Hi, I'm tsu. I build worlds.
Professional Cpellkaster is my "work" blog, collecting my essays on game design and writing.
LeviAul.com has my CV, stories, &c.
No matter what pair of words you try to use to illustrate a difference in pronunciation, there is unfortunately some commonly spoken dialect of English which is specifically designed to defeat you. There are English dialects where "dog" and "cat" have the same vowel sound in the middle. You can't win.
In my experience this doesn't actually help at all, because literally everyone who isn't a trained linguist will immediately start debating which sounds are in fact represented by which IPA symbols, particularly when dealing with pairs of IPA symbols representing phonemes among which their native dialect does not distinguish.
the real "solution" (and i acknowledge this would be a massive pain in the ass on many levels) is to instead include an audio recording of your intended pronunciation.
The trouble with falling back in audio recordings is that if you spent your formative years being exposed exclusively to languages and dialects which do not distinguish between a particular pair of phonemes, there's a good chance that you literally can't hear the difference between those phonemes without special training.
So generate the two sounds using an organic model of vocalization (e.g. Pink Trombone), and then show the diff of their audio waveforms (or perhaps even the Fourier component analysis of their waveforms, after throwing them through Melodyne.)
No, Iâm not literally suggesting that someone who just wants to explain this should do this. Iâm suggesting that a tool that makes this kind of visualization is both feasible and a thing that would be very useful to such people. Good weekend project to learn digital signal processing.
Okay, so we all know the real reason for the vampires-versus-werewolves thing in popular culture is because back in the 1930s, the same studio owned the movie rights to Dracula, Frankenstein and the Wolfman, and they decided to moosh them all together into what is arguably the first Big Stupid Cinematic Universe, but what's slightly less well known is that H G Wells' The Invisible Man was also part of that package. I want to see what the goofy we-swear-it's-personal-horror tabletop RPG based on that facet of the mythos looks like, weirdly artificial taxonomies of playable splats and all â everybody's invisible, but there are like five completely different possible reasons for that, plus a sixth, evil reason for being invisible which you're not allowed to play as because they secretly rule the world.
Realistically the basic problem you run into this is that "turns invisible" is such a flat power that it's impossible to split into multiple Discipline-like powers you can spread across your various sup-splats, unlike "being a vampire" which has tons of manifestations in pop culture.
Now, that's not an unsolvable problem; Werewolf had the same issue, where "turn into a wolf or wolf-man" could be rendered into a suite of shapeshifting with multiple forms, but was still far too narrow to justify an entire splat. Werewolf solved this by squinting really hard until it looked like 'witch/shaman' was inherently connected to the werewolf 'mythos', and proceeded to make Werewolf's actual magical power system a laundry list of various spiritual powers and weird magics obtained through pacts with spirits.
So the issue with making Invisible: the Unseen is that, if you are serious about making it playable, then you're probably going to end up with an entire other mythos shoved into your invisible people that the game swears up and down is actually entirely thematically connected to H.G. Wells's invisible people. Thus the real question is: which one?
My guess is that, because Wells's Invisible Man was a scientist who used 'chemicals' to make himself invisible, then the entire thematic space of 'alchemist'/'mad chemist' is absorbed by the Invisible People for their secondary suite of power that introduces build diversity among player characters.
That is not to say that you're looking at 'Mad Chemist: The Seruming' where the plat is based on alchemy; very importantly, the entire identity of the splat revolves around being invisible people, their alleged trauma on what turned them invisible, and their torment and 'personal horror' around what being invisible does to them. It's just that when you go to spend your XP, you're browsing through 50 pages of magical powers based on questionable late 19th century chemistry that will be the actual focus of your character building, but which the splat pretends really hard isn't what you're here for.
But also, probably youâd want to expand the definition of âinvisibleâ to cover related kinds of imperceptible-ness. True âperfectâ invisibility is only for villains; heroes are instead:
varying translucent
mostly invisible, but their skin flakes, hair, and sweat are still visible
anti-memetic (but only some of the time, as if theyâre periodically being patched in and out of the local timeline)
Too Small
guided by seeming coincidence to just never be where anyone happens to be looking
dressed so embarrassingly that itâs hard for any polite person to not reflexively look away from them
(Now consider a comedy-of-errors game where a party of differently imperceptible people were independently dispatched to a stronghold, meet one-another for the first time already within the compound, and so now must coordinate further infiltration in perfect silence, unable to explain their particular afflictions to one-another, constantly losing track of one-another.)
I love it when farm animals in strategy games have their own little pockets of visibility in the fog of war, because the fog of war is ostensibly a visual abstraction of military intelligence-gathering, which implies that the goat is submitting intel reports.
See, that's even more ridiculous, because it doesn't just have a zone of visibility â terrain it's passed through stays revealed. If we accept the abstraction that revealed terrain represents knowledge possessed by your side's commanders as a result of intel reports, this frames an implicit narrative in which the sheep drew a map.
Alternately: there are implicit civilian populations on these maps, that go un-rendered due to their strategic irrelevance. The intelligence departments of each side pay off the civilians for hearsay about "potential hostiles" â including descriptions of the area of the sighting.
The area descriptions are accurate enough to be useful for mapping.... but these civilians somehow keep confusing sheep for hostiles.
Whey you get your cat a present and it turns out they like the box it came in more, there's no great mystery why: what most cats like better than anything is to be in an Area, and a box is an Area. If you're looking for a gift that your cat is guaranteed to love, get them something that defines an Area.
I presume this is because, if some fragment of one Area becomes (clearly, objectively) delineated into being its own Area, then a cat can claim the newly-formed Area as their own territory, potentially stealing it away from the cat that claimed the original Area it was previously a part of.
Or, in other words, cat territorial relationships are defined by the same abstract concept of "rooms" that Interactive Fiction uses.
A box? New room.
A mat on the floor? Room.
The separate tiers of a cat tree? Each their own room.
Whey you get your cat a present and it turns out they like the box it came in more, there's no great mystery why: what most cats like better than anything is to be in an Area, and a box is an Area. If you're looking for a gift that your cat is guaranteed to love, get them something that defines an Area.
I presume this is because, if some fragment of one Area becomes (clearly, objectively) delineated into being its own Area, then a cat can claim the newly-formed Area as their own territory, potentially stealing it away from the cat that claimed the original Area it was previously a part of.
Or, in other words, cat territorial relationships are defined by the same abstract concept of "rooms" that Interactive Fiction uses.
A box? New room.
A mat on the floor? Room.
The separate tiers of a cat tree? Each their own room.
whenever you ask your friend whoâs really into JRPGs âhey which one do i start with is the newest one fineâ they always go âno, what you gotta do is head on over to ebay and look up the entry that completely bombed commercially. Youâll see that it goes for $800. Close ebay. Download an emulator for a console no younger than 17 years old. Download the ROM, and also this laundry list of various tweaks, retextures, and QoL tweaks. Pull up this exact spoiler-free guide put up a decade ago from GameFAQs, and youâll experience the best game this series has to offer.â Like???
Donât forget that you specifically want the unlocalised Japanese ROM so that you can apply a fan-made English translation patch published in 1998 by a guy with a dick joke for a screen handle because the official localisation was delayed for 20 years due to licensing issues and totally butchers the script by removing all the late 1990s pop culture references.
Tangent from this: why aren't modders + emulator devs working together to build pseudo-console runtimes that are backward-compatible with existing ROMs, but which also have more capabilities than the original hardware, specifically for modders (and homebrew devs) to use?
Like, I don't care if my SMW kaizo hack can run on real hardware; but I do care that it secretly has access to a Real-Time Clock so that it can kill the people playing it on release in completely different ways than the people playing it months later.
Or, for a more clear use-case: why can't we just give JRPGs multi-language softsubs â attaching an attributes file to the ROM that tells the emulator "when the game calls the code that we figured out by disassembly is RenderTextBox(ID), you should overlay this here textbox-looking emulator HUD asset on top of the game's textbox; then take the passed ID, look it up in the currently-active subtitles file, and display the resulting Unicode text in the HUD textbox."
To be clear, when you ask a linguist âis X a word?â, they fully understand that what you mean is âare you aware of any prior attested usage of X to mean Y?â. They are being annoying on purpose.
This is because people end up using the term while trying to mention it, right? But what if you intentionally avoid doing so? For example: "is there a past tense to the verb sense of 'beware'; and if so, is it, like, 'beware' plus '-ed'; or are you supposed to turn the 'ware' into 'wore'; or...?"
The most interesting thing about the present reverse Tumblr exodus is the sharp uptick in posts where somebody reblogs an ongoing discussion with a long addition that isnât actually responding to anything in particular, itâs just rambling at length about their personal feelings regarding the general topic of the discussion. Itâs like theyâve got the basic idea that youâre supposed to engage, but the distinction between commenting and conversing remains elusive, so theyâre just remarking on the thread itself as though they were posting in the comments section of a YouTube video.
People by-and-large arenât trying to converse or comment, though. Theyâre trying to editorialize. Like when someone writes a blog post inspired by a tweet, that embeds said tweet; but in this case, theyâre writing a blog post inspired by a blog post, which embeds said blog post.
Or, to put that another way:
Some people think that linking to a post just isnât good enough, so they slam the whole thing into the body of their post instead. With no intention of the people who wrote the original post being notified, any more than youâd expect writing a book review in a magazine would notify the author of the book.
The reason shitty media tends to be more conductive to robust fandoms isnât because fans have no taste. Itâs because in order to support the kind of fandom that gets loud and in your face on social media, a work needs to leave undefined spaces in its creative framework for fans to fill in.
Mediocre authors often leave such spaces by accident because they donât know what theyâre doing â but if youâve got a well-realised and internally consistent vision for your work, leaving room for fan creativity to breathe becomes an exercise in restraint, and, well, artists tend not to be so great at restraining themselves.
Note that this is not the same as the reductive take thatâs been doing the rounds on Tumblr lately asserting that fan-work merely adds what the source material lacks. Thatâs sometimes the case, to be sure, but a great deal of fan-work is aimed at providing more of what the source material already amply supplies!
The trick is that providing more of the same is subject to exactly the same creative constraints as filling in the gaps: folks are going to be discouraged from doing it if the source material ties things up so neatly that any further elaboration feels inconsequential.
This is also the reason the community gets so involved in creating an endless stream of Super Mario ROM-hacks (and Mario Maker levels), when that isn't seen so much with other game series.
The Super Mario games have a strong tendency to introduce highly generalized level-design elements (obstacles, power-ups, etc. that could in theory work in many different contexts); to then only explore them once or twice, in very constrained contexts; and then to toss them away â before the player has even had a chance to fully understand how the design element works and interacts with other design elements, let alone to have mastered those interactions.
Much of the urge being met with fanwork (including ROM-hacks) is a desire by audience members to flesh out their own incomplete mental model of something introduced in canonâwhether that be a setting element, a character dynamic, or whatever else.
The more complete a mental-model you give your readers/players â the more of the interactions and dynamics you explore of each design element â the less of an ongoing discourse you'll inspire. Thus the adage of "leave your audience wanting more." Thus the worldbuilding advice of "write it all down, in your notes, and then show maybe 5% of it."
It likely wouldnât be terribly difficult to breed domesticated raccoons, but I just canât see that working out well in practice. Like, imagine a thirty-pound parrot with hands. Imagine one living in your house.
âYou are on the island of the MAD DICTATOR. Whenever you go a block west, you are not facing the MAD DICTATORâs statue and must pay a fine. The people on this island understand English perfectly, but are only allowed to respond to questions with âbalâ and âdaâ, and you do not know which means âyesâ and which means âno.â Also, the MAD DICTATOR orders tons of executions of people who canât think of paradoxical answers to questions.â
Thereâs a type of person who loves to munchkin their way âaroundâ a logic puzzle, and who will only begrudgingly attempt to solve it directly once all routes to munchkinry have been eliminated.
âSo you get a fine whenever you arenât in sight of a certain statue...â
âWhy canât I just refuse to pay the fine? In fact, why hasnât this society come together and collectively refused to pay that fine? It seems very impractical; exactly the sort of thing that would be ignored even by the police tasked with enforcement. Can I create a petition to-â
âIndefatigable golem death squads.â
âOh, okay.â
What youâre seeing with the authoritarian framing of logic puzzles, is the result of a memetic selection process, with the versions of puzzles that best force would-be munchkins to get straight to facing the puzzle head-on, beating out the versions that leave avenues of munchkinry open.
This says something interesting, I think, about the type of person who comes up with logic puzzles, in that for this cultural evolution to have occurred, they must have been quite willing to game their puzzles out with known would-be munchkins. Logicians must have a high tolerance for people intentionally missing the point.
A recent discussion on Hacker News around Spotifyâs âmost playedâ charts, made me realize that this is, in fact, a relatively-novel way to conceptualize music popularity; one that tells us something very different about music popularity than the image thatâs been being painted for us by top-40 charts in the past.
The charts like the Billboard Hot 100 which, until 2012, only represented album sales and radio play, give us what is in retrospect a very distorted image of music popularity.
Radio play is entirely under the control of the music industry itself (i.e. the radio stations are in bed with the music labels), so using radio spins to determine a songâs popularity is a lot like using the PR publications of a totalitarian state to determine the views of its citizens. Thereâs really not much signal there.
More interesting to me (and what I didnât put together until now) is that sales-figures for albums/records/singles donât accurately represent the popularity of a song, either. Because some songs are purchased for the sake of owning them (i.e. for social-signalling reasons: to join a subculture, or because a friend recommends it, or for watercooler conversation, or as a way of giving your favorite band money); while other songs are purchased to actually listen to them. The songs people purchase to actually listen to will be listened to a lot more (obviously), but this wonât show up at all in the sales figures.
Likewise, flash-in-the-pan songs and classic songs will see equal numbers of purchases, but over their lifetimes, each purchased classic will see far more plays. A songâs staying power is not represented at all in the record-sales figures.
Combining these two measures really doesnât give us much extra information. Neither radio spins nor record sales have the crucial signal: how many times are people listening to these songs, without just tuning them out as noise?
The charting companies realized this, and so they do now incorporate streaming data into their charting algorithms. But they must still also serve entrenched corporate interests (i.e. the radio industryâs pretended impartiality, where top40 stations play what the charts say is popular, and the charts say whatever the radio industry wants them to say.) So they canât weight the streaming factor very highly. They do enough to say itâs there, and to have highly-viral indie fad songs bubble up temporarily. But they arenât about to allow e.g. a foreign-language song (published by a foreign label!) to chart, where radio stations would then be compelled to play it. Even though raw streaming data would suggest that there have been e.g. several k-pop songs that would have âchartedâ in the US.
What really interests me is not that we now have this more-accurate sense for what people are really listening to, though; but rather that this really points out the paucity of our historical/cultural-anthropological knowledge about what music people really listened to in the past.
Think about it: for the whole of human history until streaming services, the only measures we've been able to use to determine what music people actually liked listening to, was some equivalent to either "radio spins" or "record purchases."
Chamber music? All we know is what copies of musical scores survived. How many times was each score performed? No idea.
Folk songs? We only know how many times they were transcribed. Which is not-at-all representative of a folk song's popularity, given that they're oral traditions. A transcription mostly tells you how interesting the song was to elites at the time.
Hymns? Like radio spins, the "programming" for the day at a church (in most denominations) is determined by the pastorâi.e. an elite. Even when the pastor recorded in a logbook which hymns they chose for performance for each service, this says little about which of the hymns their congregation would have preferred to be singing.
This concept is so novel to me personally that I wouldn't be surprised if it was novel to many historians, despite it being their job to consider this type of thing.
As such, I wonder whether there would in fact be some "low-hanging fruit" in terms of observations that can be gathered to recover data more closely approximating "most listened" songs of the past.
For example, in the eras of wax cylinders and vinyl recordsâwhen playing the medium was itself a minutely-destructive act to the mediumâwe can use wear on the medium's grooves to let us make an educated guess about how many times the given recording was put on. (We can even observe relative wear of different sections of the medium, to see if parts of the recording were being frequently skipped!)
For chamber music, we can observe wear to the sheet music itself, but this has the confounder of how well the score's owner takes care of it. This would be best measured for one score in a book or binder, relative to the others in that same book/binder. The total amount of fingerprint residue on each score (again, relative to others in the same collection) can also be used as a measure. (I imagine this approach also works for other reference collections, like cookbooks; and so is already something cultural anthropologists do.)
I'm not-at-all sure what a valid approach would be for determining true popularity for folk-music or hymns, though. Memetic influence, potentially? Songs people like, that have strange words in them, would teach people those words, and so you'd find the word more in texts that come directly after that period?
Anyway, I just wanted to get you thinking in this mindset: about which songs people actually cared about in the past, rather than simply which songs survived.
Studying the history of locks puts the rogue class in Dungeons & Dragons in sort of a new light because it turns out that prior to around the mid 1800s, the idea of a mechanically secure lock wasnât really a thing.
Basically, the best pre-1800s locks tended to rely primarily on security through obscurity â that is, on implementing elaborate puzzle-box bullshit in order to make it difficult to determine how to access the keyhole and/or withdraw the bolt; with that knowledge in hand, picking the actual mechanism typically required very little skill, to the point that simply having a lockâs maker describe its âsecretâ to you could be enough to defeat it.
Now consider what this means in terms of what your rogue is actually doing when you roll to pick that lock.
âReal-world politics doesnât belong in Dungeons & Dragonsâ like, buddy, what exactly do you think the giant fire-breathing lizard who steals all the landâs wealth, piles it up in a cave and refuses to do anything useful with it is supposed to represent? This is not a subtle metaphor!
Sure you can take dragons as a symbol of plutocracy.
But isnât it more fun to ignore any possible cultural-crit viewpoint, and instead take the things in fantasy settings completely literally?
In a self-consistent setting, either dragons are pitiable creatures, their hoarding tendencies due to OCD-like intrusive thoughts about the safety of commodities ownership; or theyâre basically big bowerbirds, decorating their ânestsâ with gold to woo ladies. Both of these alternatives suggest productive avenues of conflict-resolution between a dragon and the towns it has settled near.
In the former case, the party can resolve the problem by just letting the dragon have all the gold, while decoupling that from any economic impact. Have the party introduce representative currency, collecting gold into a reserve, starting banks which issue promissory notes, etc. The party brings all the gold they collect through this process to the dragon, declare their hoard the âreserveâ, and declare the dragon to be employed by the new central bank as the guard. (The dragon themselves doesnât have to be aware of any of this, but maybe the party can also attempt to teach them some cognitive-behavioural therapeutic techniques for feeling less panic about dwarves mining gold out of the ground. It wonât affect their local economy, youâll assure themâyouâve implemented a stable monetary policy!)
In the latter case, the party would tell the dragon that it would be perfectly fine to borrow all the wealth in the area, if itâs just until they have successfully wooed the object of their desire (who hopefully doesnât stick around much after the wooingâdragons are solitary creatures, right?) and that theyâll see to the details, convincing townspeople to loan out all their wealth (by hook or by crook.) As dragons are more intelligent than your average bowerbird, the local dragonâs prospective mate will likely be looking for signs that the hoard is illusory (i.e. that the dragon isnât as flush as they seem), possibly even interviewing references (the people of nearby towns!) to ensure this gold was actually stolen. Theyâll use all the high-level interrogation and command magics youâd expect a dragon to know. So the party will have to use even better magics to make the illusion perfect.
Afterward, theyâll win the friendship (or perhaps reverse-fealty) of their local dragon. And though theyâd expect enmity from their new mate (who has been, presumably, bamboozled), theyâd not find it, as it turns out dragons tend to be both intelligent and wise and therefore canâat least after a good first impressionâappreciate a partner for their personality.
Honestly, the reason I canât watch a lot of video games on stream is that most streamersâ approach to puzzle-solving is so unsystematic. Like, imagine trying to solve a jigsaw puzzle by dumping it into a pile, selecting two pieces at random without looking at the designs on their faces, and checking to see if they fit together; if they donât, toss them back into the pile, select another two random pieces, and repeat. Thatâs how most video game streamers solve puzzles, and watching it makes me feel like my head is going to explode.
You say most streamers. Do you have any recs for streamers who are good systematic puzzle-solvers, and who play mostly puzzle games (or games that generate puzzle situations, like roguelikes)?
Itâd be a big bonus if they would also âshow their workâ for the audience, rather than just being mystifyingly competent.
Hot take: the authors of fantasy fiction donât actually care about capitalism or monarchies; they really just dislike big societies (which require superstructures like capitalism to function) and want to replace them with small communities/tribes (where there is an obvious winner for âwho of us is best at statecraft and diplomacyââa King Smurf, if you will.)
Note that, at the end of your average fantasy saga, the hero is usually in a close relationship with the monarch, if not a royal themselves. This is what is satisfying: not the power over tons of people that the monarch has, and their ability to be an eminence grise over it; but the shrinking of the tribal world-model the hero needs to keep loaded into their brain to understand the world. This would be just as true if the resulting âkingdomâ were a city-state of 150k people, or a defensible family farm in an anarchocapitalist dystopia, or a pirate ship. Itâs the clear delineation of a small number of âusâ, where âusâ somehow contains power of sovereigntyâand a âthemâ rendered irrelevant by the sovereignty of âusââthat does the trick.
The average fantasy author would tend to put forward the real country of Andorra as a utopia, despite it both not being a monarchy, and not being (technically) self-sovereign. It gets to mind its own business and gets to decide how it do; you can become king (for a day) just by petitioning 24 dudes; and the extremely low violent crime per capita is thought to come down to everyone there just... liking one-another, and getting along (and making bank from tourists. No real poverty.) And yet Andorra gets a seat at the table at the UN. Theyâre living the dream.
