Video summary
The speaker addresses the controversy surrounding his "Dreaming Spanish" challenge and its relationship to Automatic Language Growth (ALG), arguing that the method has been fundamentally misrepresented by its proponents. He draws a parallel between ending his language experiment and leaving a difficult relationship, noting that despite moving on to test new software like Rosetta Stone, he is still receiving aggressive messages from former supporters who are unhappy with his conclusions. The speaker contends that the toxicity of the discussion stems from how ALG was framed in his videos, which inadvertently attracted critics who then conflated his specific experiences with the broader principles of Comprehensible Input (CI) and ALG itself. He emphasizes that while he tested a specific implementation of ALG via Dreaming Spanish, this does not invalidate the core theories of CI or the entire ALG framework, as many detractors are misinterpreting his results based on academic illiteracy or vested interests.
A central argument in the video is that any attempt to scientifically test or disprove ALG using only Dreaming Spanish is invalid because the two methods differ significantly in their core mechanisms and progression models. The speaker explains that Dreaming Spanish was an adaptation of ALG that only covered the first five levels, whereas the full ALG curriculum claims to require up to 2,000 hours to reach a native-like level. He points out that Dreaming Spanish's "silent period" philosophy does not align with the complete ALG theory, which involves specific grammar analysis and pronunciation drills that were omitted in his experiment. Furthermore, he highlights that the original school teaching ALG no longer exists, making it impossible to follow the method exactly as intended, and suggests that Pablo Roman, the creator of Dreaming Spanish, misinterpreted the foundation of ALG by stopping early in the program, leading to a flawed version that cannot accurately represent the full methodology.
The speaker concludes that ALG is inherently untestable because its core pillar forbids learners from thinking about the language or analyzing grammar, yet any observation made during an experiment—such as noticing gender patterns in Spanish or realizing one understands only 20% of content—is immediately labeled as "thinking" and thus a failure. He argues that this circular logic makes it impossible to gather data without violating the method's own rules, rendering experiments like his pointless for proving or disproving the theory. While he maintains that ALG is likely the worst overall method for learning a language compared to immersion schools, he advises viewers not to abandon language learning entirely if other methods fail them. Ultimately, he urges people to seek genuine immersion experiences rather than sitting at home attempting cross-talk drills, and he plans to return to traditional tools like Rosetta Stone, acknowledging that his experiment was more of a demonstration of the method's impracticality than a definitive scientific refutation.
Read the full video transcript
What's up, grinders? Welcome back to the
channel. Have you ever had an
ex-girlfriend, ex-boyfriend, possibly?
You know, someone you spent a lot of
time with and then [ __ ] just wasn't
working out. It wasn't going as planned.
And then one day you decided to pull the
pin on the [ __ ] and you walked out. You
went your separate way. You said, "I
ain't doing this no more. I'm moving
on." Now, you might have lied about why
you're moving on. You know, it's it's
it's not you, it's me. But
realistically, at the end of the day,
you just wanted to get the [ __ ] out of
the door. You were done. Okay? And then
you finally walked out. Breathe that
fresh air. Oh, thank God. And you
started meeting other people. Well, that
is exactly what happened with me and
ALG. Automatic language growth. Okay. I
moved on. And just like an ex-boyfriend
or an ex-girlfriend, they [ __ ]
started texting you. They started
sending you messages, aggressive [ __ ]
messages. You thought the [ __ ] was
sorted. You thought you'd said
everything. You guys were done. You
moved on. And yet, out of the blue,
these messages kept coming. And that was
me after I ended my ALG experiment and I
moved on. Like I'm literally testing new
software right now. I'm testing Rosetta
Stone with Spanish. Okay. And I don't
know, maybe that just set off a couple
of [ __ ] flares or something. But
suddenly there is a very very small
number but very loud number of ALG
proponents who were very very unhappy
about these recent changes in my uh
relationship status with languages.
Okay? because uh well for one the person
who posted like thousand comments in the
ALG thread within my uh discord well
they left my discord but before leaving
they packed up all their [ __ ] deleted
every message and then left like they
[ __ ] took everything with them they
were not happy okay they didn't even say
goodbye and then then this happens
someone posts this on my Reddit my
subreddit okay now if you don't know I
have a subreddit now I don't talk about
it very often cuz mainly I hang out in
the Discord and I go over to Reddit too
often because most of the comments on
Reddit are saying [ __ ] like, you know,
Evil Deer is a [ __ ] a [ __ ]
artist. I hate that guy. [ __ ] him. I
hope he gets, you know, balding cancer
or something. You know, they say stuff
like that and it's like, I can't bother
reading that unless it's for content,
then I'll read it. But I'm not going to
read that in my downtime type thing. But
this post came in and this is like one
of those letters, okay? You know the the
long text messages, the one where they
send it, but it's so long they send you
half of it and then you see dot dot dot
and then the next half comes and then
dot dot dot and the next half comes.
This is that. We're going to be reading
through this to find out what I did
wrong. Hate dreaming Spanish all you
like, but don't drag ALG and CI down
with it. Okay, this is everything that
was unsaid now being said. I write this
because I think the entire discussion
around Evil Deer's dreaming Spanish
experiment has become toxic. I think
that Evil Deer's experiences are valid,
but the way he framed them has caused
unnecessary toxicity. Okay, here we go.
Let's find out how I [ __ ] destroyed
something else. Much of the toxicity
stems from how Evil Deer structured his
videos in a way that attracted the most
prient
This is going to be one of those English
situations, isn't it, guys? It's going
to be You know what? Let's Let's [ __ ] I
don't care. We're going to We're going
to push our way through this. Most
vested interest affected haters of ALG
and CI in general. What? [laughter]
Somehow my experiment got haters of both
ALG and CI. You know, I'm going to be
honest. I'm pretty sure the vast
majority of my audience had never heard
of ALG before I started my experiment.
ALG 2 years ago was not exactly
something that everyone knew about,
okay? People were talking about learning
like a child and there was like various
implementations of that, but ALG was not
something that, you know, every [ __ ]
language channel spoke about. I think
like Matt verse Japan spoke about it and
then like maybe a couple has dropped
some hints like other YouTubers dropped
some hints but like barely anyone's
speaking about but like haters of CI.
Who the [ __ ] is a hater of CI? It ain't
me. I got an entire course where I teach
Espano via direct method which is
literally CI. Where's the CI haters at
guys? Let me know. Like [laughter] could
you imagine being a CI hater and then
also trying to learn a language? Yeah,
guys know you can't use comprehensible
input to learn a language. How do you do
it? I do it via incomprehensible input.
It's the only way, boys. If you're not
doing it hard, you're not doing it at
all. But a great deal also stems from
these people's reactions and academic
illiteracy. Well, I do have a problem
around illiteracy. Okay, so maybe maybe
I am within this category right here.
Unsurprisingly, big words of these
vested interest attackers attracted the
zealots from the other side. And what
you get is the sacking of the Great
Library of Alexandria. [laughter]
Bro, ALG is the Great Library of
Alexandria. Jesus Christ. I didn't
realize my series had burnt down one of
Earth's greatest [ __ ] collections of
literature. Wow. If I become like a a
tyrant, a a titan of this world who
takes over and controls all power, I'm
going to organize a book burning, but
just for ALG. and all ALG proponents are
going to be burnt. I will burn you all
[laughter]
for stealing a thousand hours of my
life. Please don't clip that and take it
[laughter] out of context. There is a
proverb that says if you repeat a lie
often enough, it starts sounding like
the truth. You know who said that
proverb? Maybe there's a reason why you
didn't say who said that proverb. That
was a Nazi who said that prophet. Are
you saying that I'm a Nazi? I am half
German, but I'm half good German. I
hope. Honestly, I don't know. [laughter]
Could be half bad German. Oh [ __ ] I'm
learning Argentinian Spanish, too.
I'm just asking for problems. P1, if B
is derived from A and B is false, then A
must be false. Okay, we're getting into
the science of it. P2, ALG is derived
from comprehensible input. P3, Dreaming
Spanish is derived from AG. P4, Evil
De's experiment disproved Dreaming
Spanish, therefore ALG is false and CI
is false. I don't know who's making that
conclusion. All that happened with me is
I got to 1,100 hours or something
[ __ ] of dreaming Spanish AG and I was
just like a bro I'm not doing this I'm
not doing this for another th00and 2,000
hours to reach the mythical layer of
enlightenment you know and absolute
fluency of a language like sure being a
native speaker of a language would be
great and all but I had not seen anyone
in ALG yet reach the you know the
pinnacle of native level so I wasn't
sure how any more thousands of hours
would be necessary. And I just went, "G,
I'm I'm calling. I'm happy just being an
Australian Argentinian speaker at this
rate, you know, and claiming CI is
false." I dare you to find a clip of me
where I say CI is false. There probably
is a clip of me out there somewhere. I
say all sorts of stupid [ __ ] Probably
just then. You could probably literally
clip just that part and like farm that.
P1 commits what philosophers of science
call the Dunham Quinn problem. See, this
is the problem. You've you've already
done one problem yourself.
You're assuming that I'm claiming CI is
false. I ain't claiming CI is false.
Therefore, anything you write down here
is probably going to be based on a straw
man because I'm not claiming that CI is
false. I'm not even claiming that ALG is
false. I'm just not willing to do it for
another couple of thousand hours to get
to a B2 level, you know? Plus, I think
my audience kind of got sick of
listening to me saying the same [ __ ]
over and over, which is basically just
me, guys. Guys, I don't think I've
learned anything. [laughter] Next 50our
update, guys. Guys, pretty sure I
haven't learned anything next 50 hour
update. Guys, guys, I'm starting to lose
the plot. Like, I I think there's only
so much my audience can take of that
before they start unsubing just due to
the fact that they get angry at me for
being such a loser at Spanish. Okay, CI
does not equal dreaming Spanish. We
agree. Comprehensible Japanese, Korean
from scratch, easy English, etc., etc.,
etc. I should make that a website domain
etc etc etc and then that teaches you
Russian. These are all curated content
libraries that try to implement the CI
principle. However, you do not need
these curated content. I know I agree. I
have a I have my own content library
where I teach Espiranto via direct
method using just CI. Here is a proper
definition of CI. Oh, okay. We needed a
definition. Procrion. any content in the
L2 in which you can comprehend 70 to
80%. In that scenario, your language
skill will improve by learning the
remaining 20%. I'm pretty sure he
doesn't say 70 to 80%. He probably says
something like 80 to 90%. But
realistically, if you're going the ALG
route, you want 95 to 98% because from
memory of my readings of crashing, okay,
if you're less than the 90% range, you
need assistance with CI. Anything above
the 90% 95 98% range, you don't need
that assistance anymore. you can
actually literally just learn via
context. Okay, so ALG is trying to work
within that tiny little narrow slither
right there at the end. So for instance,
my CI for Japanese is the prime
minister's assertion speech and my CI
for Spanish is the papal ancille. I
don't know what that is. I'm not
Christian. Uh obviously you do not find
these on comprehensible Japanese and
dreaming Spanish. Content ideas for
Andreas. For reference, my native
languages are English and Mandarin
Chinese, my heritage languages which I
am A1 level up. uh Hakka Chinese and
Hyanese. Okay, so this is not the same
person as the one that was in my Discord
because his native language was uh
Portuguese. And it took about 500 to
a,000 comments to find out that he had
done ALG for Spanish.
Yeah. And he was very successful. Just
just putting that out there. And in case
some people want to howl that the
numbers are unscientific, 70 to 80%
actually lines up the human cognitive
capacity at remembering discrete facts.
See any of this, this does not matter
because ALG has to aim for the 95 to 98%
so that you can do it unassisted. So why
are we hopping on about 70 to 80%. 78%
is great, I guess, if you're using
assisted lookups and stuff, but yeah,
that's pretty much it. And then you got
here from far from being a fringe
theory, CI is the core foundation of
modern linguistics. Did I say it's not?
Pretty sure I didn't say it's not. I'm
pretty sure I just tested ALG, which by
the way is not CI. I just tested ALG.
That's pretty much it. Grammar drills,
pronunciation drills, etc. Things that
academic linguists rejected in the 1970s
before Evil Deer was born to boot. I'm
pretty sure you can find plenty of
modern linguistics. By the way, if you
guys are a linguist or a second language
acquisition, you know, expert, feel free
to drop down in the comments section a
very, very thoughtout comment on whether
this is true or not. And then I can pin
that [ __ ] for uh anyone who's watching
this video. However, the drills have
survived because modern textbooks have
rebranded the drills under communicative
learning, output practice, interactive
agency, uh leading both Swain and Long
to denounce the industry for
misappropriating their research to brand
things that they were wholeheartedly
opposed. I'm not going to go down the
route of trying to figure out if that's
true, as in go to Google, go search, go
look for that type of stuff. I could
totally imagine these guys, you know,
actually having a go at the industry
because the industry is doing basically
what ALG is doing to CI where they're
taking an idea that is true and then
taking it to the absolute extreme
because there is like you've literally
got [ __ ] Mel out here with a
spreadsheet telling you to just drill a
[ __ ] 20,000 sentences one after
another and that's how you're going to
learn a language. Okay? There's lots of
people who take things that are true and
then take it to an extreme point where
it's just complete [ __ ] at that
point. Swain would be horrified that
people are using her theory to justify
evil deer's production anky cards. Well,
they work perfectly fine for me. Okay.
So, if Swain has a problem with me, she
can get in the ring. We'll [ __ ] box
it out. I'm just going based on what's
worked. Literally what's worked.
Production cards have worked perfectly
fine for me in the languages I've been
learning. And this leads on to AG. Okay.
Funnily enough, the interaction
hypothesis from Michael Long properly
understood perfectly backs the
principles of cross talk. Cross talk
requires you to negotiate meaning with
your native interlock using your L1.
This purposeful communication that fits
the interaction hypothesis to a T. You
know what also fits it perfectly to a T?
Speaking to them in the language you're
learning. You [laughter] could negotiate
the meaning with them while also using
the language. The reason ALG likes cross
talk, okay, is only for one reason.
Because they're scared of damage. They
do not want you to accidentally try to
force out the language before you're
ready and damage your ability to reach
that absolute top level of damn near
native fluency or actually native
fluency. And if you don't believe me
about that, there's actually I'm going
to pull up something from the Dreaming
Spanish road map which they got rid of
very very long time ago. Okay? But this
is what was on the original Dreaming
Spanish road map. See this? This is
level seven. You are for all practical
purposes comparable to a native speaker.
You can understand more formal speech
and writing newspapers, novels or
technical texts in your field without
effort. You can understand any kind of
TV show or movie. Uh be it scripted or
unscripted. By this point, you also have
a good grasp of the country's pop
culture and you understand many of the
cultural references and TV shows. You
speak fluently and effortlessly and you
feel in control of the language. You may
still make some mistakes or miss a
specific word, but it doesn't hinder you
from being an effective member of
society. At 1,500
hours, that was the original claim that
you are for all purposes comparable to a
native speaker. Okay? Now, credit to
Dreaming Spanish for obviously editing
that and pulling back upon it, but
that's what basically ALG claims. And
not only do they claim it to that point,
they claim it to the point that you will
reach a level in the language that no
manual learner could ever feasibly
reach. And you will always be in this
weird state of uh there's a block
between you and the language and you
have to go via this type of like random
[ __ ] thing that's sitting there. But
you know, because you're not a native
speaker, ALG reaches native speaker, you
become a native speaker. That's the
difference between ALG and other
methods. their belief around that. Okay,
with regards to output, we hit the next
lie. The mythical view of the silent
period fluency. This is Dreaming
Spanish's big misrepresentation. Now,
this I find fascinating. When I first
looked into doing ALG, I looked into
resources for it, okay? And I remember
the ALG, some members of the ALG
community, which let's be honest is
[ __ ] tiny, okay? Were talking about
how Dreaming Spanish represents the
future of language learning because they
implement ALG. If you want to learn
Spanish, become a fluent native speaker,
basically you got to go do Dreaming
Spanish. And this was proof that the
industry, the language learning industry
was finally finally understanding what
you needed to do. And obviously since
then, there's been a lot more websites
which have popped up for different
languages implementing the same method
as Dreaming Spanish. But now after I
have done 1,000 plus hours of this
method and decided to call it quits,
suddenly [ __ ] Dreaming Spanish ain't
alg [ __ ] poof. They just threw
Dreaming Spanish under THE GODDAMN BUS.
WHAT'S GOING ON HERE? Do you Do you want
to take credit or not? Okay. Like if you
don't want if you want to say that
they're not ALG at all. Fine. Okay. But
you're going to have to fight with Pablo
over that one. He went to the original
school that invented the method and
taught the method. Okay. Now, I know the
school doesn't teach the method anymore,
but you're gonna have a bit of a hard
time fighting with the guy who literally
went there and learned there. And based
on comments on my videos of others who
literally went to that school, he's not
far off from the mark. Ag has 10 levels
across four tiers. Beginner,
intermediate, advanced, expert. Notice
where the tier is left off in Pablo's
biographies, the expert tier. Now,
because ALG tie tiers intersect with
each other, you could really reach a low
advanced level. This is actually a very,
very obvious business decision. Why
would Pablo invest heavily into making
expert level content in Spanish when
people are obviously going to stop
paying for the service at that point?
Instead, he made a super beginner level
because he noticed, "Oh, we have a
bigger problem." The bigger problem is
just jumping straight into beginner is
actually way harder for people to be
able to endure it. Like, you have to be
a diehard believer in the method to sit
through beginner content and acquire
enough of the language to be able to
move on to an intermediate. That's why
they implemented super beginner. And
also you'll notice that super beginner
is basically direct method. They lit
they almost teach via direct method. The
only aspect of direct method that they
don't utilize in their super beginner
videos is structure via a series but on
an individual videobyvideo basis it's
basically that. And for those who don't
know what direct method is. Uh it's
basically where you teach the language
via the language. And I've made videos
about direct method. I love it. It's a
CI very very very heavy input method
which is why I also find it absolutely
insane when people say I'm anti-CI
because I literally have an Esbranto
course where I teach uh Esbranto via CI
via direct method. And then we've got a
link to the Holly Bible of ALGWorld.com.
ALG's TI's official claim is that you
will be at a native adult level fluency
if you complete all 10 levels of their
program. Okay. So let's go have a look
over here. So you got level 1 2 3 4 5 6
5 10. Okay, 10 is expert. Now let's see
what that says. So to create groups of
people, we design AG to consist of 10
levels where each level is 200 hours.
These 10 levels are divided into four
groups. Basic, intermediate, advanced,
expert. The diagram below shows how each
level uh roughly coincides with the
language development of his child in his
native language. So how many hours are
we looking at? You got 200. Yeah. Times
that by 10.
You guys can do the math. Yeah. Yeah.
2,000. Just Just making sure. That means
I was halfway I was over halfway
through. Do I feel like if I slept
another 900 or actually less than that?
Another what was it like 850 hours or
something on top? Would I actually feel
like basically a native expert level at
that point? No. And that's kind of why I
pulled out. I was not prepared to spend
the extra 1,000 2,000 3,000 hours on top
to reach this mythical level. However,
that is still only half the program. Oh,
okay. And also, as per the program's
design, the actual acceleration is meant
to occur at levels 5 to 7, which are
meant to correspond to the linguistic
development stages of a native 2 to
5year-old, what is known as the blowfish
period. So, Dreaming Spanish's
fundamental design philosophy got ALG
entirely wrong. Pablo's ALG stopped at
the beginning of the middle of level 5,
and he implemented Dreaming Spanish as
an adaptation of ALG based on his
experiences of level 1 to 5 only. So,
the claim here is that Pablo [ __ ] up
ALG because Pablo only ever got to level
five of their program and that's why his
Spanish, sorry, that's why his tie is
not native level and also
why dreaming Spanish is wrong basically.
But what additional would they do after
those levels? It's this the same stuff
just over and over and over. I guess you
could claim that maybe with Spanish they
should be saying the silent period
should go for 2,000 hours rather than
1,000 hours. I don't know. Thus, any
experiment that attempts to disprove ALG
by only using Dreaming Spanish is
invalid because Dreaming Spanish
enhanced the derived experiment missed
the core mechanism of ALG claims to
follow. You know, the funny thing is
I've never claimed to disprove ALG. All
I did was show you what would happen if
you went through this methodology. And
if you say I'm doing AOG wrong, well
then everyone's doing it wrong because
the school no exists no more. If the
requirement is you go to the school in
order to do ALG properly, well then if
you can't go to the school, which by the
way doesn't exist anymore, then don't
bother with ALG. You're [ __ ] anyway.
There's no point doing it. But if the
claim is, you know, I I still got to go
for a couple of thousand more hours,
well, I'm not prepared to commit. Okay,
I'm just not prepared to commit. Now,
whether you take my experiences as me
disproving ALG or not, that's on you.
Obviously, the vast majority of people
who watched my series and left comments
on my YouTube channel said, "Yeah, evil
dear. Thousands a thousand hours is
enough to tell me that I don't want to
do this." Okay, that this seems like
crap to me. I'm sorry. Obviously, it's
not a definitive stamp of yes, it's 100%
[ __ ] or yes, it's 100% correct. I'm
just not prepared to go the extra
distance. Like I thought originally I
was going to be doing the 1,500 hours. I
did not know they would be clocking up
to 2 3,000 hours, which is what everyone
in the ALG community started saying
after a while where they're like,
"Actually, Evil Deer, if you count all
the extra stuff you got to do, you know,
the unspoken stuff
like 3,000 hours, you know, it's not
that bad. You soon you'll be halfway,
bro. I
I got other challenges I got to do,
okay? I can't be I can't just be doing
the same [ __ ] over and over. My YouTube
community will kill me. They're already
sick of listening to me speak [ __ ]
bad Spanish. You know what? Ag probably
[ __ ] me permanently with Spanish. I'm
probably going to permanently have a bad
accent in Spanish. Anyway, whatever. In
addition, ALG tie is a curriculum with a
greater progression. ALG 1 to 10 stages
in a course, not proficiency levels and
uh uh to hit like DS1 to7. The key
difference is that it is a greater
progression like ALG Thai. Each level
builds on the level before. One Dreaming
Spanish video does not necessarily build
on the next one you watch because you
may switch between proficiency levels or
switch between Augustina Shell and
Pablo. That sounds more like direct
method to me because that's what direct
method does. It builds on each previous
video introducing more and more of the
language. But the difference obviously
is with direct method they like to
actually also teach the grammar of the
language via the language as part of
that progression journey which would be
a difference to ALG. Now maybe you can
claim that's what Dreaming Spanish done
wrong but I never went to the ALG Thai
school. So, for those who did, because I
know there's a few of you in the comment
section, uh, let me know if that is
true. If every single lesson built 100%
on the previous, or was it more like
what Dreaming Spanish does? Pablo and
Diaz misinterpret the foundation of ALG
as entire the entirety of ALG. To be
clear, this is not evil deer's fault.
This is Pablo Roman's fault. Thank God
something's not my fault for once. Oh my
god, I'm so glad to not be blamed.
[ __ ] hell, Pablo. What did you do to
me, bro? You I'm I'm out here taking
[ __ ] strays for you, man. So, what is
the true purpose of the silent period?
The true purpose of the silent period is
to build the foundational fluency in the
L2 so that your L2 does not build off of
your L1 whether it is in pronunciation,
morphology or syntax. Yeah, I've already
spoken about that in the ALG. I've
spoken about that as ALG's belief around
it
which is more the pronunciation side. I
was talking about the accent and stuff
but I did not find that to be happening
for myself. Now obviously people have
made up many reasons about why I failed
at AOG and those reasons are around the
fact that apparently I'm thinking about
the language or something like people
have lots of different ways of looking
into my mind and knowing what I'm doing
but apparently that's it. So for
instance when evil deer confuses the
Spanish and Esprano words for green that
is because his brain consciously or
unconsciously built an understanding of
the Spanish word using the Espano word
that's because in Spanish it is ver and
espiranto it is vera. They semantically
cover the exact same space. So that's
why I confused them, okay, when I was
pronouncing them because the brain likes
taking shortcuts. This is an example of
morph morphological interference.
Goddamn big words, man. So therefore, if
I if that means I fail at ALG because I
speak another language, wouldn't that
discredit the claim of ALG that if you
speak a romance language and then you
learn Spanish, it's only half the time
of what I'm doing? because you would
have so much more interference, wouldn't
you? If we're going this route, then the
best thing to do would to be speaking no
language in any way related to your
target language. And if it is related
to, you know, your native language, then
don't learn that language. It's not
suitable for ALG. So maybe I should have
done North Korean. Another example is
where evil deer split the Chinese
phrase. I remember this in English
semantic chucks to match the English
translation every other week. However,
Joel in this context belongs to the
lexical class of um
that doesn't exist in English. So that
is actually an example of syntactic.
Okay, you can say that I'm using English
here to understand Chinese and that is
true. I am. But once you've practiced it
enough, that falls away. You don't have
to worry about that stuff anymore. So
yes, you're going to use your native
language.
And that's what I did with Chinese.
Okay. To help you basically demystify
Chinese, put parts to bet together, and
then once you practice it, it slowly
falls away. Dreaming Spanish's core
misinterpretation of ALG means that any
experiment that purports to use Dreaming
Spanish as a test case for ALG needs to
be furnished with several caveats. That
is not cultism. That is a proper
experimental design. The problem is you
cannot experimentally test ALG. And I've
learned this via my 1000 hours because
there is one claim of ALG that will
discredit anyone if you want to
discredit them. And the claim is they
are thinking about the language. That is
one of the core pillars of ALG is you're
not meant to think about the language.
So, at any point if you're going in a
direction that I am unhappy with, even
in like a a full-on experiment where you
lock people in a room for [ __ ] two
years with natives and you do the whole
in-classroom style teaching. Okay, for
those who fail, you can say problem is
they thought about it. They were sitting
there and they were trying to mentally
unpack what they were hearing. So, they
failed. And if anyone is successful, you
can say, well, they didn't think about
it. And yet, you have no idea of knowing
that's true or not. Okay? So we cannot
scientifically test AG. It will never be
scientifically testable. But it still
takes two hands to clap. The other
reason why Evil Deer's approach is not
following AG is because even from his
first update onwards, he's still
tracking how many words he has learned
or says things like, I would know this
word if I was using any other method. In
pure AG, words are an artifact of
grammar analysis. They don't exist.
Well, you shouldn't be doing any grammar
analysis. They don't exist in real
language. And it is also scientifically
proven the fact that by looking at a
waveform of any speech alone, you cannot
tell which section constitutes which
word. And you know the funny thing is
that's like saying color doesn't exist
because if you go down into it, that's
literally just light bouncing off of a
surface and it's bouncing off of the
surface at a different waveform. Sure,
you could debate this on some individual
level that you cannot see the words as a
waveform, but we as humans do not deal
with words on a waveform. We deal with
them as audible chunks that we hear and
we notice patterns within those chunks.
Okay. So, I don't know why you're trying
to debate this here like what what is
this? Now, this is something that even
most of the ALG supporters can't notice
because they don't have a deep
understanding of ALG. So, I don't know
this person. Maybe maybe they went to
the school where the one school that
implemented ALG. Maybe they went through
the entire program and therefore they
have a greater understand. Maybe this is
literally David Long. I have no idea.
Okay, but unless you went through the
same amount of program as Pablo, I don't
think you can be taking shots of Pablo
because uh the dude went through the
program. Okay, if anyone's got a good
understanding of it, it's him. And if
anyone else has got a secondary good
understanding understanding of it, it's
me because I did the Dreaming Spanish
implementation of it. So, Dreaming
Spanish still holds the majority of the
blame. In conclusion, when these two
approaches fuse together, they create an
experiment that is good for testing
dreaming Spanish perhaps, but not for
testing ALG as a method in general, and
definitely not CI as a principle. I'm
going to say this. If anyone out there
decides that they want to redo my
experiment and test ALG, don't. It is
untestable. You cannot test it because
you will just have people in the
comments section saying, "You're
thinking about it. That's why you
failed." You couldn't even show proof of
you doing the method without
accidentally failing. Because what would
you do? You'd watch a video and then you
would say, "Yeah, guys, I understood
about 20% of that."
That's the best you could do. Because if
you tried to translate and show your
audience what you understand and allow
them to check you on that, you would
fail. That is deemed as basically a form
of interference, damage, because you are
mentally translating. you are now seeing
that and you are doing a mental
translation to your native language.
That's damage. You fail. So literally,
if you were to do this experiment, you
would have to sit there every 50 hours
or whatever time period you wanted to do
and just say, "Yeah, I understand like
[ __ ] 30% of that seems good." And
that would be it. You could not talk
about any observations. You couldn't
say, you know, that there's an
interesting thing about Spanish. You
know, I've noticed that the endings seem
to change on words like just I think
it's based around gender or something.
You're kind of like a gendered language
maybe. Boom. You failed. That's a
failure right there. You shouldn't have
noticed that. Like it's it's not
testable. So don't try to [ __ ] it. I
maybe I should have been more literate
at the beginning and read into the
comments of all the ALG people before I
tried doing my test because then I would
know it's not testable and it would have
been a pointless exercise. A bunch of
you guys asked me for a final analysis
on ALG what I think about the method. It
is by far the worst method to learn a
language. For the listening portion, it
may be the easiest method if
you have a lot of time available. But
for just learning a language overall, it
is by far the worst method. Now, does
that mean we should just chuck it out
and not do it? No. If you've tried
everything else and it doesn't stick and
at the end of the day, you like doing
ALG, [ __ ] do it, bro. I'd rather you
learn a language than not. That's pretty
much my thoughts on it. If anyone comes
up to me and they're like, "Hey, Evil
Deer, so um I'm thinking about going to
this uh this immersion school where they
use textbooks and stuff and they they
get you to do out practice or I'm
thinking about, you know, um sitting at
home and hiring cross talk teachers for
a thousand hours while I just speak in
English. Which method do you recommend?
Go to the [ __ ] immersion school. What
are you on about, you [ __ ] idiot?
What are you doing? What are you doing?
Why would you sit at home doing cross
talk for a thousand hours? You're not
even speaking." Like that's what I'd
say. And if the person went, "But evil
dear, I want to sound like a native."
I'd be like, "Then find me a native ag
because I am yet the fine white." Cuz
that's the only claim that they have at
the end of the day. That is that's the
reason why you should go through all of
this is that you reached this mythical
native level. Maybe that's going to be
my last video, but who knows? Maybe uh
this is going to set off a new
firestorm. Uh back to Rosetta Stone for
me. I'll see you all in the next video.