The Mystery of S., a Man with an Impossible Memory (2017) | Hacker News Reader

The Mystery of S., a Man with an Impossible Memory (2017)

newyorker.com

61 pointskeiferski43 comments

For a slightly more modern occurrence of something similar, the actress Marilu Henner also has total, perfect recall: https://en.m.wikipedia.org/wiki/Marilu_Henner

There was a 60 minutes from several years ago with her and a handful of others with this ability.

https://www.cbsnews.com/news/the-gift-of-endless-memory/

This explains a lot about S. The mystery book with an additional story in the margins and various artefacts tucked in the pages.

https://www.goodreads.com/book/show/17860739-s

As a memory researcher, I'd be interested in conducting case studies on such superior (not perfect) recall. In particular, I'd like to examine relational memory under incidental encoding conditions with materials not easily verbalized with relations that are arbitrary. hat is, I am interested whether their hippocampal dependent memory is superior, and/or whether it depends mostly on extrahippocampal processes such as semantic-based organizational mnemonics.

Do you have an example of something like that?

I'm not sure what you mean exactly. If you mean what kind of paradigm may be used to try to get performance to inordinately depend on hippocampally-mediated binding processes? See Konkel and Cohen, 2008 Hippocampal amnesia impairs all manner of relational memory or my own paper, Lee et al 2016: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4733390/ ..although these were not under incidental encoding.

> incidental encoding conditions with materials not easily verbalized with relations that are arbitrary

An example of such conditions and materials.

https://i.imgur.com/dqBFwBp.png from Konkel et al, 2008

Objects which are hard to name and have little meaning (very novel), and the relations between them (e.g. spatial position, temporal order) are arbitrary. These tend to impede organizational encoding or retrieval strategies. Not impossible, but harder. You might be able to tag a color, but after seeing a few dozen, that is hard.

Interesting. Thanks for the examples, those do seem challenging. I suspect I'd do astoundingly well on that test in the medium to long term, at least compared to population norms, and quite abysmal in the very short and short term without time to process and encode the memories.

> The editor picked up a newspaper and read at length from it, challenging S. to repeat everything back to him. When S. did so verbatim, the editor sent him to have his head examined.

Not sure whether intentional but this is the funniest thing I've read in a long time.

There are all sorts of hilarious little absurdist anecdotes in the book. Take this one:

S: I'm sitting in a restaurant—there's music. You know why they have music in restaurants? Because it changes the taste of everything. If you select the right kind of music, everything tastes good. Surely people who work in restaurants know this.

Well, that's not really absurd or far for the truth. Music is selected for its effects on the consumer in all kinds of commercial establishments, and besides trial-and-error (this kind of music works best for people to drink more in our bar etc) there are lots of commercial research into it.

In the context of the story and Shereshevsky's personality, it is. To him, a person with extremely strong synesthesia, musical sounds literally have flavors, flavors that are as strong as actual real food to a normal person.

He isn't saying, "Restaurants play certain music to push customers to feel a certain way," he's literally saying that the flavors of music can match and enhance the flavors of food.

It could still be argued that the difference is in degree rather than in kind- that the 'extra bandwidth connecting his senses' allowed him to clearly articulate as sensation a phenomenon much more subconscious in the typical patron's mind.

No, not really. If you read the book, you will see that his mind was very idiosyncratic and the associations he made were highly unique and imagination-based, and not some sort of heightened sensory input.

Very interesting. I find myself in the odd position of having an extraordinary memory akin to (though perhaps not as extreme as) Shereshevsky that is not especially useful to me.

I often put it as "I don't think my memory is that good, but I've never met someone with a better memory to check it against". I often find myself reminding people of interactions we've had, conversations, what we had for dinner, what they were wearing... Largely ephemeral recollections of no particular relevance.

A small discussion from last year: https://news.ycombinator.com/item?id=14997702

A friend of mine has a very, very good memory. He says it not always an advantage as he can’t forget the bad things either.

Am I the only one thinks that The New Yorker takes 5 pages to give 1 page of information? Their articles are so frustrating to read!

The New Yorker is a literary magazine, not an encyclopedia. If you're reading it purely for gathering information, I suggest the Wikipedia article [1] or Luria's book itself instead. [2]

[1] https://en.wikipedia.org/wiki/Solomon_Shereshevsky

[2] http://arteflora.org/wp-content/uploads/2018/05/Luria-The-Mi...

I've always wished "long form" articles would put an abstract at the beginning.

Me too.

I even tried to see if such a "summary" would be a viable business: https://github.com/simonebrunozzi/MNMN/blob/master/Weekly-Su...

No, no, no, there's literature and then there's literature that gets paid by the word and has to be written fast. I'm sure they never met an adjective they didn't like. I get it, people need to make a living but it is still frustrating to read.

Literature does not have to be fluff. One of the best non-fiction books I've read that's both informative and has beautiful writing is, "On Immunity: An Inoculation " by Eula Biss. It's both informative and a pleasure to read.

As a friend one remarked, "The New Yorker makes fiction out of nonfiction."

What's your hurry, friend?

The inevitable march of death?

Ugh don’t remind me.

You're able occasionally to forget? Lucky you!

Am I the only irritated by these inevitable questions from people with no sense of literary craft any time an article from the New Yorker is posted on HN?

Am I the only one thinks that The New Yorker takes 5 pages to give 1 page of information

I’m pretty sure The New Yorker thinks that. It’s called “literature”, and a lot of it is pretty good. Sounds like you’ll want to stay away from Russian authors, though.

“Literature” is definitely not the opposite of “concise.”

Good thing, too, because that wasn’t my point. If you want the TL;DR, The New Yorker probably isn’t for you. It is, of course, more nuanced than that and leaving less room for snapping one-liner comebacks, but such is the world of low attention spans that we live in.

True but the point of literature is not just to deliver factual information as quickly as possible.

By analogy, you can drive to the top of Mt. Washington, but people still hike it. Either way will deliver the view from top, but each will be a distinct experience.

I think Michael Crichton is a good example of this. He has (had) an incredibly terse style that still manages to communicate quite a lot of information/imagery.

Exactly, literature can be both artful and to the point or at least insightful.

I think a lot of Russian authors are pretty to-the-point. Turgenev, Chekhov and Tolstoy come to mind, as opposed to e.g. Gogol who has a more playful style and can write pages about anything and everything.

Tolstoy’s best-known work is the poster child for long reads, and at the forefront of my mind when I wrote the sentence. I feel I’m being trolled. :-)

I didn't know we were discussing "long" I thought it was about writing 5 pages on something that can be written in 1 page (did you read OC?). I would say Tolstoy's writing style cannot be characterized as such. Tolstoy's work are long but they cover a whole lot, information density is very high.

I, and I think probably many others, would disagree. Tolstoy is a lot of things, but to-the-point is not one of them.

Do y'all think folks with such memory will make better programmers?

From what I'm reading on Wikipedia it isn't equivalent to being able to memorize information. The skill is limited to being able to recall autobiographical details...so what you had for lunch in 1992 but not what you read in this CS textbook.

I think about this question a lot. I think the article (and Borges) touch on the core issue: abstraction. I find that many of my employees find it very easy to program based on their lived experience programming similar tasks, but hard to take a step back and answer the question "What is the appropriate generalization of what I am doing right now?"

My hypothesis these days is that a strong autobiographical memory has little to do with programming. Rather, a good memory for semantic information as well as the ability to creatively rotate one's perspective seem to be much more important.

Reply on news.ycombinator.com