So the 2016 Angry Video Game Nerd Halloween episode just came out, and it was one of my favourite yet! I wasn't that big a fan of the Beavis and Butt-Head episode, but I think that's more because I don't know too much about Beavis and Butt-Head at all.

If you haven't seen it yet, I suggest you do. It's a good watch, and I'm about to spoil quite a bit of it.

The Berenstein Bears. Or Berenstain, universe depending.

I somehow missed this conspiracy theory when people started talking about it, but very quickly it goes like this:

  1. Many people remember a thing one way.
  2. All physical evidence currently proves thing is actually another way.
  3. We all jumped from a parallel universe.

This phenomenon has been dubbed the "Mandela Effect", coined apparently by Fiona Broome, author of paranormal research guides. It describes situations where a large number of people have memories of something that is apparently different. This can include C3PO having a silver leg in Star Wars: A New Hope (he actually does), KitKat not having a hyphen (people remember "Kit-Kat"), and Coke Zero actually being labelled as "Coca-Cola Zero".

People describe this like experiencing a glitch in the Matrix, and evidence of something fishy going on in our world. Fiona Broome argued that this effect could be explained by the existence of parallel universes.

Steins;Gate Time Leap

It's worth noting that "Mandela Effect" can be found on Wikipedia under the page for Confabulation, which is a disturbance of memory through the production of false, distorted, or misinterpreted memories.

Honestly, it's weird. For sure. And it could be pretty convincing to me... if it wasn't for one major thing—I don't put that much stock into memory.

Basically you have an idea that challenges our understanding of Physics and reality based on human memory. Not just a single person's memory, but many people's memories. Ockham's Razor tells us that the simplest theory, the one with the fewest assumptions, is most often the best one. Nobody's memory is reliable, not out of opinion but as just a hard fact of how memory works, and so it's most likely everybody really is remembering incorrectly.

Maybe that's not enough to satisfy you though. I know, I know, you need more. But that's fine, because it's always good to get more people to understand how memory works!

My Explanation of the Berenstain-Mandela Effect

One take away from memory research is that memory is a reconstructive process. The fact that you think you can have a vivid, specific memory of something is an illusion. Memories can be distorted stupidly (and scarily) easy. And no, that doesn't just include old memories but recent ones as well. Even something that happened five minutes ago is subject to deterioration and change.

It's reconstructive, based on semantics and expectations. So every time you remember something you subject that memory to change. You can easily tweak it, either intentionally or by error, but then it becomes likely for you to remember it that way the next time.

Just how easy is it? For that, I recommend you check out this awesome TED talk by Elizabeth Loftus on her research on False Memories. I'll leave it embedded below, but basically you can get someone to remember something incorrectly (without them realizing it!) with very simple priming through leading questions. In her famous study, she had people watch a video of a car accident and then answer some questions immediately. The leading question was "How fast were the cars going when they verb into each other. That verb was exchanged for words like smashed, crashed, bumped, hit, etc. They found that more impactful words had people estimate faster speeds, but it gets better than that.

One week later participants came in and answered more questions. A key one was "did you see any broken glass?". In reality there wasn't, but people who got words like "smashed" were much more likely to say yes. They remembered there being glass in the video, even though there wasn't! This is an example of a false memory, and basically this happens all the time. This is what I believe is going on with the Mandela effect.

REF: Loftus, E. F., & Palmer, J. C. (1974). Reconstruction of automobile destruction: An example of the interaction between language and memory. Journal of verbal learning and verbal behavior, 13(5), 585-589.

"But how can everyone make the same error?", I hear you asking, "Isn't that evidence that it's not a simple memory problem?". Excellent point, but honestly this I don't think this is as big a problem as you think. While we're all different, we're also all pretty similar in how we process things. I think this effect is an example of two things: People making similar encoding errors coupled by false memory priming (which Loftus demonstrates is indeed very possible).

For one group of people at least, I think they're being implanted by a kind of "contact" false memories. Basically you can be primed by others when you hear about this phenomenon into believing you remember Berenstein. Now again, memory is reconstructive. You hear about the effect and you probably immediately try to generate a memory of what the name is. You aren't pulling this out from anywhere, you're vaguely recreating it. And thus, it's really easy to generate it with errors.

You have in your mind that the name could be spelt with an "e". How easy is it to imagine an e as an a? I think it's the easiest error you could make given that they're both basically the same shape. Even without the priming, you can make that error.

Mama Berenstain know's best.

So then what about the people who propagate this false memory? The originals? Well, how about you try to say "Berenstain" out loud. Not slowly or carefully pronounced, say it normally. Don't look at the word while doing this either, or that'll manipulate what you hear (perception is weird). Or even better, just listen to the theme song:

Was that an "ai" sound, or an "ee" sound? I'd argue that it's pretty ambiguous.

I would also argue that you don't spend the time to memorize the classic "Berenstain Bears" title/logo every time you read the books or watch the show. You see it once, but every other time you don't really need to process it again. It's familiar, it hasn't changed, so you don't pay much attention to it. But then when you go to reconstruct that logo in your head, not only is vision involved, but audition too. You know what "Berenstain Bears" sounds like, and that info is used when constructing the logo.

So here is the crux of my argument.

The pronunciation of "Berenstain" is ambiguous, and more likely to sound like it's spelt "Berenstein". When you remember the logo, that mispronunciation contaminates your memory.

That's it. That's all.

So again, I think this is an example of people making similar encoding errors (stemming from an ambiguous pronunciation) coupled by false memory generation as the story gets around.

And this is with very few assumptions as all the ideas I'm talking about here arise from Cognitive Psychology. Definitely fewer assumptions than the massive amount the parallel universe theory would need once you actually start to delve into how that could be possible. So I wouldn't worry about World Lines and universe hopping.

Steins;Gate World Lines

I think many of the Mandela Effect situations can be explained with a similar reasoning to this, though I'm not going to deconstruct every one here.

If you've made it this far, thank you! You can feel free to stop reading now, but if you're interested more in how memory works I've got more explanations down below.

How Memory Works

The first thing you should know is that memory probably isn't what you think it is. Many people think of memory as like some kind of data store, like a computer harddrive, or a video camera. People think of memory as being some kind of tangible object. For example you have a memory of your 10th birthday party which is stored in some storage space as-is. This is absolutely not the case at all.

Memory is a reconstructive and re-interpretive process. When you remember something, you are actually recreating that memory from semantics, ideas, and expectations. There is no place that stores an episodic snapshot of your birthday party. Rather, when you "remember" the party, you call upon semantic facts, your expectation of what a birthday party was, and put it together. So say you know you were living at your first house at the time, you invited your best friends from school, you played Super Smash Bros, dad bought a big square pizza, there's one photo of your friend that came out weird because he was moving and he looks like a ghost, things like that. Combine that stuff with how you expect a birthday party to go and you generate a "memory".

It was probably a lot lamer than you remember.

"Hold on", you may say "but then something is being stored, right? Otherwise what are memories being created out of?". And you're right, but it's not being stored how you'd expect them to. I must admit that our understanding of memories is incomplete. Don't get me wrong, we know what memories are not. But there's still active research trying to discern every facet of how memory works. Nevertheless, I can give you some really cool theories.

Note that when I say "you" here, I mean everyone—scientists and non-scientists alike. Many researchers have explored ideas like these for the past few decades.

Imagine you see a red ball on the ground. You get sensory input that is processed up to and through the Visual Cortex before it is sent to higher order cognitive areas around the frontal cortex for further processing and Working Memory (your kind of actively-thinking-about space). If I stick you in an fMRI, I can look at your brain and see specific networks of activity.

Networks of Neurons

But now if I bring you back into an fMRI and ask you to remember that ball, I would see a similar pattern of activation! Simply put, the memory of what you perceive is stored in the networks that perceived them.

Picture that you have this complex system that can reorganize itself, form new connections, prune connections away, make connections stronger, and so on. Whenever you perceive something a specific pattern arises. But when that happens, some system goes in and takes that specific network and strengthens the connections a bit. Basically somehow makes a way that when you try to retrieve that red ball, that specific network will likely fire. THAT could be a piece of memory.

Hebbian Law: Neurons that fire together wire together

So basically just expand that idea all over your brain in impossibly complicated ways and you can start to get a picture of how memory could work. You have bits and pieces that are kind of connected to each other. You have some system (which is probably mediated by the Hippocampus) that interacts with those bits and pieces to reactivate them when you want to retrieve them, bring them into Working Memory, and then you essentially build around that. Fill in the blanks.

That's how memory works, and as you can probably guess, it's not actually that accurate.

I actually made a video a year back trying to talk about this stuff a year back. It's kind of silly, it was in the context of Kingdom Hearts: Chain of Memories (I thought it'd be a good idea to tack on a science discussion to a game review), but I did some nice visuals that might help.