MACHINE ASR ACCESSIBILITY AID

The_battle_for_Maithili_web_journalism.mp4

Not an editorially verified transcript. This text was generated automatically from the preserved recording and may contain recognition, language-detection, spelling, segmentation or name errors. Consult the source recording for authoritative content.

Collection: Part 1 · VIDEHA MITHILA MAITHILI DISCUSSION CRITICISM SERIES PART 1
Status: asr-draft
Human verified: No
Editorial review: not-reviewed
ASR model: small
Detected language: en (0.998152)
Duration: 33:16
Source: Open preserved recording

Timestamped machine output

0:00–0:11So, picture this, it's 2018 and there's this absolutely vicious, just highly public digital fistfight breaking out on Facebook.
0:11–0:12Yeah, a full on brawl.
0:12–0:17Right. And it's over what, I mean, on the surface seems like a really simple question, right?
0:17–0:20Like, who published the very first Mephilim language blog?
0:20–0:24Which sounds so niche, but the stakes for them were incredibly high.
0:24–0:29Exactly. And to prove his case, one of these guys essentially has to become
0:30–0:31like a digital forensic detective.
0:31–0:34I mean, he's digging into the invisible architecture
0:34–0:35of the internet itself.
0:35–0:39Yeah, he's exposing fake URL timestamps,
0:39–0:41forged publication dates, all that stuff.
0:41–0:43Yeah, and even debunking these really technical claims
0:43–0:46about like 1990s font software.
0:46–0:47It is wild.
0:47–0:48So welcome to the deep dive everyone.
0:48–0:51Today we are exploring the surprisingly cutthroat,
0:51–0:55fiercely passionate, and just incredibly complex world
0:55–0:56of Mathilli web journalism.
0:56–0:57It really is a fascinating world.
0:57–1:01And it perfectly illustrates this much larger phenomenon
1:01–1:01because what we're looking at
1:01–1:03is a transnational language community.
1:03–1:05Right, because Mathilius spoken in a few different areas.
1:05–1:09Yeah, primarily in Bihar and Harkhand in India
1:09–1:12and then also the Eastern Terai region of Nepal.
1:12–1:15And we're basically watching this entire community
1:15–1:18step onto the completely uncharted,
1:18–1:20volatile territory of the early internet.
1:20–1:22And that is exactly our mission today.
1:22–1:25We're diving into a really grounded,
1:25–1:27super detailed source text today.
1:27–1:30It's called A History of Methili Web Journalism,
1:30–1:32and it's written by Ashish Anshar.
1:32–1:34It's a great book, really meticulous.
1:34–1:36It is, and we're gonna use it to figure out
1:36–1:40how this linguistic community literally fought for space
1:40–1:43online, you know, the immense technological brick walls
1:43–1:46they kept slamming into, and the sheer grass roots
1:46–1:48unpaid labor that was required
1:48–1:50to digitize an entire culture.
1:50–1:52Because honestly, if you think about it,
1:52–1:54this isn't just about one language in South Asia.
1:54–1:55Right.
1:55–1:56It's a blueprint.
1:56–2:00It's a case study for how any traditional culture survives the digital age.
2:00–2:01Yeah, exactly.
2:01–2:05Oh, let's untack this because before we get into the forge timestamps and the Facebook
2:05–2:08showdowns, we kind of need to understand the environment these pioneers were walking
2:08–2:09into.
2:09–2:10Yeah, you have to set the stage.
2:10–2:11Right.
2:11–2:15And the author, Anshinhar, he spends a good chunk of time just setting up the sociology
2:15–2:16of the internet itself.
2:16–2:20He does, and I think it's completely necessary because you really can't understand the
2:20–2:26magnitude of putting a regional language on the web without understanding what a chaotic
2:26–2:31paradigm shift the web actually was in the first place. Oh for sure. So Anshin Har establishes this
2:31–2:36massive global and Indian timeline just to show how quickly everything changed. I mean we go from
2:36–2:43the US Defense Department tying computers together with Arpanet in 1969. Right, to super early days.
2:43–2:50Yeah and then to Tim Berners-Lee creating the World Wide Web in 1989. But the Indian context
2:50–2:55is really where the clock starts ticking for our specific story. Yeah the book notes that the
2:55–3:00the internet arrived in India via Urnet, which is the educational and research network, around
3:00–3:061986. But I imagine that wasn't something like a teenager in Bihar was just logging into
3:06–3:07from their bedroom.
3:07–3:11No, not at all. No. Urnet was strictly a closed loop. It was for the government and top academic
3:11–3:12institutions only.
3:12–3:14Okay. So highly restricted.
3:14–3:19Very. The real turning point, the one they author highlights as the big bang is August
3:19–3:2115th, 1995.
3:21–3:221995.
3:22–3:28Yeah, that is when the general public finally got access through VSNL, which is the Vadesh
3:28–3:30Sanjharni Gump Limited.
3:30–3:31August 1995.
3:31–3:34I mean, it's wild to think that's barely three decades ago.
3:34–3:36No, it feels like ancient history, but it's really not.
3:36–3:37Right.
3:37–3:41And the author points out how the internet in India didn't just grow, it basically
3:41–3:44detonated, especially as we moved into the 2000s and 2010s.
3:44–3:46Oh yeah, an absolute explosion.
3:46–3:50It went from this very slow, you know, dial-up novelty where you couldn't use the landline
3:50–3:54phone to the absolute backbone of daily life, especially for the youth.
3:54–3:59I mean, banking, education, booking a train ticket, suddenly it all migrated online.
3:59–4:00It became essential.
4:00–4:04But what's really fascinating here is that Incheonhard doesn't write this as a purely
4:04–4:05triumphant tech bro narrative.
4:05–4:07No, he really doesn't.
4:07–4:12He takes a highly critical, deeply sociological look at what this new frontier was actually
4:12–4:13doing to people.
4:13–4:19Like, he explicitly brings up internet addiction disorder as a societal crisis that just
4:19–4:25Arrived right alongside all the convenience. Yeah, he really doesn't gloss over the ugly stuff at all, which I appreciate it
4:25–4:29So he talks about people essentially living their entire lives through screens
4:30–4:36Spending up to 15 hours a day online and he goes into specific societal damages, too
4:36–4:41Yeah, he pulls no punches like he notes how the massive proliferation of online pornography started
4:41–4:45You know destroying marriages and creating deep psychological rifts in society, right?
4:45–4:51And he even points to the rise of cyberterrorism, where warfare suddenly became invisible.
4:51–4:56You didn't need a physical army anymore, you just needed bugs, viruses, and a keyboard.
4:56–5:00And he also delves into the political weaponization of the internet, which is a huge theme.
5:00–5:02Oh, the IT cells.
5:02–5:03Exactly.
5:03–5:08The text specifically calls out the creation of political IT cells.
5:08–5:13And for the listener, these are basically organized groups using these massive call
5:13–5:17center-like setups to intentionally manufacture and spread fake news.
5:17–5:21Wow. Yeah, they use it to incite riots and manipulate
5:21–5:26public perception on a massive scale. He's really painting a picture of a digital space
5:26–5:29that is incredibly hostile. Hustle, yeah. Yeah.
5:29–5:32And he also talks about the psychological shift in how we value things, right? Like
5:32–5:36the whole validation economy. Yes. This is one of his most crucial
5:36–5:40points, I think. And to illustrate this, the author brings up Justin Rosenstein.
5:40–5:42Wait, isn't he the guy who invented the Facebook like button?
5:42–5:44He co-created it, yes.
5:44–5:44Okay, yeah.
5:44–5:46And the author points out the heavy irony here.
5:46–5:49Rosenstein later actually deleted the Facebook app
5:49–5:51from his own phone because he was so terrified
5:51–5:54of the psychological feedback loop he helped create.
5:54–5:55That is so telling.
5:55–5:56Right.
5:56–5:58Anshinhar observes that in this new digital space,
5:58–6:02the like button basically created a brand new currency.
6:02–6:05Digital validation literally became wealth.
6:05–6:07Which completely changes human behavior.
6:07–6:10I mean, the author mentions how this metric driven
6:10–6:14culture turned artists, poets, writers just into clout chasers.
6:14–6:21Precisely. Like, think about it. If you are a Mathili poet in, say, 1990, you write a poem
6:21–6:26based on your internal artistic drive, right? Or maybe the feedback of a very small peer
6:26–6:27group in your village.
6:27–6:28Yeah, it's intimate.
6:28–6:33Exactly. But Anshinhar notes that once you put that poet on the internet, the quality
6:33–6:37of the art is suddenly judged by a highly visible public number of likes.
6:37–6:39Right. The metric is right there for everyone to judge.
6:39–6:41And it causes genuine anxiety.
6:41–6:45It actually alters what people choose to create because they start favoring whatever
6:45–6:47gets the most immediate viral reaction.
6:47–6:51Okay, I have to pause here because reading this brought up a massive contradiction for
6:51–6:52me.
6:52–6:53Oh, what's that?
6:53–6:58Well, if the author recognizes that the internet is this, you know, this toxic casino of fake
6:58–7:04news cyber warfare and desperate validation seeking, then why even put Mathili on it?
7:04–7:07Like it seems totally counterintuitive.
7:07–7:14If you have this rich, nuanced, centuries-old literary tradition, isn't digitizing it just
7:14–7:15throwing it to the wolves?
7:15–7:21I mean, it's like building a beautiful, quiet library in the middle of a chaotic amusement
7:21–7:22park.
7:22–7:23That is the ultimate paradox, isn't it?
7:23–7:26And what's great is the author faces it head-on.
7:26–7:29He doesn't shy away from that contradiction.
7:29–7:30How does he justify it?
7:30–7:33Well, he argues that the internet has a dual nature.
7:33–7:37makes people incredibly knowledgeable but simultaneously incredibly ignorant.
7:37–7:41Wait, how does that actually work in practice, knowledgeable and ignorant at the same time?
7:41–7:43Do you think about the mechanics of a search engine?
7:43–7:43Okay.
7:43–7:48When you look up a historical fact, the algorithm doesn't just hand you a beautifully bound,
7:48–7:50peer-reviewed encyclopedia.
7:50–7:51Right, it gives you millions of links.
7:51–7:53Exactly, it hands you the truth.
7:53–7:58But right next to it, maybe one link down, it hands you highly visible, extremely well-funded,
7:58–8:00fake news or wild conspiracy theories.
8:00–8:01Oh, I see.
8:01–8:06And because both results are formatted exactly the same way, you know, a clean black headline,
8:06–8:10a blue link, a little snippet of text critical thinking is naturally eroded.
8:10–8:15The digital authority looks identical for both the truth and the absolute lie.
8:15–8:18Wow. Yeah. That's terrifying when you put it like that.
8:18–8:22Okay, so if the environment is that corrosive to the truth, I have to go back to my question.
8:23–8:25Why build your cultural home there?
8:25–8:29Because simply put, the alternative is extinction.
8:29–8:29Extinction.
8:29–8:36Yes. Anshinhar makes it very clear. Avoiding the internet is just no longer a viable option for any culture.
8:36–8:44If a young person in 2010 searches for a Mathili poem or the history of a specific region in Bihar and the internet returns zero results...
8:44–8:46Then it just doesn't exist for them.
8:46–8:52Exactly. To that young person, that history simply ceases to exist. It drops entirely out of the modern consciousness.
8:52–8:58Therefore, establishing a pristine, factual, and robust digital presence wasn't about, you know,
8:58–9:04wanting to join the chaotic amusement park. It was basically a hostage situation. Digitize or disappear.
9:04–9:10Wow. Okay. So the stakes literally couldn't be higher. It's a literal battle for cultural
9:10–9:14memory. Exactly. And that brings us right back to the drama we started the deep dive with.
9:15–9:19Because once people realize this digital land grab is essential for survival,
9:19–9:21the race to plant the first flag begins.
9:21–9:22The rush is on.
9:22–9:27And human nature dictates that everyone wants to be recognized as the first explorer in the history books.
9:27–9:32Manually. And Anshan Har spends a significant portion of his research definitively sorting out
9:32–9:37this timeline because in a digital void establishing who laid the very first brick
9:37–9:40is a matter of immense historical pride.
9:40–9:43Right. So according to his digital archaeology, the absolute first
9:43–9:47The Thilly Presence on the Internet was a website called Balsaric Gah.
9:47–9:49Balsaric Gah, yes.
9:49–9:52And it was hosted on Yahoo Geo Cities in the year 2000.
9:52–9:57Yahoo Geo Cities. Man, for listeners who might not have been around for that era,
9:57–10:00you really have to understand what a wild west that was.
10:00–10:01Oh, it was chaos.
10:01–10:06It was a platform where absolutely anyone could claim a little plot of digital city space
10:06–10:10and build a very rudimentary, usually visually chaotic website.
10:10–10:13Oh yeah, we're talking neon green backgrounds.
10:13–10:17spinning globe gifs, the sparkly tax, the whole nine yards.
10:17–10:19Exactly.
10:19–10:22So this site, Palsarkak, was built by a man named
10:22–10:24Gajendra Thakur.
10:24–10:25Gajendra Thakur.
10:25–10:26Yes.
10:26–10:27And the book notes that this initial site
10:27–10:29was followed by a few other pioneers.
10:29–10:33There was a site called Palav Mithila in 2003,
10:33–10:36and an early news blog named Samadiyah in 2004.
10:36–10:38Okay, so that's the established timeline.
10:38–10:40But anytime you have someone claiming
10:40–10:42to be the first in history,
10:42–10:44you're inherently going to have challengers.
10:44–10:45Oh, always.
10:45–10:46Which brings us to the controversy.
10:46–10:49Here's where it gets really interesting.
10:49–10:52A rival blog called Kutek Rasbat,
10:52–10:54which was operated by Kumar Pabmanab and a few associates,
10:54–10:57they stepped forward and basically said, no, we were first.
10:57–10:58Right.
10:58–11:01And to prove it, they pointed to posts on their blog
11:01–11:04that explicitly displayed publication dates of 1999 and 2004.
11:04–11:05Now, if that were true,
11:05–11:07it would completely rewrite the timeline
11:07–11:09of the language's digital history.
11:09–11:11Tucker wouldn't be first anymore.
11:11–11:16Right. But Anshinhar, he doesn't just look at the dates typed on the screen and not his head, you know.
11:16–11:20He pulls the curtain back on how blog platforms actually function.
11:20–11:25He investigates the underlying architecture of Blogger, which is the platform Kedekross bot was using.
11:25–11:28And the forensic breakdown he uses here is just fascinating.
11:28–11:33He starts by looking at the features Google actually offered on Blogger at that specific time.
11:33–11:39Right. He points out that the ability to schedule a future post or back data post,
11:39–11:43Basically that little feature that lets you manually change the display date of a piece of content
11:43–11:47that wasn't even introduced by Google until years later around 2008.
11:47–11:48It didn't exist.
11:48–11:49Didn't exist.
11:49–11:53But Anshanar goes even deeper than just feature updates.
11:53–11:54He looks at the URLs.
11:54–11:59And for the listener, we really should explain why a URL is so damning in this specific context.
11:59–12:02Yeah please do, because I was a little lost on this at first.
12:02–12:04How does a web address prove a forgery?
12:04–12:08Okay, so when you publish a post on a platform like Blogger,
12:08–12:12The system automatically generates a unique web address, a URL.
12:12–12:15So the internet knows exactly where to find that specific page.
12:15–12:17Right, the www.whatever.
12:17–12:18Exactly.
12:18–12:21And the way Blogger's automated system worked back then
12:21–12:25was by embedding the actual year and month of creation
12:25–12:27directly into that URL string.
12:27–12:27Oh, automatically.
12:27–12:28Completely automatically.
12:28–12:30So it would look something like, you know,
12:30–12:36blogspot.com, for-pression 21-307, name of the post.
12:36–12:40That URL is hard-coded by the server at the exact moment you hit the publish button.
12:40–12:46Ah, so anyone can go into the text editor, type published on July 1st, 1999,
12:46–12:51in like huge bold font at the top of the page, and the public will see that display date.
12:51–12:52Yes.
12:52–12:53But the server knows the truth.
12:53–12:58Exactly. It's the digital equivalent of forging a medieval manuscript, but forgetting that the
12:58–13:02paper you use has a 2013 watermark visible when you hold it up to the light.
13:02–13:03That's a perfect analogy.
13:03–13:07The display date is manual and easily faked, but the URL time stamp is systemic.
13:07–13:10And Anshanhar just caught them red-handed with this.
13:10–13:11Yeah.
13:11–13:20I mean, he proves that a post on Catacross Bot that proudly displayed a 1999 date actually had a URL time stamp of July 2013 embedded in his web address.
13:20–13:22It's undeniable proof.
13:22–13:24And he went post by post.
13:24–13:29Another one claiming a 2004 publication date had a 2005 URL time stamp.
13:29–13:35The server architecture completely 100% exposed the manual manipulation of the historical record.
13:35–13:39Which naturally led to that massive confrontation you mentioned at the start of the deep dive.
13:39–13:41Yeah, the 2018 Facebook showdown.
13:41–13:42Yeah.
13:42–13:45And I love that Anshin Haar actually includes screenshots of this public debate in the book.
13:45–13:46He brings the receipts.
13:46–13:47He does.
13:47–13:53So Kumar Padmanab, defending his claim, tries to use this highly technical argument to invalidate
13:53–13:56to invalidate Gijendra Thakur's year 2000 site.
13:56–13:59Padmanabh argues that it was functionally impossible
13:59–14:02to build a Mathili website in 2000
14:02–14:04because according to him,
14:04–14:07Devanagari typing tools didn't even exist before 2003.
14:07–14:09With Devanagari being the primary script
14:09–14:11that Mathili is often written in today.
14:11–14:13Right, so Padmanabh is essentially saying
14:13–14:15you couldn't have built a site in 2000
14:15–14:17because you literally couldn't type the alphabet
14:17–14:19on a computer yet, checkmate.
14:19–14:21And how does the author systematically dismantle
14:21–14:22that defense?
14:22–14:23Because he does?
14:23–14:25Oh, he destroys it.
14:25–14:29NGR points out a very specific piece of software history.
14:29–14:32He casually notes that fonts like KrutiDev
14:32–14:34had been around since 1997.
14:34–14:36Ah, KrutiDev.
14:36–14:38Wow, that is a name that brings back a lot of memories
14:38–14:40for early Indian internet users.
14:40–14:42Walk us through that, because I don't know it.
14:42–14:44Why is KrutiDev the smoking gun here?
14:44–14:46How did it actually work?
14:46–14:47Oh, okay, so before we had Unicode.
14:47–14:48What's Unicode?
14:48–14:50Unicode is the modern universal standard
14:50–14:53that allows computers to natively understand
14:53–14:59thousands of different alphabets. Before that, computers were stubbornly, completely English-centric.
14:59–15:04They really only understood the Latin alphabet. So software developers in India had to create a
15:04–15:10hack, and Krudi Dev was a non-unicode font. What it did was visually map Devanagari characters onto
15:10–15:14the standard English QWERTY keyboard. Okay, so wait, the computers still thought you were
15:14–15:19typing in English? Exactly. If you press the letter K on your keyboard, the computer internally
15:19–15:24registered in English K. It had no idea about Hindi or Mathili, but the Kruti Dev font file
15:24–15:30told the monitor to draw the Hindi character, Ela, on the screen. It was entirely a visual
15:30–15:34illusion. In fact, if you sent that document to someone who didn't have the Kruti Dev font
15:34–15:39installed on their computer, they would open it up and just see a string of random English gibberish.
15:39–15:43That is brilliant, but also sounds so incredibly frustrating.
15:43–15:48It was. But because Crudy Dev existed and was widely used in 1997,
15:48–15:55Anshinor proves that people absolutely had the tools to type and publish in Dev and Agari by the year 2000.
15:55–15:58So Padmanov's technical alibi completely falls apart.
15:58–15:59Completely.
15:59–16:02You know, this raises such an interesting behavioral question, though.
16:02–16:02What's that?
16:02–16:07Why go through all the trouble? Like, why forge a digital time stamp,
16:07–16:12backdate your blog posts and then get into a public bitter highly technical
16:12–16:17debate on Facebook over a blog. Well because in the fluid constantly shifting
16:17–16:20landscape of the early internet being the undisputed first grants you an
16:20–16:24incredible amount of legacy. Yeah, yes. You become the founding father of a
16:24–16:29digital movement you get the Wikipedia page the citations but more importantly
16:29–16:32I think Anshunhar is highlighting a massive vulnerability in how we record
16:32–16:36history today. Because it's so easy to fake. Because it relies entirely on the
16:36–16:39digital literacy of the public.
16:39–16:40Think about it.
16:40–16:42If the general public doesn't know the mechanical difference
16:42–16:46between a display date that someone just typed on a page
16:46–16:50and an automated URL timestamp generated by a server,
16:50–16:52they will just believe the lie.
16:52–16:54Yeah, if it says 1999, they believe 1999.
16:54–16:55Right.
16:55–16:58It's a stark warning about how easily digital history
16:58–16:59can be rewritten by bad actors
16:59–17:02if nobody checks the underlying code.
17:02–17:03That is genuinely unsettling.
17:03–17:06It really means our history is only as secure
17:06–17:08our understanding of the software that hosts it.
17:08–17:09Exactly.
17:09–17:14But you know, this actually transitions perfectly into the next major hurdle the book covers
17:14–17:19because while proving who published first is great drama, there's a much bigger, much
17:19–17:21quieter struggle happening in the background.
17:21–17:22We have to talk about the infrastructure.
17:22–17:27Yes, the invisible wall because you can't have a vibrant web journalism scene if the
17:27–17:30keyboard literally does not recognize your language.
17:30–17:31Right.
17:31–17:34We just talked about crudy dev and how it hacked the English keyboard to show Devan
17:34–17:38But Mathili has its own unique historical script, right?
17:38–17:39It isn't just Devanagari.
17:39–17:40It does.
17:40–17:45While it is commonly written in Devanagari today, its traditional historical script is called
17:45–17:47Terhuta, or sometimes Mythilakshar.
17:47–17:48Okay.
17:48–17:51And Terhuta is visually and functionally distinct.
17:51–17:55It has different curves, different ways the consonants combine with vowels, which are
17:55–17:56called conjuncts.
17:56–18:00You can't just slap a Devanagari phon hack on it and call it a day.
18:00–18:02It needs its own bespoke architecture.
18:02–18:09And the author points out that the very first digital Trujuta font was created in 2003 by a man named C.K. Rout
18:09–18:16Which is a fascinating footnote by the way because Anshun Har mentions that C.K. Rout is now a very prominent youth political leader in Nepal
18:16–18:17Oh really?
18:17–18:24Yeah, it just shows you the caliber of people who are dedicating their time to this early deeply technical foundational work. Wow
18:24–18:28But you know creating a font so you can type a word document is really only step one
18:28–18:38The internet is navigated through browsers and the book details this massive milestone in 2012 when Mozilla Firefox was fully localized from a Thilly.
18:38–18:43And this was spearheaded by a team led by Sangeeta Kumari and Rajesh Ranjan.
18:43–18:48And we really need to pause and think about what localizing a browser actually means mechanically.
18:48–18:52Yeah, it's not just translating the word file or edit, right?
18:52–18:57Far from it, a web browser has thousands and thousands of strings of text.
18:57–19:02Every single menu, every dropdown, every obscure settings page, every privacy warning, every
19:02–19:07error code like for a whore page not found or SSL certificate invalid.
19:07–19:08Oh my God.
19:08–19:11All of that had to be manually translated into monthly so that a user could actually
19:11–19:13navigate the internet natively.
19:13–19:14That sounds exhausting.
19:14–19:15Yeah.
19:15–19:19And that introduces a really fascinating dynamic that Ankeenhar explores throughout the text,
19:19–19:23which is the difference between how these massive translation projects are actually funded and
19:23–19:24organized.
19:24–19:25Yes.
19:25–19:28He draws a very sharp line between paid projects and community projects.
19:28–19:32It's a crucial distinction for understanding how digital infrastructure is actually built
19:32–19:34in the developing world.
19:34–19:40The author uses the Microsoft Bing Translator Project as a prime example of a paid project.
19:40–19:41Right.
19:41–19:46And this was led by Dr. Girishnath Ha, a JNU, Jawaharlal Nehru University.
19:46–19:47Correct.
19:47–19:53in a paid institutional model, linguists, developers, data entry workers, they are all compensated
19:53–19:55financially, they are doing a job.
19:55–19:56Which is great.
19:56–20:02It is, but the trade off is that the massive corporation, Microsoft in this case, ultimately
20:02–20:07owns the proprietary algorithm and keeps the overarching credit.
20:07–20:09It's a corporate product at the end of the day.
20:09–20:13But on the other side of the spectrum, you have the community projects, and these are
20:13–20:19just staggering to read about because they rely entirely on unpaid crowd sourced volunteer labor.
20:19–20:24It's just people doing grueling technical work purely out of a sense of cultural duty.
20:24–20:28The prime example of this, which the author details quite extensively, is the creation
20:28–20:33of the Mathilli Wikipedia. The Wikipedia epic. This blew my mind. Because I think we take
20:33–20:37Wikipedia for granted. You just assume every language has one. You go to the drop-down
20:37–20:43menu and there it is. But getting an official Mathy Wikipedia approved wasn't just a matter
20:43–20:47of sending an email to the Wikimedia Foundation and asking nicely.
20:47–20:47Not at all.
20:47–20:53It started in 2008, again led by Gojendra Thakur, the guy who made the first GeoCity site.
20:53–20:58They had to enter what Wikipedia calls the incubator phase.
20:58–21:00I'll explain that because it sounds intense.
21:00–21:05It is. The Wikimedia Foundation doesn't just hand out empty domains to anyone who asks.
21:05–21:10To get an official language Wikipedia, a community has to prove two things.
21:10–21:14First, that the language has a dedicated active base of contributors.
21:14–21:19And second, that they can generate a significant volume of high quality standardized content.
21:19–21:22Right. You have to build a house before they give you the address.
21:22–21:24Exactly. You have to prove you can maintain it.
21:24–21:27And the sheer volume of work required is mind-numbing.
21:27–21:30The author points out one individual, Humech Mandal,
21:30–21:35who was responsible for creating roughly 70% of the early pages
21:35–21:40during this incubator phase. 70% and that brings up a massive mechanical challenge.
21:40–21:44How do you translate the sum total of human knowledge into a language that might not have
21:44–21:49established vocabulary for modern concepts? Right. I was wondering about this. If Omish
21:49–21:55Mandal is sitting at his computer, translating a Wikipedia article about, say, quantum physics
21:55–22:01or the internal combustion engine or even just internet terminology like hyperlink or download,
22:01–22:04what does he do if there is no traditional methodally word for it?
22:04–22:08This is where translators essentially become linguistic engineers.
22:08–22:09They have to make choices.
22:09–22:14They either borrow words from English or Hindi and standardize how they're spelled in Miphili,
22:14–22:18or they have to look at the root structures of their own language and coin entirely new words
22:18–22:20that make grammatical sense.
22:20–22:21That is a huge responsibility.
22:21–22:22It is.
22:22–22:24Ohmesh Mandal wasn't just translating facts.
22:24–22:30He was actively expanding the modern functionality of the language in real time.
22:30–22:31And doing it all for free.
22:31–22:34Just sitting there for hours, days, years.
22:34–22:39It took six years of this relentless, unpaid petitioning and translating before the
22:39–22:42mathily Wikipedia finally received official approval in 2014.
22:42–22:46It is the ultimate example of grassroots digital survival.
22:47–22:52And honestly, it perfectly bridges the gap to what the author sees happening at the absolute
22:52–22:54cutting edge today.
22:54–22:56Which is the artificial intelligence revolution.
22:56–22:56Exactly.
22:56–22:59Yeah, Anshinghar doesn't just leave us in 2014.
22:59–23:01He brings the history right up to the present.
23:01–23:08He discusses the work being done by AI for her at specifically their work out of the Naila Khani Center at IIT Madras
23:08–23:14Right. They have created this AI model called Indie trans to which is incredible
23:14–23:19It handles translation speech recognition and transliteration for 22 different Indian languages
23:19–23:26Including maythili. I really want to drill down into the transliteration aspect because I think people often confuse that with translation
23:26–23:27And they are very different
23:27–23:32How does Indie trans too handle transliteration, say, between Davenigari and the traditional
23:32–23:33T'houda script?
23:33–23:38Okay, so translation is changing the meaning from one language to another, like changing
23:38–23:42the English word hello to the French word bonjour.
23:42–23:45Transliteration is keeping the exact same spoken words, but swapping the alphabet they're
23:45–23:46written in.
23:46–23:50Okay, give me an analogy, because that sounds tricky.
23:50–23:54Imagine you have a live interpreter listening to an English speech.
23:54–23:56instead of translating it into Russian,
23:56–23:58they are writing down exactly what the English speaker
23:58–24:02is saying phonetically using the Russian Cyrillic alphabet.
24:02–24:04So if you read it out loud, it still sounds like English,
24:04–24:07but visually it's a completely different script.
24:07–24:08Oh, that makes total sense.
24:08–24:12So in Detrans2, using integrated tools like Aksharamuka
24:12–24:14can take an entire news article written
24:14–24:17in the Devanagagari script and instantly
24:17–24:19transliterate it, swap out the visual architecture
24:19–24:21into the traditional Tirhuta script.
24:21–24:21Wow.
24:21–24:24It automates the agonizing manual labor
24:24–24:28that people like CK Rao and Umesh Mandal had to do by hand a decade ago.
24:29–24:31It's a quantum leap in infrastructure.
24:31–24:32It really is. Yeah.
24:32–24:33I mean, think about the progression.
24:33–24:37First, you have the chaotic Wild West where nobody can even type the language.
24:37–24:42Then you have the manual painstaking era of hacking English fonts and crowd
24:42–24:44sourcing encyclopedias page by page.
24:44–24:47And now you have AI that can seamlessly translate and
24:47–24:50transliterate across dozens of languages instantly.
24:50–24:52So the infrastructure is finally robust.
24:52–24:54The train tracks are officially laid down.
24:54–24:59Exactly, but laying the tracks is really only half the story.
24:59–25:05The question then becomes, once the trains can run, what do these pioneers actually build?
25:05–25:07Which brings us to the empires.
25:07–25:12The massive publications that define the era of Methyli Web Journalism.
25:12–25:15And according to the author, there was one absolute juggernaut that towers over the
25:15–25:18rest and it's called Vedea.
25:18–25:21Vedea is the foundational pillar of this entire history.
25:21–25:25And it was born out of that very first year 2000 website, Balsaric Gac.
25:26–25:31Videa evolved and eventually launched its first official issue as a comprehensive digital journal
25:31–25:35in 2008. Again, under the leadership of Kajendra Thakur.
25:35–25:40And what makes Videa so critical isn't just its longevity, though that is impressive. It's the
25:40–25:45aggressive pursuit of global academic legitimacy. The author highlights that Videa became the
25:45–25:49the first-monthly e-journal to obtain an ISSN.
25:49–25:51An international standard serial number.
25:51–25:53Now, for the listener who isn't, you know, a librarian,
25:53–25:56why is an ISSN such a big deal for a digital blog?
25:56–25:59It is the difference between a self-published Zeme
25:59–26:02and a recognized academic journal.
26:02–26:05An ISSN is an eight-digit code used globally
26:05–26:08by libraries, universities, and researchers
26:08–26:12to identify and catalog serial publications.
26:12–26:15Getting one is a strict bureaucratic process.
26:15–26:17By securing an ISSN,
26:17–26:19Vidya was essentially forcing the global
26:19–26:22academic establishment to recognize them.
26:22–26:23They planted a flag saying,
26:23–26:26we aren't just a niche corner of the internet,
26:26–26:29we exist in the same global registry as the New York Times
26:29–26:30or the Harvard Business Review.
26:30–26:32That is such an incredible power move.
26:32–26:33It really is.
26:33–26:36And their ambition didn't stop at bureaucracy.
26:36–26:38The book details their commitment to inclusivity,
26:38–26:40which is honestly staggering.
26:40–26:43But they didn't just publish in one standard script.
26:43–26:45They eventually published in 12 different scripts.
26:45–26:4712 scripts for a single language journal.
26:47–26:48Right.
26:48–26:51Devanagari, Terhuda, Kathi, Noori.
26:51–26:53And they even created a Braille blog
26:53–26:54for the visually impaired.
26:54–26:57Which is a fascinating technical achievement in itself.
26:57–26:57Yeah.
26:57–26:59How does a digital Braille blog actually work?
26:59–27:02Cause obviously you can't touch a computer monitor.
27:02–27:02Right.
27:02–27:05It relies on screen reader software interfacing
27:05–27:08with a physical, refreshable Braille display.
27:08–27:09What's that?
27:09–27:13piece of hardware that sits in front of the keyboard and actually pushes physical pins
27:13–27:19up and down to form braille characters as the user navigates the digital text on the
27:19–27:20screen.
27:20–27:21Oh wow, that's amazing.
27:21–27:26So by formatting their content specifically to be cleanly read by these systems, the
27:26–27:32DAO was ensuring that even the most marginalized members of their linguistic community weren't
27:32–27:34left out of the digital revolution.
27:34–27:37And speaking of the marginalized, Angie and Harp points out that this desire to
27:37–27:42bring the margins to the center wasn't just technical. It was editorial too.
27:42–27:42Yes.
27:42–27:46Videa created an initiative called the Videa Mathila Ratna.
27:46–27:50This is such a prime example of how digital media can disrupt historical
27:50–27:51gatekeeping.
27:51–27:55Exactly. The author explains that traditional print critics and historians
27:55–27:58in the region had largely ignored lower caste historical figures.
27:58–27:59Yeah.
27:59–28:02There was a very tight grip on the visual history of the culture.
28:03–28:06Like if the print magazines didn't commission a portrait of someone,
28:06–28:09Their face was essentially lost to history.
28:09–28:15But the DEA used their digital platform to bypass those print gatekeepers entirely.
28:15–28:17They didn't need permission.
28:17–28:23They actively commissioned and published portraits of historically ignored figures, like Banthach M.R.
28:23–28:28They used digital art as a tool for social justice and historical correction.
28:28–28:32This is the true promise of the internet that we talked about at the very beginning.
28:32–28:34Yes, it has a dark side.
28:34–28:41But it also completely removes the middleman if the traditional publishers won't print your script or tell your history
28:41–28:42You don't need their printing press anymore
28:42–28:48You just build your own digital platform and do it yourself and this explosion of independent platforms led to a massive
28:49–28:54Cultural boom especially in literature and I have to mention the author of our source text Ashish Ankenhar
28:55–28:58He wasn't just an observer and all this he was an active participant
28:58–28:59Yes
28:59–29:01He ran a highly influential blog called Unchinhar Aker
29:01–29:06which became the absolute epicenter for a very specific literary movement.
29:06–29:07Mateli Ghazals.
29:07–29:11Right. The text notes that his blog hosted online, Musharas.
29:11–29:16And a Mushara for context is a traditional poetic symposium.
29:16–29:20It's usually held in person where poets gather to recite and the audience responds.
29:20–29:25Right. But Anshin Haar digitized it. They held these gatherings entirely online.
29:25–29:29And it was so impactful that traditional critics actually started referring to this
29:29–29:35period as the Anshin Har era of poetry. It's amazing. By removing the geographic barriers,
29:35–29:40they brought in dozens of new poets who had never published before, simply because this digital
29:40–29:46stage now existed. It created a completely self-sustaining ecosystem of art and criticism.
29:46–29:50And we can't forget the hard news side of the spectrum either. The author points to
29:50–29:55platforms like Mathilomir, which was a major news portal started by Pradeep Shah, Lola Narayan Shah,
29:55–30:00and Rahul Rey and the book includes this incredible human detail. Rahul Rey
30:00–30:04literally donated his own personal laptop just to give the project a physical
30:04–30:07machine to launch from. I mean that is the literal definition of grassroots
30:07–30:13journalism. It is. And these early news blogs like Samadhiya back in 2004, they
30:13–30:18were basically inventing the format of online regional reporting on the fly.
30:18–30:21And Antion Har makes a really crucial point about the broader impact of
30:21–30:27of these news and literary portals. They weren't just posting updates, they were actively shaping
30:27–30:29the language itself.
30:29–30:30I don't know.
30:30–30:35Well, by bringing readers and writers from India and Nepal onto the exact same platforms,
30:35–30:40they helped standardize the grammar across regional variations, and they also acted as
30:40–30:41a watchdog.
30:41–30:44Oh, the plagiarism exposing. I love this part.
30:44–30:47Exactly. In the early days of the internet, some traditional print writers thought the
30:47–30:48the web was invisible, right?
30:48–30:50Like nobody was looking.
30:50–30:53So they would blatantly plagiarize content from blogs
30:53–30:55and publish it in print or vice versa.
30:55–30:58But these digital platforms ruthlessly exposed
30:58–31:00the literary thieves,
31:00–31:01proving that the digital archive
31:01–31:03was permanent and searchable.
31:03–31:05You know, when you zoom out and look at the whole timeline
31:05–31:08Anshinar presents, it's just a wild, exhausting
31:08–31:10and honestly beautiful trajectory.
31:10–31:11It really is.
31:11–31:14You start in the chaotic neon days
31:14–31:16of Yahoo Geo Cities in 2000.
31:16–31:20You navigate the petty, forged timestamp dramas
31:20–31:23and the Facebook fistfights.
31:23–31:25You watch volunteers like Umesh Mandal
31:25–31:28sit at a computer translating thousands of Wikipedia pages
31:28–31:31for free because there was literally no other way.
31:31–31:33And you finally arrive at today,
31:33–31:36an era of AI translation and sprawling,
31:36–31:39multi-script digital empires
31:39–31:41with international serial numbers.
31:41–31:43Every single step was a fight against friction.
31:43–31:45And I think that's the ultimate takeaway
31:45–31:46from this deep dive.
31:46–31:50the internet with all its addiction and fake news didn't automatically save
31:50–31:54Mithili. Passionate, stubborn people use the internet as a tool to save it,
31:54–31:57because a language only dies if its people stop innovating.
31:57–31:58Well said.
31:58–32:02We want to thank you, the listener, for joining us on this exploration.
32:02–32:05It's a niche corner of internet history, sure, but the mechanics of what they
32:05–32:08achieved apply to absolutely everything we do online today.
32:08–32:11But as we close, I want to leave you with one final thought to mull over,
32:11–32:15building on that transition to artificial intelligence we just discussed.
32:15–32:22We've just spent this entire deep dive marveling at the sheer, messy, passionate human grit
32:22–32:25required to force a language onto the internet over the last two decades.
32:25–32:29The Facebook fights, the late nights translating Wikipedia, the donated laptops.
32:29–32:36Exactly. It was a deeply human struggle. But right now, models like Indeed Trans 2 and ChatGPT
32:36–32:41are becoming perfectly fluent in almost every language instantly. So looking forward,
32:41–32:45Will this dramatic human struggle to preserve language become a thing of the past?
32:46–32:50Will AI be the ultimate frictionless savior of global linguistic diversity?
32:51–32:57Or by removing the struggle, will it strip away the beautiful, petty, deeply human drama that
32:57–32:59makes a language truly alive in the first place?
32:59–33:03Wow. Remember that ancient beautiful manuscript we talked about at the beginning?
33:03–33:05The one that needed to be encoded into ones and zeros?
33:05–33:08It turns out the ones and zeros were the easy part.
33:08–33:12The hardest part and the most fascinating part was the people fighting over who got to press the keys.
33:12–33:16Keep exploring and we'll see you on the next Deep Dive.

Plain text

So, picture this, it's 2018 and there's this absolutely vicious, just highly public digital fistfight breaking out on Facebook. Yeah, a full on brawl. Right. And it's over what, I mean, on the surface seems like a really simple question, right? Like, who published the very first Mephilim language blog? Which sounds so niche, but the stakes for them were incredibly high. Exactly. And to prove his case, one of these guys essentially has to become like a digital forensic detective. I mean, he's digging into the invisible architecture of the internet itself. Yeah, he's exposing fake URL timestamps, forged publication dates, all that stuff. Yeah, and even debunking these really technical claims about like 1990s font software. It is wild. So welcome to the deep dive everyone. Today we are exploring the surprisingly cutthroat, fiercely passionate, and just incredibly complex world of Mathilli web journalism. It really is a fascinating world. And it perfectly illustrates this much larger phenomenon because what we're looking at is a transnational language community. Right, because Mathilius spoken in a few different areas. Yeah, primarily in Bihar and Harkhand in India and then also the Eastern Terai region of Nepal. And we're basically watching this entire community step onto the completely uncharted, volatile territory of the early internet. And that is exactly our mission today. We're diving into a really grounded, super detailed source text today. It's called A History of Methili Web Journalism, and it's written by Ashish Anshar. It's a great book, really meticulous. It is, and we're gonna use it to figure out how this linguistic community literally fought for space online, you know, the immense technological brick walls they kept slamming into, and the sheer grass roots unpaid labor that was required to digitize an entire culture. Because honestly, if you think about it, this isn't just about one language in South Asia. Right. It's a blueprint. It's a case study for how any traditional culture survives the digital age. Yeah, exactly. Oh, let's untack this because before we get into the forge timestamps and the Facebook showdowns, we kind of need to understand the environment these pioneers were walking into. Yeah, you have to set the stage. Right. And the author, Anshinhar, he spends a good chunk of time just setting up the sociology of the internet itself. He does, and I think it's completely necessary because you really can't understand the magnitude of putting a regional language on the web without understanding what a chaotic paradigm shift the web actually was in the first place. Oh for sure. So Anshin Har establishes this massive global and Indian timeline just to show how quickly everything changed. I mean we go from the US Defense Department tying computers together with Arpanet in 1969. Right, to super early days. Yeah and then to Tim Berners-Lee creating the World Wide Web in 1989. But the Indian context is really where the clock starts ticking for our specific story. Yeah the book notes that the the internet arrived in India via Urnet, which is the educational and research network, around 1986. But I imagine that wasn't something like a teenager in Bihar was just logging into from their bedroom. No, not at all. No. Urnet was strictly a closed loop. It was for the government and top academic institutions only. Okay. So highly restricted. Very. The real turning point, the one they author highlights as the big bang is August 15th, 1995. 1995. Yeah, that is when the general public finally got access through VSNL, which is the Vadesh Sanjharni Gump Limited. August 1995. I mean, it's wild to think that's barely three decades ago. No, it feels like ancient history, but it's really not. Right. And the author points out how the internet in India didn't just grow, it basically detonated, especially as we moved into the 2000s and 2010s. Oh yeah, an absolute explosion. It went from this very slow, you know, dial-up novelty where you couldn't use the landline phone to the absolute backbone of daily life, especially for the youth. I mean, banking, education, booking a train ticket, suddenly it all migrated online. It became essential. But what's really fascinating here is that Incheonhard doesn't write this as a purely triumphant tech bro narrative. No, he really doesn't. He takes a highly critical, deeply sociological look at what this new frontier was actually doing to people. Like, he explicitly brings up internet addiction disorder as a societal crisis that just Arrived right alongside all the convenience. Yeah, he really doesn't gloss over the ugly stuff at all, which I appreciate it So he talks about people essentially living their entire lives through screens Spending up to 15 hours a day online and he goes into specific societal damages, too Yeah, he pulls no punches like he notes how the massive proliferation of online pornography started You know destroying marriages and creating deep psychological rifts in society, right? And he even points to the rise of cyberterrorism, where warfare suddenly became invisible. You didn't need a physical army anymore, you just needed bugs, viruses, and a keyboard. And he also delves into the political weaponization of the internet, which is a huge theme. Oh, the IT cells. Exactly. The text specifically calls out the creation of political IT cells. And for the listener, these are basically organized groups using these massive call center-like setups to intentionally manufacture and spread fake news. Wow. Yeah, they use it to incite riots and manipulate public perception on a massive scale. He's really painting a picture of a digital space that is incredibly hostile. Hustle, yeah. Yeah. And he also talks about the psychological shift in how we value things, right? Like the whole validation economy. Yes. This is one of his most crucial points, I think. And to illustrate this, the author brings up Justin Rosenstein. Wait, isn't he the guy who invented the Facebook like button? He co-created it, yes. Okay, yeah. And the author points out the heavy irony here. Rosenstein later actually deleted the Facebook app from his own phone because he was so terrified of the psychological feedback loop he helped create. That is so telling. Right. Anshinhar observes that in this new digital space, the like button basically created a brand new currency. Digital validation literally became wealth. Which completely changes human behavior. I mean, the author mentions how this metric driven culture turned artists, poets, writers just into clout chasers. Precisely. Like, think about it. If you are a Mathili poet in, say, 1990, you write a poem based on your internal artistic drive, right? Or maybe the feedback of a very small peer group in your village. Yeah, it's intimate. Exactly. But Anshinhar notes that once you put that poet on the internet, the quality of the art is suddenly judged by a highly visible public number of likes. Right. The metric is right there for everyone to judge. And it causes genuine anxiety. It actually alters what people choose to create because they start favoring whatever gets the most immediate viral reaction. Okay, I have to pause here because reading this brought up a massive contradiction for me. Oh, what's that? Well, if the author recognizes that the internet is this, you know, this toxic casino of fake news cyber warfare and desperate validation seeking, then why even put Mathili on it? Like it seems totally counterintuitive. If you have this rich, nuanced, centuries-old literary tradition, isn't digitizing it just throwing it to the wolves? I mean, it's like building a beautiful, quiet library in the middle of a chaotic amusement park. That is the ultimate paradox, isn't it? And what's great is the author faces it head-on. He doesn't shy away from that contradiction. How does he justify it? Well, he argues that the internet has a dual nature. makes people incredibly knowledgeable but simultaneously incredibly ignorant. Wait, how does that actually work in practice, knowledgeable and ignorant at the same time? Do you think about the mechanics of a search engine? Okay. When you look up a historical fact, the algorithm doesn't just hand you a beautifully bound, peer-reviewed encyclopedia. Right, it gives you millions of links. Exactly, it hands you the truth. But right next to it, maybe one link down, it hands you highly visible, extremely well-funded, fake news or wild conspiracy theories. Oh, I see. And because both results are formatted exactly the same way, you know, a clean black headline, a blue link, a little snippet of text critical thinking is naturally eroded. The digital authority looks identical for both the truth and the absolute lie. Wow. Yeah. That's terrifying when you put it like that. Okay, so if the environment is that corrosive to the truth, I have to go back to my question. Why build your cultural home there? Because simply put, the alternative is extinction. Extinction. Yes. Anshinhar makes it very clear. Avoiding the internet is just no longer a viable option for any culture. If a young person in 2010 searches for a Mathili poem or the history of a specific region in Bihar and the internet returns zero results... Then it just doesn't exist for them. Exactly. To that young person, that history simply ceases to exist. It drops entirely out of the modern consciousness. Therefore, establishing a pristine, factual, and robust digital presence wasn't about, you know, wanting to join the chaotic amusement park. It was basically a hostage situation. Digitize or disappear. Wow. Okay. So the stakes literally couldn't be higher. It's a literal battle for cultural memory. Exactly. And that brings us right back to the drama we started the deep dive with. Because once people realize this digital land grab is essential for survival, the race to plant the first flag begins. The rush is on. And human nature dictates that everyone wants to be recognized as the first explorer in the history books. Manually. And Anshan Har spends a significant portion of his research definitively sorting out this timeline because in a digital void establishing who laid the very first brick is a matter of immense historical pride. Right. So according to his digital archaeology, the absolute first The Thilly Presence on the Internet was a website called Balsaric Gah. Balsaric Gah, yes. And it was hosted on Yahoo Geo Cities in the year 2000. Yahoo Geo Cities. Man, for listeners who might not have been around for that era, you really have to understand what a wild west that was. Oh, it was chaos. It was a platform where absolutely anyone could claim a little plot of digital city space and build a very rudimentary, usually visually chaotic website. Oh yeah, we're talking neon green backgrounds. spinning globe gifs, the sparkly tax, the whole nine yards. Exactly. So this site, Palsarkak, was built by a man named Gajendra Thakur. Gajendra Thakur. Yes. And the book notes that this initial site was followed by a few other pioneers. There was a site called Palav Mithila in 2003, and an early news blog named Samadiyah in 2004. Okay, so that's the established timeline. But anytime you have someone claiming to be the first in history, you're inherently going to have challengers. Oh, always. Which brings us to the controversy. Here's where it gets really interesting. A rival blog called Kutek Rasbat, which was operated by Kumar Pabmanab and a few associates, they stepped forward and basically said, no, we were first. Right. And to prove it, they pointed to posts on their blog that explicitly displayed publication dates of 1999 and 2004. Now, if that were true, it would completely rewrite the timeline of the language's digital history. Tucker wouldn't be first anymore. Right. But Anshinhar, he doesn't just look at the dates typed on the screen and not his head, you know. He pulls the curtain back on how blog platforms actually function. He investigates the underlying architecture of Blogger, which is the platform Kedekross bot was using. And the forensic breakdown he uses here is just fascinating. He starts by looking at the features Google actually offered on Blogger at that specific time. Right. He points out that the ability to schedule a future post or back data post, Basically that little feature that lets you manually change the display date of a piece of content that wasn't even introduced by Google until years later around 2008. It didn't exist. Didn't exist. But Anshanar goes even deeper than just feature updates. He looks at the URLs. And for the listener, we really should explain why a URL is so damning in this specific context. Yeah please do, because I was a little lost on this at first. How does a web address prove a forgery? Okay, so when you publish a post on a platform like Blogger, The system automatically generates a unique web address, a URL. So the internet knows exactly where to find that specific page. Right, the www.whatever. Exactly. And the way Blogger's automated system worked back then was by embedding the actual year and month of creation directly into that URL string. Oh, automatically. Completely automatically. So it would look something like, you know, blogspot.com, for-pression 21-307, name of the post. That URL is hard-coded by the server at the exact moment you hit the publish button. Ah, so anyone can go into the text editor, type published on July 1st, 1999, in like huge bold font at the top of the page, and the public will see that display date. Yes. But the server knows the truth. Exactly. It's the digital equivalent of forging a medieval manuscript, but forgetting that the paper you use has a 2013 watermark visible when you hold it up to the light. That's a perfect analogy. The display date is manual and easily faked, but the URL time stamp is systemic. And Anshanhar just caught them red-handed with this. Yeah. I mean, he proves that a post on Catacross Bot that proudly displayed a 1999 date actually had a URL time stamp of July 2013 embedded in his web address. It's undeniable proof. And he went post by post. Another one claiming a 2004 publication date had a 2005 URL time stamp. The server architecture completely 100% exposed the manual manipulation of the historical record. Which naturally led to that massive confrontation you mentioned at the start of the deep dive. Yeah, the 2018 Facebook showdown. Yeah. And I love that Anshin Haar actually includes screenshots of this public debate in the book. He brings the receipts. He does. So Kumar Padmanab, defending his claim, tries to use this highly technical argument to invalidate to invalidate Gijendra Thakur's year 2000 site. Padmanabh argues that it was functionally impossible to build a Mathili website in 2000 because according to him, Devanagari typing tools didn't even exist before 2003. With Devanagari being the primary script that Mathili is often written in today. Right, so Padmanabh is essentially saying you couldn't have built a site in 2000 because you literally couldn't type the alphabet on a computer yet, checkmate. And how does the author systematically dismantle that defense? Because he does? Oh, he destroys it. NGR points out a very specific piece of software history. He casually notes that fonts like KrutiDev had been around since 1997. Ah, KrutiDev. Wow, that is a name that brings back a lot of memories for early Indian internet users. Walk us through that, because I don't know it. Why is KrutiDev the smoking gun here? How did it actually work? Oh, okay, so before we had Unicode. What's Unicode? Unicode is the modern universal standard that allows computers to natively understand thousands of different alphabets. Before that, computers were stubbornly, completely English-centric. They really only understood the Latin alphabet. So software developers in India had to create a hack, and Krudi Dev was a non-unicode font. What it did was visually map Devanagari characters onto the standard English QWERTY keyboard. Okay, so wait, the computers still thought you were typing in English? Exactly. If you press the letter K on your keyboard, the computer internally registered in English K. It had no idea about Hindi or Mathili, but the Kruti Dev font file told the monitor to draw the Hindi character, Ela, on the screen. It was entirely a visual illusion. In fact, if you sent that document to someone who didn't have the Kruti Dev font installed on their computer, they would open it up and just see a string of random English gibberish. That is brilliant, but also sounds so incredibly frustrating. It was. But because Crudy Dev existed and was widely used in 1997, Anshinor proves that people absolutely had the tools to type and publish in Dev and Agari by the year 2000. So Padmanov's technical alibi completely falls apart. Completely. You know, this raises such an interesting behavioral question, though. What's that? Why go through all the trouble? Like, why forge a digital time stamp, backdate your blog posts and then get into a public bitter highly technical debate on Facebook over a blog. Well because in the fluid constantly shifting landscape of the early internet being the undisputed first grants you an incredible amount of legacy. Yeah, yes. You become the founding father of a digital movement you get the Wikipedia page the citations but more importantly I think Anshunhar is highlighting a massive vulnerability in how we record history today. Because it's so easy to fake. Because it relies entirely on the digital literacy of the public. Think about it. If the general public doesn't know the mechanical difference between a display date that someone just typed on a page and an automated URL timestamp generated by a server, they will just believe the lie. Yeah, if it says 1999, they believe 1999. Right. It's a stark warning about how easily digital history can be rewritten by bad actors if nobody checks the underlying code. That is genuinely unsettling. It really means our history is only as secure our understanding of the software that hosts it. Exactly. But you know, this actually transitions perfectly into the next major hurdle the book covers because while proving who published first is great drama, there's a much bigger, much quieter struggle happening in the background. We have to talk about the infrastructure. Yes, the invisible wall because you can't have a vibrant web journalism scene if the keyboard literally does not recognize your language. Right. We just talked about crudy dev and how it hacked the English keyboard to show Devan But Mathili has its own unique historical script, right? It isn't just Devanagari. It does. While it is commonly written in Devanagari today, its traditional historical script is called Terhuta, or sometimes Mythilakshar. Okay. And Terhuta is visually and functionally distinct. It has different curves, different ways the consonants combine with vowels, which are called conjuncts. You can't just slap a Devanagari phon hack on it and call it a day. It needs its own bespoke architecture. And the author points out that the very first digital Trujuta font was created in 2003 by a man named C.K. Rout Which is a fascinating footnote by the way because Anshun Har mentions that C.K. Rout is now a very prominent youth political leader in Nepal Oh really? Yeah, it just shows you the caliber of people who are dedicating their time to this early deeply technical foundational work. Wow But you know creating a font so you can type a word document is really only step one The internet is navigated through browsers and the book details this massive milestone in 2012 when Mozilla Firefox was fully localized from a Thilly. And this was spearheaded by a team led by Sangeeta Kumari and Rajesh Ranjan. And we really need to pause and think about what localizing a browser actually means mechanically. Yeah, it's not just translating the word file or edit, right? Far from it, a web browser has thousands and thousands of strings of text. Every single menu, every dropdown, every obscure settings page, every privacy warning, every error code like for a whore page not found or SSL certificate invalid. Oh my God. All of that had to be manually translated into monthly so that a user could actually navigate the internet natively. That sounds exhausting. Yeah. And that introduces a really fascinating dynamic that Ankeenhar explores throughout the text, which is the difference between how these massive translation projects are actually funded and organized. Yes. He draws a very sharp line between paid projects and community projects. It's a crucial distinction for understanding how digital infrastructure is actually built in the developing world. The author uses the Microsoft Bing Translator Project as a prime example of a paid project. Right. And this was led by Dr. Girishnath Ha, a JNU, Jawaharlal Nehru University. Correct. in a paid institutional model, linguists, developers, data entry workers, they are all compensated financially, they are doing a job. Which is great. It is, but the trade off is that the massive corporation, Microsoft in this case, ultimately owns the proprietary algorithm and keeps the overarching credit. It's a corporate product at the end of the day. But on the other side of the spectrum, you have the community projects, and these are just staggering to read about because they rely entirely on unpaid crowd sourced volunteer labor. It's just people doing grueling technical work purely out of a sense of cultural duty. The prime example of this, which the author details quite extensively, is the creation of the Mathilli Wikipedia. The Wikipedia epic. This blew my mind. Because I think we take Wikipedia for granted. You just assume every language has one. You go to the drop-down menu and there it is. But getting an official Mathy Wikipedia approved wasn't just a matter of sending an email to the Wikimedia Foundation and asking nicely. Not at all. It started in 2008, again led by Gojendra Thakur, the guy who made the first GeoCity site. They had to enter what Wikipedia calls the incubator phase. I'll explain that because it sounds intense. It is. The Wikimedia Foundation doesn't just hand out empty domains to anyone who asks. To get an official language Wikipedia, a community has to prove two things. First, that the language has a dedicated active base of contributors. And second, that they can generate a significant volume of high quality standardized content. Right. You have to build a house before they give you the address. Exactly. You have to prove you can maintain it. And the sheer volume of work required is mind-numbing. The author points out one individual, Humech Mandal, who was responsible for creating roughly 70% of the early pages during this incubator phase. 70% and that brings up a massive mechanical challenge. How do you translate the sum total of human knowledge into a language that might not have established vocabulary for modern concepts? Right. I was wondering about this. If Omish Mandal is sitting at his computer, translating a Wikipedia article about, say, quantum physics or the internal combustion engine or even just internet terminology like hyperlink or download, what does he do if there is no traditional methodally word for it? This is where translators essentially become linguistic engineers. They have to make choices. They either borrow words from English or Hindi and standardize how they're spelled in Miphili, or they have to look at the root structures of their own language and coin entirely new words that make grammatical sense. That is a huge responsibility. It is. Ohmesh Mandal wasn't just translating facts. He was actively expanding the modern functionality of the language in real time. And doing it all for free. Just sitting there for hours, days, years. It took six years of this relentless, unpaid petitioning and translating before the mathily Wikipedia finally received official approval in 2014. It is the ultimate example of grassroots digital survival. And honestly, it perfectly bridges the gap to what the author sees happening at the absolute cutting edge today. Which is the artificial intelligence revolution. Exactly. Yeah, Anshinghar doesn't just leave us in 2014. He brings the history right up to the present. He discusses the work being done by AI for her at specifically their work out of the Naila Khani Center at IIT Madras Right. They have created this AI model called Indie trans to which is incredible It handles translation speech recognition and transliteration for 22 different Indian languages Including maythili. I really want to drill down into the transliteration aspect because I think people often confuse that with translation And they are very different How does Indie trans too handle transliteration, say, between Davenigari and the traditional T'houda script? Okay, so translation is changing the meaning from one language to another, like changing the English word hello to the French word bonjour. Transliteration is keeping the exact same spoken words, but swapping the alphabet they're written in. Okay, give me an analogy, because that sounds tricky. Imagine you have a live interpreter listening to an English speech. instead of translating it into Russian, they are writing down exactly what the English speaker is saying phonetically using the Russian Cyrillic alphabet. So if you read it out loud, it still sounds like English, but visually it's a completely different script. Oh, that makes total sense. So in Detrans2, using integrated tools like Aksharamuka can take an entire news article written in the Devanagagari script and instantly transliterate it, swap out the visual architecture into the traditional Tirhuta script. Wow. It automates the agonizing manual labor that people like CK Rao and Umesh Mandal had to do by hand a decade ago. It's a quantum leap in infrastructure. It really is. Yeah. I mean, think about the progression. First, you have the chaotic Wild West where nobody can even type the language. Then you have the manual painstaking era of hacking English fonts and crowd sourcing encyclopedias page by page. And now you have AI that can seamlessly translate and transliterate across dozens of languages instantly. So the infrastructure is finally robust. The train tracks are officially laid down. Exactly, but laying the tracks is really only half the story. The question then becomes, once the trains can run, what do these pioneers actually build? Which brings us to the empires. The massive publications that define the era of Methyli Web Journalism. And according to the author, there was one absolute juggernaut that towers over the rest and it's called Vedea. Vedea is the foundational pillar of this entire history. And it was born out of that very first year 2000 website, Balsaric Gac. Videa evolved and eventually launched its first official issue as a comprehensive digital journal in 2008. Again, under the leadership of Kajendra Thakur. And what makes Videa so critical isn't just its longevity, though that is impressive. It's the aggressive pursuit of global academic legitimacy. The author highlights that Videa became the the first-monthly e-journal to obtain an ISSN. An international standard serial number. Now, for the listener who isn't, you know, a librarian, why is an ISSN such a big deal for a digital blog? It is the difference between a self-published Zeme and a recognized academic journal. An ISSN is an eight-digit code used globally by libraries, universities, and researchers to identify and catalog serial publications. Getting one is a strict bureaucratic process. By securing an ISSN, Vidya was essentially forcing the global academic establishment to recognize them. They planted a flag saying, we aren't just a niche corner of the internet, we exist in the same global registry as the New York Times or the Harvard Business Review. That is such an incredible power move. It really is. And their ambition didn't stop at bureaucracy. The book details their commitment to inclusivity, which is honestly staggering. But they didn't just publish in one standard script. They eventually published in 12 different scripts. 12 scripts for a single language journal. Right. Devanagari, Terhuda, Kathi, Noori. And they even created a Braille blog for the visually impaired. Which is a fascinating technical achievement in itself. Yeah. How does a digital Braille blog actually work? Cause obviously you can't touch a computer monitor. Right. It relies on screen reader software interfacing with a physical, refreshable Braille display. What's that? piece of hardware that sits in front of the keyboard and actually pushes physical pins up and down to form braille characters as the user navigates the digital text on the screen. Oh wow, that's amazing. So by formatting their content specifically to be cleanly read by these systems, the DAO was ensuring that even the most marginalized members of their linguistic community weren't left out of the digital revolution. And speaking of the marginalized, Angie and Harp points out that this desire to bring the margins to the center wasn't just technical. It was editorial too. Yes. Videa created an initiative called the Videa Mathila Ratna. This is such a prime example of how digital media can disrupt historical gatekeeping. Exactly. The author explains that traditional print critics and historians in the region had largely ignored lower caste historical figures. Yeah. There was a very tight grip on the visual history of the culture. Like if the print magazines didn't commission a portrait of someone, Their face was essentially lost to history. But the DEA used their digital platform to bypass those print gatekeepers entirely. They didn't need permission. They actively commissioned and published portraits of historically ignored figures, like Banthach M.R. They used digital art as a tool for social justice and historical correction. This is the true promise of the internet that we talked about at the very beginning. Yes, it has a dark side. But it also completely removes the middleman if the traditional publishers won't print your script or tell your history You don't need their printing press anymore You just build your own digital platform and do it yourself and this explosion of independent platforms led to a massive Cultural boom especially in literature and I have to mention the author of our source text Ashish Ankenhar He wasn't just an observer and all this he was an active participant Yes He ran a highly influential blog called Unchinhar Aker which became the absolute epicenter for a very specific literary movement. Mateli Ghazals. Right. The text notes that his blog hosted online, Musharas. And a Mushara for context is a traditional poetic symposium. It's usually held in person where poets gather to recite and the audience responds. Right. But Anshin Haar digitized it. They held these gatherings entirely online. And it was so impactful that traditional critics actually started referring to this period as the Anshin Har era of poetry. It's amazing. By removing the geographic barriers, they brought in dozens of new poets who had never published before, simply because this digital stage now existed. It created a completely self-sustaining ecosystem of art and criticism. And we can't forget the hard news side of the spectrum either. The author points to platforms like Mathilomir, which was a major news portal started by Pradeep Shah, Lola Narayan Shah, and Rahul Rey and the book includes this incredible human detail. Rahul Rey literally donated his own personal laptop just to give the project a physical machine to launch from. I mean that is the literal definition of grassroots journalism. It is. And these early news blogs like Samadhiya back in 2004, they were basically inventing the format of online regional reporting on the fly. And Antion Har makes a really crucial point about the broader impact of of these news and literary portals. They weren't just posting updates, they were actively shaping the language itself. I don't know. Well, by bringing readers and writers from India and Nepal onto the exact same platforms, they helped standardize the grammar across regional variations, and they also acted as a watchdog. Oh, the plagiarism exposing. I love this part. Exactly. In the early days of the internet, some traditional print writers thought the the web was invisible, right? Like nobody was looking. So they would blatantly plagiarize content from blogs and publish it in print or vice versa. But these digital platforms ruthlessly exposed the literary thieves, proving that the digital archive was permanent and searchable. You know, when you zoom out and look at the whole timeline Anshinar presents, it's just a wild, exhausting and honestly beautiful trajectory. It really is. You start in the chaotic neon days of Yahoo Geo Cities in 2000. You navigate the petty, forged timestamp dramas and the Facebook fistfights. You watch volunteers like Umesh Mandal sit at a computer translating thousands of Wikipedia pages for free because there was literally no other way. And you finally arrive at today, an era of AI translation and sprawling, multi-script digital empires with international serial numbers. Every single step was a fight against friction. And I think that's the ultimate takeaway from this deep dive. the internet with all its addiction and fake news didn't automatically save Mithili. Passionate, stubborn people use the internet as a tool to save it, because a language only dies if its people stop innovating. Well said. We want to thank you, the listener, for joining us on this exploration. It's a niche corner of internet history, sure, but the mechanics of what they achieved apply to absolutely everything we do online today. But as we close, I want to leave you with one final thought to mull over, building on that transition to artificial intelligence we just discussed. We've just spent this entire deep dive marveling at the sheer, messy, passionate human grit required to force a language onto the internet over the last two decades. The Facebook fights, the late nights translating Wikipedia, the donated laptops. Exactly. It was a deeply human struggle. But right now, models like Indeed Trans 2 and ChatGPT are becoming perfectly fluent in almost every language instantly. So looking forward, Will this dramatic human struggle to preserve language become a thing of the past? Will AI be the ultimate frictionless savior of global linguistic diversity? Or by removing the struggle, will it strip away the beautiful, petty, deeply human drama that makes a language truly alive in the first place? Wow. Remember that ancient beautiful manuscript we talked about at the beginning? The one that needed to be encoded into ones and zeros? It turns out the ones and zeros were the easy part. The hardest part and the most fascinating part was the people fighting over who got to press the keys. Keep exploring and we'll see you on the next Deep Dive.

← Transcript index