Open threads at the Open Thread tab every Sunday and Wednesday

Extreme Mnemonics

[Very minor Game of Thrones spoilers]

From one of my LW comments, discussing how hard it is to learn biochemistry:

I’m having the same problem with molecular biology right now, and I agree with the track you’re taking. The issue seems to be the large amount of structure totally devoid of any semantic cues. For example, a typical textbook paragraph might read:

JS-154 is one of five metabolic products of netamine; however, the enzyme that produces it is unknown. It is manufactured in cells in the far rostral region of of the cerebrum, but after binding with a leukocynoid it takes a role in maintaining the blood-brain barrier – in particular guiding the movements of lipid molecules.

I find I can read paragraphs like this five or six times, write them on flashcards, enter them into Anki, and my brain still refuses to understand or remember them after weeks of trying.

On the other hand, my brain easily remembers vastly more complicated structures when they’re loaded with human-accessible meaning. For example, just by casually reading the Game of Thrones series, I know an extremely intricate web of genealogies, alliances, locations, journeys, battlesites, et cetera. Byte for byte, an average Game of Thrones reader/viewer probably has as much Game of Thrones information as a neuroscience Ph.D has molecular biology information, but getting the neuroscience info is still a thousand times harder.

Which is interesting, because it seems like it should be possible exploit isomorphisms between the two areas. For example, the hideous unmemorizable paragraph above is structurally identical to (very minor spoilers) :

Jon Snow is one of five children of Ned Stark; however, his mother is unknown. He was born in a castle in the far northern regions of Westeros, but after binding with a white wolf companion he took a role in maintaining the Wall – in particular serving as mentor to his obese friend Samwell.

This makes me wonder if it would be possible to produce a story as enjoyable as Game of Thrones which was actually isomorphic to the most important pathways in molecular biology. So that you could pick up a moderately engaging fantasy book – it wouldn’t have to be perfect – read through it in a day or two, and then it ends with “By the way, guess what, you now know everything ever discovered about carbohydrate metabolism”. And then there’s a little glossary in the back with translations about as complicated as “Jon Snow = JS-154” or “the Wall = the blood-brain barrier”. I don’t think this could replace a traditional textbook, but it could sure as heck supplement it.

This started a sobering line of thinking regarding the social structure of my medical school class. I can remember the names, appearances, voices, place of origin, specialties, and personalities of about a hundred of my classmates and faculty. It seems entirely possible that (again, byte for byte) I left medical school with more social data than medical data, and the social data seems less susceptible to decay from disuse. I forget which heart murmurs sound like what all the time, but the chance that I will forget my friend Leo was Japanese, liked Frisbee, and was dating my friend Angela seem next to nil, even though I haven’t talked to him in a year.

The more I think about this sort of thing, the more I feel like I should be finding a way to exploit this kind of idea. Trouble is the isomorphism just isn’t good enough. If I continue the conceit in the example of representing metabolic pathways by families, the story risks turning into a geneology, which is famously difficult to remember in itself (who but the most devout remember all the begats in the Bible?) And metabolism has a bad habit of looping in on itself, such that any attempt to historicize it is going to sputter into bizarre time travel arcs where people become their own grandfathers.

Still, if history has taught us anything, it’s that people are really good at ridiculous constrained writing tasks. Maybe I could abandon the idea of there being “a” mapping, and just do something vaguely suggestive, so that glucose could be a character in one part and an artifact in another, so long as the story helped give a general sense of the glucose pathway.

Of course, the real benefits would have to wait for the blockbuster movie adaptation. I’m quite serious here. I thought about this when Ozy mentioned how ze doesn’t watch Game of Thrones because then the actors will forever imprint themselves on zir mind and replace zir own imaginary view of the characters. If we had actors, people would remember what each character looked like, which would allow another level of encoding. All blonde people are lipids. All brunettes are amino acids. Everyone with long hair has a methyl group. Taller people are larger molecules. Et cetera.

Maybe there’s another field of study more amenable to this kind of treatment than biology. But as far as solutions in search of problems go, this is one of my favorites.

This entry was posted in Uncategorized. Bookmark the permalink.

40 Responses to Extreme Mnemonics

  1. DB says:

    One thing that makes this trickier (and ties into one of your other recent posts): while there are general trends, people are somewhat different when it comes to what kind of information they absorb easily. I’m really good with practical computer algorithms/optimizations and somewhat bad with theoretical papers, for instance.

  2. I never got into Game of Thrones on TV because they introduced too many characters at once. How am I supposed to enjoy the sexy incest if I don’t know who’s incesting whom? I might accidentally be aroused by outbreeding.

    I don’t know if an entire roman a clef is necessary. Maybe a textbook that comes with ready-made mnemonic doggerel as full of tawdry incident as the most beardy epic.

    Jess 154
    That motherless whore
    Her dad is some netamine
    Always high on ketamine
    Born on a rostrum
    Married to Luke Wolfman (leucocynoid, see?)
    Works at the brain wall
    Watch her lipids move y’all

    Obviously I have chosen to render the story of JS-154 as if it’s about a drug addicted stripper, but different biochemicals could inspire verse about pimps/hookers/bank robbers etc

    • Randy M says:

      “I never got into Game of Thrones on TV because they introduced too many characters at once”
      Not sure you’d enjoy biochem much, either, in that case 😉

      • amuchmoreexotic says:

        I actually have a natural science degree – I can memorise biochemical pathways if I have to. But I used to do it with stupid mnemonics (as above).

        I don’t think the Game of Thrones method would work for me because there are things wrong with me such that I don’t find family trees very interesting or memorable, and I can never remember how cousin removal works – but it’s an ingenious idea.

        Ironically my Part II (final year) was in genetics, so I actually needed some knowledge of non-metaphorical incestuous family trees to pass my degree.

        It’s strange that we have societally approved mnemonics for relatively simple information, like My Very Educated Mother Just Served Us Nine (Pumpkins), but when you’re trying to learn something vastly more complicated, you’re left to devise your own. I remember the medical students at my university had special colouring books to help them learn anatomy. There’s definitely an untapped market for learning materials which are designed to help you remember the information they’re imparting.

        Have you ever actually taken one of these free online courses? I started doing one, assuming it would be a mixture of online text and Anki-style flashcards and so on, but it mainly consisted of VIDEOS of COMPUTER SCIENCE PROFESSORS EXPLAINING THINGS, which must be one of the least enjoyable ways to convey information ever.

        A tame computer scientist or bioinformatician could probably generate the metabolism-based fantasy novel Scott proposes programmatically from the right databases.

  3. sixes_and_sevens says:

    You may be overestimating the typical Game of Thrones watcher. I remember wondering why a TV show about feuding dynasties engaging in underhanded political machinations over claims to a throne had such widespread appeal. Then I realised that underneath that layer it was basically a medieval soap opera with full-frontal nudity.

  4. There was a fairly recent post at LW about making science articles more interesting by turning them into drama– amping up the molecules into space battles or some such. I can’t find it in a reasonable amount of time, and I’m hoping someone will remember it.

  5. suhrob says:

    A related idea is employing the brain’s superior face recognition circuits for multivariate data visualization and pattern recognition by isomorphic mapping of high-dimensional datasets into facial features [1]. A very neat idea that has only one minor problem – it doesn’t seem to work.


  6. eyecolour says:

    Objectification? Maybe you’re just writing BDSM fiction. (And the next 50 Shades of Grey will be about molecules…)

  7. eyecolour says:

    Oh my….
    The part I meant to quote was:
    “glucose could be a character in one part and an artifact in another”

  8. Aris Katsaris says:

    “Jon Snow is one of five children of Ned Stark”

    There’s a small irony here in that the legitimate children of Ned Stark and Catelyn are actually 5 (Robb, Sansa, Bran, Arya, Rickon), that making Jon a 6th. So he didn’t actually know the story well enough either. 🙂

  9. Randy M says:

    I don’t know, maybe that’s a bad example, but I find the biochemistry not that much more difficult to understand than the Game of Thrones paragraph. I mean, I don’t know what a Leukocyanoid is or how it accomplishes it’s binding, but I know have a decent understanding of what JS-154 does and who would be most interested in it (doctors/researchers specializing in Alzheimers or strokes, for instance).

    And while I don’t remember so many details of my semester of undergrad biochemistry (that I haven’t used in a decade), when I look at online conversations about, for instance, who Jon Snow’s real parents are, I don’t recognize almost any of the people being discussed (who the heck was Rheagar?) and I have a tendency to confuse Tywin and Tyrion. Sure, I remember the Starks and many of the others, but only those that matter to the plot and occur in many scenes. The trouble with Biochemistry is that it isn’t just one coherant story, but hundreds going on at once with some overlap and reoccurance, but not nearly so much as a coherant, good selling narative, unless you are able to dial it down and focus on one niche area, like the krebs cycle.

    That I did remember for at least a couple years, and probably would still if I’d picked up a sequel every 2 years to refresh, discussed it on-line, speculated on what might happen to it in the future, etc.

    But I think there is a good point that if you could use narrative tricks in the telling of the chemistry to get people to have emotional connections to it (like one does commonly wiht classmates or beloved/hated characters) then it would be much more memorable, and might be able to be done in a stretch.

    Smell is also highly tied to memory formation, but I doubt you could have enough different scents to make a scratch and sniff biochem book, rad as that would be.

    Another thing I always wanted was a biochem themed trading card type game, which actually might be quite doable considering some of the interactions are siilar (use and enzyme card to generate ATP, attach to activate another enzyme to transcript some DNA to RNA, generate a new protein card, etc.)

  10. Alex says:

    The Jon Snow parallel was hilarious.

    I think that it is quite remarkable how much information a typical person can absorb when the information is entertaining. Fantasy novels are a very good example of this. Another example is the card game Magic the Gathering. Without actively trying, I have memorized probably over 500 cards each one about this level of complexity in the past year. A quick glance at the artwork for a card is all I need to determine what the card does.

    But I think its probably not simply the fact that the information in story form is more entertaining that makes it easier to learn. I think it has to do with the templates that we use to make sense of new information. If we don’t have a ready made template for something, then the information is harder to learn.

    I must admit find mnemonics difficult to use because while it does make learning easier, you have to decode mnemonic back into the encoded meaning. Maybe I’m not doing it right though.

  11. Evan Guiney says:

    I think theres a lot to this idea-
    I just got my PhD in molecular biology, and i know a good fraction of the inner workings of a yeast cell- pathways, structure, physiology. And I definately know it in a narrative way, almost like an epic poem.
    FYI, I find that the best isomorphism is between a cell and a city. Genes are professions, individual proteins are the individual workers. Its not perfect, but ive used it with several non biologists and it helps a lot.

  12. Strinka says:

    This reminds me of an idea I once had. There are city building/resource management computer games, in which you have to, for example, have a farmer grow wheat, to be ground by a miller into flour, to be baked by a baker into bread, and if you don’t have enough bread, your people start starving. I think it would be cool to see a similar game that modelled a cell rather than a city.

    • Sniffnoy says:

      Yes, if you have to do the transformations yourself, and have to make sure you have the appropriate resources, that should make a good way of internalizing it! (As well as such things as e.g. what things are reagents vs. what things are catalysts — it makes quite a difference to the game/puzzle obviously!)

      Randy M seems to have suggested a similar idea above. It’s not clear to me that trading cards would really add anything to it though.

    • amuchmoreexotic says:

      I’d like to play an action game where you’re a Fantastic Voyage style explorer, shrunk down to nanoscopic size and trapped inside a single-celled organism in a pond, and you’re moving around inside its vacuoles and endoplasmic reticulum and what have you. You can’t steer yourself directly, but you can manipulate the local biochemical conditions (pH, signalling molecules, etc) to trick the cell into moving you where you want to go. To level up, you need to seize control of the organism’s flagellae and steer it into the mouth of a predator (say, a water flea); then you have to work out how to control the water flea and get it eaten by a tadpole.

  13. Sarah says:

    Memory is my main talent. (Not a savant or anything, it’s just that I can retain more than >99% of people I’ve met, with next to no effort.) I find that things automatically get filed in my brain with colors, genders, and sometimes other sensory qualities. If I need to brute-force memorize something that wouldn’t just stay put by default, I put it to music (ten years later I can still sing the acids from weakest to strongest) or just repeat it focusing on the rhythm and the sound of the words.

    I definitely think there’s a rhythm/meter/music/verbal pathway, because all the things that come easily to me, as opposed to things I have to work on, feel like doing the same thing. (Languages, both vocab and accent; learning music by ear; memorizing things verbatim; etc.) It’s the “would make a good preacher” skillset. In the First Great Awakening they would find a twelve-year-old kid who knew a hundred verses of Barbary Ellen and throw him up on a pulpit to preach.

    Advanced memory feats tend to be spatial (memory palace) or combinatorial (Brahmin scholars would take apart a poem and learn it backwards or alternating lines). But I think you can get a long way on oral stuff alone.

    I wonder if it would be possible to revive oral culture? Reciting poems, singing in groups from memory, etc?

  14. I failed out of Chemistry quite early, when it became obvious that I cannot remember large amounts of largely disconnected information. I _can_ remember lots of information if it slots together, but not when there’s no link between the different parts that my brain can latch onto.

    What GoT does is give you something to care about (a character or two) and then build up a web around them, linking the bits together in ways that make sense (logically and emotionally). If Chemistry could do likewise, then I would probably have found it a lot easier. I suspect that good teachers are ones who can turn their subject into a story that most humans can take in.

  15. Doug says:

    Cool idea!

    I think translating directly from biochemistry won’t do it. To be really memorable, you need the gripping human narrative and (occasionally lurid) detail.

    My reasons for thinking this include:

    – People famous for memorising lots of dull stuff (phone books, arbitrary lists of objects, decks of cards) seem to all report some sort of very-vivid imagery and narrative to their mnemonics, often using all forms of human perception (not just how it looks, but spatial orientation, how it sounds, smells, etc)

    – The genealogies in the Hebrew Bible are desperately tedious and unmemorable. Mercifully I never had to learn them by heart, but those that have complain in very similar terms to people struggling with biochemistry, A&P, etc.

    – Vocabulary lists and conjugation/inflection/declension tables are desperately tedious and unmemorable. This is why modern teaching of languages very rarely uses individual words in the target language out of context.

    – The details of Tolkien’s Middle Earth are desperately tedious and unmemorable, apart from the bits where there is gripping human (or elvish) narrative and (occasionally lurid) detail. Which, mercifully, is the bits people generally read (The Hobbit and Lord of the Rings)

    The vivid human bit is the key

    I have long had a theory that one of the most important parts of early clinical experience for medical students is to expose them to lots and lots of situations where desperately tedious and unmemorable facts will play key roles in gripping human narratives with lurid details.

    … the upshot of this is that the whole of preclinical medicine should be replaced by a a load of specially-commissioned series of House, MD. And you’ll still have a better memory for which of the Cottages made a pass at whom than for whether fumarate goes to malate or the other way round.

  16. I remember parts of the song in this protein synthesis musical (“Twas brillig, and the 30S-ribosome did gyre and gimble in the wabe”) after having heard it once over 35 years ago, but I’ve found it inadequate for simple things like remembering which is mRNA and which is tRNA.

  17. Elissa says:

    So I’m an amino acid, low MW, no methyl group… I’m glycine?

  18. Doug S. says:

    I have the same problem with foreign language vocabulary.

  19. nemryn says:

    One big problem I immediately noticed is, this leaves you wide open to confusing yourself via anthropomorphization and the pathetic fallacy.

  20. Nestor says:

    The Japanese already hav educational manga and anthopomorphic personifications of the most outlandish things, just put them together and make the adventures of Carbon-tan and friends..

  21. Eric Rall says:

    This sort of mneumonic techniques were heavily used in the middle ages, when written records were rare, expensive, and unreliable (because literacy was rare and library management was a lost art), so most stored information was memorized. The most common tricks were composing the information into poetic forms, and what was called “memory theater”: making up vivid mental images (often punny) in a familiar setting to remind the memorizer of the information.

    The illustration that I was given of memory theater (from James Burke’s documentary The Day the Universe Changed) concerned the Seven Liberal Arts that made up the basic educational curriculum.

    Enter a house. In the foyer, there’s a suit of armor holding an open book. A knight, studying. Knight study -> astronomy.

    Continue down a hallway. On the wall, there’s a portrait of the Queen. Queen = ruler. Rulers are used for measuring. Measurement -> geometry.

    In the living room, there’s a jigsaw puzzle laid out on the coffee table. You solve puzzles with logic.

    At the foot of the stairs, there’s a balance scale. I pick it up and scale the stairs. Scales -> music.

    At the top of the stairs, I find the bathroom. Inside the bathroom, the toilet lid is up and for some bizzare reason there’s a megaphone in the toilet bowl. Megaphone -> public speaking.

    Next door down the upstairs hall is a bedroom. I turn back the covers of the bed to reveal a mannequin wearing a British judge’s wig. Judges hand out sentences. Sentences -> grammar.

    On the nightstand, a snake is curled up ready to strike. This particular snake is an adder. Adder -> arithmatic.

  22. Rob says:

    I think a key thing to consider is the almost total failure of “educational videogames”, which are a similar idea. I remember there was an enormous drive for them in the 90s, and intuitively there is an *enormous* amount of potential there – think about how much kids those days knew about Pokemon. But that potential really never materialised. I don’t know exactly what the failure mode whas, but it’s possible that this might suffer from it as well.

  23. Rob says:

    I’m also reminded of The Story About Ping:

  24. Galuruthaz says:

    Id suggest looking into the F.Yates “Art of Memory” and works of J. Campbell for.those interested why this works.. (:

  25. Avantika says:

    This is something a lot of people are thinking about! In fact I’m part of a project to teach cell biology by writing a murder mystery set inside a cell, with molecules – lots of them – as characters. It presents a lot of challenges – we started out anthropomorphizing proteins, but it soon became very clear that a society set in a cell would have to have utterly bizarre social rules. Anyway, I believe it can work.

    Meanwhile, we do try to make our research understandable by presenting it in story form. I occasionally open my discussions by asking people to imagine that RNA Polymerase is Henry VIII.

  26. Pawel Aleksander Fedorynski says:

    I know nothing of neuroscience, but I have studied a field or two and I think you got things backwards here. Unless neuroscience is very exceptional, anyone who’s into it can relate more strongly to JS-154, blood-brain barrier etc. than to all that medieval babble. I’m not talking about people who need to cram some molecular biology for the test, but about those who actually find it interesting. Most PhD students fall in that latter category and so there’s no way they’d know more byte for byte about Game of Thrones. For an N=1 sample, my dad who is a chemistry professor often has trouble following the plot of a movie and generally likes for someone to sit next to him and remind him who is who in the show, but obviously never has any such problem with molecules.

    (This answers amuchmoreexotic’s question about mnemonics: a lot of people who don’t give a shit about planets need to memorize planets, hence My Very Educated Mother, but you generally only learn complicated things when they fascinate you, and then you don’t need mnemonics.)