journalism in those respective 00:04:15.840 --> 00:04:18.319 regions. Meaning the U .S. is the birthplace. 00:04:18.439 --> 00:04:20.259 Right, the United States is the birthplace of 00:04:20.259 --> 00:04:22.319 the ARPANET and the subsequent internet protocols. 00:04:23.259 --> 00:04:25.639 Much of the foundational funding came from the 00:04:25.639 --> 00:04:28.639 U .S. Department of Defense, and the core engineering 00:04:28.639 --> 00:04:31.439 was driven by American universities and research 00:04:31.439 --> 00:04:33.560 institutions. So there's a sense of ownership 00:04:33.560 --> 00:04:36.800 there. Exactly. Consequently, there is a deep 00:04:36.800 --> 00:04:39.079 -seated cultural and historical attachment in 00:04:39.079 --> 00:04:41.680 the U .S. to treating the Internet as a formal 00:04:41.680 --> 00:04:44.939 domestic invention, a proper noun requiring a 00:04:44.939 --> 00:04:46.920 capital letter out of historical respect. And 00:04:46.920 --> 00:04:49.779 the U .K. In contrast, the U .K. adopted the 00:04:49.779 --> 00:04:52.040 technology more rapidly as a consumer utility. 00:04:52.350 --> 00:04:54.290 particularly following the development of the 00:04:54.290 --> 00:04:57.449 World Wide Web, which, while invented by a Briton, 00:04:57.629 --> 00:05:00.509 Tim Berners -Lee, accelerated the network's transition 00:05:00.509 --> 00:05:03.589 into a global commercial medium. That historical 00:05:03.589 --> 00:05:06.529 proximity theory carries a lot of weight. If 00:05:06.529 --> 00:05:09.129 you examine American tech culture, particularly 00:05:09.129 --> 00:05:11.470 the narrative emanating from Silicon Valley throughout 00:05:11.470 --> 00:05:15.129 the 1990s and 2000s, there is a distinct reverence 00:05:15.129 --> 00:05:18.100 for the infrastructure. The network is often 00:05:18.100 --> 00:05:20.680 mythologized. It really is. Meanwhile in the 00:05:20.680 --> 00:05:23.560 UK, the journalism surrounding technology historically 00:05:23.560 --> 00:05:26.839 tends to be slightly more pragmatic. less reverent 00:05:26.839 --> 00:05:29.459 toward the underlying infrastructure. They were 00:05:29.459 --> 00:05:31.819 much quicker to strip away the monumentality 00:05:31.819 --> 00:05:34.279 of the term and integrate it into the generic 00:05:34.279 --> 00:05:37.079 lexicon. It also highlights a critical reality 00:05:37.079 --> 00:05:40.120 for anyone communicating across borders. If you 00:05:40.120 --> 00:05:42.339 are drafting a document today for an international 00:05:42.339 --> 00:05:45.300 audience, your autocorrect or your ingrained 00:05:45.300 --> 00:05:48.100 typing habits are telegraphing a specific cultural 00:05:48.100 --> 00:05:50.100 legacy. That's a great point for you listening 00:05:50.100 --> 00:05:52.360 to keep in mind. Depending on who you're emailing, 00:05:52.519 --> 00:05:54.480 an American colleague versus a British friend, 00:05:54.980 --> 00:05:57.180 you might be signaling your geographic bias without 00:05:57.180 --> 00:06:00.180 realizing it. Leaning into the capital I historically 00:06:00.180 --> 00:06:02.600 aligns with the American deference to the engineering 00:06:02.600 --> 00:06:05.519 origins. While the lowercase i aligns with the 00:06:05.519 --> 00:06:08.860 British acceleration toward utility. So To truly 00:06:08.860 --> 00:06:11.800 understand those engineering origins the American 00:06:11.800 --> 00:06:14.360 standard was deferring to, we have to examine 00:06:14.360 --> 00:06:16.939 the actual inception of the word because it was 00:06:16.939 --> 00:06:19.399 not conceived by a marketing team or a branding 00:06:19.399 --> 00:06:22.360 agency. No, definitely not. The term was born 00:06:22.360 --> 00:06:24.800 strictly as an engineering shorthand in the early 00:06:24.800 --> 00:06:28.100 1970s. Specifically, we can trace it to 1974 00:06:28.100 --> 00:06:32.199 within the text of RFC 675. Request for comment 00:06:32.199 --> 00:06:36.370 675. It is a foundational document in computer 00:06:36.370 --> 00:06:39.529 science, authored by Vint Cerf, Yogan Dalal, 00:06:39.689 --> 00:06:42.470 and Carl Sunshine. This document provided the 00:06:42.470 --> 00:06:45.149 first detailed specification of the transmission 00:06:45.149 --> 00:06:47.750 control program. And the context here is vital, 00:06:47.949 --> 00:06:50.189 right? Absolutely vital. The engineers working 00:06:50.189 --> 00:06:52.209 on the ARPANET were gravelling with the complex 00:06:52.209 --> 00:06:54.560 problem of internet work routing. they were not 00:06:54.560 --> 00:06:56.920 building a single homogeneous network. Right. 00:06:57.259 --> 00:06:59.040 They were trying to create a standard by which 00:06:59.040 --> 00:07:01.399 entirely disparate networks, satellite links, 00:07:01.720 --> 00:07:03.939 radio networks, and varied packet switching systems 00:07:03.939 --> 00:07:06.839 could successfully exchange data without centralized 00:07:06.839 --> 00:07:09.980 control. The challenge was heterogeneity. And 00:07:09.980 --> 00:07:11.800 in describing the process of connecting these 00:07:11.800 --> 00:07:15.000 diverse networks, earlier documents like RFC 00:07:15.000 --> 00:07:19.379 604, back in 1973, they used this really cumbersome 00:07:19.379 --> 00:07:21.540 phrase. They called it the internetworking protocol. 00:07:21.720 --> 00:07:24.319 It's a bit of a mouthful. It is. So by the time 00:07:24.319 --> 00:07:28.379 Serf, Dalal, and Sunshine drafted RFC 675, they 00:07:28.379 --> 00:07:31.319 needed a more efficient terminology. They began 00:07:31.319 --> 00:07:33.740 using the term internet packet and internet as 00:07:33.740 --> 00:07:36.139 a functional abbreviation for internetworking. 00:07:36.379 --> 00:07:39.360 It was entirely utilitarian. Entirely. It was 00:07:39.360 --> 00:07:42.579 a noun adjunct modifying the packet or the protocol. 00:07:42.660 --> 00:07:45.540 It was not a proper noun denoting some global 00:07:45.540 --> 00:07:48.639 entity. Exactly. The lowercase functional usage 00:07:48.639 --> 00:07:50.860 was the original standard because the global 00:07:50.860 --> 00:07:54.540 entity did not yet exist. However, as the TCPIP 00:07:54.540 --> 00:07:57.579 suite matured and began to be adopted by a wider 00:07:57.579 --> 00:07:59.779 array of educational, military, and commercial 00:07:59.779 --> 00:08:02.740 networks in the early 1980s, the internet standards 00:08:02.740 --> 00:08:04.879 community encountered a taxonomical problem. 00:08:04.980 --> 00:08:07.199 They needed a precise vocabulary. Yes, they needed 00:08:07.180 --> 00:08:09.540 differentiate between the general concept of 00:08:09.540 --> 00:08:12.339 connected networks and the specific rapidly expanding 00:08:12.339 --> 00:08:14.519 global network that was coalescing around their 00:08:14.519 --> 00:08:17.279 protocols. This is where the formal capitalization 00:08:17.279 --> 00:08:19.720 rule was born. It was an elegant solution to 00:08:19.720 --> 00:08:23.319 a really complex routing reality. The technical 00:08:23.319 --> 00:08:26.360 community established a strict binary. If you 00:08:26.360 --> 00:08:28.879 wrote an internet with a lowercase i, you were 00:08:28.879 --> 00:08:31.259 referring to any set of interconnected networks, 00:08:31.519 --> 00:08:34.240 regardless of the protocols they used. Two corporate 00:08:34.240 --> 00:08:36.860 lands bridged together constituted an Internet. 00:08:37.000 --> 00:08:39.279 Right. But if you wrote the Internet with a capital 00:08:39.279 --> 00:08:42.980 I, you were referring exclusively to the specific 00:08:42.980 --> 00:08:46.580 global IP based network descended from ARPANET. 00:08:46.659 --> 00:08:49.200 It was a necessary semantic distinction for the 00:08:49.200 --> 00:08:51.480 engineers maintaining the infrastructure. You 00:08:51.480 --> 00:08:53.600 see this formalized in the Request for Comments 00:08:53.600 --> 00:08:57.500 series. Documents like RFC 871, which was published 00:08:57.500 --> 00:09:01.240 in in 1982 explicitly discussed the generic terminology. 00:09:01.480 --> 00:09:03.340 What did they call it again? They noted that 00:09:03.340 --> 00:09:05.200 a network composed of subnets could be called 00:09:05.200 --> 00:09:07.620 a catnet, but was more often simply referred 00:09:07.620 --> 00:09:10.179 to as an internet. The lower casing denoted the 00:09:10.179 --> 00:09:12.500 generic architecture, while the capitalization 00:09:12.500 --> 00:09:15.559 was reserved for the singular global instantiation 00:09:15.559 --> 00:09:18.399 of that architecture. There is a remarkably precise 00:09:18.399 --> 00:09:22.220 articulation of this taxonomy in a 1989 IBM manual. 00:09:22.679 --> 00:09:25.899 It's called the TCP -IP Tutorial and Technical 00:09:25.899 --> 00:09:28.440 Overview. I love this quote. It's slightly mind 00:09:28.440 --> 00:09:31.240 -bending. The manual states, the word internet, 00:09:31.399 --> 00:09:33.379 also internetwork, is simply a contraction of 00:09:33.379 --> 00:09:36.019 the phrase interconnected network. However, when 00:09:36.019 --> 00:09:37.980 written with a capital I, the internet refers 00:09:37.980 --> 00:09:40.179 to the worldwide set of interconnected networks. 00:09:40.519 --> 00:09:42.679 Hence, the internet is an internet, but the reverse 00:09:42.679 --> 00:09:44.799 does not apply. The internet is an internet, 00:09:44.879 --> 00:09:47.610 but the reverse does not apply. That IBM quote 00:09:47.610 --> 00:09:50.690 is the definitive distillation of the 1980s engineering 00:09:50.690 --> 00:09:54.169 consensus. It operates on the exact same categorical 00:09:54.169 --> 00:09:57.570 logic we apply to celestial taxonomy. Oh, the 00:09:57.570 --> 00:10:00.809 planet analogy. Yes. Think about it. We inhabit 00:10:00.809 --> 00:10:03.549 the Earth, which is a specific proper noun, but 00:10:03.549 --> 00:10:05.509 the Earth belongs to the broader category of 00:10:05.509 --> 00:10:08.610 planets. Our planet has the moon. a capitalized 00:10:08.610 --> 00:10:11.350 proper noun, but Jupiter possesses many generic 00:10:11.350 --> 00:10:13.309 moons. Right, like how the Earth is a planet 00:10:13.309 --> 00:10:15.289 but not every planet is the Earth. That makes 00:10:15.289 --> 00:10:18.169 the dense 1980s network logic instantly understandable 00:10:18.169 --> 00:10:20.769 for you. The technical community demanded this 00:10:20.769 --> 00:10:24.299 level of precision because At that time, configuring 00:10:24.299 --> 00:10:26.700 a gateway between a generic internet and the 00:10:26.700 --> 00:10:29.259 internet required distinct technical parameters 00:10:29.259 --> 00:10:32.259 and security considerations. But that strict 00:10:32.259 --> 00:10:35.120 technical taxonomy was bound to face intense 00:10:35.120 --> 00:10:37.419 pressure once the network expanded beyond the 00:10:37.419 --> 00:10:40.279 purview of systems administrators and DARPA researchers. 00:10:41.080 --> 00:10:43.539 As the commercial web exploded in the 1990s and 00:10:43.539 --> 00:10:46.320 early 2000s, the general public wasn't concerned 00:10:46.320 --> 00:10:48.840 with the distinction between a local interconnected 00:10:48.840 --> 00:10:52.259 subnet and the global IP backbone. interacted 00:10:52.259 --> 00:10:55.019 with the singular global entity. Exactly. So 00:10:55.019 --> 00:10:57.320 for a long time the capital I became the universal 00:10:57.320 --> 00:10:59.940 standard in popular media. It was treated as 00:10:59.940 --> 00:11:02.659 a monolithic destination. You went on the internet. 00:11:02.960 --> 00:11:05.360 Yet linguistic history demonstrates that technologies 00:11:05.360 --> 00:11:08.500 rarely retain their capitalized proper noun status 00:11:08.500 --> 00:11:11.440 indefinitely. There is a predictable trajectory 00:11:11.440 --> 00:11:14.240 from technological miracle to mundane utility. 00:11:14.559 --> 00:11:17.159 A 2002 column by John Schwartz in the New York 00:11:17.159 --> 00:11:20.120 Times presciently identified this exact phenomenon. 00:11:20.240 --> 00:11:22.740 What did he compare it to? He compared the linguistic 00:11:22.740 --> 00:11:25.360 future of the internet to the historical trajectory 00:11:25.360 --> 00:11:28.820 of the phonograph in the 19th century. The phonograph 00:11:28.820 --> 00:11:31.820 analogy is incredibly instructive. When the phonograph 00:11:31.820 --> 00:11:34.940 was introduced, the ability to capture and reproduce 00:11:34.940 --> 00:11:37.580 sound mechanically was perceived as bordering 00:11:37.580 --> 00:11:40.899 on the supernatural. It was magic. It was a singular 00:11:40.899 --> 00:11:44.539 awe -inspiring invention. Texts from that era 00:11:44.539 --> 00:11:48.120 often capitalized it. as the phonograph. It commanded 00:11:48.120 --> 00:11:51.360 orthographic reverence. But over decades, as 00:11:51.360 --> 00:11:54.360 the technology proliferated and became a standard 00:11:54.360 --> 00:11:57.299 fixture in living rooms, the awe dissipated. 00:11:57.759 --> 00:12:00.519 It became an environmental constant. Linguistically, 00:12:00.700 --> 00:12:02.799 it was stripped of its capitalization and became 00:12:02.799 --> 00:12:06.059 a generic Common noun. We see this pattern repeatedly. 00:12:06.480 --> 00:12:08.940 The capitalization serves as a linguistic buffer 00:12:08.940 --> 00:12:11.919 for the shock of the new. It isolates the technology, 00:12:12.259 --> 00:12:14.379 marking it as something outside the normal bounds 00:12:14.379 --> 00:12:17.000 of everyday reality. But as integration deepens, 00:12:17.340 --> 00:12:19.740 that buffer is no longer necessary. We saw the 00:12:19.740 --> 00:12:21.799 exact same compression happen with the term website. 00:12:22.440 --> 00:12:24.220 In the early days of the commercial web, it was 00:12:24.220 --> 00:12:26.080 standard practice to write it as two words with 00:12:26.080 --> 00:12:28.480 a capital W. Acknowledging its specific location 00:12:28.480 --> 00:12:31.620 on the worldwide web. Precisely. As the novelty 00:12:31.620 --> 00:12:34.320 wore off, it rapidly collapsed into the single 00:12:34.320 --> 00:12:36.899 loader case word website. The shift from proper 00:12:36.899 --> 00:12:39.480 noun to common noun isn't just about familiarity, 00:12:39.620 --> 00:12:42.299 though. There is a deeper philosophical argument 00:12:42.299 --> 00:12:45.259 regarding agency and control. The researchers 00:12:45.259 --> 00:12:47.879 Annette Markham and Nancy Bame explored this 00:12:47.879 --> 00:12:50.940 in their 2009 work, Internet Inquiry. Yes, their 00:12:50.940 --> 00:12:53.799 work is pivotal here. They documented a deliberate 00:12:53.799 --> 00:12:56.179 movement within the academic field of internet 00:12:56.179 --> 00:12:58.720 studies to utilize the uncapitalized internet. 00:12:59.200 --> 00:13:01.600 Their rationale wasn't based on casual formatting. 00:13:01.980 --> 00:13:04.559 It was a theoretical stance on power dynamics. 00:13:04.840 --> 00:13:07.740 Markham and Bame's analysis is crucial. They 00:13:07.740 --> 00:13:09.799 argue that persistently capitalizing internet 00:13:09.799 --> 00:13:12.549 reifies the network. It treats the infrastructure 00:13:12.549 --> 00:13:14.970 as a monolithic sovereign actor with its own 00:13:14.970 --> 00:13:17.429 agency, something that acts upon society. Deving 00:13:17.429 --> 00:13:20.169 it too much power. Exactly. By intentionally 00:13:20.169 --> 00:13:22.870 lower casing the term, academics and technologists 00:13:22.870 --> 00:13:25.269 were symbolically dismantling that monolith. 00:13:25.870 --> 00:13:27.750 They were asserting that the internet is not 00:13:27.750 --> 00:13:30.529 a self -directing force of nature. It is a human 00:13:30.529 --> 00:13:33.490 -built infrastructure. It is a tool. That reframing 00:13:33.490 --> 00:13:36.990 is profound. If you write, the internet is changing 00:13:36.990 --> 00:13:39.360 how we communicate. You grant the network the 00:13:39.360 --> 00:13:42.019 power. It sounds inevitable. But if you write, 00:13:42.639 --> 00:13:45.220 we are using the internet to change how we communicate, 00:13:45.580 --> 00:13:48.000 you place humanity back at the center of the 00:13:48.000 --> 00:13:50.779 equation. It's a subtle but potent shift. Lower 00:13:50.779 --> 00:13:53.240 casing the eye is a way of reminding ourselves 00:13:53.240 --> 00:13:55.740 that the infrastructure serves us, not the other 00:13:55.740 --> 00:13:58.320 way around. It takes the power away from the 00:13:58.320 --> 00:14:01.220 medium as this monolithic entity and gives it 00:14:01.220 --> 00:14:03.700 back to the developers and users. It removes 00:14:03.700 --> 00:14:07.049 the mystique. It demystifies the protocol. And 00:14:07.049 --> 00:14:09.490 that demystification is ultimately what drove 00:14:09.490 --> 00:14:11.830 the massive orthographic realignment we witnessed 00:14:11.830 --> 00:14:15.750 in 2016. For years, the major arbiters of English 00:14:15.750 --> 00:14:18.429 style, the institutions that govern how journalism, 00:14:18.830 --> 00:14:21.289 academia, and corporate communications are formatted, 00:14:21.610 --> 00:14:24.190 held fast to the capitalized standard. They maintained 00:14:24.190 --> 00:14:26.250 the reverence long after the public had begun 00:14:26.250 --> 00:14:29.210 to view the network as a utility. But in 2016, 00:14:29.610 --> 00:14:32.169 the consensus finally fractured. The dam broke, 00:14:32.450 --> 00:14:34.370 and it was primarily because of the Associated 00:14:34.370 --> 00:14:37.480 Press. For the media landscape, the AP style 00:14:37.480 --> 00:14:40.279 book dictates the standard. When AP changes a 00:14:40.279 --> 00:14:42.200 rule, it immediately ripples through thousands 00:14:42.200 --> 00:14:45.940 of newsrooms globally. In April 2016, the Associated 00:14:45.940 --> 00:14:48.419 Press officially announced that they were downgrading 00:14:48.419 --> 00:14:51.240 Internet to lowercase alongside web. This wasn't 00:14:51.240 --> 00:14:54.000 a minor tweak. It was a formal declaration that 00:14:54.000 --> 00:14:56.620 the era of technological reverence had ended. 00:14:56.840 --> 00:14:59.320 The cascade effect was immediate. The New York 00:14:59.320 --> 00:15:01.720 Times quickly followed suit, adopting the lowercase 00:15:01.720 --> 00:15:04.460 style in June of that same year. The AMA Manual 00:15:04.460 --> 00:15:06.659 of Style, which governs medical and scientific 00:15:06.659 --> 00:15:09.720 publishing, made the transition. By 2017, the 00:15:09.720 --> 00:15:12.299 Chicago Manual of Style, a titan in book publishing, 00:15:12.840 --> 00:15:15.500 announced their shift to lowercase. And in 2019, 00:15:15.720 --> 00:15:18.539 the APA Style, heavily used in the social sciences, 00:15:19.059 --> 00:15:21.539 also capitulated. Within a remarkably condensed 00:15:21.539 --> 00:15:23.820 time frame, the architectural rules of written 00:15:23.820 --> 00:15:26.600 English were fundamentally rewritten. The timing 00:15:26.600 --> 00:15:29.460 of this is what requires analysis, because publications 00:15:29.460 --> 00:15:31.460 like Wired had officially made the switch to 00:15:31.460 --> 00:15:34.440 lowercase way back in 2004, recognizing early 00:15:34.440 --> 00:15:36.360 on that the internet was simply another medium 00:15:36.360 --> 00:15:38.500 for delivering information. They were ahead of 00:15:38.500 --> 00:15:41.679 the curve. They were. So why did it take institutions 00:15:41.679 --> 00:15:44.100 like the AP and the New York Times another 12 00:15:44.100 --> 00:15:47.580 years to catch up? Was there a specific technological 00:15:47.580 --> 00:15:50.639 tipping point around 2016 that forced their hand? 00:15:50.889 --> 00:15:53.370 The delay represents the natural conservatism 00:15:53.370 --> 00:15:56.629 of institutional style guides. They require overwhelming 00:15:56.629 --> 00:15:58.970 evidence of a permanent societal shift before 00:15:58.970 --> 00:16:02.750 altering their rules. By 2016, that evidence 00:16:02.750 --> 00:16:06.049 was undeniable. The smartphone era had reached 00:16:06.049 --> 00:16:08.769 absolute saturation. It was everywhere. The network 00:16:08.769 --> 00:16:11.210 was no longer a destination you actively traveled 00:16:11.210 --> 00:16:14.200 to by sitting at a desktop computer. It had become 00:16:14.200 --> 00:16:17.259 an ambient, invisible layer of daily existence, 00:16:17.740 --> 00:16:19.840 powering everything from our communication to 00:16:19.840 --> 00:16:22.799 our thermostats. The New York Times explicitly 00:16:22.799 --> 00:16:25.080 cited this ubiquity in their rationale, stating 00:16:25.080 --> 00:16:27.580 that adopting the lowercase minimizes distraction. 00:16:28.159 --> 00:16:30.820 Minimizes distraction. That is a highly revealing 00:16:30.820 --> 00:16:33.539 phrase. When a word is capitalized mid -sentence, 00:16:33.860 --> 00:16:36.820 it acts as a visual anchor. It demands the reader's 00:16:36.820 --> 00:16:39.679 attention. It slows the eye down. Right. The 00:16:39.679 --> 00:16:41.679 Times was effectively acknowledging that the 00:16:41.679 --> 00:16:44.019 network was no longer special enough to warrant 00:16:44.019 --> 00:16:47.059 that cognitive interruption. It had become as 00:16:47.059 --> 00:16:50.059 fundamental and generic as electricity or telephone 00:16:50.059 --> 00:16:52.940 both of which also lost their capital letters 00:16:52.940 --> 00:16:55.600 as they integrated into society. It is also worth 00:16:55.600 --> 00:16:58.600 examining the orthography used by the corporations 00:16:58.600 --> 00:17:01.059 that actually construct and maintain this infrastructure. 00:17:01.289 --> 00:17:04.950 Long before the 2016 media shift, tech giants 00:17:04.950 --> 00:17:07.869 like Apple, Microsoft, and Google had already 00:17:07.869 --> 00:17:10.690 standardized the lowercase internet within their 00:17:10.690 --> 00:17:13.309 internal style guides and developer documentation. 00:17:13.630 --> 00:17:15.970 The engineers and product designers view it strictly 00:17:15.970 --> 00:17:19.380 through a utilitarian lens. To Apple or Microsoft, 00:17:19.619 --> 00:17:22.000 the internet is not a philosophical concept or 00:17:22.000 --> 00:17:24.940 a monumental entity. It is layer three of the 00:17:24.940 --> 00:17:27.339 OSI model. It is the plumbing. Exactly, it's 00:17:27.339 --> 00:17:29.839 the plumbing. It is a raw material they manipulate 00:17:29.839 --> 00:17:32.599 to build applications and services. You don't 00:17:32.599 --> 00:17:34.420 capitalize the pipes in a building, and you don't 00:17:34.420 --> 00:17:36.299 capitalize the routing protocols that deliver 00:17:36.299 --> 00:17:38.980 your data. However, we must highlight that this 00:17:38.980 --> 00:17:41.359 utilitarian view is not universally accepted. 00:17:41.690 --> 00:17:44.170 not even within the technical community. There 00:17:44.170 --> 00:17:46.829 is a passionate, highly technical resistance 00:17:46.829 --> 00:17:49.309 that refuses to abandon the capital lie. The 00:17:49.309 --> 00:17:51.809 holdouts. The resistance. This is where it gets 00:17:51.809 --> 00:17:54.269 spicy. Not everyone gave up the capital letter. 00:17:54.609 --> 00:17:57.289 The Modern Language Association, the MLA, which 00:17:57.289 --> 00:17:59.549 governs much of humanity's writing, still mandates 00:17:59.549 --> 00:18:03.450 capitalization. But more significantly, foundational 00:18:03.450 --> 00:18:05.930 engineering bodies like the Internet Engineering 00:18:05.930 --> 00:18:09.150 Task Force, the IETF, and the Internet Society 00:18:09.150 --> 00:18:12.609 adamantly maintain the capital letter. The IETF 00:18:12.609 --> 00:18:15.650 stance is entirely logical. They are the direct 00:18:15.650 --> 00:18:17.789 descendants of the original ARPANET engineers 00:18:17.789 --> 00:18:20.329 who established an Internet versus the Internet 00:18:20.329 --> 00:18:23.470 taxonomy in the early 1980s. For them to abandon 00:18:23.470 --> 00:18:25.869 the capital I would be to abandon their own historical 00:18:25.869 --> 00:18:28.009 definitions and technical precision. But the 00:18:28.009 --> 00:18:30.609 resistance goes beyond mere historical stubbornness. 00:18:31.109 --> 00:18:33.230 There is a profound ideological argument being 00:18:33.230 --> 00:18:35.269 made by those who manage the deep infrastructure. 00:18:35.849 --> 00:18:38.509 We see this vividly articulated in a 2021 Founders 00:18:38.509 --> 00:18:41.039 Letter from Cloudflare. Cloudflare is a critical 00:18:41.039 --> 00:18:43.900 piece of the modern Internet architecture. They 00:18:43.900 --> 00:18:46.700 manage a massive percentage of global web traffic, 00:18:47.079 --> 00:18:50.740 DNS resolution, and DDoS protection. Their perspective 00:18:50.740 --> 00:18:53.400 is informed by the daily reality of keeping the 00:18:53.400 --> 00:18:56.299 global network operational. In their 2021 letter, 00:18:56.420 --> 00:18:59.119 the founders explicitly rejected the media's 00:18:59.119 --> 00:19:02.640 shift to lowercase. They wrote, we always capitalize 00:19:02.640 --> 00:19:05.180 the eye and Internet in spite of what the AP 00:19:05.180 --> 00:19:08.099 style guide has said since 2016, because it's 00:19:08.099 --> 00:19:10.880 a proper noun. We believe there is and only should 00:19:10.880 --> 00:19:13.559 be one. And we have an enduring respect for what 00:19:13.559 --> 00:19:16.140 a miracle it is that it exists. That statement 00:19:16.140 --> 00:19:18.759 is a powerful counter narrative to the idea of 00:19:18.759 --> 00:19:21.569 the Internet. as an invisible utility, an enduring 00:19:21.569 --> 00:19:23.609 respect for what a miracle it is. It's beautiful, 00:19:23.609 --> 00:19:26.230 really. It is. The media organizations and the 00:19:26.230 --> 00:19:28.490 consumer tech companies want to minimize distraction. 00:19:28.809 --> 00:19:30.549 They want the network to fade into the background. 00:19:31.250 --> 00:19:33.730 Cloudflare, however, is arguing against complacency. 00:19:34.190 --> 00:19:35.890 Because they sit at the infrastructure level, 00:19:36.289 --> 00:19:38.829 managing BGP routes and mitigating massive network 00:19:38.829 --> 00:19:41.509 failures, they know firsthand how fragile the 00:19:41.509 --> 00:19:44.559 system is. They know it is not a given. The capital 00:19:44.559 --> 00:19:47.519 I is a monument to the immense ongoing human 00:19:47.519 --> 00:19:50.039 effort required to keep disparate global networks 00:19:50.039 --> 00:19:53.299 unified. It sets up a fascinating tension. Is 00:19:53.299 --> 00:19:55.319 Cloudflare simply being romantic about their 00:19:55.319 --> 00:19:58.440 industry, or are they technically more accurate? 00:19:59.119 --> 00:20:01.039 Because from a strict networking perspective, 00:20:01.440 --> 00:20:04.880 there really is only one global BGP routing table 00:20:04.880 --> 00:20:07.579 that makes the whole thing work. They are defending 00:20:07.579 --> 00:20:10.500 the singularity of the network against the fragmentation 00:20:10.500 --> 00:20:13.440 of generic utility. They are defending the architectural 00:20:13.440 --> 00:20:16.859 reality. While you, the user, experience the 00:20:16.859 --> 00:20:19.299 internet as a generic environment, the network 00:20:19.299 --> 00:20:22.240 engineer experiences it as a highly specific, 00:20:22.599 --> 00:20:25.859 singular construction. The orthography you choose 00:20:25.859 --> 00:20:28.140 depends entirely on which of those two realities 00:20:28.140 --> 00:20:30.220 you inhabit. Which brings us to the practical 00:20:30.220 --> 00:20:32.579 application for you, the listener. Given this 00:20:32.579 --> 00:20:35.339 deep historical division, the competing philosophies 00:20:35.339 --> 00:20:37.859 of agency, and the conflicting style guides, 00:20:38.180 --> 00:20:39.880 how are you supposed to write it? So what are 00:20:39.880 --> 00:20:41.420 you supposed to do next time you write an email? 00:20:41.660 --> 00:20:44.440 The guidance ultimately relies entirely on context 00:20:44.440 --> 00:20:47.759 and audience. Precisely. If your writing is directed 00:20:47.759 --> 00:20:50.519 at a general audience, if you are drafting standard 00:20:50.519 --> 00:20:52.859 business communications, writing for a mainstream 00:20:52.859 --> 00:20:55.720 media outlet, or creating marketing copy, the 00:20:55.720 --> 00:20:58.240 consensus has firmly shifted to the lowercase 00:20:58.240 --> 00:21:02.339 i. Adhering to the AP or Chicago style minimizes 00:21:02.339 --> 00:21:05.240 visual friction and aligns with modern colloquial 00:21:05.240 --> 00:21:08.160 usage. However, if you are operating in specialized 00:21:08.160 --> 00:21:10.839 environments, the rules remain rigid. If you 00:21:10.839 --> 00:21:13.319 are submitting an academic paper in the humanities 00:21:13.319 --> 00:21:16.839 subject to MLA guidelines, the capital I is required. 00:21:17.119 --> 00:21:19.599 And crucially, if you are authoring highly technical 00:21:19.599 --> 00:21:22.079 documentation regarding network protocols or 00:21:22.079 --> 00:21:24.480 communicating with organizations like the IETF 00:21:24.480 --> 00:21:27.119 or infrastructure purists, utilizing the capital 00:21:27.119 --> 00:21:29.920 I signals a necessary respect for the technical 00:21:29.920 --> 00:21:32.900 taxonomy. There is also a highly specific grammatical 00:21:32.900 --> 00:21:35.180 loophole that historically bridged the gap between 00:21:35.180 --> 00:21:37.799 these two camps. It's the differentiation between 00:21:37.799 --> 00:21:40.400 the term's use as a noun and its use as an attributive 00:21:40.400 --> 00:21:42.839 noun or a noun adjunct. It's quite funny how 00:21:42.839 --> 00:21:44.920 language tries to compromise when it can't make 00:21:44.920 --> 00:21:48.440 up its mind. The noun adjunct variation is a 00:21:48.440 --> 00:21:51.900 brilliant example of linguistic compromise. Even 00:21:51.900 --> 00:21:55.680 before the mass shift in 2016, many strict style 00:21:55.680 --> 00:21:58.259 guides recognized that the word functioned differently 00:21:58.259 --> 00:22:00.910 depending on its placement in a sentence. The 00:22:00.910 --> 00:22:02.990 rule stated that when referring to the network 00:22:02.990 --> 00:22:05.950 itself as a standalone noun, for example, I am 00:22:05.950 --> 00:22:08.490 on the internet, it retained the capitalization. 00:22:09.210 --> 00:22:11.309 But when the word was pressed into service to 00:22:11.309 --> 00:22:13.789 modify another noun functioning as an adjective, 00:22:14.289 --> 00:22:16.349 as in I am using internet resources or I have 00:22:16.349 --> 00:22:18.549 an internet connection, it was stripped of its 00:22:18.549 --> 00:22:20.890 capitalization. It demonstrates the incredible 00:22:20.890 --> 00:22:23.430 elasticity of grammar when forced to accommodate 00:22:23.430 --> 00:22:26.220 rank of technological evolution. The language 00:22:26.220 --> 00:22:28.259 attempts to preserve the monumentality of the 00:22:28.259 --> 00:22:31.200 network as an entity, while simultaneously acknowledging 00:22:31.200 --> 00:22:33.660 its mundane function as a delivery mechanism. 00:22:34.380 --> 00:22:36.859 The cognitive load required for a writer or copy 00:22:36.859 --> 00:22:39.220 editor to constantly evaluate the syntactic role 00:22:39.220 --> 00:22:41.059 of the word before deciding to press the shift 00:22:41.059 --> 00:22:43.980 key is immense. It is unsurprising that general 00:22:43.980 --> 00:22:46.500 style guides eventually opted for a blanket lowercase 00:22:46.500 --> 00:22:49.140 mandate to simplify the editorial process. It 00:22:49.140 --> 00:22:51.940 is a perfect encapsulation of how messy the intersection 00:22:51.940 --> 00:22:54.910 of engineering and linguistics can be. To summarize 00:22:54.910 --> 00:22:57.589 this entire trajectory, the debate over this 00:22:57.589 --> 00:23:00.009 single keystroke tracks the complete narrative 00:23:00.009 --> 00:23:03.650 arc of the digital age. In the 1970s and early 00:23:03.650 --> 00:23:06.670 80s, the capitalization was a strict functional 00:23:06.670 --> 00:23:09.470 necessity for DARPA engineers building internetwork 00:23:09.470 --> 00:23:12.190 routing protocols. It was the crucial distinction 00:23:12.190 --> 00:23:14.890 between a generic local network and the emerging 00:23:14.890 --> 00:23:17.490 global backbone. As we transitioned into the 00:23:17.490 --> 00:23:20.470 1990s and 2000s, that engineering distinction 00:23:20.470 --> 00:23:23.589 evolved into a cultural reverence. The internet, 00:23:23.680 --> 00:23:26.519 capitalized became a monolithic destination, 00:23:27.000 --> 00:23:29.180 a digital frontier that commanded respect and 00:23:29.180 --> 00:23:31.799 awe. It was a proper noun that fundamentally 00:23:31.799 --> 00:23:34.319 altered global communication. And finally, as 00:23:34.319 --> 00:23:36.480 the infrastructure achieved total saturation 00:23:36.480 --> 00:23:40.140 in the 2010s, that reverence dissolved into utilitarian 00:23:40.140 --> 00:23:42.519 ubiquity. The loss of the capital letter across 00:23:42.519 --> 00:23:45.019 major style guides in 2016 was not a demotion. 00:23:45.180 --> 00:23:47.220 It was the ultimate proof of the network's total 00:23:47.220 --> 00:23:49.740 integration into our reality. It became lowercase 00:23:49.740 --> 00:23:52.220 because it became essential and visible and ubiquitous, 00:23:52.559 --> 00:23:54.980 as common as the air we breathe. It serves as 00:23:54.980 --> 00:23:57.799 a reminder that the language we deploy is a living 00:23:57.799 --> 00:24:00.680 record of our relationship with our tools. Every 00:24:00.680 --> 00:24:03.380 time you type the word, you are actively participating 00:24:03.380 --> 00:24:06.519 in this living history of orthography. The choices 00:24:06.519 --> 00:24:08.480 you make are not just matters of compliance with 00:24:08.480 --> 00:24:11.160 a style guide. They are active declarations of 00:24:11.160 --> 00:24:13.160 how you perceive the technological environment 00:24:13.160 --> 00:24:15.839 you inhabit. As we conclude this deep dive, consider 00:24:15.839 --> 00:24:17.960 the trajectory we've just mapped and apply it 00:24:17.960 --> 00:24:20.400 to the horizon of emerging technologies. Think 00:24:20.400 --> 00:24:23.160 about what we capitalize today. If capitalizing 00:24:23.160 --> 00:24:26.769 a technology signifies novelty and awe, And lower 00:24:26.769 --> 00:24:29.509 -casing it signifies it has become an invisible 00:24:29.509 --> 00:24:32.190 utility. What capitalized words are we using 00:24:32.190 --> 00:24:34.390 right now, like artificial intelligence, virtual 00:24:34.390 --> 00:24:37.210 reality, or the metaverse, that will inevitably 00:24:37.210 --> 00:24:39.190 lose their capital letters in the decades to 00:24:39.190 --> 00:24:43.069 come? When AI simply becomes AI, operating silently 00:24:43.069 --> 00:24:45.289 in the background of your existence, what will 00:24:45.289 --> 00:24:48.289 that orthographic shift reveal about how deeply 00:24:48.289 --> 00:24:50.309 it is integrated into your mind and your life?