The Moses Illusion

How many animals of each kind did Moses take on the Ark? In this episode of Stuff to Blow Your Mind, Robert and Joe discuss The Moses Illusion, one persistent example from a class of mental phenomena known as “knowledge neglect.”

Learn more about your ad-choices at https://www.iheartpodcastnetwork.com

See omnystudio.com/listener for privacy information.

Episode Transcript

Speaker 1 (00:03):
Welcome to Stuff to Blow Your Mind, the production of
My Heart Radio. Hey, welcome to Stuff to Blow Your Mind.
My name is Robert Lamb and I'm Joe McCormick. And
today we're going to be talking about an interesting observation
in cognitive psychology that deals with language that starts off

(00:26):
as kind of just a funny little quirk about the
way we process certain kinds of sentences, but ends up
having some some broader and more interesting implications about knowledge
and language and thought. But I thought the best way
to start here would just be to illustrate the prime
example of the effect we're going to be talking about.
And to do that, I think we need to do

(00:47):
a bit of Bible trivia. Rob Are you ready to
go to Sunday School. Let's do it. Let's go to
Sunday School. Okay, if you I'm gonna ask you a
few questions about the Bible. If you get one wrong,
you are going to get a paddling. WHOA, what the
domination is this one of the ones that means business? Okay?

(01:08):
So how let's see it? So um in in the
garden of Eden, what type of animal is it that
tempts Eve to eat from the tree. Oh, that's a snake,
that's right, the serpent, it is, uh. Okay. When after
God created the world, on which day of the week
did he rest? Oh, that was the seventh day. He

(01:29):
got that one, right, Okay, next one, How many animals
of each kind did Moses take on the arc? Too?
Of course? And there you go. That is the prime example. Now, Rob,
I know you were playing along because you already know
the trick in here what I actually said when I
asked that question. Hopefully you were playing along at home
as as you're listening, or maybe you're not at home,

(01:50):
wherever the heck you are. Um, you may have thought
the same thing, right, Moses took two of each animal
on the arc. But in fact, in the Bible story
which maybe not everybody knows, but maybe you do know
this story of of the arc and in the Book
of Genesis, and you do, in fact to know that
it was not Moses who did that. It was Noah
in the story who took animals on the arc. And

(02:12):
yet you thought, after I said the question that the
answer is too and didn't even register the fact that
the name was wrong. Yeah, it's it's an interesting uh
phenomenon to uh, to encounter you others, but also in yourself,
because you there's several different ways to look at it,
and we'll get into a number of these here. But

(02:33):
like even just now, when you ask me those questions
like the serpent one, I'm totally firm on that, Like
I know that I know that aspect of the story
inside and out. Uh, And of course I know it's
the seventh day that he rested on the God rested
on he. She Yet however you want to look at it,
But there was still like this moment of hesitation because
I was like, it's seven, right, it is seven. I
don't want to come come off with the wrong answer

(02:54):
on the podcast. Um. But then when one encountering granted
already knew the answer to the third one, but there
is this temptation though to like when you when you
know why, when you know the answer to something, like
you just you can just jump in without hesitation, Like
there's a certainty that just propels you. Um, you're excited
to get your answer in and then you know, get

(03:15):
the acclaim and the praise for getting it right. Yeah,
there's a certain kind of way in which a question,
especially a question posed in quiz format where you feel
you are under performance pressure and you're being evaluated for
whether or not you're going to get the right answer.
It sort of takes away some amount of critical thinking
that would normally go into reading a sentence and causes

(03:36):
you to focus more exclusively on like, just can I
get the right answer? And so it's not hard to
see how now, and this of course might not be
the only explanation for why this is happening, but it's
not hard to see why you could pretty easily miss
a major error in a question that is not you know,
that is not necessarily something that you're fuzzy on to
begin with, Like you could know perfectly well that it's

(03:58):
Noah in the story, and yet it just goes completely
over your head. Yeah, and you know, we've we've been
doing this podcast quite a while at this point, and
occasionally this this comes up in our data, not so
much in things that we've researched for the podcast, because
I feel like if we've been crunching the facts or
the numbers, you know, or the you know, we're we're
more likely to be putting a lot of thought into

(04:19):
the situation and we're maybe just a you know, a
little hesitant anyway. But the times where I've personally like
said something that was absolutely incorrect, it would be something
that I felt so sure about that I just belt
it out without fact checking it at all. You know, Uh,
something you generally it's something not directly related to the episode,
but something that just kind of comes up in organic conversation.

(04:39):
That's exactly right. Yeah, it's when you feel so confident
that you're not even being careful, you know that that
you can really make some big blunders. Uh. There was
some other questions I was reading about in one of
the I think the earliest study on this phenomenon we're
talking about today. Some of the other questions were in
the Biblical story, what is Joshua swallowed by? Of cour

(05:00):
that's Jonah that is swallowed by the whale or the
great fish, the sea monster. Joshua, of course, is the
the conquering leader of the Israelites as they go about Kanaan.
Another one I really liked was in the novel Moby Dick.
What color was the whale that Captain Nemo was after?
I think I think I might have fallen for that one. Yeah,
I mean, I wonder how much of the ego is

(05:20):
involved here because it's like you're kind of like, all right,
let's get to the part of this where I get
to talk and get to be the one is correct,
you know, like like fast forward through all this other stuff.
I don't care. I have an answer and it is
the correct one. Yeah. Um, that's that's quite perceptive and
I think that's right. Um. But but anyway, so this
question that we're looking at today, that this effect of

(05:42):
not noticing that the question says Moses and just barreling
right on through to the answer, even if you know
that it's actually Noah and the story and not Moses.
This effect has a name, and it's known as the
Moses illusion. It's a particular type of semantic illusion that
occurs when we are trying to process certain kinds of sentences.

(06:03):
And this was first explored in a classic study in psychology.
It was a study called from Words to Meaning, a
Semantic Illusion, published in the Journal of Verbal Learning and
Verbal Behavior in nineteen eighty one by Thomas D. Ericsson
and Mark E. Mattson. And I think it's interesting that
this original observation about this this question about Moses, it
comes out of a mysterious question about how we process

(06:28):
the meaning of sentences. Uh, the authors of this study ask, quote,
how are the meanings of individual words combined to form
a more global description of meaning? And if you start
to think hard about this question about the human capacity
for language, I would argue it is absolutely astonishing. It's

(06:48):
almost baffling the way that we're not only able to
associate symbolic meaning with certain sounds coming out of our
mouths or glyphs on a page, but you're able to
combine those things endlessly to form and comprehend infinite variations
of combinations of those sounds, to create sentences that actually

(07:09):
means something and other people can understand what you mean
when you say them. Like, I think this type of
capacity for language is one of the features of the
natural world that to me seems closest to magic. Yeah. Absolutely,
And I feel like that the mostest illusion is one
of those things that that reveals the magic, that makes
you more aware of the magic trick that is inherent

(07:31):
to your just everyday perception of reality and how you
engage with facts and information and the fact that you're
just like, it's crazy that we're just constantly throwing together
sentences almost effortlessly that are combining all these words together.
Each word has a huge range of of possible meanings
and associations, and that we are able to do this

(07:52):
with such fluency, I mean, sometimes with more fluency than
other times. But uh, but yeah, it is truly a
staund unding to me. And so the authors here are
sort of talking about this process and some of the
question marks that existed at the time in science about
how we form sentences and how we comprehend sentences. So
they start in their introduction by talking about how quote,

(08:14):
a central process in language comprehension is the construction of
a global description of the sentence meaning from the meanings
of individual words which make up the sentence. Right, So
you know what individual words mean, but somehow, like we're
just talking about, you can combine them into these overall
gist forms of what somebody's getting at, you know, like

(08:35):
like what kind of answer is being requested by a
question that might be made up of ten different words
that are all, you know, throwing your brain in ten
different directions. Yet you can get the gist of the
question and figure out what is getting at pretty quickly actually,
And they talk about how there's been a lot of
work on how language processing works in the realm of
artificial intelligence, but at the time of this paper, there

(08:57):
was still a lot that we didn't know about the
globe whole meaning of of a sentence and and how
that's constructed in the brain. And so they summarize the
way they're starting this paper by saying, uh, it has
become widely assumed that sentences are subject to exhaustive analysis
and consistency checks during processing, but this is not the case.

(09:18):
People do not always understand what is said to them.
Sometimes they fail to understand, sometimes they misunderstand. And while
these failures of comprehension are sometimes due to lack of
appropriate knowledge or error on the part of the speaker,
there are other cases in which such failures occur when
the understander possesses all the knowledge necessary for correct understanding.

(09:40):
This paper explores such a phenomenon, and then they give
the example of the Moses illusion that we already talked about.
The question that they pose is how many animals of
each kind did Moses take on the arc? And so
what the authors here found in their original study in
eighty one was that the majority of people fail to
notice a problem with the question and simply answer to

(10:02):
despite later displaying knowledge that it was in fact Noah
and the story and not Moses, and so that it's
not that they just don't know that much about the Bible,
like they can answer the question correctly when it's posed like, hey,
was it Noah or Moses who took animals onto the ark?
They can answer that correctly and yet still fail to
notice a problem in the question, And studies find that

(10:25):
people do this even when they're not rushed. They still
make the mistake when they are given unlimited time to
think about it. Another interesting thing here they found was
that the effect is not caused by people misreading or
mishearing the question, because people still make the Moses illusion
mistake even if they themselves read the question out loud,

(10:45):
including the name Moses, so they are saying Moses out
of their own lips and they still might not notice it. Now.
In this first study, the authors conclude that what's very
important because they're getting at things about the semantics of
words and a sentence and how the meanings of sentences
are formed. They conclude that shared semantic features of the
mix up are probably significantly contributing to the effect. In

(11:08):
other words, this effect would probably not be nearly as pronounced,
maybe not even maybe totally non existent if the items
were not in some way closely related in the way
that's a two Bible characters are. If you ask, you know,
how many of each kind did Captain Hook take into
the arc, the effect probably vanishes. Another study, I was

(11:29):
looking at side at an example I found really funny,
which was how many animals of each kind did Nixon
take on the arc? And yeah, and and I like
that because they were saying, Okay, well, what if it's
just like phonological similarities, like Nixon and Noah have some similarities.
They start with the same sound, they've got the same
number of syllables. But clearly when you put Nixon in

(11:49):
the sentence, people notice. And so the Moses illusion is
just one persistent example from a class of mental phenomena
that could be called knowledge neglect act. This is a
term used by a couple of authors that will cite
later in the episode, But knowledge neglect and simplified terms,
is when you behave as if you don't know something

(12:10):
even though you definitely do know it, And the Moses
illusion is of course an example of knowledge neglect because
the problem isn't that people think Moses was the biblical
character who built the arc. You can know that it
was Noah, not Moses. If you're asked directly, you'll get
the answer right, But you don't notice the problem when
it's phrased in a question like this. And of course

(12:32):
it's not just Moses and Noah. There are plenty of
other sentences in studies that have shown the same thing.
Though it is interesting that Moses and Noah are like,
sort of the perfect example of it. I think there
might be particular characteristics of these two names and characters
that make it like that make people especially prone to
the mix up in this case, though it is true

(12:52):
for lots of other types of you know, words and objects. Well,
speaking of that, let's do a quick breakdown. I'm just
especially for folks who are not up on Moses and Noah. Uh,
just to give a little you know, basic information about
each of them, and give me the give me the
magic the gathering card on each one. Okay, well, let's start.
Let's start with with Noah. Okay, certainly the the older

(13:14):
of the two, the first that in the chronological order.
So Noah was is written as as A was an
antediluvian patriarch in Jewish, Christian, and Islamic traditions. The basic story,
God grows sick of humanity, so he tells Noah to
round up his family and two of every animal and
get them on a big old boat the arc, uh,
the first of two arcs we're going to discuss here,

(13:35):
so they alone can survive the global flood that's about
to happen. Yeah. Now, one interesting variation. I think most
people probably wouldn't even their brains wouldn't go this far
into the question. Uh, it is actually more complicated than
two of every kind, because it also says in the
Noah's ark story that I think they're supposed to bring
more of every kind of like certain types of animals,

(13:58):
like certain clean animals and just two of the unclean
animals or something. But but yeah, when you get into
the nitty gritty of it, he gets a little more complicated, right,
I mean it's all kinds of animal management, Yeah, which
I would love to see somebody fail the test of
the the Noah illusion the Moses illusion here by by
going into a lot of detail about the you know,

(14:20):
the the actual biblical text while still failing. I think
that was right. Well it was fourteen of every kind
of clean animal. Alright. Well, anyway, Noah strengths megaproject management
and animal handling. Obviously weakness, alcoholism. That's a major part
of the story. Um. Actors of note who have betrayed him. Uh.

(14:41):
This is not a complete list, but these are the
main main ones. John Houston, Russell Crowe, David thrill Full
This is a guy on Shameless. He also played Dr
John d and Elizabeth the Golden Age. John Voight, David Rentall. Uh,
David Rentall is the guy who played Aries Targarian on
the Game of Thrones show. Oh interesting, wait a Aries

(15:03):
to the Mad King. I believe. So that's the main areas, right, Okay, yeah,
well maybe I guess for some reason I thought there
was another one. I am wrong. Um, okay, so the
I've got a really funny story about John Voight playing Noah.
I remember seeing this one, oh I have. It was
made for TV. I think came out when I was
in like middle school, and it is not at all

(15:26):
faithful to the Bible and to say, very hollywooded up
version of the Noah's arks story. John Voight does play
Noah and the arc is attacked by pirates. What. Yeah,
it's attacked by like water World pirates. I mean, it
might as well be Dennis Hopper and the Smokers, but
it's actually I think they get attacked by pirates led
by the biblical character Lot. Okay, alright, Well, if that

(15:50):
is in the Bible, at least they're they're playing around
with it. Was this brought up at all when um Uh,
when Darren Arnosky was being criticized for the plot of
his Noah movie, which has like um Uh giants and
nephelim in it. Oh he I kind of liked his
Noah movie. It was way more more faithful to I

(16:11):
think it included stuff from non canonical ancient texts, but
was actually inspired by ancient texts. Okay, alright, I still
haven't seen it. It's it's been on the list for
a while. All right. Let's talk about Moses real quick, Okay.
So Moses comes later. He's an Old Testament prophet um
central figure in the narrative of the Exodus. In the account,
he helps the Jewish people in their liberation from Egypt

(16:33):
Egyptian captivity and following tim the tin Plegs of Egypt,
he assists them in the Exodus, and he also is
involved with an arc. But it's the Ark of the
Covenant that we've discussed on the show before. Not a boat,
but a golden vessel that contains sacred items. Yeah. I
would assume that the words are related because they're both
like a container of kinds, like a big box. Okay,

(16:56):
So Moses his strength community organizing of course, and sorcery
his weaknesses. This is this is kind of interesting, I guess,
because it's either not obeying God and everything or obeying
God and everything depending on who you ask, right, Uh,
I mean, if you ask God, he would say, well,
he didn't obey me in everything. That's why I didn't
get to go into the Promised Land. But especially modern

(17:18):
critics are like, it seems like he he may be
followed the letter of the law a little bit too.
Uh uh too. Seriously, I seem to recall at one
point him commanding the death penalty for a dude who
was working on the Sabbath. That seems a little harsh. Yeah,
it seems it seems a little harsh. Um. Okay, So
actors of note who have betrayed Moses, well, Charlton Heston obviously,

(17:38):
Bert Lancaster, mel Brooks, Ben Kingsley, Val Kilmer though that
that may have just been a voice role. And Christian Bale. Now,
the last one is interesting because as I was looking
at these actors was one of the interesting things is
even though they're basically interchangeable, like the same. Um, you know,
in most of these cases, you're dealing with the same

(17:59):
white dude that could play either of these characters in
a big Hollywood production. Um, but it's interesting that I
don't think anyone has actually played both Moses and Noah,
though Christian Bale reportedly came very close because Darren Aronovski
originally wanted Christian Bale to play the title role in
his Noah film, but scheduling conflicts prohibited that from happening.

(18:22):
Oh he couldn't because he was filming like Terminator Mick
g or whatever. Yeah. I don't know, but um, but
imagine if if Bale had played both Noah and Moses.
What would that have meant for the Moses illusion? Would
it have made the would would it just destroyed our
semantic understanding of reality? Maybe there's a secret council. There's

(18:42):
like no Hollywood no actor can play both of these
roles because it will totally tear our understanding of of
of facts and fiction apart. I could see that. I mean, so,
I think what some of the authors here are proposing
is that the the fact that it's not just that
Moses and Noah are words that kind of sounds similar.
They've got some similar consonants and in the same number

(19:04):
of syllables, similar vowel sounds. That's all true, and that
does seem to matter, but it's also very important that
they are semantically related, that they are both characters from
the Torah, from the Old Testament, and that sort of
links them together. And I think the more you could
do to link them even further together and associate them
in in our minds, like yes, having one actor play both,

(19:25):
I think that would actually probably make people even more susceptible. Yeah. Um,
I was thinking about this too, Like obviously we've already
touched on a few extra examples of this, but I
was trying to come up with with other examples that
would play on the same energy here, and one that
came to mind would be, Uh, if we were to
look to Chinese mythology, if we were to say, hey,
how did the Yellow Emperor decide how to order the

(19:47):
animals of the zodiac? And you might respond with, oh, well,
there's this cool little story about a race for the animals, etcetera. Um,
but it wasn't the Yellow Emperor. It was the Jade Emperor,
who's an even more primordial god ruler than the low Whimperer. Um.
So I don't know. That seems like it could be
could play in the similar could work in a similar

(20:07):
way to the Moses and Noah illusion. Or how about
this in Return of the Jedi, what was Django fet
swallowed by? Oh? I just see. For some reason, I
feel like that one doesn't work because then learn as
you as soon as you say the word django, like
people's alarms go off and like, wait a minute, what
are we talking about? Yeah, yeah, yeah, yeah, well maybe
it would. Okay, here's one for Avatar the last Airbender

(20:29):
fans out there, Um, we're hearing from several of them,
which nation was the Avatar Apa born into. I don't
know if that one worked or not, but of course
Ang is the last and not the last Avatar. A
is the avatar uh Appa is the sky by sin
that he rides on. Ah, I see, so I don't know,
ang Appa, maybe that works. Not sure? Well that went

(20:50):
over my head anyway. So you might think, well, now
that we have told told you there is such a
thing as the Moses illusion, uh you know you would
never fall for it, right because you know you will
now always having this knowledge in your mind. Notice when

(21:10):
there will be substitutions of this kind in a question
orous sentence. But it turns out that's not necessarily true.
Uh So there was this original research from nineteen eighty one,
but there have been a bunch of studies in the
decades since then replicating the original finding and further probing
the effect to figure out what's going on in our brains.
Uh So, I wanted to talk about some typical findings.

(21:32):
First of all, some things that were summarized in h
in a few literature reviews I was looking at. One
was in a book chapter by Elizabeth J. Marsh and SHARDA. Umanath.
It was a book called Processing Inaccurate Information published by
M I. T. Press inteen. That book sounds like a scream,
but their chapter is called knowledge neglect Failures to notice

(21:55):
contradictions with stored knowledge will revisit this chapter a few
times later in the episode. But but they summarize some
things about the Moses solustion. Uh. So they say that
most of the time people will fall for the Moses
solution even though they actually know the difference between Moses
and Noah, as demonstrated with later interrogation. So you can
ask people questions like who built the arc or who

(22:17):
took the animals into the arc, and they'll get the
answer right, but they still fail to notice that it's
Moses in the question. And this can be accomplished with
other similar switcher Us actually included rob a list for
you to look at of questions like this one. I like,
is um, what did Goldilocks eat at the Three Little
Pigs house? And a lot of people will just answer Porridge,

(22:38):
even though you can later ask them like, hey, whose
house did Goldilocks go into? The Three Bears or the
three Little pigs? And they of course know that it
was the bears. Now that one's interesting because for me anyway,
there's a there's an associated mental image of the bears
or the pigs. Uh. They they look rather different uh
and and ultimately they have different functions in this stories.

(23:01):
Whereas Moses and Noah are more interchangeable, and it is
the same sort of character, and there of course the
same species, because the pigs are there to be the
victims of the big bad wolf story and to get eaten,
and the bears are there too. I don't know what
just hang out in their house, I guess right. But
I can still imagine someone um falling for this or
or you know, having airing in answering this question, because

(23:24):
in a way, again you're you're racing into the finish line,
you're picking up on the you know, the basics of
the question, even though you're you're you're skipping over this,
this this this misinformation that's embedded in the middle of it. Right. Uh.
Though it's interesting that you mentioned racing to get to
the answer, I do think you're basically right about that,
Except it doesn't really seem that time is a factor here,

(23:45):
because giving people extra or even unlimited time to think
about the question does not eliminate the effect, does it.
So it doesn't seem to result from people being in
a hurry in terms of time, though I think you
could still think about it as people being in a
hurry in terms of just like wanting to get to
the part where they answer the question I don't know.
Maybe that could be like self imposed time limits, even

(24:08):
if they're not imposed by somebody externally trying to rush
you through. Now, Also, in a typical setup for these
Moses illusion experiments, readers will be warned that some questions
will contain incorrect presuppositions, so it's not just like a
trick question where they don't know this is coming. They'll
be told, Okay, some of these questions will be valid questions,

(24:28):
in which case you should just answer them, But other
questions will have incorrect presuppositions, and when you come across
one of those, you should note that the question is
not valid. Now, the interesting thing is, I would think
something like that would almost completely erase the effect, because
you're putting people on guard to be like interrogating the questions.
But it doesn't. You can put people on guard like

(24:50):
that and they still fall for the Moses illusion. In
these experiments, it does seem to be a very robust effect,
like a substantial number of people will fail to detect
to errors in questions, even though they later showed that
they possessed the knowledge to answer them correctly. Uh. The
exact percentages of the effect, though very a good bit
uh from that chapter by Martian Umanov the they right quote. Overall,

(25:14):
the Moses ilusion is robust, with readers answering from fourteen
percent to forty percent to fifty two percent to seventy
seven percent of distorted questions depending on the particular experiments.
So they're citing a number of different results there. The
fourteen percent was by Van Jarsveld Dikstra and Herman's was

(25:35):
Hannon and Donovan in two thousand, one percent was Ericson
and Mattson in N one, and seventy seven percent was
Barton and Sandford in three. And I would imagine these
differences have a lot to do with, like, what what
exactly types of warnings you're giving people ahead of time,
what exactly what exact examples are used? As we've said,

(25:56):
you know, it's it's clear that different questions are more
prone than others. So like, I think more people would
probably fall for the Moses Noah confusion than for the
three Little Pigs Three Bears confusion. Yeah, I have to
say some of the the examples that you included on
a list here, it's it's interesting to run through this
because even though I'm not encountering them as actual questions,

(26:18):
like one and someone in one of these studies would
be I can certainly pick up on the ones that
I feel like would have been more likely to fool me, like,
for instance, what kind of treated Lincoln chop down? What
kind of treated Washington chop down? Um? Like, I I
can imagine myself sort of this being a story I'm
not tremendously invested in, but have a version off stored away.

(26:40):
I can instantly skip, or even not instantly, but even
with some thought, would be like I think, yeah, cherry Tree,
Cherry Tree, that's the one, you know, even if said Lincoln. Yeah,
even if it's said Lincoln, because also I don't know Lincoln.
Something about like their stories about him, you know, we
also have sort of tall tales about him and his exploits,
and uh, the one about him, there's one about him

(27:03):
answering a duel. So I challenged him to a duel,
and he says, well, I get to choose the place
and the weapon. So I choose, Uh, let's sledge hammers
and five ft of water or something. Right there, David,
he's tall and the other person was short, something like that.
I have no idea if that's even a legitimate story,
but I have it in my head. So I have
an image of Lincoln holding some sort of a long

(27:24):
handled tool. So it fits in nicely into the story,
like I can easily overlay one over the other. Yeah.
One of the examples that I feel extremely confident that
I would not fall for is the one of what
is the name of the Mexican dip made with mashed artichokes?
I definitely, I mean, I just know artichokes. No, that

(27:47):
is not what it is. You don't mash artichokes, do you?
I mean, I haven't seen him match. He could, like
can make an artichoke paste, but artichoke guacamole. That sounds gross.
I mean, but yet arti choke death is amazing but
sound right? But anyway, So Marcia and Umanov also note

(28:09):
that um that that error detection is lower when items
uh that items are swapped are similar in a couple
of ways. We've already mentioned these, but they reiterate that
it helps when there's phonological similarity. So do the words
sound close to each other? I feel like uh, avocados
and artichokes like they have some similar vowel sounds, and
they start with the same letter, but they sound different

(28:31):
enough to me that I'm immediately strong. I think somehow
like the hard K sound coming towards the end of
the word artichoke, but coming towards the beginning of or
I guess in the middle of avocado. Somehow that makes
a big difference in my brain. And then, of course,
as we've been saying, semantic similarity, are the concept somehow
similar or related? Would we put them in a kind

(28:53):
of meaning next us together in the brain? Uh? And
and of course it's notable that the Moses versus Noah
one meets both with the criteria. They sound similar and
they're related. So anyway, it's just this interesting fact about
our brains, that something about being asked a question like this.
So trying to process a sentence like the questions in
these studies causes us to ignore the fact that the

(29:16):
contents of the sentence conflict with things that we know
to be true. And I wanted to mention one other
study I was looking at that. This one is by
Hadency Bottoms, Andrea N. Slick, and Elizabeth J. Marsh from
published in the journal Memory called Memory and the Moses
illusion failures to detect contradictions with stored knowledge yield negative

(29:37):
memorial consequences. Now we can revisit some of the things
and this more as we go on, but I just
wanted to note a few things that they bring up.
Uh So, first of all, they note some other previous
findings in their introduction. One is that um error detection improves,
so people are less likely to fall for the Moses
illusion when the error appears in what they called the

(30:00):
cleft phrase or the main focus of the sentence. So
there are ways that you can basically ask the same
question but just sort of rearrange the words to make
people more likely to notice the problem. So, if you
take the sentence how many animals of each kind did
Moses take on the arc? The word Moses is kind
of syntactically de emphasized in that sentence, you know, it's

(30:23):
not like the main focus of the way the sentence
is phrased. You can reorient the words to make Moses
more prominent in the sentence, in which case people are
more likely to catch the problem. Yeah, Like, I also
feel like having the word show up so late in
the sentence. I'm I'm, I'm like you're always predicting where
sentences are going, you know, so you've kind of already

(30:44):
filled it in to a certain extent, like you know,
you know who we're talking about. Uh, even if you
end up using the wrong name. Um, yeah, I think
you're exactly right about that. Like that, once you've heard
I don't know, you get like four or five words
into the sentence, you sort of are like you already
know what it's going to be, and you're just sort
of like okay, you like mostly ignoring the words that
come after that. Another thing that they point out that's

(31:06):
interesting is that error detection improves when questions appear in
a difficult to read font and they say this is
because it reduces processing fluency, which in turn makes material
seem less familiar and less true. And this was found
by Song and Schwartz in two thousand and eight. And this,
of course, this comes back to our old friend. Processing

(31:28):
fluency a cognitive factor that I believe is one of
the most underappreciated influences on our thoughts and beliefs and behavior.
We talked about it in our episode on the Illusory
truth Effect. Basically, processing fluency means how easy is it
for this stimulus to be processed by the brain. And uh,
and it came up in the illusory truth effect episode

(31:50):
because I remember. The illusory truth effect is where statements
you've encountered before seem more true than statements that are
new to you. And one possible explanation for this is
that familiar statements are easier for the brain to process
than unfamiliar ones are, and at some level, the brain
makes an equivalence between that processing fluency, how easy it

(32:13):
is to process this incoming sentence because it's familiar and
factual trustworthiness. They actually have nothing to do with one another,
but the brain maybe uses a little bit of shortcut there.
So are you saying that in the future for our
our shared notes, Joe, we should use chiller font instead
of whatever we're using now. Yeah, that would that make

(32:35):
it less like I mean, I think that would generally
slow us down and make it harder to do the podcast,
But it also might make it less likely that we
would just like flub words here and there, because it
would be a like really effortful, laborious process to get
through every single thought, which you know sometimes it is anyway,
but that that's on us, um but anyway, so Song

(32:58):
and Shorts here in two thousand and eight found that
simply by making statements harder to read, so you put
them in, you said Chiller, I was thinking Papyrus. I
don't know what what actual thought they used, but it
would just make people more likely to spot errors in
the questions instead of just rolling right over them without noticing.
And you know that makes sense to me, Yeah, yeah,

(33:18):
it does. It is interesting that that's how our brains work, though, Yeah,
it is sort of counterintuitive at the same time, like
you might just assume that if something's harder to read,
you would be less likely to catch errors in it.
But yeah, I think there's some kind of process where
it's like slowing you down. It's not allowing you to
just like skip over the parts that that seemed like yeah, yeah, okay,

(33:39):
Moses whatever. Yeah, it's like like a bit of food
that's extra chewy, so you're going to really taste this,
You're really going to get a feel for the text here.
There's no just wolf in this down. Yeah. Now. In
the study by Bottoms at All, they were looking at
the question of whether participants can detect errors in questions
better if there are just more errors overall in the
sample of quest gens. So, if I give you a

(34:01):
bunch of questions and like, I don't know, seventy of
them contain errors of this kind in them, are people
more likely to catch them? And it looks like the
answer is yes. Like if you if you've got people
on guard because they were just constantly problems with these questions,
their guard goes up, and they do seem to make
the Moses illusion mistake less often. And it strikes me

(34:21):
that that could be possibly, or at least partially because
once you start, you know, showing people questions where most
of them contain a problem, or even just a large
minority of them contain a problem, people probably start uh
interacting with the questions less as questions and becoming less
focused on just getting the answer and start looking at

(34:44):
them more like a puzzle where you're you're trying to
parse the sentence very clearly. Yeah. Yeah, it's like, how
is this trying to trick me? Yeah? But then there's
one kind of scary implication from this paper the author's
right quote. More generally, the failure of detect errors had
negative memorial consequences, increasing the likelihood that errors were used

(35:05):
to answer later general knowledge questions. Methodological implications of this
finding are discussed, as it suggests that typical analyzes likely
underestimate the size of the Moses illusion. Overall, answering distorted
questions can yield errors in the knowledge base. More importantly,
prior knowledge does not protect against these negative memorial consequences.

(35:28):
And Robert, I think you had a note about that.
We can talk a little bit more about that in
a bit, but yeah, basically, there there is some evidence
that just steamrolling over an incorrect fact in a sentence,
even when you know otherwise, can can later damage your
ability to recall that fact correctly. Yeah, yeah, so it

(35:49):
yeahs as we'll discuss here. It's it's not just a
situation where oh, well this is a quirk, this is interesting.
The brain does this. I mean, it is that, but
it it has it has greater implications. Yeah. Now, I
want to go back on the other side and say that, uh,
when we encounter things like this, you know, illusions that
humans often fall for when you read about a certain

(36:10):
type of I don't know, cognitive bias or or something.
I think our tendency is often to at first react
like wow, our dumb brains were so stupid. But but
I think there's another way to think about it, and
that's this, How amazing is it that we have such
a powerful command of language based reasoning that we can

(36:32):
answer questions even though key elements of the sentence do
not match with our knowledge base. I mean, think about
the trouble that a computer would run into trying to
do the same thing. Like, While it's an interesting case
of an illusion failing to notice facts that conflict with
our existing knowledge, it's also a demonstration of an absolutely

(36:55):
amazing capacity for language comprehension, even when there are severe
errors in the questions or sentences that we're trying to comprehend.
Like somehow our brains are so good at getting what
seems to be the gist the intended global meaning of
a sentence, even when pivotal items in that sentence are
wrong and should be pointing you off in the wrong

(37:17):
direction and make you totally confused. Yeah, yeah, um, you know.
I can't help but be reminded in all this about
the drawing of the bicycle that we've touched on before
about often, I mean, it's different. We're not dealing with language,
We're dealing with a uh, like a mental image. Like
we all think we have the mental image of a
bicycle pretty firm in our heads, and yet when put

(37:37):
to the test, when acts asked to draw a bicycle, um,
we're often floored. Yeah. That was a different one of
our cognitive Allusions episodes. That was the the illusion of
explanatory depth. Yeah, the issue where people they tend to
think like that they understand how something works until they're
asked to explain it. So somehow the brain has a

(37:57):
way of representing a sort of pat Tempken comprehension. You
know that it puts up this facade of yeah, you
know how that works. Yeah, I I I know how
I know the parts of a bicycle. I know all
the parts of a can opener. I could make one basically.
But then if you are asked to like explain the
steps of how it works or draw all the parts,
you're like, uh, yeah. I thought about this a lot

(38:20):
watching the Outlander TV show about the time traveler goes
back in time and she's recreating various things that she
knows about from the future, and I'm like, God, like,
how many of us, you know, we go if we
were to do that, if we were to go back
in time, we might tell somebody about all these marvelous
things like oh yeah, penicillin and uh, you know, bicycles
and whatnot, and they'd be like, oh great, how does
it work? And be like, um, yeah, no, no idea.

(38:41):
I have some some vague So I have some of
the facts in my head, but not near enough to
reproduce anything that I'm talking about. Thank thank Coming back
to this thing about how the Moses solution is, it
is and could be looked at as an example of
how amazingly adaptive at comprehension our brains are actually found

(39:06):
a book chapter discussing this very aspect of the effect.
So the authors here were hik Young Park and Lynn M.
Rader uh And this was a chapter in a book,
and the chapter was called the Moses Illusion. I think
it was published in two thousand four. And so they're
talking about different potential explanations for the Moses illusion, what's
going on in the brain, and they conclude that they

(39:29):
or at least they argue that the most likely explanation
for what's going on when we fall for this is
something they call the partial match hypothesis. So I just
want to read from their conclusion that's along the lines
of what we've just been talking about. Quote. Research on
the moses ilusion demonstrates that people have difficulty in detecting
distortions or inaccuracies when a distorted element is semantically related

(39:52):
to the theme of the sentence. Why should our cognitive
system be so tolerant of distortions and find it so
difficult to do careful matches to memory. It might seem
that partial matching is a less than ideal way to
process information. However, the partial match process is not only
common and normal, but also a necessary mechanism of our

(40:13):
cognitive system. This partial match process enables useful communication and comprehension.
Very few things that we see or here will perfectly
match the representation that we already have stored in memory.
In order to answer questions, we need to be able
to use an acceptable match. In order to understand a
new situation and map it onto something we have already

(40:35):
seen or done, we must accept slight variations every day.
At many levels, we accept slight distortions without even noticing
the process. Occasionally we notice a distortion and choose to
ignore it, but more frequently we do not even realize
that distortions have occurred, a rigid comprehension system would have

(40:56):
a difficult time. Indeed, many of our cognitive operations are
driven by familiarity based heuristics rather than careful matching operations.
The Moses illusion is an example of how the adaptive
human cognitive system works. Everyday, cognitive processing must be based
on simple heuristics, such as matching sets of features, rather

(41:17):
than exact matches, as very few tasks require exact matches.
Sentences do not match stored information, faces change, voices may
change slightly, even our pets and friends change over time. Therefore,
it makes sense that people do use partial matches in
the normal course of matching to memory. Partial matching is

(41:38):
immutable because it is the most efficient way for memory
to operate given the nature of the environment in which
we live. And so, yeah, this really makes me think
along the lines of what we were just saying a
few minutes ago. Like the Moses illusion is kind of
funny when you notice yourself doing it, but it's also
it's also kind of a superpower. Like I'm imagine if

(42:01):
you went to a video store, which we still have
one in Atlanta. Imagine you went there and you were
to say, um, yeah, I'm looking for a particular movie. Um,
it started Anthony Hopkins and it had a puppet in it.
And instead of being able to piece that together and
tell you which movie you're talking about, what if they
were to say, Okay, keep listening, I need you to
list the entire cast. I need all of the details.

(42:23):
We have to make a match here. Or yeah. Imagine
somebody comes into the video store and they say, I'm
looking for The Godfather too, and they say, sorry, we
don't have that. What they actually have is The Godfather
col In part two. Oh man, that that's not completely unbelievable,
not with our video store, but just sort of like
the cliche video store. You mean, the Godfather Part two? Philistine.

(42:50):
I mean, that's a kind of silly example. But I
think the authors of this chapter are exactly right that
every basically every single moment of our lives, we are
testing reality against our memories, and we have to do
so in a fast and loose way, And our ability
to do so in a fast and loose way, without
relying on every detail to be an exact correct match

(43:11):
is is what allows us to live adaptively, to sort
of like be thinking creatures. Looking for exact matches between
the current case you're observing and what's stored in your memory.
Like I made the comparison to a computer earlier. Today,
I guess we're more familiar with more adaptive types of
computer functions that are based on like AI or like

(43:33):
huge amounts of machine learning or something like that. It
makes me think about like the early old days of
dealing with the you know, computer programming, where like if
you slightly misspelled, like you know, um, you're you're playing
Zork or something and you type will like woolke north
the w o l K, it's not is to be
like that is not a valid action, Like, yeah, it's

(43:54):
amazing nowadays, how just like how much thumb fumbling I
can put into typing something in search and it still
knows what I'm talking about. I still um able to
floor it every now and then because I'll get really
reckless and u and it'll just have no clue. But
but more often than not, it'll it'll guess what I'm
going for. But that is amazing because that is the

(44:15):
the the the input receiver whatever, you know, this piece
of technology it's called AI. Because it's becoming more like
our brains. It's becoming usefully sloppy and and loose in
the way our brains are. Now, I guess we could
talk about a couple of other possible examples of knowledge
neglect or implications of knowledge neglect. One that I came
across that I thought was pretty funny is something that

(44:37):
is seems fairly narrow. But it's known as the yolk phenomenon. Uh,
So it goes like this apparently was originally described in
an article in the Psychological Review by Gregory Kimball and
Lawrence Pearl Mutter. Uh. This was in the year nineteen seventy,
if I didn't already say that. But it consists of
asking somebody a list of questions and and it's designed

(45:00):
to produce a certain answer. So you say, what do
we call the tree that grows from acorns? And you
say an oak? And then you say, what do you
call a funny story joke? What's the sound made by
a frog croak? What's another word for a cape cloak?
What do we call the white of an egg? And
most people say yolk um, which is obviously wrong. And

(45:22):
people are not confused about the white of an egg
being called the yolk, But it seems like instead, the
implication is that there's a certain kind of pattern seeking
that overtakes semantic processing here, like the brain starts to
conclude while you're answering these questions because of the established
pattern that rhyming is more important than the actual meaning

(45:43):
of the word that rhymes and you know it rhymes march. Yeah, exactly,
it's the rhyme as reason effects sort of. I mean, uh,
which I think we talked with that in our episode
on anti metaboli. But I was wondering, I wonder how
many items in a list like this it takes before
the majority of respondents will give the yolk type answer,
will ignore the known meaning of a word and just

(46:05):
supply the nonsensical rhyming match. I don't know. I feel
like I'm very susceptible to this one because I I
recently was trying to do a recipe and it got
kind of confusing, and I had a moment where I
had to ask myself, wait, which part is the yolk
and which is eg white. Um. It was only a
momentary lapse, but there were a lot of things going on.

(46:26):
There was a lot I was like having to take
them apart, you know, as one of those we have
to have the egg white and one bowl and the
yolks and the other. And it was I was making
a suflat, That's what it was. And yeah, and I
did had I had not had coffee yet either, so
I had that going for me. Um. It was successful.
But yeah, there was that moment where I'm like, okay,
I have to have so many egg whites and then

(46:47):
a different number of yolks and which ones which now,
so I would totally fall for this, I mean, did
you succeed? Did it rise? Yeah? I had rise. It
was good. Yeah, I don't think I want to put
it in regular their weekly rotation, but it was. It
was good for a special treat. I feel like the
soufle a that is just one of the most notorious, tricky,
tricky dishes for people who aren't I guess, like working

(47:09):
in you know, kitchens or bakeries every day. Yeah, it
was still it was tricky. It was tricky for me,
even though I went with a very what seemed like
a very simple recipe that that didn't steer me too wrong,
but still I got lost a little bit for a moment. Well,
I'm impressed, So I was. I was reading through this
book chapter as well, um on knowledge neglect by marsh
and Humana, and uh yeah, this was this was very interesting.

(47:33):
Um So yeah, they point to a couple of other misconceptions.
I don't think we've mentioned these on on the episode
thus far, but one of them was Toronto is the
capital of Canada, and a blow to the head cures amnesia,
which I guess is like a TV you know, cartoon
kind of a thing. But these are all like examples
of misconceptions that you might have in your head that

(47:54):
are are not true. They point out that you know
it tries we might misconceptions are impossible to ignore, and uh,
your best hope if you can't avoid hearing misconceptions altogether,
which again is probably impossible, uh, is to have them
immediately corrected. But that would be difficult, Like you'd have
to have like a standing conversation with somebody who would

(48:16):
not fall for your miscommunication, you know, or you'd have
to just be constantly, uh, like with with paranoia, just
fact checking everything you come across. Otherwise, some of them
are going to get past your your guard and they're
not going to be instantly corrected, and then they're just
kinda they're just kind of in there, like even if
you hear otherwise later, you might still fall back to

(48:37):
the earlier misconception. Yeah, or it's just or it's something
that doesn't come up in daily life, you know, so
you just there's never been an opportunity for it to
be corrected. I'm reminded of that episode of This American
Life where they started off by talking about this, uh,
this this particular individual who had just grown up thinking
that unicorns existed, like it had never been corrected for,

(48:59):
and so just had that misconception in her head until
finally she's at a party and there in a conversation,
like just a random chatter about hey, what are your
favorite animals or something, and she she mentions the unicorn,
and there's like this awkward silence. So why would that
be all that awkward? I mean, would she like the unicorn?
Which is real? Well, I think it was. It was

(49:20):
probably why, if I'm remembering it correctly, it was. There's
a certain bit of ambiguity where people are like if
she joking or oh my goodness, she's not joking, she thinks,
But it also makes all of us I think wonder,
which what what misconceptions do we have just rattling around
in our brain right now, we have no idea, but
they're just they're ready to go at any moment, you know,
they can be loaded into the torpedo tube of conversation

(49:43):
or podcasting or the next job interview, just just just
ready to go when you have no idea. I'd say
one of the most common edits I have to make
to this show before we release it as I realized
that I just sort of said something that I knew
was true, And then later I'm listening back to it,
I'm like, wait a minute, it I don't think that's right. Yeah, yeah,
I've definitely definitely done that before. But well, I mean

(50:05):
when I said it, I wasn't even wondering, you know,
just now now. The authors here they they touch on,
of course, the fact that the prior knowledge seems like
it should be able to protect us, uh, you know,
and and yet quote surprisingly, the effects of exposure to
misconceptions are not limited to cases where people are ignorant
of the true state of the world. We touched on
that already. Um. Another great example they bring bring out

(50:27):
is a plane crashed, where did they bury the survivors? Okay,
which you know obviously you're not going to bury survivors.
You were going to bury the dead. But again, this
is another question where you've kind of filled in all
the blanks, you know. Uh, they by the time the
survivors is the last word in the sentence. Uh, and
you fall for it, right, So it's not like you

(50:48):
think that the survivors get buried, but you could be
trying to answer the questions just because like that's gone
straight past you. Yeah, And they really drive home in
this that knowledge neglect isn't just a momentary lapse in memory,
but rather something with real consequences for memory. If you
don't recognize the error, the error can become coded into
your memory, into your worldview as fact. Uh. And because

(51:10):
that error was recently encountered, it's more easily accessed. So
again we have to remember that items in our memory
are not made of stone, they're made of clay. Merely
accessing them can change them. And our most accessed memories
are the most changed memories of all are the ones
we can trust the least. Um. So an air that
pops to mind quickly is more likely to be thought

(51:33):
of it as fact, not oh I heard once that X,
I'm not sure about X, but I think X, but
rather just X is true. X is the answer yes,
So I guess this is this is connecting back to
that finding we talked about earlier that you know, um
that even against your existing prior knowledge, like misconceptions or
errors that get by you unnoticed in one of these

(51:55):
Moses solution type sentences can later damage your ability to
remember the actual fact of that sentence correctly. Um, it
can undermine your knowledge that it was in fact Noah, potentially,
And this makes me think about the broader phenomenon of
people who are really trying to argue a point will
often structure sentences to try to get something past you

(52:18):
really quickly. In the non pivotal part of the sentence.
It's almost like we have an intuitive grasp of the
Moses illusion type thing, where like a, I don't know,
you see people like like arguing about politics on TV
or something, and like so one person will pose a
question to the other person, and the the pivotal part
of the sentence that's supposed to be in dispute maybe

(52:40):
is is one part of the sentence, But then in
a different part of the sentence, there's also like a
disputable claim that's just like shoved in there and goes
by real quick, right right, Yeah, If you end up
with a statement that has some some mistruths sort of
sprinkled in there, they're not key to the like the
main you know, talking point, or even the main untruth.

(53:00):
You know, that's that can oft often be the nefarious
thing too. It's like you catch the larger um misconception
or lie in the statement, but then there are other
lies in there that you're not paying attention to because
of the big one. Now, the authors here they point
out that improved monitoring can help you know, this is
stuff like we're talking about, like putting things into a
different font, etcetera. Um, But drawing attention to errors can

(53:23):
have the opposite effect, increasing suggestibility, which is is weird
therefore to it as an ironic effect. Um. Plus, many
manipulations designed to promote monitoring may actually fail to do so,
and they say it's difficult to predict which manipulations will
actually work. So again there's no there's no like one
guy like, here are the three steps you need to

(53:44):
take to uh to keep this misinformation from leaking into
your brain. I think a lot of what I take
away from this is that, uh, I don't know, being
well informed is an ongoing process that last your entire life.
And it's not a question of like just getting the
right facts in the bank one time and then you're set,
you know. Yeah, there's a lot of upkeep involved and

(54:07):
a lot of just continual pruning and not just new weeds,
weeds that have been in there your whole life sometimes
or seeing right they happen, and the very least um, yeah,
the authors who they also drive home that ultimately we
know a lot more about how people come to misremember
events versus misremember facts, especially when errors are are the

(54:28):
errors involved contradict stored knowledge. So uh you know, you know,
again we get into the complexity of memory, the different
types of memory that we have going on in the brain. Um,
and we we still have a lot more to learn
about just how this all comes together. Yeah. Now, you know,
here's a question that comes to mind. Um, I wonder
if anyone has constructed a Moses illusion statement using Bilbo

(54:51):
and frodo. Oh, yes, that might work. So, um, like
what was Bilbo carrying into the fires of Mountain Doom? Yeah,
that's sort of thing I don't know, of course, I
guess you would want to. You'd want to try and
construct it right so that you get Bilbo there at
the very end or Frodo at the very end, depending
on how you're you're you're messing around with it. Um,

(55:12):
who was who was the dragon whose lair was infiltrated
by Frodo Baggins. Yeah, yeah, that sort of thing that
might work. Yeah, I said, Bilbo and Frodo or even
closer together than Noah and Moses. Yeah, I mean they
are certainly that they actually overlap, as opposed to being
separated by by long stretches of time. Very very similar characters,

(55:33):
actually related, right, they are related? Yeah? Um? Yeah, what uncle,
great uncle Uncle? I always forget what happened to Frodo's parents.
I've read it and I still forget it. I'm gonna say, uncle,
all the all the Hobbits are cousins. Yeah, they're all related. Actually, yes,
all right, Well, there you have it. We'd love to

(55:54):
hear from everybody about this, because, of course, this just
touches on how our brains work and how it now
they are brains work with with new information, be it
accurate or or or or a misconception. Uh so, I
think everybody out there has something to share. Which of
these Moses illusions worked the most on you? Which ones
I've worked on you in the past. Uh, we'd love

(56:15):
to hear from you, all right. If you want to
check out other episodes of Stuff to Blow your Mind,
you know where to find it. You can find the
Stuff to Blow your Mind podcast feed wherever you get
your your podcast, and we'll have core episodes of Stuff
to Blow your Mind on Tuesdays and Thursdays. You've got
listener mail on Monday's, you've got them, We've got the
Artifact on Wednesdays. You've got Weird House Cinema on Fridays
in a vault episode on the weekends. Huge things. As

(56:38):
always to our excellent audio producer Seth Nicholas Johnson. If
you would like to get in touch with us with
feedback on this episode or any other, to suggest a
topic for the future, or just to say hello, you
can email us at contact. That's Stuff to Blow your
Mind dot com. Stuff to Blow Your Mind is production

(57:01):
of I Heart Radio. For more podcasts for my heart Radio,
this is the I Heart Radio app, Apple Podcasts or
wherever you listening to your favorite shows,