The fading memories of youth

I have memories of my second birthday, of the kiddie pool I splashed in and the toy boat I played with in it. I remember the neighbors' kid Paul, that he was blonde, and that we each had a rubber ball. Mine had rainbow stripes on it, and Paul's was transparent green with silver shiny bits.

I remember sitting on my dad's shoulders and looking at the babies in the maternity ward when my sister was born.

I even have an infancy memory: closing my eyes and crying, confused as to where the sound was coming from.

These are not things I was told about. They're pictures in my mind. I don't have a very strong episodic memory of that time, but I do remember things from well before age 3, and it pisses me off when scientists say "that just doesn't happen". It's like when my first grade teacher, a Ph.D. candidate in education at the time, told me I hadn't learned things I damn well had learned, like how to read.

Lots of memories are accessed by association. But when your entire experience is small (very young), the associations are by necessity fragmentary or coincidental. I love iced cream, which I first encountered when visiting Uncle Roger. Is good ol' Roger gonna be 'the ice cream man' forever? It's an 'incidental association' that is not useful or helpful.

I think that alone is enough reason for evolution to 'downgrade' early memories. To reset the 'weights' between experiences because they are often wrong and misleading. Never mind adult humans find this annoying or disturbing. Nature doesn't give a flying fart how we feel about what hacks it uses on us.

I stand by thinking that the problem with memories fading is that (experiential) memories mostly work by replaying a frame of mind and checking the result for internal consistency, and if your frame of mind deviates too much from when the memory was formed, this fails; so memories you retain are memories you rehearse and thereby iteratively bring forward in a similar manner to if you walked a large software project that has had incremental "upgrade" steps from 1.0 to 1.1 all the way up to 39.75 or something. (This would also help underscore why being in the same "place" or "frame of mind" for memories makes your recall scores better - or the stories about people who take mind-altering substances and have memories that they don't recall when sober but can recall reliably the next time they take said substance.)

So killing neurogenesis might preclude this problem by preventing your mind from changing enough to make this failure mode happen.

(This would also explain why we freak out when other kinds of "memory loss" happen, e.g. from dementia or old age - it's not that we found it but it doesn't import, it's that we literally can't find it, and we don't have great handling for actual memory loss, so we freak the hell out. This could also be the reverse, where we load something and don't notice the consistency checks failing, and then as those are pointed out we become increasingly distressed at the cognitive dissonance from different parts disagreeing on the sanity of what we said, but...)

Maybe because memorization is also learnt, some learn faster than others. Some people don’t remember well party physiological but sometimes they don’t care to remember.

Trauma faded a lot of my fondest memories.. I'm grateful that this article made it to HN, I didn't know what to look for on pubmed.

I don't think people tend to believe me, but I remember learning how to walk. There was a fairly short area of our house between two corners, and I distinctly remember that standing in the corners was generally easy -- I had two places to put my hands and old myself. The corner provided quite a bit of stability. But, walking in between the two corners was quite difficult because I only had one wall within to balance myself. I would confidently stand in the corner, and walk to the next corner somewhat quickly and awkwardly to minimize the amount of time I had to maintain my balance.

When I was very young, I was somewhat neglected, and also quite introverted. As a result, I spent a lot of time by myself, thinking and introspecting. I don't remember if I realized this myself, or if someone taught it to me -- but at some point perhaps age 5-9, I was aware that memories were not permanent, and drifted over time. I spent a lot of time thinking, and trying to remember the clear facts of my earliest memories, and do my best to commit the the discrete facts of my memory to what I considered my "permanent" memory. I was distressed by the idea of degradation and loss, and so the idea of making permanent memories was especially appealing. I vaguely recall selecting three different memories for "permanence," but I only remember one clearly. The 2nd one I remember vaguely, and it's generally too crude and embarrassing to bring up when discussing these sorts of topics. (it was about diaper changing) The "learning to walk" memory is the one I bring up the most, and is the most interesting, since it can be roughly dated. I was just a bit over a year old. I don't recall the possibly third memory whatsoever.

People will tell you that revisiting a memory is the same as "rebuilding" a memory, and therefore that you're modifying the memory in an unconscious way every time you do this. I disagree somewhat: you have partial control over how much a memory drifts. You can declare the facts (in the case above: the corners, learning to walk, and how precarious my walking was between the corners) firmly an indefinitely. It's only the the less important details that drift. What color were the walls? Was there carpet? Which house was I in when I made this memory. Not only can I not answer any of those questions, but I'm fairly confident that the hallway I "remember" forming this memory in is not the hallway it actually occurred in. The hallway I remember wasn't available to me until I was about 3 years old, or so. And so, it is possible to compare facts and introspect in at least some cases and attempt to correct the record. I would also argue that introspection itself is a skillset which exists on its own bell curve. Are of aware of what facts you know vs. which you don't? Do you have a sense for which memories you can be certain of? How well do you know your own thoughts and motivations? Like any mental skillset, these are things at which people can be worse or better.

Lastly, I would note that that you do retain all sorts of things which might be called memories from youth: how to articulate your hands. How to see silhouettes, how to move your mouth in order to make the correct sounds. These aren't things you have visual memories, but these are things your brain learned to do at a young age, and has never forgotten. Certainly, these patterns have been reinforced this whole time, which must help significantly. By replaying my memory of learning how to walk, I have done the same: reinforced it over decades, so that it could survive into my 40s, where it otherwise would have been permanently lost. If I had the foresight to reinforce other memories, who knows what else I could remember.

Rabbit-hole ahead: is there a connection between tooth-loss and amnesia? Is it just coincidence that forgetting early memories coincides with a total tooth change-out?

Edit: I'm an idiot who didn't read the article. This literally is the research I'm complaining it's not.

> Ryan and Frankland propose that the rapid birth of new neurons, known as neurogenesis, in infants might be overwriting memories and that infant amnesia disappears once neurogenesis slows. When Frankland’s team used a drug to suppress neurogenesis in the hippocampus in infant mice, the juveniles performed as well as adults on memory tests.

Original comment below, for completeness.

---

> The main goal, Power says, is to figure out exactly when the developing brain switches on the ability to form accessible long-term memories.

I don't get this. Why do seemingly all the academics in this field believe there's a threshold, before which there's no long-term memory, after which there is long-term memory? I mean, that might be true, but nobody seems to be considering the "memory is contextual, and context changes"-type hypotheses?

Piotr Wozniak (2017) wrote about this in https://supermemo.guru/wiki/Semantic_aspects_of_childhood_am...:

> Frankland's hypothesis on the origins of childhood amnesia is that prolific neurogenesis hinders the establishment of long-term memories. This hypothesis keeps garnering evidence. In one of his seminal experiments, Frankland slowed down neurogenesis and achieved better learning. This simulates the effects of daycare: seeming acceleration due to stress (see: Stress and neurogenesis, and Stress and learning).

> In this text I would like to provide some evidence for the dramatic impact of prior knowledge on the possibility of forming memories in childhood. Again the problem derives from a simple observation: the lesser the extent of prior knowledge, the lesser the opportunity to associate new pieces of knowledge with the existing semantic structure.

There are lots of possibilities that don't seem to be considered. Maybe they've all been discredited since I last looked into this? But more likely, they'll only be seriously investigated once the current avenue loses steam (if it does: I have no special insight here).