If you'd like an essay-formatted version of this post to read or share, here's a link to it on pluralistic.net, my surveillance-free, ad-free, tracker-free blog:
Six years ago today, after 19 years with Boing Boing, during which time I wrote tens of thousands of blog posts, I started a new, solo blog, with the semi-ironic name "Pluralistic." I didn't know what Pluralistic was going to be, but I wasn't writing Boing Boing anymore, and I knew I wanted to keep writing the web in some fashion.
Six years and more than 1,500 posts later, I am so satisfied with how Pluralistic is going. I spent a couple of decades processing everything that seemed interesting or significant through a blog, which created a massive database (and mnemonically available collection of partially developed thoughts) that I'm now reprocessing as a series of essays that make sense of today in light of everything that I've thought about for my whole adult life, which are, in turn, fodder for books, both fiction and nonfiction. I call this "The Memex Method":
"The Memex Method" is also the title of a collection of essays (from this blog) that I've sold to Farrar, Straus and Giroux, but that book keeps getting bumped because of other books I end up writing based on the work I do here, starting with last year's Enshittification. I'm now fully two books ahead of myself, with The Reverse Centaur's Guide to Life After AI coming in June, and The Post-American Internet in early 2027 (in addition to two graphic novels and a short story collection). Professionally speaking, these are the most successful books I've written, in a long, 30+ book career with many notable successes. Intellectually and artistically speaking, I'm incredibly satisfied with the direction my career has moved in over my six Pluralistic years.
Blogging is ā and always has been ā a lot of work for me, but it's work that pays off, even if I don't always know what form that payoff will take.
One essential part of this blog is my daily retrospective of posts from this day through my blogging history ā 25 years ago, 20 years ago, 15 years ago, 10 years ago, 5 years ago, and last year. I used to call this "This day in history" but now I call it "Object permanence," for the developmental milestone when toddlers gain the ability to remember and reason about things that have recently happened (roughly, it's the point at which "peek-a-boo" stops being fun).
The daily business of reviewing and selecting blog posts from different parts of my life started as a trivial exercise, but it's become one of the most important things I do. I liken it to working dough and folding the dry crumbly edges back into the center; in this case, I'm folding all the fragments that are in danger of escaping my working memory back into the center of my attention.
Six years ago, I didn't know what Pluralistic was going to be. Today, I still don't know. But because this is a labor of love, and a solo project, I get to try anything and either give it up or carry it on based on how it makes me feel and what effect it has on my life. I'm always tinkering with the format: this year, I also added a subhead to the Object Permanence section that tries to call out (in as few characters as possible) the most important elements of the day's list.
I also dropped some things this year, notably, my "linkdump" posts. A couple years ago, at the suggestion of Mitch Wagner, I added a new section called "Hey look at this," which featured three bare links to things I thought were noteworthy but didn't have time or inclination to delve into in depth. Later, I expanded this section to five.
However, even with five bare links per edition, I often found myself with a backlog of noteworthy things. So I started writing the occasional Saturday "linkdump" essay in which I wove together the whole backlog into a giant, meandering essay. These made for interesting rhetorical challenges, as I found elegant ways to bridge completely disparate subjects ā a kind of collaging, perhaps akin to how a mashup artist mixes two very different tracks together. Mentally, I thought of this as "ringing the changes," but ultimately, I decided to drop these linkdump posts (for now, at least). They ended up being too much work, and of little value to me, because I found myself unable to remember what I wrote in them and thus to call them up to refer to them for future posts. Here's all 33 linkdumps; they're not gone forever (not so long as the links pile up in my backlog), but when they come back, they'll be in a different form:
https://pluralistic.net/tag/linkdump/
This really is a labor of love, in the sense that I love doing it, and because it's hard work. The fact that it's hard work is a feature, not a bug. Working hard on stuff is really important to me, because when I am working hard, I gain respite from both physical and mental discomfort. As a guy with serious chronic pain living through the Trump years, I've got plenty of both kinds of discomfort. I can't overstate how physically and mentally beneficial it is to me to have an activity that takes me out of the moment. This year, I wrote several editions of Pluralistic from an infusion couch at the Kaiser Sunset hematology center in LA, where I was receiving immunotherapy for a cancer diagnosis that I'm assured is very treatable, but which ā to be totally honest ā sometimes gets my old worrier running hot:
Making Pluralistic is several kinds of hard work. Over the past six years, I've become an ardent collagist, spending more and more time on the weird, semi-grotesque images that run atop every edition. Anything you devote substantial time to on a near-daily basis is something that gives you insight ā into yourself, and into the thing you're doing. I've always had a certain familiarity with computer image editing (I think I got my start writing Apple ][+ BASIC programs that spat out ASCII art, before graduating to making pixel-art for Broderbund's "Print Shop"), but I've never applied myself to any visual field in a serious way, until now.
Amazingly, after 50 years of thinking of myself as someone who is "bad at visual art," I find myself identifying as a visual artist. I find myself pondering visual works the same way I think about prose ā mentally tearing it apart to unpick how it is done, and thinking about how I could productively steal some new techniques for my own work. I'm also privileged to have some accomplished visual artists in my circle, like my pal Alistair Milne, who generously share technical and aesthetic tips. It's got to the point where I published a book of my art, and I think I'll probably do it again next year:
There's also a ton of technical work that goes into publishing each edition of this newsletter. Things have moved on somewhat since I published an in-depth process-post in 2021, though I'm still totally reliant on Loren Kohnfelder's python scripts that help me turn the XML file I compose every day into files that are (nearly) ready to publish:
This means that after I write the day's post, I reformat it and republish it as a text-only newsletter, a Medium post, a Tumblr post, a Twitter thread and a Mastodon thread. This involves a ton of manual work, because none of the services I post to are designed to facilitate this, so I'm always wrestling with them. This year, all of them got worse (incredibly).
Medium ā where I used to have a paid column ā has dropped its free-flag for my account, which now limits me to how many posts I can schedule. This doesn't come up often, but when I do schedule a post, it's generally because I'm going to be on a plane or a stage and won't be able to do it manually. There's no way I'm going to pay for this feature: I'm happy to give Medium my work gratis, but I will not and do not pay anyone to publish my work, and I never will.
Tumblr did something to its post-composing text editor that completely broke it and I've given up on fixing it. I can't even type into a new post field! I have to paste in some styled text, then delete it, then start typing. It's ghastly. So now I just have a text file full of formatted HTML snippets and I work exclusively in the Tumblr HTML editor, pasting in blobs of preformatted HTML (including the florid, verbose HTML Tumblr uses for its own formatting) and then laboriously flip back and forth to the "visual" editor to see the parts that went wrong. Here's how busted that visual editor is: searching for a word then double-clicking on it does not select it. You have to click once, wait about 1.5 seconds, click again, wait again, and then you can select the word.
Twitter has entered a period of terminal technical decline. I know, I know, we always talk about how fucked Twitter's content moderation is, for obvious and good reasons, but from a technical perspective, Twitter just sucks. If I make a post with an image and alt text in anticipation of later using it to start a thread, it often goes "stale" and will not publish until I delete the image and re-attach it and re-paste the alt text. Meanwhile, the thread editor is also decaying into uselessness. Fill in a 25-post thread and hit publish and, the majority of times, the thread publication will die midway through, displaying lots of weird failure modes (phantom empty posts at the end of the thread that need to be individually selected and deleted are a common one, but not the only one). The old Twitter's ability to add a new thread to an existing one has been dead for at least a year, so every post after the 25th stanza has to be manually tacked on to the previous one, which is made far harder by the fact that Twitter no longer reliably shows you the post you just made after it publishes.
Mastodon still lacks a decent thread editor, one that has even the minimal functionality of Twitter circa 2020. Meanwhile, the Fediverse HOA continues to surface from time to time, with someone who's had a Masto account for ten seconds scolding me for posting threads ā from my account whose bio starts "I post long threads." It's genuinely tedious to be shouted at for "using Mastodon wrong" by someone who started using Mastodon yesterday (I opened my first Mastodon account in 2018!), and even worse when they double down after I point them to the essay I've written to explain why I post the way I do, and what to do if you want to read my work somewhere that's not your Mastodon timeline ("Can you believe this asshole wrote a whole essay to explain why he posts his stupid Mastodon threads?"):
Then there's email: I continue to love email, but email doesn't love me back. After years of being blackholed by AT&T and then Google, this turns out to be the year that Microsoft bounces thousands of messages to its Hotmail and Outlook users because they have arbitrarily and without warning added my mail-server to a blacklist. Thank you to the Fediverse friends who escalated my trouble ticket ā but man, this is a headache I could certainly do without:
https://pluralistic.net/2021/10/10/dead-letters/
My sysadmin, the incomparable and tireless Ken Snider, tells me that he's got the long-overdue new hardware installed at the colo and he's nearly ready to stand up my long-anticipated personal Mastodon server, which will let me solve all kinds of problems. He's also going to stand up my own Bluesky server, at which point I will part ways with Twitter. I wish I could have used the regular Bluesky service while I waited, but just setting up an account permanently binds you to totally unacceptable and dangerous terms of service:
What's the point of a service that has account- and data-portability if signing up for it makes you permanently surrender your rights, even if you switch servers? This might be the stupidest social media unforced error of the post-zuckermuskian era.
There is one technology that has made my POSSE life better, and it might surprise you. This year, I installed Ollama ā an open-source LLM ā on my laptop. It runs pretty well, even without a GPU. Every day, before I run Loren's python publication scripts, I run the text through Ollama as a typo-catcher (my prompt is "find typos"). Ollama always spots three or four of these, usually stuff like missing punctuation, or forgotten words, or double words ("the the next thing") or typos that are still valid words ("of top of everything else").
The reason this is so valuable to me is that errors magnify through each stage of POSSE. Errors that make it through the python publication script take 10x the time to fix that they would if I caught them beforehand. Errors that I catch after running the scripts and publishing the posts take 10x time more. Errors that I have to fix later on ā once I've closed all the relevant tabs and editors ā take 10x again more time. Some POSSE channels (email, Twitter) can't be fixed at all.
So catching these typos at the start of the process is a huge time-saver. I have some very generous readers who have the proofreader's gene and are very helpful in catching my typos (hi, Gregory and 9o6!), and I feel bad about depriving them of their fun, but there's still the odd error that slips through, and they always catch it.
Ollama is a pretty good typo-catcher. Probably half of the "errors" it points out are false positives, which is better than the false positive rate for Google Docs' grammar-checker. As someone who uses a lot of jargon, made up words, etc in his prose, I'm used to overriding my text-editor. I wouldn't simply trust an LLM's edits any more than I would accept every suggestion from a spell-checker. Hell, yesterday I sent back a professionally copyedited manuscript (the intro for the paperback of Enshittification) and marked "STET" on about a third of the queries.
Doubtless some of you are affronted by my modest use of an LLM. You think that LLMs are "fruits of the poisoned tree" and must be eschewed because they are saturated with the sin of their origins. I think this is a very bad take, the kind of rathole that purity culture always ends up in.
Let's start with some context. If you don't want to use technology that was created under immoral circumstances or that sprang from an immoral mind, then you are totally fucked. I mean, all the way down to the silicon chips in your device, which can never be fully disentangled from the odious, paranoid racist William Shockley, who won the Nobel Prize for co-inventing the silicon transistor:
Further, we wouldn't have the packet-switched network that delivered these words to you without the contributions of the literal war-criminals at the RAND corporation:
https://en.wikipedia.org/wiki/ARPANET
Refusing to use a technology because the people who developed it were indefensible creeps is a self-owning dead-end. You know what's better than refusing to use a technology because you hate its creators? Seizing that technology and making it your own. Don't like the fact that a convicted monopolist has a death-grip on networking? Steal its protocol, release a free software version of it, and leave it in your dust:
That's how we make good tech: not by insisting that all its inputs be free from sin, but by purging that wickedness by liberating the technology from its monstrous forebears and making free and open versions of it:
Purity culture is such an obvious trap, an artifact of the neoliberal ideology that insists that the solution to all our problems is to shop very carefully, thus reducing all politics to personal consumption choices:
I mean, it was extraordinarily stupid for the Nazis to refuse Einstein's work because it was "Jewish science," but not merely because antisemitism is stupid. It was also a major self-limiting move because Einstein was right:
Refusing to run an LLM on your laptop because you don't like Sam Altman is as foolish as refusing to get monoclonal antibodies because James Watson was a racist nutjob:
Or to refuse to communicate via satellite because they were launched into space on a descendant of a rocket designed by the Nazi Wernher von Braun and built by slaves in a death camp:
There's plenty of useful things people can do with AI. There's plenty of useful things people will do with AI. AI is bad because it's an economic bubble and a grift, but not because we've created a bunch of utilities that would ā under normal circumstances ā be called "plug-ins":
I started blogging 25 years ago, just before the dotcom bubble popped. That bubble-pop inflicted a lot of pain on people who didn't deserve it, including the normie investors who'd been suckered into blowing their life's savings on dogshit stocks, and everyday workers who found themselves out of a job. But the world was better off. So was the web. With the bubble popped, real, good stuff could access talent, servers and office space.
In the six years I've been doing this, I've seen several bubbles come and go: crypto, web3, metaverse. Now it's AI. But those bubbles were like Enron, frauds that left nothing good behind. AI is like the dotcom bubble, awash in sin and inflicting untold misery, but it will leave something useful behind:
Anya is live and ready to show you everything. Watch her strip, dance, and perform exclusive shows just for you. Interact in real-time and make your fantasies come true.
ā Live Streamingā Interactive Chatā Private Showsā HD Quality
Anya is LIVE right now
FREE
Free to watch ā¢ No registration required ā¢ HD streaming
like, when i'm in this mode of trying to make a thing and it's going sort of well i wind up sort of subsuming most of my attention to it
and i keep accidentally hijacking conversations IRL and god i am struggling not to do that but literally everything I encounter IRL Gives Me An Idea For That Book
(yes, i did send a message on instagram, the only platform i still have contact with him, to my friend the former Navy diver, to ask him for more dolphin insights. yeah. we'll see. it's been years and he got politically Weird [not that way!!! the other way actually don't worry] so i'm not sure how that'll go.)
but anyway, on Character Creation
one of the problems i have is that i have to kind of write things to find them. and so i was going along writing about a particular B-plot problem, and I had a character say something snippy that I then needed an uninvolved person to overhear and react to, for scene pacing reasons (idk, it just seemed right)
and i had needed an original character for another role in the A plot so i was like ah yes, Placeholder can do that and then we segue flawlessly back into the A plot, so I wrote a bunch of that but the character was such a bland placeholder (i devoted zero thought to their name, and was like i guess they'd call themself a consultant ok, and then my brain filled in "Rin" for their name, and I'm a thousand words in before I'm like. this person cannot be named Consultant Rin.) But I got the A plot sequence done and it holds up, and reading back over it, mm this person needs more personality.
I stepped away from the computer and was eating a meal with Dude and talking to him about something unrelated, and then I was musing on how various of his coworkers sound on Zoom calls, and the only one I can tell apart is the woman with a mild speech impediment, and he was telling me more about her and I suddenly in midsentence was like "oh I don't think I've ever written a character with a speech impediment" and like.
there's technical challenges to that, which was interesting to contemplate-- just like writing a character with an accent, where you don't want to descend into like, exotifying dialect but you do want to convey something of the uniqueness of this person's diction whenever they speak. so that's a fun and interesting constraint to put on a character, and can help with a real problem I have always had in my writing, where I love writing dialogue but everyone just talks like me unless I put a ton of attention into making sure they don't.
but this character immediately morphed in my mind, from Placeholder With Stolen Name to
Extremely Autistic Technician With Like, Absolutely No Rizz Who Within Five Minutes Becomes Everyone's Favorite Person
and is not based on anyone I know but I immediately knew and loved them and that is a much better standpoint from whence to create a character.
I also need to come up with names for everybody, and that's on the list, but I'm getting there. I'm just glad I have a mental image now. I don't do visual imagination stuff very well-- fanfiction is nice because I can usually find an image of a character and refer back to that, because I'm not exactly faceblind but I don't hold images well in my mind. With original characters it's so so so so hard, because I can't imagine that well, but it really helps me keep a voice distinctive if I can look at the person and think about them talking. So.
At least having some vague notion rather than like one of those blank wooden poseable artist figure model things helps.
I am committed to lounge piano vibes from here on out.
I was really motivated to continue working on it after I did the first version - lyrics and editing notes under the cut. Tagging @orcusnox as promised - this is now properly song-shaped!
Lyrics
[you're brand new to me
now I'm snared in your smile
you see me like it's my first time] [1]
[and it rings true to me
like we're not out of context
can I call it in if you want this?] [2]
[do you even know me, yes you do
stuck my head in a crowded room
cause every universe has a me and you
and every me got your number on my phone] [3]
[and if I got all this time here with you
and the crowd in my head tonight] [4]
[then it's simple, blink and you'll miss it
thoughts they stick like glitter
maybe it's not the same but
can I tell you a secret?] [5]
[bye bye (bang bang) gone today, but tomorrow---
you'll look like someone I know
someone I know] [6]
[so hey stranger, you know what they say
"came for the money and stayed for the fun"
you had your pick---where do you belong?] [7]
[and if I got the time, I can drift away
but I'll be here when you get home] [8]
[cause it's simple, blink and you'll miss it
thoughts they stick like glitter
maybe it wasn't real but
can I tell you a secret?] [9]
[bye bye gone today, but tomorrow---
you'll look like someone I know
someone I know] [10]
[1] [4-6] improved rhythm and reduced verbosity to let the lines breathe. bane of my existence. the first line of [6] was one of the first lines I originally wrote for this song, but discarded in the previous iteration
[2] new melodic material
[3] unchanged. this was the first full phrase I wrote for this song and I'm attached to it
[7] uses melody from [2] for symmetry. also replaced placeholder lyrics
[8] uses melody from [4] for symmetry
[9-10] slight variation in the chorus
intro/outro is ad-libbed because I needed to start and end it somehow, and lyrically doesn't fit (it's not about leaving, it's about belonging).
given the lyrics, the title for this version cannot be anything other than "someone I know" given the lyrics. if the lyrics change again, title will also change.
Daisy Bell feat. Eleanor Forte AI Lite and ANRI Lite
here it is!! half a year in the making, and it's finally up. i'm really proud that i finally put this together after a long couple months of not being able to work on it.
I also wrote some notes on the costume design of the daisy belles on my ko-fi! because i put a good bit of thought in including elements of both their official character designs and 1890s western women's fashion.
image descriptions in the alt text and under the cut.
ID: two digital collages with yellow backgrounds and brown borders.
(1) eleanor forte as she appears in the daisy bell video, next to an image of her official synthesizer v studio character art and a cropped image of a woman in a shirtwaist, a slim skirt, and a jacket with gigot sleeves. eleanor has long white hair done up in two buns with long pigtails. she is wearing a jacket closed with a ribbon brooch over a pleated blouse, long white gloves, and a slim, cone-shaped skirt, all trimmed with red ribbon. she has a small daisy button pinned to her jacket.
(2) anri as she appears in the daisy bell video, next to an image of her official character art and a cropped image of a woman in a bicycle suit. anri has short brown hair with a fringe. she is wearing a wide red hat and a cycling suit with a full skirt, wide lapels, and daisy buttons. she has a small red ribbon around the collar of her blouse.
wanted to share some of my process from the last art I posted...
The theme started as the future looking at the past, but then I remembered the medals used in the mobile game; that had representations of heroes and villains from the future. So then I imagined an art show event opening up in daybreak town, featuring artistic interpretations of the future as foretold by the foretellers.
It sort of became a sense of the future, past, and present gazing directly at each other, while the other is blissfully unaware.
Also hereās the full detail of the frame. So much work to then cover a big chunk of detail upš I wouldāve wanted to design more options but just making one took so long.
For this tiny part, Iāve joked that this was my favorite detail because of the effort attempting the look of engraved lettering and was so happy when I finally achieved the effect! The caption/label next to a painting can also be called a ātombstoneā I recently discovered š¤
I forgot to note this, so I had to decipher the scala language here again. It says:
Agent of Darkness, as foretold by the Masters.
This was a last minute detail but Iām glad I put it in, or otherwise that spot beneath the painting wouldāve been an awkward blank space š
These details also took a bit of time, but honestly the most soothing parts to work on compared to everything else! I deliberately invoked those āmomento moriā and āvanitasā paintings to add to Marluxiaās death motifs and a sense of foreboding for Lauriam.
Anya is live and ready to show you everything. Watch her strip, dance, and perform exclusive shows just for you. Interact in real-time and make your fantasies come true.
ā Live Streamingā Interactive Chatā Private Showsā HD Quality
Anya is LIVE right now
FREE
Free to watch ā¢ No registration required ā¢ HD streaming
I usually make clips for one episode at a time when Iām restless rewatching, but Iāve arrived at Episode 31 now, so I think Iām going to go ahead and clip 31-33 all together, because these are the ouchiest episodes, for me. (Yi City arc is super ouchy but I really, REALLY love looking at Li Bowen as Song Lan so it balances out)Ā Anyway I figure itāll be like quickly pulling off a bandaid, for 4 hours in a row. Whee!
When it comes time for a critique, there is a good chance that youāre going to give your presentation talking about your most recent model/iteration andĀ the iteration before it. Thereās also a good chance that one of the jurors is going to ask about the origin of your project:Ā āwhere did this action come from? Why this and not this?ā And if youāre like me, thereās a good chance youāre going to blank on everything.Ā
Not everyone will have that issue, some people are able to answer that question immediately, and can draw connections at a momentās notice. I am not that person.Ā
There are three main things I keep in my process notes:Ā
1. Critique notes: Pretty much anything said during my critiques, I write it down, both good and bad. These kinds of notes give me launching points for the next phases of each project.
2. Planning Notes: I generally write these out right after a critique (once my desk is clean, of course). With these, I take a look at my previous work, my critique notes, and come up with a plan of action for the next iteration of the project. Sometimes I include little diagrams, showing what parts of previous models I want to keep, and which ones I need to alter.Ā
Iāve personally included my process journal right within my bullet journal, so thereās also a couple pages I put at the end of it where I can write down different works that I feel I could use for inspiration in my own future work.
Sketching has been good at breaking up the misery of staring at a manuscript and being stuck. At least with the drawing I'm roadblocked by my lack of skill rather than my lack of ideas.
There are things from an adventure I am currently writing for Colin Le Sueurās We Deal In Lead. It began as a homage to Wisit Sasanatiengās tomyamgong western Fa Thalai Chon / Tears Of The Black Tiger.
+++
WIDOW GON'S PALANQUIN
A broad teak throne: canopied, curtained, cushioned. Stinks of tobacco.
Its bearers: the captive brothers Khol. Every night Lady Sao Rai visits their garage, selects a brother, and fucks him in her grandmother's palanquin.
The Khols are too afraid to refuse her.
+
The Widow is matriarch of House Gon. It will be her fiftieth birthday, soon. An elaborate fete is planned.
Captives are found across the sea, created through poverty, criminal sentences, or legal abduction. By Admiralty law, a captive must go free once they earn their owner their original price, a hundred times over.
In practice, few owners obey.
+++
It should be obvious what captives are. I ding-donged with myself about the nomenclature, here.
A simple reason for avoiding the word "slave" is because most people think "transatlantic slave trade" as soon as you say it. If nothing else I want to avoid the association because it is inaccurate.
On the other hand: annoying to have to decenter Southeast Asia in this way! The equivalent of having to say "chai tea" when I should be able to say "tea", because that is what the word means to me!
(I strain against this specific problem often.)
Finally I decided "captive" was good, after all. This kind of legalistic euphemism ("Oh, they aren't slaves, they are indentured servants.") is exactly in character for rich assholes bending language to assuage their consciences.
+++
HOUSEHOLD PSYCHOPHONE
Listening room: settees; shelf of wax-cylinder records; a podium on which sits a psychophone.
Pop a cylinder into the psychophone, point its antenna at a servant wearing the receiving brooch, listen to them sing in an alto entirely not their own.
+
Psychophones have been ruinous to local performers. Once-celebrated local singers have been reduced to glorified loudspeakers: vessels for the voices of famous chanteuses from across the Ocean.
This home entertainment system requires at least two to operate:
One servant (or more commonly a servitor) to turn the crank;
One servant to serve as a receiver-singer.
A receiver-singer's health eventually suffers. When you have somebody else's voice (and soul) forced into you over and over, and you begin to lose your own ā¦
+++
This one was troublesome. Felt like production design. Appliance design.
Had several goals:
The core mechanism has to look like it makes sense, to its own internal logic. No greebling; every bit needs to look like it has a purpose.
Lots of ornamentation. This is a luxury device belonging to aristocrats from a rococo Indochinese-inspired society. It needs to be a jewelbox.
Genteel normalisation of vicious magic. The needle made of bone; the antenna that is basically a massive needle pointed at your head---but disguised as a pretty bird.
The receiver-brooch is something I discovered while sketching. Seems gameable? Also, in the spirit of point 3: the brooch has a pin you stick in your forehead.
+++
GHOST WATER
Auw Yin Yan, the Sea of Sorrows---of Sighs.
Imagine bodies in a mass grave the size of a country. Imagine them luminescent, in motion. Pulled by the moon, waved by the wind, clawing at the quay.
+
Always forms into human shapes: when poured into a bowl, ghost water sits as a balled fist.
Like saltwater in most respects. The Sea teems with marine life, though these are cunning and cruel in human ways. Humans cannot swim ghost water. Do not fall in.
Ghosts wear the outfits and injuries they had at death. Rarely, one will crawl onto land, eyes open, a hungry ghast.
+++
Yeah, so: the wider campaign setting for this adventure is defined by the Sea of Sorrows. It has whales and islands and pirates. It is filled with ghosts instead of water.
I saw the Sea in my mind as a vast Escher-esque tangle of interlocking ghost-bodies.
A wave would be bodies flinging themselves on a beach; their arms and hands dragging on the sand as they pull back into the surf.
I drew a way simpler visual. And the ghost's hair is cheating: it already looks like water.
Still: very pleased with this sketch. Gentle, sort of sweet, quietly creepy. Also it is a modest bailing bucket, which contrasts with the material excess of the palanquin and psychophone.
